1 DEPARTMENT OF HEALTH AND HUMAN SERVICES FOOD AND DRUG ADMINISTRATION CENTER FOR DRUG EVALUATION AND RESEARCH PERIPHERAL AND CENTRAL NERVOUS SYSTEM DRUGS ADVISORY COMMITTEE Wednesday, June 6, 2001 8:15 a.m. Holiday Inn Bethesda, Maryland 2 PARTICIPANTS Claudia H. Kawas, M.D., Consultant and Acting Chairman Sandra Titus, Ph.D., Executive Secretary MEMBERS: Ella P. Lacey, Ph.D., Consumer Representative, LaRoy P. Penix, M.D. Richard D. Penn, M.D. Gerald Van Belle, Ph.D. CONSULTANTS: Gustavo C. Roman, M.D. Jerry S. Wolinsky M.D. XYREM CONSULTANTS: VOTING: Pippa Simpson, Ph.D. Carol Falkowski, Ph.D. NON-VOTING: Christine A. Sannerud, Ph.D. Jerry Frankenheim, Ph.D. Jo-Ellen Dyer, Ph.D. ON PONE-LINK - NON-VOTING: Ronald Chervin, M.D. Christian Guilleminault, M.D. FDA: Robert Temple, M.D. Russell Katz, M.D. Ranjit Mani, M.D. John Feeney, M.D. Deborah B. Leiderman, M.D. 3 C O N T E N T S Call to Order and Introductions 4 Conflict of Interest 6 FDA Overview, Russell Katz, M.D. 8 Orphan Medical Presentation: Introduction, David Reardan, Ph.D. 19 Medical Need, Emmanuel Mignot, M.D. 25 Efficacy, William Houghton, M.D. 36 Polysomnographic Effects of Xyrem, Jed Black, M.D. 55 Safety and Summary of Risk/Benefit Assessment, William Houghton, M.D. 61 FDA Response to the Presentation, Ranjit Mani, M.D. 84 Committee Discussion and Deliberations 89 FDA Invited Speakers on Risk Management Issues: Epidemiology of GHB Abuse Issues, Carol Falkowski 131 Adverse Medical Effects with GHB, Jo Ellen Dyer 148 Sponsor Presentation on Risk Management and Abuse Liability, Bob Balster, Ph.D. 162 Risk Management, Patti Engel, R.N., BSN 176 Open Public Hearing: Sharon Fitzgerald, Littleton, Colorado 187 Richard L. Gelulla, MSW, National Sleep Foundation 191 Abbey S. Meyers, National Organization for Rare Disorders, Inc. 197 Robert L. Cloud, Narcolepsy Network 200 Cindy Pekarick 204 Eric C. Strain, M.D., College on Problems of Drug Dependence 208 Deborah Zvorsec, Ph.D., Hennepin County Medical Center 213 Trinka Porrata, LAPD 218 Matt Speakman 223 Charles F. Cichon, National Association of Drug Diversion Investigators 227 Debbie Alumbaugh, Michael's Message Foundation, Inc. 230 Brian A. Hunter, Young Adults with Narcolepsy 233 Joe Spillane, Pharm.D., ABAT 237 Mali Einen 241 Sandra Jones 246 Continued Committee Discussion and Deliberations 249 4 1 P R O C E E D I N G S 2 Call to Order and Introductions 3 DR. KAWAS: Good morning, everyone, and 4 welcome to the Wednesday, June 6, 2001 meeting of 5 the Peripheral and Central Nervous System Advisory 6 Committee. My name is Claudia Kawas, and I think 7 we can begin with introductions, please, perhaps 8 over by Dr. Temple's side. 9 DR. TEMPLE: Bob Temple, I am the Office 10 Director. 11 DR. KATZ: Russ Katz, Division of 12 Neuropharmacological Drug Products, FDA. 13 DR. FEENEY: John Feeney, neurology team 14 leader, FDA. 15 DR. MANI: Ranjit Mani, medical reviewer, 16 Neuropharm., FDA. 17 DR. LEIDERMAN: Deborah Leiderman, 18 Director, Controlled Substance Staff, FDA. 19 DR. SIMPSON: Pippa Simpson, University of 20 Arkansas Medical Sciences, biostatistician. 21 DR. FALKOWSKI: Carol Falkowski, drug 22 abuse researcher, Hazelden Foundation. 23 DR. ROMAN: Gustavo Roman, Professor of 24 Neurology at the University of Texas, San Antonio. 25 DR. WOLINSKY: Jerry Wolinsky, Professor 5 1 of Neurology, University of Texas, Houston. 2 DR. TITUS: Sandy Titus, FDA, the 3 administrator of the Peripheral and Central Nervous 4 System Committee. 5 DR. PENN: Richard Penn, neurosurgeon at 6 the University of Chicago. 7 DR. LACEY: Ella Lacey, professor emerita, 8 Illinois University, Carbondale, Illinois. 9 DR. VAN BELLE: Gerald Van Belle, 10 Department of Biostatistics, from the University of 11 Washington. 12 DR. PENIX: LaRoy Penix, Associate 13 Professor of Neurology at Moorehouse School of 14 Medicine. 15 DR. SANNERUD: Christina Sannerud, Drug 16 and Chemical Evaluation Section, Drug Enforcement 17 Administration. 18 DR. DYER: I am Jo Dyer, with the 19 University of California, San Francisco and the San 20 Francisco Poison Control System, California. 21 DR. FRANKENHEIM: Jerry Frankenheim, 22 pharmacologist, National Institute on Drug Abuse. 23 DR. KAWAS: Today we have met to discuss 24 the consideration of Xyrem, proposed to reduce the 25 incidence of cataplexy and to improve the symptom 6 1 of daytime sleepiness for persons with narcolepsy. 2 The main focus of the deliberations will also be on 3 risk management issues. 4 If we could ask Dr. Titus to begin with 5 the conflict of interest statement? 6 Conflict of Interest Statement 7 DR. TITUS: Before I begin the conflict of 8 interest statement, I just want to announce that we 9 have two people on line with us, Dr. Chervin and 10 Dr. Guilleminault. They are both in a room 11 listening to us and will participate with us on the 12 mikes. 13 The following announcement addresses the 14 issue of conflict of interest with regard to this 15 meeting and is made a part of the record to 16 preclude even the appearance of such at this 17 meeting. 18 The special government employees 19 participating in today's meeting have been screened 20 for interests in Orphan Medical's Xyrem and for 21 interests in the products and sponsors deemed by 22 the agency to be competing. Based on the agency's 23 review of each participant's response to the 24 conflict of interest screening, it has been 25 determined that there is no potential for a 7 1 conflict of interest with regard to this meeting. 2 With respect to FDA's invited guests, 3 there are reported affiliations which we believe 4 should be made public to allow the participants to 5 objectively evaluate their comments. 6 Dr. Ronald Chervin would like to disclose 7 for the record that he has a contract with Cephalon 8 to study Provigil, but not for use in narcolepsy. 9 He is the principal investigator, however, no funds 10 from Cephalon, present or past, have contributed to 11 his personal salary and none have been made 12 available for his non-research related use. 13 Further, in previous years Dr. Chervin was a 14 co-investigator with Cephalon in a narcolepsy 15 clinical trial. 16 Christian Guilleminault has been the 17 administrator of the Sleep Disorder Clinic in Palo 18 Alto, California, where the study of Xyrem was 19 performed by a team of researchers. 20 In the event that the discussions involve 21 any other products or firms not already on the 22 agenda for which an FDA participant has a financial 23 interest, the participants are aware of the need to 24 exclude themselves from such involvement and their 25 exclusion will be noted for the record. 8 1 With respect to all other participants, we 2 ask in the interest of fairness that they address 3 any current or previous involvement with any firm 4 whose products they may wish to comment upon. 5 Thank you. 6 DR. KAWAS: Thank you very much, Dr. 7 Titus. We will begin with Dr. Russell Katz, of the 8 FDA, who will give us the FDA overview of the 9 issues. I want to point out to the committee 10 members that they have much of the materials that 11 they will be seeing during this meeting in front of 12 them. 13 FDA Overview 14 DR. KATZ: Thanks, Claudia. First, I 15 would like to welcome the committee back. You were 16 here just a few months ago so I appreciate your 17 coming back so soon. 18 We have a number of invited guests who are 19 augmenting the committee today, and many of them 20 are experts in the evaluation of issues related to 21 drug abuse, and I would just like to welcome them, 22 in particular Drs. Simpson, Sannerud and 23 Frankenheim. 24 We have two other experts who will 25 actually be speakers later this morning. Dr. Dyer 9 1 will speak on her experience with GHB use and 2 misuse in cases she has seen, and Dr. Falkowski 3 will talk about the epidemiology of GHB abuse in 4 the United States. 5 Finally, as Dr. Titus mentioned, we have 6 two acknowledged experts in sleep disorders who are 7 attending the annual sleep meetings in Chicago, but 8 who have agreed to sit in a hotel room for however 9 long this takes and participate by phone. So, Drs. 10 Guilleminault and Chervin, wherever you are, thank 11 you. Thanks for being here. 12 As you know and as you have heard, today 13 we will ask you to discuss NDA 21-196, which was 14 submitted by Orphan Medical for the use of Xyrem, 15 gamma hydroxybutyrate or better known as GHB, for 16 the treatment of cataplexy and excessive daytime 17 sleepiness in patients with narcolepsy. 18 GHB is a simple molecule and it is 19 ubiquitous in mammalian tissues, its function 20 though is not really well known. Its relevant 21 regulatory history goes back to about 1990, and 22 prior to that date it was freely available in 23 health food stores. But in 1990 the agency began 24 to receive reports of widespread recreational use 25 in a number of different types of folks, for a 10 1 number of different types of reasons, or GHB and 2 began to get numerous reports of serious adverse 3 events associated with its misuse. 4 It was not entirely clear that all of 5 these events were necessarily related to GHB. It 6 was difficult to interpret some of these reports 7 because there were concomitant medications that 8 were unreported and it wasn't entirely clear 9 whether or how much GHB was in a particular 10 preparation that someone had taken. Those sorts of 11 issues made it difficult to completely interpret 12 the reports, but many of the reports were of events 13 that were known to be consistent with GHB's effect 14 as a potent CNS depressant, including things like 15 respiratory depression, coma and other decreased 16 levels of consciousness. So, it was reasonable to 17 believe that GHB was at least in part responsible 18 for some of these reports. 19 As a result of these reports, the agency 20 withdrew GHB from health food shelves and made it 21 illegal to use. However, illicit use continued and 22 continues to this day, not only with GHB but with 23 two related drugs which are precursors, GBL and 24 1,4-butanediol, and there have been similar reports 25 of serious adverse events associated with the use 11 1 of those products. 2 So, against this background of use, the 3 investigation of GHB as a treatment for cataplexy 4 began. Based on the results of a single trial 5 performed by the sponsor and their commitment to 6 perform additional trials, the sponsor was granted 7 a treatment IND in December of 1998. For those of 8 you unfamiliar with a treatment IND, it is 9 basically a mechanism to permit use of an 10 investigational drug outside the context of a 11 controlled trial for a serious disease for which 12 there aren't other available treatments. It is 13 usually granted relatively late in the development 14 of a drug so that by the time you grant it you have 15 some reasonable idea, based on controlled data, 16 that the drug is probably effective and reasonably 17 well tolerated. 18 Just another relevant piece of history, in 19 2000 Congress passed a law which placed GHB in 20 Schedule I and also placed it into Schedule III for 21 any approved uses that may be granted. 22 The NDA that we are discussing today was 23 submitted in September of 2000 by the company, and 24 it contains the results of four controlled trials 25 which the sponsor believes establish substantial 12 1 evidence of effectiveness for cataplexy and 2 excessive daytime sleepiness in patients with 3 narcolepsy. It also contains, obviously, safety 4 experience. 5 I just want to talk about the safety 6 experience for just a little bit. As you know from 7 the briefing documents, much of the safety data in 8 the application was not generated by the company 9 but by an individual investigator under his own 10 individual investigator IND. This is Dr. Scharf, 11 and he is an acknowledged expert in the use of GHB 12 and he has been treating patients under his IND for 13 about 16 years. His data comprise almost 30 14 percent of the patient safety database in the NDA. 15 If one looks at patient time, his experience 16 constitutes about 70 percent of the total patient 17 exposure. 18 As part of a routine investigation of the 19 NDA to look at source documents, the agency 20 investigators found that they were unable to locate 21 some critical source documents of Dr. Scharf's IND, 22 and it was difficult to confirm the sponsor's 23 submission of Dr. Scharf's data. However, 24 subsequent to that, Dr. Scharf has made extensive 25 efforts to provide the additional source documents 13 1 and agency investigators have reinspected that 2 data. I believe the conclusion of that 3 investigation is that we find that the records, for 4 the most part, do support the sponsor's 5 descriptions of Dr. Scharf's data. And, we believe 6 we can make certain statements about that data at 7 this point. 8 We were particularly interested in the 80 9 or so patients that Dr. Scharf treated that did not 10 move on into the company's treatment IND. He 11 treated a total of 143, or thereabouts, patients, 12 60 of whom went into the sponsor's treatment IND. 13 So, we had a good idea of what was happening to 14 those patients but there were about 80 that didn't 15 and who were basically discontinued from treatment 16 under Dr. Scharf's own IND. 17 So, except for a handful of patients, we 18 believe we know why those 80 patients discontinued 19 and their status. I believe we can say reasonably 20 comfortably say that nothing catastrophic that we 21 don't know about happened to those patients but, 22 unfortunately, we have relatively little 23 well-documented data regarding other less serious 24 adverse events in that cohort of 80. Other than 25 patient diaries, we have essentially no 14 1 documentation about exactly what dose those 2 patients took and for how long. 3 I have gone into this at some depth 4 because the safety experience in the NDA is 5 relatively small as compared to a typical NDA, and 6 that is by agreement. This is an orphan product. 7 Based on the sponsor's estimated prevalence of 8 cataplexy of about 25,000, it received orphan 9 designation and one wouldn't necessarily expect 10 that a safety database of a typical size, which is 11 somewhere in at least 10000 to 2000 patients in the 12 typical NDA, would be submitted in an orphan 13 application. So, we agreed with the sponsor that 14 about 500 patients treated for appropriate 15 durations, at appropriate doses would be 16 acceptable. 17 But, given the relatively small database 18 and some of these residual questions about a 19 reasonable proportion of it, that is to say Dr. 20 Scharf's data, that may take on some additional 21 meaning and we would like you to think about that 22 as the day goes on. 23 In addition to the safety and the 24 effectiveness data which is required in an NDA of 25 course, the sponsor has proposed a detailed risk 15 1 management program, and that has three goals: to 2 inform patients and physicians about the risks of 3 GHB; to minimize the risks to those patients; and 4 also to minimize the likelihood that subjects for 5 whom the drug has not been prescribed will be 6 exposed to it. This latter point not only refers 7 to diversion and its use illicitly by folks who 8 shouldn't be taking it, but also to the accidental 9 use of GHB in the home, perhaps by small children, 10 and you will hear how GHB is administered and what 11 form it is prepared in, and we think that is a 12 potential risk. So, we would like you to think 13 about that as the day goes on too. 14 As far as the risk management program, you 15 will hear about it in great detail from the company 16 but, in brief, it consists of a couple of sort of 17 major components. One is that the product will be 18 made available through a central pharmacy and will 19 be shipped directly to the patient at home. 20 Physicians and patients will also receive detailed 21 materials about the risks and the appropriate use 22 of the drug after the first prescription is filled. 23 Actually, they will receive those materials 24 initially and all subsequent refills of 25 prescriptions will be contingent upon patients and 16 1 physicians documenting that they have read these 2 materials, and they understand the risks and how to 3 take the drug appropriately. 4 All patients and physicians will be 5 entered into a registry, and there will be close 6 surveillance instituted to ensure that untoward 7 events are minimized, for example, to ensure that 8 patients don't go from doctor to doctor trying to 9 get refills of prescriptions that are 10 inappropriate. 11 So, with these data and against the 12 background of misuse of GHB out in the population 13 at large, we bring you today's application and we 14 will ask you to formally vote on three questions. 15 One is whether or not you think that substantial 16 evidence of effectiveness has been submitted for 17 the indications that the sponsor has proposed, that 18 is to say, cataplexy and excessive daytime 19 sleepiness in patients with narcolepsy. If you 20 find that they haven't, we would be very interested 21 to know whether or not you feel that substantial 22 evidence has been submitted for either of those two 23 indications. 24 While you listen to the effectiveness 25 data, we would like you to pay particular attention 17 1 to the question of dose and for which dose you 2 think evidence of effectiveness has been submitted. 3 If you find there is substantial evidence of 4 effectiveness for a particular indication, we need 5 to ask you whether or not GHB can be considered 6 safe in use given appropriate labeling. Now, we 7 are not going to discuss necessarily the specifics 8 of proposed labeling but, nonetheless, we ask you 9 to think of it in that context. 10 Again, in assessing the safety of the 11 product, we ask you to concentrate on at least the 12 question of what dose you have found to be 13 effective and whether or not there is sufficient 14 safety experience at that dose for the drug to be 15 approved. 16 Finally, we want to take a formal vote on 17 the question of whether or not you think it is 18 required or should be required that the drug be 19 approved only with the risk management program of 20 some type, not necessarily the one specifically 21 proposed by the company. Obviously, the company 22 has proposed a risk management program but we need 23 to know whether or not you think it is mandatory 24 that it be approved with such a program in place. 25 If you do, we have a number of questions that we 18 1 would like you to discuss -- not necessarily take a 2 formal vote on but discuss with regard to a risk 3 management program and some of the provisions that 4 the sponsor has proposed. 5 There are some aspects of the program that 6 they have proposed that we would like you to pay 7 particular attention to and discuss. For example, 8 there is some considerable sympathy in the agency 9 for including a provision in the risk management 10 program that would restrict the use of the drug to 11 patients with whatever indication you believe has 12 been supported, that is to say, to restrict as much 13 as possible off-label prescribing. That is one 14 possibility. 15 There is also some enthusiasm internally 16 for physicians and patients to document that they 17 have reviewed the relevant materials before the 18 first prescription is filled. So, we would like 19 you to think about that as well as we talk about 20 the risk management program. 21 So, as you can see from the agenda, the 22 company is going to present the safety and 23 effectiveness data, after which Dr. Mani, from the 24 Division, will come up and present briefly some of 25 our views about the data you will have just heard. 19 1 Specifically, I believe we have some different 2 views about the evidence submitted for establishing 3 a claim for excessive daytime sleepiness in 4 narcolepsy, and there may be other additional 5 safety issues that we would like to bring up at 6 that time, in particular the question of an event 7 that has been called sleep walking. 8 I think with that as background, I will 9 turn it back to Dr. Kawas. Thank you. 10 DR. KAWAS: Thank you, Dr. Katz. Orphan 11 Medical presentation is to follow. Dr. David 12 Reardan, Orphan Medical? 13 Orphan Medical Presentation 14 DR. REARDAN: Hi. Good morning. Good 15 morning, ladies and gentlemen, members of the 16 committee and FDA. 17 [Slide] 18 My name is David Reardan, and I represent 19 Orphan Medical as head of regulatory affairs. 20 Orphan Medical is a small, 60-person firm, 21 dedicated to the development of orphan drugs. We 22 have obtained marketing approval for six orphan 23 products from FDA since we were founded, in 1994. 24 The firm became involved with Xyrem when 25 approached by FDA that same year, and Xyrem was 20 1 designated an orphan drug in 1994. Today we will 2 share with you the data that has been collected 3 with respect to the efficacy and safety since our 4 IND was submitted, in 1996. 5 [Slide] 6 Dr. Mignot, director of the Narcolepsy 7 Institute at Stanford University, will present a 8 picture of a narcoleptic patient and the serious 9 medical need such patients have for new therapeutic 10 treatments. 11 Dr. Houghton is the chief medical officer 12 and chief operating officer at Orphan Medical, and 13 he will present next on the efficacy that has been 14 collected. Dr. Houghton was chair of anesthesia 15 and critical care in Australia. 16 Dr. Black, director of the Stanford Sleep 17 Clinic and an investigator for several trials, will 18 share with you the EEG pharmacology of Xyrem. Dr. 19 Houghton will then present the safety data and 20 finish up with a benefit/risk assessment. 21 Following presentations by two FDA invited 22 speakers with respect to GHB abuse, Dr. Balster, 23 director of the Institute for Drug and Alcohol 24 Studies at the Medical College of Virginia, will 25 share with you his views on abuse liability. 21 1 Since there is public abuse of GHB and its 2 analogs, the company has developed a risk 3 management program for Xyrem that will be presented 4 by Patti Engel, our vice president of marketing and 5 sales. 6 [Slide] 7 In addition to those presenting today, the 8 following experts are available in the audience to 9 answer questions from the committee or FDA: Dr. 10 Emsellem, Dr. Hagaman and Dr. Ristanovic are all 11 directors of their respective sleep institutes, and 12 have been investigators in our clinical trials. 13 Dr. Okerholm is a consultant in the area of 14 pharmacokinetics and drug metabolism; Dr. Reno in 15 the area of toxicology; and Dr. Richard Trout, who 16 is a professor emeritus in statistics from Rutgers, 17 is here if there are any statistical questions. 18 [Slide] 19 This is the chemical structure of sodium 20 oxybate, more commonly known as gamma 21 hydroxybutyrate, or GHB. Notice that it is a 22 simple 4-carbon hydroxy fatty acid and, as such, 23 quite easy to synthesize. In fact, kits have been 24 illegally promoted on the Internet for its 25 manufacture. If an amino group were to replace 22 1 this alcohol functional group at position 4, you 2 would have GABA, gamma aminobutyric acid, another 3 CNS active chemical. Oxybate is a natural compound 4 in the human body. 5 [Slide] 6 Gamma hydroxybutyrate was first discovered 7 in the 1960's by Dr. Labore, in France, and was 8 investigated as an analog for GABA. It was found 9 to have hypnotic properties and was first approved 10 in France, and later a few other countries of 11 Europe, as an adjunct in anesthesia. It was used 12 in labor and delivery for quite a few years. The 13 injectable form is still available today in parts 14 of Europe. 15 In the 1970's initial work was begun in 16 Canada to test its properties in narcolepsy. 17 Following initial promise for use in patients with 18 narcolepsy two controlled trials were conducted by 19 independent investigators, one in the U.S. and one 20 in The Netherlands. In 1994, due to the promising 21 investigator trials, FDA Office of Orphan Products 22 approached Orphan Medical to consider the compound 23 for development. 24 Since there was no patent protection and 25 the market was very small, no other firms were 23 1 willing to consider the development of GHB for 2 narcolepsy at the time. Orphan Medical agreed to 3 sponsor this medication. Our new drug application 4 was submitted in October of 2000 and was designated 5 by FDA for priority review. 6 The clinical development has been fairly 7 straightforward and all controlled trials conducted 8 to date have shown sodium oxybate to be effective 9 and safe for the treatment of narcolepsy. This 10 project has been made more difficult because of the 11 abuse situation. 12 [Slide] 13 Let me explain why Xyrem is not going to 14 be a factor in the abuse of GHB and its precursors. 15 Orphan Medical was aware abuse existed at the time 16 the company agreed to sponsor development of Xyrem. 17 At this same time, Internet was burgeoning. Due to 18 its ease of synthesis and ready availability of 19 precursor chemicals, GHB was initially an easy 20 target for promoters of illegal drugs. 21 But GHB is not the only problem. GBL and 22 1,4-butanediol are precursor chemicals that can be 23 easily converted to GHB and are, in fact, converted 24 to GHB in the human body. These precursors are 25 widely available as bulk chemicals and are being 24 1 illegally used in the United States, and the abuse 2 problem is growing. 3 Federal legislation, enacted in 2000, 4 helped to control the availability of GHB and GBL 5 but not 1,4-butanediol and other precursor 6 chemicals that can be used for the same purpose. 7 In many states, even with GHB schedules, GBL and 8 1,4-butanediol are not controlled. 9 We believe that approval of Xyrem for use 10 by patients with narcolepsy will not add to the 11 general abuse problem of GHB and its numerous 12 precursors. 13 [Slide] 14 The proposed indication for which we are 15 asking FDA for marketing approval is to reduce the 16 incidence of cataplexy and to improve the symptom 17 of daytime sleepiness in patients with narcolepsy. 18 [Slide] 19 Narcolepsy fits the definition of orphan 20 disease in the United States, with less than 21 200,000 patients. There are estimated to be about 22 135,000 patients, of which 55 percent are 23 diagnosed, with about 24,000 seeking treatment for 24 cataplexy. 25 [Slide] 25 1 I would now like to introduce you to Dr. 2 Emmanuel Mignot, from Stanford. Dr. Mignot has 3 been widely published in this area and is 4 considered one of the premiere international 5 experts on narcolepsy. He has not participated in 6 any of our clinical trials. 7 Medical Need 8 DR. MIGNOT: It is my privilege to talk to 9 you today about narcolepsy. I have been working on 10 narcolepsy for about 15 years, both at the level of 11 basic research as well as clinical care. I am a 12 medical doctor and I see patients with narcolepsy. 13 [Slide] 14 I am going to try to summarize in a few 15 minutes really a lot of data about narcolepsy and 16 how it impacts people. 17 [Slide] 18 First, I would like to start briefly by 19 reviewing the symptoms of narcolepsy. Narcolepsy 20 is usually associated with 5 different symptoms. 21 The most disabling and the most problematic in 22 patients with narcolepsy is sleepiness. Patients 23 with narcolepsy are sleepy all the time; tired; 24 they have sleep attacks; they cannot stay awake for 25 a long period of time, and it is usually why they 26 1 come to see the doctor. They just cannot live a 2 normal life. Especially in work conditions, as you 3 probably know, it is very difficult -- you have to 4 be awake all day long and it is a major problem in 5 narcolepsy. 6 Now, it is not enough to diagnose 7 narcolepsy. Narcolepsy is not just sleepiness and 8 there are a lot of other medical conditions that 9 are associated with sleepiness. Patients with 10 narcolepsy also have a series of symptoms that 11 correspond to the fact that they go very quickly 12 into rapid eye movement sleep. As probably many of 13 you know, rapid eye movement sleep is a stage of 14 sleep that only occurs 1.5 or 2 hours after you 15 fall asleep where you are actively dreaming but 16 your body is completely paralyzed and you have 17 these rapid eye movements. 18 Patients with narcolepsy go into REM sleep 19 extremely quickly, sometimes in a few minutes, and 20 that leads to a series of symptoms where patients 21 sometimes are half way through REM sleep, being 22 still awake. Consequently, they may experience odd 23 symptoms that we call the dissociated REM sleep 24 event, abnormal REM sleep event. Those are 25 cataplexy, hypnagogic hallucinations and sleep 27 1 paralysis. 2 An example is cataplexy. When a patient 3 gets emotionally excited, typically when they are 4 happy, they meet a good friend, sometimes when they 5 are angry but most often when they are joking, in a 6 nice environment and happy about something, they 7 may feel suddenly weak; they become paralyzed; 8 sometimes they fall down to the ground, completely 9 paralyzed and they cannot move. In very rare cases 10 they may even go into REM sleep. We believe 11 somehow being emotionally excited stimulates the 12 paralysis of rapid eye movement sleep that every 13 one of us experiences during sleep, except that in 14 patients with narcolepsy it may occur in the middle 15 of the day in response to emotion. 16 Also, when they fall asleep they sometimes 17 have hallucinations because they go so quickly into 18 REM that sometimes they dream while they are still 19 awake. I remember a patient, for example, who 20 every night would fall asleep and he would see 21 someone coming and strangling him. Or, they may 22 hear people talking; or see people walking in the 23 room. It can be very frightening and it can be a 24 very terrible experience for patients with 25 narcolepsy. 28 1 Another symptom of abnormal REM sleep that 2 patients with narcolepsy have as well is called 3 sleep paralysis. When they wake up from a nap or 4 when they fall asleep, sometimes they again go so 5 quickly into REM and disassociated REM sleep events 6 that sometimes they may be paralyzed from REM but 7 still be awake. Basically, they would wake up from 8 sleep and they cannot move, not even their little 9 finger. It can be very scary. It lasts a few 10 minutes and then finally they can move. Some 11 patients with narcolepsy have multiple episodes of 12 sleep paralysis when they nap during the day, and 13 so forth, and that is another very bothersome 14 symptom. 15 Finally, patients with narcolepsy, 16 contrary to what people way, don't sleep too much; 17 their main problem is that they just cannot stay 18 awake. They fall asleep very quickly in many 19 circumstances, but they are unable to stay asleep 20 for a long period of time. In fact, patients with 21 narcolepsy don't sleep 20 hours a day. What 22 happens is that at night they don't sleep well. 23 Often that is another symptom that is very 24 bothesome. They fall asleep very quickly at night 25 but after one hour they cannot sleep again. They 29 1 are just awake and cannot sleep. 2 Then, all these symptoms are quite severe 3 and, of course, affect the lives of patients. And, 4 since GHB is recommended in cataplexy, which is 5 muscle atonia triggered by emotion, I will just 6 show you a quick video of a patient with cataplexy. 7 This is a boy, a 9-year old. Narcolepsy 8 usually starts during adolescence and here the 9 clinicians are trying to make him laugh to just try 10 to elicit the symptom, and you see he is falling 11 down and he is completely paralyzed and he is 12 losing his muscle tone. Some of these patients 13 have that many time per day and it can be extremely 14 socially disabling. You can imagine being at a 15 party or being with some friends and having this 16 happen to you. In this kid it was particularly 17 severe. 18 Most cases of narcolepsy start during 19 adolescence but occasionally it starts as early as 20 5 years of age. It peaks around 15 years of age. 21 It is often extremely problematic because I am sure 22 you realize when you have this type of thing 23 happening to you and sleepiness at school, 24 especially when you are 15 years old, when you are 25 an adolescent, it really wrecks your life apart, 30 1 especially when it is not properly diagnosed. 2 [Slide] 3 There have been a number of studies, and I 4 won't have time to review them, that have shown 5 that the quality of life of patients with 6 narcolepsy is extremely impaired, as much as 7 depression, epilepsy or other reference conditions 8 in almost all the scales that you look at. 9 Clearly, it is a very socially disabling disorder. 10 [Slide] 11 It is also, of course, a disorder that 12 impacts just your daily life. For example, driving 13 -- patients with narcolepsy have a very increased 14 rate of accidents and sometimes many of them refuse 15 to drive just because of falling asleep or having 16 cataplexy while driving. 17 [Slide] 18 We have objective tests for diagnosing 19 narcolepsy. In fact, it is not just a 20 psychological disorder. You can actually use a 21 test like the Multiple Sleep Latency Test, where 22 you ask patients to come to the sleep lab. You 23 check that they sleep normally and the following 24 day you ask them to nap every two hours and you 25 measure how fast they fall asleep. You see, 31 1 normally people won't fall asleep or nap in the 2 middle of the day, or they would fall asleep with a 3 15-minute latency in the dark. A patient with 4 narcolepsy, as soon as you switch off the light, 5 they are sleeping. In a few minute latency, they 6 are asleep. So, we have objective ways to show 7 that these people have a problem. 8 [Slide] 9 Also, in this nap you see that they go 10 very quickly into REM sleep. Normal people won't 11 have REM sleep before one hour after falling 12 asleep, but patients with narcolepsy will go 13 straight into REM. You can actually demonstrate -- 14 we call that sleep onset REM period -- that 15 patients with narcolepsy have all this sleep 16 abnormality and REM abnormality using sleep 17 testing. 18 [Slide] 19 Current treatment for narcolepsy is 20 completely symptomatic. We don't treat the cause 21 of the disease; we only treat the symptoms. 22 Typically, the treatment now uses two drugs, two 23 lines of drug. A patient with cataplexy will be 24 treated usually with two drugs. One is a stimulant 25 which would be a classical amphetamine-like 32 1 stimulant or this more recent drug that was just 2 approved that is called modafinil, Provigil, which 3 works on sleepiness. It will keep a patient awake 4 but will never normalize him; it only improves him. 5 And, they all have a lot of side effects. You 6 know, the stimulants can even produce psychosis in 7 some rare cases but, of course, they raise blood 8 pressure. They produce psychological changes. 9 They have a lot of other side effects. 10 We all know now that they all increase 11 dopamine in the brain. We have done a series of 12 studies which have shown that. Even modafinil, the 13 most recent drug -- we know now that it works by 14 increasing dopamine in the brain. And, they don't 15 have anything different from each other so some of 16 them are definitely safer than others. 17 For the antidepressants, for the treatment 18 of cataplexy -- this works well on sleepiness but 19 it doesn't work on cataplexy or nightmares, or 20 hallucination or sleep paralysis. For this you use 21 antidepressants. Why? Because antidepressants 22 depress REM sleep and they also suppress cataplexy 23 and all the other abnormal dreaming that patients 24 with narcolepsy have. The problem is they also 25 have a lot of side effects. Actually, the new 33 1 SSRI, they don't work as well as the old 2 tricyclines. Often you even have to use the old 3 tricycline antidepressants because norepinephrine 4 uptake inhibition seems to be the mode of action of 5 these drugs, more than serotonin. They don't 6 really work that well and, of course, they have a 7 lot of side effects and a lot of different 8 problems. 9 [Slide] 10 Finally, I want to stress again that we 11 need new treatments for narcolepsy just because all 12 the treatments we have now just don't make people 13 normal. They just help them to be better. You can 14 best illustrate that using the MSLT/MWT, which is a 15 slightly different test where, instead of measuring 16 how fast people fall asleep in the dark, you ask 17 people to try to stay away in the dark and you see 18 that normal people can stay awake. They don't fall 19 asleep in 20 minutes, whereas patients with 20 narcolepsy fall asleep very dramatically after a 21 few minutes in the dark. 22 Even if you treat them with modafinil 23 which is a very good treatment for narcolepsy, 24 which was recently approved, you improve them but 25 they never become normal. Then, it is clear that 34 1 what we have is not enough. We just need better, 2 and this would be the same for amphetamines. Even 3 high dose amphetamines don't normalize these 4 patients. That has been shown by multiple studies. 5 [Slide] 6 We have worked for more than 15 years 7 trying to find the cause of narcolepsy, and 8 recently we have isolated the gene for narcolepsy 9 in a canine model where the disease is genetically 10 determined, and we found that it was a receptor for 11 a norpeptide that is called hypocretin. We found 12 that in humans with narcolepsy it is not like dogs 13 with narcolepsy; it is not the receptor but a 14 peptide called hypocretin which is expressed in 15 about 10,000 cells in the brain, here in the 16 hypothalamus, which is missing in patients with 17 narcolepsy. 18 This is brain tissue of a patient with 19 narcolepsy. You see here is the normal; everything 20 is gone. If you measure in the cerebrospinal 21 fluid, this is a normal level in a normal person, 22 or in patients with MS or other neurological 23 symptoms, and you see in all patients with 24 narcolepsy that this hypocretin molecule is gone. 25 We know now that the cause of narcolepsy is not 35 1 dopamine or norepinephrine, which is the current 2 treatment for narcolepsy, which are stimulants and 3 antidepressants acting through these 4 neurotransmitters, and probably replacing this 5 hypocretin would be an ideal treatment for 6 narcolepsy. But this finding was only made one 7 year ago and it is going to take probably 10 years 8 or many years before we actually have a treatment 9 based on this new discovery. 10 [Slide] 11 To summarize the medical need, I think I 12 have convinced you that narcolepsy is a serious and 13 disabling condition that needs treatment, and these 14 patients are in desperate need of better treatment. 15 As you will see from the presentation afterwards, 16 GHB is one of the effective treatments which helps 17 a lot of people. So, current treatments like 18 amphetamines and antidepressants don't work well in 19 terms of efficacy. They have a lot of side 20 effects. They all work the same way but they don't 21 act on the cause of the disease and, clearly, we 22 know that GHB, even though it probably doesn't act 23 on hypocretin, acts differently from other drugs. 24 And, it is one more drug that would be available to 25 help a lot of patients with narcolepsy. 36 1 Finally, even though there have been 2 numerous, very recent developments that are very 3 exciting in the hypocretin area, unfortunately, you 4 all know it takes a long time until drugs are 5 available and it is going to take probably many 6 years until this available. 7 This is a very quick summary of what we 8 know about narcolepsy to date. Thank you. 9 DR. REARDAN: Thank you, Dr. Mignot. Dr. 10 Houghton will now present the data which has been 11 assembled in support of the efficacy of Xyrem. Dr. 12 Houghton is a qualified anesthesiologist, with 18 13 years of clinical experience in critical care 14 medicine and numerous years experience in 15 pharmaceutical drug development. Bill? 16 Efficacy 17 DR. HOUGHTON: Good morning. 18 [Slide] 19 I am sorry to start with such a complex 20 diagram but this just outlines the pattern of 21 studies that we will be talking about this morning. 22 On the left-hand side here are the 4 controlled 23 studies on which the assessment of efficacy will be 24 based, but what is unusual about this program is 25 that patients, in an uncommon way, move to 37 1 extension protocols. So, as Dr. Katz pointed out, 2 even though the total database may be small, the 3 total duration of exposure of patients is quite 4 promising. 5 The first study that I will talk about is 6 entitled OMC-GHB-3, and the patients, at the 7 completion of this short-term treatment study did 8 progress to a long-term, open label study and then 9 had the opportunity to move into one of the 10 treatment IND protocols, with some of them still 11 participating in that study. 12 A second contributor to that protocol was 13 the patients who completed the first 6-month safety 14 treatment IND protocol, and the significance of all 15 of that is that it was from this protocol that the 16 patients are represented in the long-term pivotal 17 blinded efficacy study that supports the long-term 18 efficacy of Xyrem. 19 [Slide] 20 The first and pivotal study is a 21 randomized, double-blind, placebo-controlled, 22 parallel group, multi-center trial comparing the 23 effects of three doses, 3 g, 6 g and 9 g of orally 24 administered Xyrem with placebo for the treatment 25 of narcolepsy. As I mentioned, this was a study 38 1 conducted in 136 patients in 16 centers. 2 [Slide] 3 The primary efficacy parameter was the 4 change in the number of total cataplexy attacks in 5 the last two weeks of the treatment period compared 6 to the two weeks of the baseline period. 7 Secondary efficacy parameters that were 8 considered included complete and partial cataplexy 9 attacks; daytime sleepiness; inadvertent sleep 10 attacks during the day; hypnagogic hallucinations; 11 sleep paralysis; and a clinical global impression 12 of change. 13 [Slide] 14 Patients naive to sodium oxybate therapy 15 were chosen with a bona fide diagnosis of 16 narcolepsy for at least 6 months. They were 17 required to have a record of a polysomnograph or 18 Multiple Sleep Latency Test within the last 5 years 19 to exclude other causes of daytime sleepiness, and 20 particularly sleep apnea. 21 They were required to have a history of 22 daytime sleepiness and cataplexy for at least 6 23 months, and recurrent daytime naps that occurred 24 almost daily in the preceding 3 months. 25 [Slide] 39 1 The overall study design was divided into 2 5 stages. Firstly, there was a screening period in 3 which the patients were required to qualify for 4 entry criteria and then withdrawn from their 5 existing anti-cataplectic medications over a 4-week 6 period to avoid rebound phenomena which were 7 considered a safety consideration. At the end of 8 this withdrawal period they entered a washout 9 period, which was determined by at least 5 times 10 the half-life of their preceding drug to remove any 11 effects of those drugs. However, if patients 12 weren't on any cataplectic medications, they were 13 still required to remain 5 days in that washout 14 period to familiarize themselves with the use of 15 diaries. 16 They then proceeded to a baseline period 17 of 2 to 3 weeks, using daily diary recording to 18 establish the severity of their disease and to 19 confirm that they had reached a stable stage in 20 their disease. They then entered a 4-week blinded, 21 randomized treatment period, with a visit at 2 22 weeks, a telephone call the day after commencing 23 treatment, and then safety telephone calls 3 times 24 a week during the treatment period, at the end of 25 which they were abruptly withdrawn from drug and 40 1 followed up 3 to 5 days later to assess any rebound 2 phenomena and any adverse experiences that may have 3 ensued. 4 [Slide] 5 As is shown here, the patient groups were 6 very evenly balanced at baseline. They represented 7 a fairly severe group of narcoleptics, with an 8 average incidence of cataplexy of around 34 per 9 week at baseline. 10 There was a dose-response relationship 11 across the doses based on median change in the 12 total number of cataplexy attacks that, when 13 compared to placebo, approached significance at the 14 9 g dose, with a p value of 0.0529, and achieved 15 highly significant change at the 9 g dose. 16 [Slide] 17 This dose relationship is clearly shown in 18 the plot of median change from baseline in the 19 number of cataplexy attacks per week, and the 20 spread of the data is demonstrated as the quartile 21 lines around these median values. 22 [Slide] 23 A more clinically relevant presentation of 24 the data is the percentage change in the number of 25 cataplexy attacks from baseline. This was 41 1 calculated as the distribution of percentage change 2 values for each individual patient and is again 3 presented as the medians. This representation 4 clearly shows that the major change in cataplexy 5 occurs in the first 2 weeks, but with ongoing 6 change in the subsequent 2 weeks, as represented in 7 2 of the dose groups. 8 [Slide] 9 Secondary efficacy variables included 10 assessment of excessive daytime sleepiness using 11 the validated Epworth Sleepiness Scale which rates 12 the patient's feeling of daytime somnolence by 13 scoring on a scale of 0-3 the probability of 14 falling asleep in the circumstances of 8 common 15 life scenarios. This results in a potential 16 maximum score of 24. 17 [Slide] 18 This slide demonstrates a clear 19 dose-related reduction in the Epworth Sleepiness 20 Scale, reaching a significant level of 0.0001 in 21 the 9 g group compared to placebo. This change was 22 incremental beyond the effects of stable dosing of 23 stimulants because stimulant medications were 24 maintained constant throughout the study. In all 25 Xyrem-treated groups some patients improved beyond 42 1 the defined narcolepsy range, with some patients in 2 the 6 g and 9 g groups actually improving into the 3 normal range as rated by the Epworth Sleepiness 4 Scale. 5 The second component of daytime 6 sleepiness, the number of inadvertent naps during 7 the day, was also significantly reduced compared to 8 placebo in the 6 g group and 9 g dosing. 9 [Slide] 10 The severity of the disease at baseline 11 was rated by the principal investigator according 12 to the following validated scale. Then, at the end 13 of the treatment period a blinded global impression 14 of change according to the rating shown here was 15 made, rating from very much improved through no 16 change to very much worse. 17 [Slide] 18 Assignment of these modal values indicated 19 a primary distribution of the placebo patients 20 mainly to no change or minimally improved, but 21 there is an obvious predominance of assignment in 22 the 9 g dose to very much improved and much 23 improved. 24 [Slide] 25 Because of the complexity of presenting 43 1 these assigned categories, a post hoc 2 simplification was applied to group the patients 3 that showed clear clinical improvement into a 4 responder group, and all others were called 5 non-responders. This again displays the 6 dose-response trend in the categorical data, with a 7 clear statistical difference between the 9 g group 8 and the placebo group. 9 [Slide] 10 Other secondary measures that achieved 11 significant change included the number of 12 awakenings at night, subjective sleep quality, 13 morning alertness, the ability to concentrate. 14 Hypnagogic hallucinations and sleep paralysis, 15 which had a much lower incidence at baseline, 16 showed a non-significant trend towards improvement. 17 [Slide] 18 The next study that I would like to 19 present is the study that was suggested by the FDA 20 to provide evidence of long-term efficacy of Xyrem 21 based on the return of cataplexy following the 22 cessation of long-term treatment with the active 23 drug. 24 [Slide] 25 Patients entered this blinded, randomized 44 1 study from the long-term open-label study I showed 2 you initially having completed the GHB-2 protocol 3 and proceeded into the GHB-3 protocol for periods 4 up to 2 years, or from the initial treatment IND 5 protocol. This provided assessment of potential 6 adverse consequences of the abrupt withdrawal of 7 long-term therapeutic doses of Xyrem as well. 8 Patients having taken the drug for 6 9 months to 3.5 years were screened, and after 10 blinded randomization entered a single blind 11 baseline period in which daily diaries were used to 12 record the severity of their cataplexy. They then 13 entered a double-blind phase of 2 weeks wherein 14 they were randomized in a 50 percent ratio to 15 either continued, unchanged dose of Xyrem in a 16 blinded fashion or to placebo. Randomization was 17 performed in a centralized manner to ensure equal 18 representation of dosing in the comparative groups. 19 [Slide] 20 The primary efficacy variable was the 21 change in the number of cataplexy attacks in the 22 double-blind period compared to baseline. There 23 was a median change of zero in the Xyrem group but, 24 as seen, there was a marked increase in the 25 incidence of cataplexy in those randomized to 45 1 placebo. This was highly significant. 2 [Slide] 3 When the median change from baseline by 4 week was calculated, you can see that there was a 5 step-wise increase in cataplexy which supported the 6 long-term efficacy of the drug in a statistically 7 significant manner, but they represent a gradual 8 return of cataplexy rather than an acute rebound 9 phenomenon. 10 [Slide] 11 I will now present very briefly some 12 supportive data from 2 early controlled, crossover 13 design studies that have been published, and for 14 which Orphan Medical purchased the databases and 15 included in the NDA submission. 16 [Slide] 17 The first was a study conducted by Dr. 18 Lawrence Scrima, then of the University of 19 Arkansas, in 20 patients, 10 males and 10 females, 20 using a dose of 50 mg/kg, much lower than some of 21 those in the previous studies and equivalent to 22 about 3.5 g per day in a 70 kg man. 23 Following the withdrawal of 24 anticataplectic medications, he recorded a baseline 25 period during which the patients were required to 46 1 have a minimum of 10 cataplexy attacks, then were 2 randomized into an initial treatment period of 29 3 days, followed by a washout period of 6 days, and 4 then crossed over to the alternate treatment, again 5 followed by a washout of 6 days. Stimulants were 6 continued throughout this study and all patients 7 were actually transferred to methylphenidate as 8 their stimulant. 9 [Slide] 10 The primary efficacy measures are 11 identified, with the average number of cataplexy 12 attacks compared to baseline and objective 13 sleepiness index as determined by the Multiple 14 Sleep Latency Test. This was to represent a 15 measure of daytime sleepiness. 16 Because of logistic issues in the study 17 conduct and methodologic issues in design and 18 definition, this is presented as supporting data 19 only to represent cataplexy response at a lower 20 dose. As can be seen, this patient group again 21 represented a reasonably severe narcoleptic 22 population. They had a baseline measure of 20 23 cataplexy attacks per week. There was an initial 24 fairly significant placebo response, as was shown 25 in the previous studies, but by week 3 and week 4 47 1 statistically significant differentiation between 2 placebo and active treatment was shown, and there 3 was a statistically significant overall response in 4 the study. There was no significant change in the 5 sleepiness index as the measure of daytime 6 sleepiness, however, in this study. 7 [Slide] 8 The second study that I will present very 9 briefly was conducted by Dr. Lammers, in The 10 Netherlands. It is, again, a randomized, blinded, 11 crossover design study in 24 narcoleptics. The 12 other significant difference in this study was that 13 concomitant medications for both cataplexy and 14 excessive daytime sleepiness were continued 15 throughout the study. 16 Following a 1-week baseline to establish 17 disease severity, the patients were randomized to a 18 4-week treatment period at a dose of 60 mg/kg in 19 divided nightly doses, followed by a washout period 20 of about 3 weeks, and then a baseline period of 1 21 week again preceding a second treatment period of 4 22 weeks. 23 [Slide] 24 As is obvious here, the severity of 25 cataplexy during the baseline period was much lower 48 1 in this study, potentially the consequence of 2 continued anticataplectic medication in some 3 patients. But, again, there is a significant 4 response. According to the statistical plan which 5 was very scant that was represented in the 6 published study, and agreed to by the FDA, there 7 was an incorrect or unsatisfactory statistical 8 management of this study. The change in cataplexy 9 was not statistically significant. When the 10 results of this study were submitted by Orphan, 11 they were reanalyzed with an ANCOVA analysis as had 12 been applied in the GHB-2 study, and this change 13 was significant according to the ANCOVA analysis. 14 [Slide] 15 Other measures that showed significant 16 improvement included hypnagogic hallucinations and 17 daytime sleep attacks again. 18 [Slide] 19 Although not eligible for determination of 20 efficacy since it is an open-label study, I would 21 like to briefly mention three aspects of the 22 follow-on study to the pivotal GHB-2 study. And, 23 117 patients chose to participate entering the 24 study at the 6 g per day dose and then slowly 25 titrating to clinical efficacy between the doses of 49 1 3 g and 9 g. This study, therefore, represents the 2 proposed clinical use of the drug and, although 3 primarily a safety study, represents some important 4 dynamic information. 5 [Slide] 6 This slide shows the response in cataplexy 7 over the 12-month period. What is surprising is 8 that the maximum nadir occurred at about 8 weeks, 9 and then the sustained efficacy was maintained 10 across the 12 months in all dose groups. 11 [Slide] 12 A similar pattern was seen in the Epworth 13 Sleepiness Scale, which shows the same time frame 14 with maximum response at about 8 weeks, and then 15 maintained efficacy over the course of 12 months in 16 this open-label study. What is also interesting to 17 note is that most of the patients in most dose 18 groups were maintained beyond the defined 19 narcolepsy range. 20 [Slide] 21 When the distribution of doses to which 22 the patients were titrated is shown, it is seen 23 that 6 g per day is the most common dose, followed 24 by the 9 g dose group. 25 [Slide] 50 1 This represents the pattern of dosing seen 2 in other open-label studies where doses were 3 titrated to clinical response. What is important 4 to note is that there is not a change in dosing 5 between the 6-month and the 12-month dosing groups, 6 suggesting no tolerance development to maintain the 7 dynamic effects shown. 8 [Slide] 9 This slide represents the cohort of 10 patients that entered the SXB-21 protocol via the 11 GHB-2 and then GHB-3 protocol. Represented here is 12 the incidence of cataplexy for each individual 13 patient at the baseline in GHB-2. They were then 14 maintained in the study I have just shown you over 15 the course of up to 2 years, and this is the 16 incidence of cataplexy of each of the individual 17 patients in the single-blinded baseline in the 18 SXB-21 protocol. When the paradigm of random 19 assignment to placebo is shown, then there is 20 certainly a demonstration of efficacy between those 21 who were randomized to the placebo group in SXB-21 22 versus those that maintained their Xyrem treatment, 23 which certainly helps to support the efficacy 24 statement in the GHB-3 protocol. 25 [Slide] 51 1 Finally and to summarize, we have 2 presented data to show efficacy of sodium oxybate 3 to reduce cataplexy in 4-week treatment periods in 4 a dose-related manner that is highly statistically 5 significant at the 9 g dose, and approaching 6 statistical significance at the 6 g dose. 7 We have presented supportive data 8 demonstrating statistically significant efficacy of 9 the lower doses, and demonstrated statistically 10 significant efficacy in terms of daytime 11 sleepiness, using the Epworth Sleepiness Scale, 12 again at 9 g. In a scale used in the Lammers study 13 at 60 mg/kg daytime sleep attacks were 14 statistically significantly reduced in all 3 15 studies. We supported the long-term efficacy of 16 Xyrem with return of cataplexy when blindedly 17 assigned to placebo in the SXB-21 protocol. 18 [Slide] 19 I would now like to very briefly summarize 20 the pharmacokinetics studies that were conducted by 21 Orphan Medical. 22 [Slide] 23 In total, we conducted 8 clinical 24 pharmacokinetic studies, including 2 studies in 25 narcoleptic patients and 6 in healthy human 52 1 volunteers. This slide lists the 8 pharmacokinetic 2 studies by their primary objective. 3 The studies included a single dose pilot 4 study in 6 narcoleptics, and a second study in 5 narcoleptic patients comparing acute and chronic 6 dosing over an 8-week period. Normal volunteer 7 studies were conducted to examine the kinetics of 8 Xyrem with respect to gender differences, dose 9 proportionality and the effects of food. Also, 3 10 drug interaction studies were performed with 11 Zolpiden, protriptyline and modafinil as 12 representatives of the 3 classes of drugs used 13 commonly to treat the symptoms of narcolepsy. 14 Lastly, an in vitro study, using human hepatic 15 microzymes, was conducted to assess the effects of 16 oxybate. 17 [Slide] 18 I will only present the studies that have 19 a significant message, and in very brief summary 20 form. This slide displays the results of the dose 21 proportionality study that compared nightly dose of 22 4.5 and 9 g given in 2 equally divided doses at 23 bedtime and 4 hours later. A randomized, 2-day 24 crossover design was utilized, and doubling the 25 dose from 4.5 to 9 g resulted in a nearly 4-fold 53 1 increase in the area under the time concentration 2 curve. The peak plasma concentration and the time 3 to peak concentration changed significantly with 4 doubling the dose, the latter suggesting 5 capacity-limited absorption. C max was higher after 6 the second dose than with the first nightly dose, 7 as has been seen in other studies with divided 8 dosing. 9 These findings indicate non-linear 10 kinetics and capacity-limited elimination and 11 absorption, as reported in previously published 12 studies. 13 [Slide] 14 The results of the effect of food study 15 are displayed graphically on this slide. In this 16 randomized, crossover study 34 healthy subjects 17 were dosed with 4.5 g of Xyrem on 2 occasions 1 18 week apart, either after an overnight 10.5 hour 19 fast or immediately following a high fat 20 standardized breakfast. After the high fat meal 21 the peak plasma concentration decreased by almost 22 60 percent. The median time to achieve peak levels 23 increased from 45 minutes to around 2 hours, and 24 the AUC decreased by 37 percent. All of these 25 differences were statistically significant. The 54 1 apparent half-life was not significantly altered. 2 Thus, the presence of food significantly reduces 3 systemic exposure to GHB, a finding not previously 4 reported. 5 In the 3 volunteer kinetic studies the 6 urinary excretion of Xyrem was measured, and renal 7 excretion was shown to be a minor pathway of 8 elimination, accounting for less than 5 percent of 9 the administered drug. 10 [Slide] 11 As an example of the drug interaction 12 studies, on this slide we present the modafinil 13 results. The upper graph indicates that 14 co-administration of 200 mg of modafinil had no 15 impact on the kinetics of Xyrem. The lower graph 16 demonstrates that 4.5 g of Xyrem had no clinically 17 significant effect on the kinetics of a standard 18 dose of modafinil. 19 Likewise, in the Zolpiden protriptyline 20 interaction studies, no significant kinetic 21 interactions were found. In the separate in vitro 22 study using human hepatic microzymes, sodium 23 oxybate was found to have no effect on 6 cytochrome 24 p450 enzymes either to inhibit or induce their 25 activity. 55 1 [Slide] 2 So in summary, Xyrem oral solution is 3 rapidlyh absorbed and eliminated with a half-life 4 of about one hour. The drug displays non-linear, 5 dose-dependent kinetics, indicative of 6 capacity-limited absorption and elimination. Xyrem 7 kinetics are similar in men and women and do not 8 change with chronic administration at therapeutic 9 doses. 10 [Slide] 11 Chronic dosing did not change the kinetics 12 of Xyrem in a patient population, and a high fat 13 meal appreciably delayed absorption and reduced 14 total systemic exposure to the drug. Three 15 separate in vivo drug interaction studies, as well 16 as the in vitro p450 enzyme study, would suggest 17 the probability of significant drug-drug 18 interaction with Xyrem is minimal. Thank you very 19 much. 20 DR. REARDAN: Thank you. I would now like 21 to introduce Dr. Jed Black, from Stanford 22 University Sleep Center, and he will present on the 23 polysomnographic effects of Xyrem and GHB. 24 Polysomnographic Effects of Xyrem 25 DR. BLACK: Good morning, ladies and 56 1 gentlemen. I would like to summarize the body of 2 data that has been collected over the past 25 years 3 which characterizes the effects of gamma 4 hydroxybutyrate or sodium oxybate on sleep 5 parameters. I will then speculate briefly on a 6 possible mechanism whereby these effects on sleep 7 result in a robust improvement in daytime 8 narcolepsy symptoms seen with this agent. 9 This has been a particular focus of my 10 research in sleep over the past years. That is, 11 how does what happens in the brain at night affect 12 various aspects on daytime function and alertness? 13 It is unexpected that a medication that 14 objectively markedly improves sleep quality also 15 improves measures of daytime alertness as this 16 finding has never been observed with traditional 17 hypnotics or sleep aids. To pursue an 18 understanding of this possible interaction, 6 19 investigations have been conducted in humans. 20 These studies explored the effect of sodium oxybate 21 on a variety of nocturnal sleep parameters, using 22 electroencephalography during sleep and a 23 laboratory test known as polysomnography. 24 The first 3 studies found an increase in 25 slow wave sleep. Slow wave sleep, also known as 57 1 stages 3 and 4 sleep, is the deepest portion of 2 sleep and correlates positively with functions of 3 daytime concentration, attention and alertness in 4 normal subjects. These studies also reveal a 5 reduction in nocturnal awakenings with GHB. 6 The more recent studies of Scrima, Lammers 7 and Orphan Medical explored both measures of 8 nocturnal sleep as measured by polysomnography, or 9 PSG, and measures of daytime sleepiness with the 10 Multiple Sleep Latency Test, or daytime alertness 11 with the Maintenance of Wakefulness Test. 12 [Slide] 13 These 2 studies, the design of which has 14 been reviewed by Dr. Houghton, again found 15 significant reductions in slow wave sleep, that is 16 to say stage 3-4 sleep or slow wave sleep, and 17 reductions in nocturnal awakenings. Additionally, 18 the Scrima group reported a reduction in stage 1 19 sleep, a very light stage of sleep, and the Lammers 20 group noted significant reduction in the percentage 21 of time patients spent awake during nocturnal 22 polysomnography. 23 [Slide] 24 The most recent study, a multi-center 25 trial performed at 4 sites with an enrollment of 25 58 1 patients, was designed to further explore the 2 effects of sodium oxybate on nocturnal sleep 3 parameters and daytime measures of sleepiness and 4 alertness. In this open-label study patients were 5 kept at a stable stimulant dose throughout the 6 protocol. Cataplexy medications were tapered, 7 followed by a 2-week washout and baseline period. 8 Sodium oxybate was initiated at 4.5 g in a divided 9 nightly dose for 4 weeks, then increased to 6, then 10 7.5, then 9 g for 2 weeks each. Nocturnal 11 polysomnography and the Maintenance of Wakefulness 12 Test, or MWT, were obtained at the time points 13 noted here. 14 [Slide] 15 This study revealed the expected increase 16 in slow wave, or stages 3-4 sleep, and increase in 17 delta power. Delta power is the measure of the 18 depth of sleep. It incorporates the combination of 19 the amplitude of the slow frequency waves and the 20 prevalence of those waves through the night to 21 produce a single number called delta power. Delta 22 power is another measure found in a variety of 23 animal and human studies to correlate positively 24 with sleep quality. The calculation of this value 25 requires sophisticated processing which was 59 1 unavailable for the prior studies. The increments 2 in slow wave sleep and delta power were found to be 3 dose related. Dose-related improvements in daytime 4 alertness and subjective sleepiness were also 5 observed. 6 [Slide] 7 The dose-response increase in the number 8 of minutes of slow wave sleep is illustrated in 9 this slide, with an increase from 6 g up to the 9 g 10 dose. The total duration of slow wave sleep 11 increased to over 5-fold that of baseline at the 9 12 g dose. 13 It is important to note that while these 14 results are predicted to be dose related, time on 15 medication cannot be factored out as a potential 16 contributor to these increments. 17 [Slide] 18 Delta power, which characterizes slow wave 19 activity throughout the entire sleep period, not 20 just during stages 3 and 4, was also found to 21 increase in a dose response fashion with a 50 22 percent increase noted at the 9 g dose over 23 baseline. 24 [Slide] 25 The Maintenance of Wakefulness Test, or 60 1 MWT, is a daytime evaluation which places the 2 patient in a dimly lit room in a semi-recumbent 3 position, with nothing to do and with the 4 instruction to remain awake. The duration of 5 sustained wakefulness was measured in this study 6 over 40-minute intervals across 4 periods, spaced 2 7 hours apart during the day. Substantial 8 dose-related increases in the ability to remain 9 awake were observed at both the 4.5 g and 9 g 10 doses. 11 [Slide] 12 As previously noted, the MWT was not 13 performed at the 6 g nor 7.5 g doses in this 14 protocol. Similar marked reductions were found in 15 the Epworth Sleepiness Scale scores. In this 16 measure the individual rates their own potential to 17 fall asleep in a variety of more sedentary daytime 18 activities. 19 [Slide] 20 A post hoc analysis of the possible 21 correlations between sodium oxybate-related changes 22 in nocturnal parameters with changes in daytime 23 measures revealed the strongest correlation 24 occurring with delta power and Epworth Sleepiness 25 Scale scores. This was a negative correlation, 61 1 such that the greater the delta power, the lower 2 the daytime sleepiness. In addition, trends toward 3 significant correlations between delta sleep and 4 MWT scores, and between slow wave sleep and Epworth 5 and MWT scores were observed. 6 [Slide] 7 In conclusion, studies of sodium oxybate's 8 effects on sleep demonstrate increases in measures 9 of restorative sleep, including dose-related 10 increments in slow wave and delta sleep, coupled 11 with and correlated with improvements in measures 12 of daytime alertness and sleepiness. 13 It is postulated that sodium oxybate works 14 directly to enhance brain neurochemical activity 15 critical to the restorative mechanisms of slow wave 16 sleep and of slow wave activity during the total 17 sleep period. Such enhanced activity may be the 18 cause of substantial improvement in both subjective 19 and objective measures of sleepiness and alertness 20 observed with sodium oxybate in narcolepsy. 21 DR. REARDAN: Thank you, Dr. Black. Dr. 22 Houghton will now present the safety summary 23 overview of Xyrem and finish up with a benefit/risk 24 assessment. 25 Safety Overview and Summary of 62 1 Risk/Benefit Assessment 2 DR. HOUGHTON: Thank you. 3 [Slide] 4 I am sorry to horrify you with this 5 complex diagram again but it is just to outline the 6 15 studies that will be referred to today as the 7 updated safety database. The Lammers study was 8 excluded because adverse events were not recorded 9 in the classical way and, as Dr. Katz explained, 10 the Scharf study was separated and will be 11 explained again later. 12 [Slide] 13 The safety profile was reported based on 14 exposure of 479 narcoleptic patients and 125 15 healthy volunteers from the pharmacokinetic 16 studies. This represents an exposure of greater 17 than 6 months in 360 patients in total, and greater 18 than 12 months in 296 patients, which represents a 19 total patient-year exposure of 1328 years with the 20 Scharf database included. 21 [Slide] 22 When exposures were restricted to the 23 studies other than the Scharf database, 399 24 narcoleptics and 125 subjects represent exposure in 25 524 persons. This represents exposure of greater 63 1 than 6 months in 296 patients and greater than 12 2 months in 223 patients, for a total exposure of 330 3 patient-years. 4 [Slide] 5 In the open-label studies patients were 6 titrated between the doses of 3-9 g in divided dose 7 at night. This slide represents the distribution 8 of patients across this defined dose range and, 9 again, identifies the 6 g dose as the most commonly 10 used, followed again by the 9 g dose. In fact, 11 approximately 80 percent of patients were titrated 12 within the 6-9 g range. 13 [Slide] 14 In the updated integrated safety database, 15 composed of 402 patients, 399 of whom were treated 16 with active drug and 3 patients received placebo 17 only, it can be seen that 65 percent of patients 18 completed therapy or were ongoing in the treatment 19 IND study. Thirty-five percent have discontinued 20 treatment for the reasons noted here, with 13 21 percent discontinuing due to adverse events; 2 22 percent discontinuing because of lack of efficacy; 23 and there were 2 deaths that occurred in the 24 treatment IND studies, both due to suicide. 25 [Slide] 64 1 Across all of these studies, 82 percent of 2 treated patients reported any adverse event, as did 3 70 percent of patients exposed to placebo. It is 4 important to note that the placebo exposure 5 represents 4 weeks as compared to active drug 6 treatment over a much longer period of up to 4 7 years. Hence, severe adverse event 8 discontinuations and serious adverse events are 9 significantly greater in the active treatment 10 groups. 11 [Slide] 12 When considered in terms of dose at onset, 13 there seemed to be a slight preponderance of 14 incidence in the 9 g group. 15 [Slide] 16 This slide represents the most frequent 17 adverse events reported across the integrated 18 database. There was a consistent pattern of events 19 across the study. Nausea, dizziness, sleep 20 walking, are represented here as a partial 21 representation of the term sleep disorder, enuresis 22 and confusion were most frequently considered dose 23 related, while others represent intercurrent 24 illness. 25 [Slide] 65 1 This profile is reinforced by 2 consideration of the controlled trials in which 3 there is represented a balanced exposure to placebo 4 and active medication. Again, dizziness, nausea, 5 pain, sleep disorder, confusion, infection, 6 vomiting and urinary incontinence separate. A dose 7 relationship was shown introduction eh GHB-2 trial 8 for confusion, nausea, dizziness and urinary 9 incontinence. 10 [Slide] 11 In the SXB-21 trial the most common 12 adverse events that were reported are shown here. 13 The incidence was very low in this study of 14 patients on long-term treatment, but what is 15 relevant is the data that looks at the possible 16 presentation of a withdrawal syndrome with the 17 abrupt cessation of long-term therapy. 18 [Slide] 19 This is in marked contrast to a severe 20 syndrome that is being described in the abuser 21 population who have significantly escalated both 22 dose and frequency of dosing. When we looked at 23 symptoms that could relate to a withdrawal 24 phenomenon, we saw only 2 patients with anxiety in 25 a circumstance of escalating cataplexy, 1 patient 66 1 with dizziness, 1 insomnia, 1 sleep disorder that 2 actually in verbatim terms, was increased 3 awakenings, and 1 patient with somnolence as their 4 narcolepsy worsened. 5 [Slide] 6 I would like to now address the Scharf 7 database. This was conducted under an investigator 8 IND commencing about 10 years before Orphan's 9 involvement, without any of the rigors of external 10 monitoring, and really represents over 16 years 11 experience in the use of the drug rather than drug 12 development clinical research with regulatory 13 disciplines. 14 Patients were scattered all over the 15 country and, hence, the data is based primarily on 16 diary recordings without medical review and 17 interpretation, leading to a significant 18 discontinuation rate for lack of compliance. Dose 19 accountability and titration were less clearly 20 defined and less controlled. Patients had less 21 defined entry criteria and represent a broader 22 profile of associated pathologies. On this basis, 23 the study data has been reported separately to the 24 integrated database, as Dr. Katz had suggested. 25 [Slide] 67 1 We will address the Scharf open-label 2 experience in terms of dosing exposure, patient 3 disposition, adverse event incidence over 16 years, 4 and then to try and establish some parity with the 5 integrated database. We have considered the 6 adverse event experience reporting in just the 7 first 6 months of the study. 8 [Slide] 9 Patient disposition in the Scharf database 10 is represented in this slide. At the time of 11 database closure 63 patients transferred into the 12 SXB-7 protocol. The FDA expressed concern 13 regarding the accountability of the 80 patients 14 that did not continue. We provided a narrative 15 account for each individual patient, with updated 16 status where possible, in the form of a major 17 amendment. In addition, FDA requested further 18 clarification of adverse events initially deemed 19 uaevaluable, which we have also provided. 20 Of these 80 patients, 8 continued in the 21 Scharf trial under his treatment IND. The 71 22 patients who withdrew had received oxybate for from 23 5 days to 10 years, and the reasons for early 24 withdrawal of the 71 patients were primarily 25 classified into non-compliance, adverse event and 68 1 cost. 2 [Slide] 3 The adverse event profile reflects the 4 length of the study. The relatively large numbers 5 of viral infection, flu syndrome, pharyngitis, etc. 6 shouldn't be worrisome considering the 16 years 7 duration of the study. However, of particular 8 interest is the unusual incidence of sleepwalking 9 and urinary incontinence and these will be 10 discussed in some detail later. 11 [Slide] 12 The most frequent adverse events in the 13 first 6 months of the Scharf trial are shown here. 14 When compared to the integrated safety database, 15 few adverse events separate in incidence. Most 16 notable are somnolence, infection, viral infection 17 and malaise. There were few new adverse events 18 reported after the first 6 months. 19 The FDA requested further information 20 regarding the following adverse events of 21 particular interest. They were represented by 22 incontinence and convulsions, confusion, 23 neuropsychiatric events and sleepwalking. 24 [Slide] 25 I will address incontinence first. In 69 1 their review of the GHB-2 trial, submitted in 2 October, 1998, the FDA requested an analysis of 3 adverse event terms for incontinence in association 4 with central nervous system adverse events 5 suggestive of seizure. 6 [Slide] 7 We responded by initiating the following: 8 a questionnaire to all investigators to review the 9 history of abnormal nocturnal observations that 10 could be suggestive of seizures; a detailed 11 urologic history preceding oxybate therapy and any 12 new neurologic symptoms. 13 Examination of the databases for potential 14 correlation between central nervous adverse events 15 that could be related to seizures and incontinence, 16 either urinary or fecal, was undertaken. Review of 17 both preclinical and clinical data in the 18 literature was performed and an overnight EEG 19 recording after a 9 g dose was conducted in 6 20 patients who had reported incontinence during their 21 oxybate therapy. An expert opinion was provided by 22 Dr. Nathan Chrone, a neurologist of Johns Hopkins 23 University. 24 [Slide] 25 The issue as represented is shown here. 70 1 Urinary incontinence was presented by 8 patients 2 reporting 15 events in the GHB-2 study, by 13 3 patients reporting 51 events over the 2-year period 4 of GHB-3, and in the Scharf study by 33 patients 5 reporting 140 events. 6 When central nervous system events were 7 analyzed for contemporaneous reporting, 2 patients 8 in each of the GHB-2 and -3 trials recorded such 9 events corresponding to episodes of incontinence, 10 as did 7 patients in the Scharf database. 11 Relatively few incontinence events were temporally 12 associated with the CNS adverse events suggestive 13 of seizure. No potential seizure genesis was 14 reported by bed partners in response to specific 15 questions, and many of the partners reported 16 relevant urinary symptoms such as frequent nocturia 17 preceding the Xyrem treatment. 18 [Slide] 19 Single events of fecal incontinence 20 occurred in 4 patients in 4 different trials. 21 Association between these incontinence events and 22 central nervous system adverse experiences were 23 present only in 1 patient in the Scharf trial and 1 24 in the pharmacokinetic SXB-11 trial. In this 25 patient the event of fecal incontinence was 71 1 definitely associated with a seizure in a patient 2 with a known pre-study history of seizures. The 3 subject in the SXB-11 effect of food study was a 4 patient who, while significantly obtunded and with 5 respiratory obstructive symptoms, had a brief 6 episode of fecal incontinence. 7 [Slide] 8 In conclusion, there was limited support 9 for a relationship between incontinence and 10 seizures from the clinical trials, the prospective 11 EEGs or from the literature. 12 [Slide] 13 The vast majority of events that could 14 have been coded as convulsions were actually 15 recorded under the COSTART dictionary as cataplexy 16 events. One patient in the integrated trial 17 database did not represent this classification and 18 he has been investigated by a neurologist for 19 seizure genesis. His fugue state and automatic 20 behavior episodes have been deemed part of his 21 narcolepsy syndrome. 22 In the Scharf database two patients with 23 definite seizures recorded history of preexisting 24 disease, and two other patients recorded seizure 25 events without definitive diagnosis but with 72 1 complicated polypharmacy. 2 [Slide] 3 To now address confusion, in the 4 integrated safety database 30 patients or 70 5 percent reported 48 events recorded as confusion, 6 leading to discontinuation from study in 3 7 patients. A possible dose relationship was 8 suggested by a review of the entire database. In 9 the Scharf database, again 7 percent of patients 10 reported 15 such events, with no discontinuations 11 and no dose relationship pattern observed. 12 [Slide] 13 The coding of confusion embodied a wide 14 range of verbatim terms, as shown here. These do 15 not represent confusion based on a standard medical 16 status examination. They do not differentiate 17 between nighttime events from those of awakening or 18 arousal parasomnias. These events led to no dosage 19 adjustment in 37 instances, but dose was reduced in 20 4 events, led to temporary discontinuation 21 following 4 events, and 3 patients discontinued 22 permanently because of a side effect of confusion. 23 [Slide] 24 When the GHB-2 controlled trial was 25 considered with respect to confusion, the highest 73 1 incidence in the databases is represented in this 2 4-week study by 10 patients. The highest incidence 3 was seen in the 9 g dose, and 6 of the 10 developed 4 during the first week of treatment. Seven of these 5 10 events were in patients over the age of 50. The 6 difference in this study, of course, was the 7 assigned doses rather than dose titration. It is 8 important to note that 1 event was reported in a 9 placebo patient. 10 [Slide] 11 In conclusion, the term represents a 12 symptom report rather than confusion defined in a 13 medical sense by formal mental status examination, 14 and all resolved usually without interruption of 15 therapy or dose modification. Confusion and other 16 associated symptoms are not unexpected with 17 sedating medications. The blinded, controlled 18 trial results suggest that a higher incidence may 19 result without dose titration. 20 [Slide] 21 Neuropsychiatric events will now be 22 reviewed. The adverse event database was searched 23 for terms that could represent neuropsychiatric 24 symptoms, and this led to the classification shown 25 in this slide. Fifty-two patients reported 57 such 74 1 events in the integrated safety database, of whom 2 12 discontinued as a result of these events. In 3 the Scharf database 41 patients reported 84 such 4 events, leading to 2 patient discontinuations. 5 [Slide] 6 Of these 57 events, 1 occurred while a 7 patient was on placebo. This slide lists the terms 8 examined and some, such as stupor and coma, failed 9 to represent neuropsychiatric events. Many 10 represented symptoms of narcolepsy such as 11 hypnagogic hallucinations COSTART-coded to the term 12 hallucinations. The most frequent was clinical 13 depression, and this represents a symptom rather 14 than a diagnosis of major depressive disorder. 15 Depressive symptoms are frequent accompaniments in 16 narcolepsy, and this is well recorded in the 17 literature. Suicide was attempted in 4 patients 18 with major preexisting psychiatric history, and 19 resulted in death in 2 of these patients. The 20 other representations of psychotic disorders and 21 the patient with manic depressive disorder also 22 occurred in patients with preexisting major 23 psychiatric disease. As is shown, a similar 24 profile of reported symptoms is found in the Scharf 25 database. 75 1 [Slide] 2 In conclusion, most patients with major 3 events had a preexisting psychiatric disorder. 4 Many events do not qualify as neuropsychiatric 5 disorders, as was represented by the terms pointed 6 out. Assignment of causality is very difficult 7 because narcolepsy is associated with depression 8 and even mechanistically there has been an 9 association between psychosis and the central 10 processes in narcolepsy. As Dr. Mignot mentioned, 11 stimulant medications are associated with central 12 nervous system side effects that are represented by 13 neuropsychiatric symptoms. And, it is true to say 14 that in many patients, particularly in the Scharf 15 database, pre-study screenings were deficient. 16 [Slide] 17 To lastly address sleepwalking, in the 18 integrated safety database 7 percent of patients 19 reported such events, whereas in the Scharf 20 database 32 percent of patients reported events 21 that were listed as sleepwalking. In the Scharf 22 trial, however, these reports were primarily data 23 listings in patient diaries in response to a 24 specific leading question, listed as a line item in 25 the diary. 76 1 [Slide] 2 The listing of this term did not receive 3 the benefit of medical consideration of a 4 differential diagnosis of somnambulism, and since 5 most patients were not seen by the investigator no 6 clarification was provided. Post hoc consideration 7 was rendered impossible given the lack of 8 information regarding sleep stage, time of night, 9 relationship to drug dosing, and could be 10 representative of any of the differential diagnoses 11 listed on this slide. 12 [Slide] 13 In the controlled trials only 3 14 sleepwalking events were reported, 2 of which 15 occurred on active treatment and 1 occurred in a 16 patient during placebo treatment. 17 [Slide] 18 Hence, in conclusion, the incidence in the 19 integrated safety database of 7 percent is not 20 particularly dissimilar to the range reported in 21 the literature for normal patients. This was 22 reported by Dr. Mahowald, of Minneapolis, as 23 between 4-10 percent in a publication in 1998, and 24 between 1-7 percent by Dr. Roger Broughton of 25 Canada. 77 1 Diary recording without medical 2 classification represents a potential increased 3 reporting in the Scharf trial. The slight increase 4 in incidence over the general population may 5 certainly be representative of Xyrem effects with 6 increase in slow wave sleep, but REM behavior 7 disorder, common in narcolepsy, mayou be a separate 8 consideration. 9 [Slide] 10 To summarize the safety profile of this 11 drug, we based our assessment to date on 604 12 patients, which represents 524 patients excluding 13 the Scharf database. Dosing was between 3-9 g per 14 day in divided nightly dosing. The common adverse 15 events were certainly headache, unspecified pain, 16 nausea, dizziness, and less common but important 17 adverse events were vomiting, confusion, 18 restlessness, agitation, sleepwalking and enuresis. 19 [Slide] 20 All events have been reversible. There 21 were no significant changes in lab values or vital 22 signs identified across the studies. There was no 23 evidence of organ toxicity outside the 24 pharmacologic effects in the central nervous 25 system. There was no diversion or consumption of 78 1 clinical trial supplies by any family members 2 during the trials, and there was certainly no 3 evidence of Xyrem diversion in our database. 4 [Slide] 5 I would like to conclude with the 6 statement that Xyrem was generally well tolerated. 7 [Slide] 8 To commence a risk/benefit assessment, I 9 would like to remind you of the indication proposed 10 by Orphan Medical for the use of Xyrem. That is, 11 to reduce the incidence of cataplexy and to improve 12 the symptom of daytime sleepiness in patients with 13 narcolepsy. 14 [Slide] 15 As has been pointed out, narcolepsy is an 16 uncommon disease, with an incidence of around 0.05 17 percent and, as such, has been qualified for orphan 18 designation. There are no therapies approved for 19 the treatment of cataplexy. Because of this, the 20 FDA were very kind to apply a priority review to 21 our submission and we are very appreciative of that 22 recognition. Current off-label therapies, so well 23 described by Dr. Mignot, are unsatisfactory. 24 Excessive daytime sleepiness has approved therapies 25 but these do not address cataplexy. There is 79 1 clearly a medical need existing beyond the 2 therapies available. 3 [Slide] 4 The benefits of Xyrem in the trials 5 presented were based on patient diary recordings, 6 investigator ratings of overall clinical 7 improvement in overall disease severity, and 8 objective measures of changes in sleep architecture 9 and daytime response. 10 [Slide] 11 Clinical benefit in the short-term 12 reduction in cataplexy was shown by the 13 dose-related reduction in cataplexy in the GHB-2 14 and Scrima studies and in the long-term efficacy in 15 the SXB-21. Subjective changes in the Epworth 16 Sleepiness Scale have been well demonstrated, and 17 reduction in daytime sleep attacks have accompanied 18 this change. Early objective Maintenance of 19 Wakefulness Test data supported these changes in 20 daytime sleepiness. The global impression of the 21 investigators for overall changes in disease 22 severity also showed a significant dose 23 relationship. 24 [Slide] 25 Xyrem was generally well tolerated when 80 1 used in the proposed dose range, with the most 2 common side effects reported including nausea, 3 dizziness, headaches, pain and confusion. Less 4 common but important associated effects include 5 enuresis and sleepwalking, with a possible dose 6 relationship suggested. Although there were 11 7 deaths in the Scharf trial over 16 years and 2 8 deaths by suicide in the integrated database, no 9 deaths were associated with Xyrem. 10 [Slide] 11 In relation to the specific FDA inquiries, 12 there is a possible relationship between Xyrem 13 therapy and somnambulism but further definition is 14 required. There is a marked discrepancy between 15 the reported incidence in the Scharf study of the 16 32 percent, recorded solely by diary entry in 17 response to a leading question, and the 7 percent 18 in the integrated database, which is really in the 19 range in public literature for the normal 20 population. In the controlled trials there were 21 only 3 such reports in total, 2 recorded in active 22 treatment and 1 during placebo treatment. 23 [Slide] 24 Confusion is also an adverse accompaniment 25 of sedative hypnotic drugs and has been identified 81 1 as an occasional side effect of Xyrem. Dose 2 titration may assist in limiting this side effect 3 but it remains an important component of patient 4 and physician education. 5 [Slide] 6 The incidence of enuresis with Xyrem 7 treatment supports an association that may be dose 8 related, but any association of these events with 9 seizure activity is very weak. In terms of Xyrem 10 causing seizures at the therapeutic doses, there 11 was no reliable support for such causality. In 12 this regard, the coding to the COSTART dictionary 13 terms of cataplexy as convulsion was confusing. 14 However, there were 2 patients recording seizures 15 with preexisting causes. Two further patients in 16 the Scharf database reported seizures where 17 confounding contributions rendered assignment very 18 difficult. One patient in the Orphan studies 19 represented a complex history of symptoms 20 characterized by fugue state and these symptoms 21 have been attributed to his narcolepsy syndrome. 22 [Slide] 23 No significant measures were seen in 24 laboratory measures, vital signs or ECG measures 25 and these changes were comparable across the 82 1 treatment groups. There was no evidence of organ 2 toxicity at therapeutic doses that were not part of 3 the central nervous system pharmacology of the 4 drug. 5 [Slide] 6 We did not identify any evidence of 7 kinetic or dynamic tolerance in the narcoleptic 8 populations studied and the absence of drug-drug 9 interactions in the 3 classes of drugs commonly 10 used in narcolepsy, along with the absence of 11 either induction or inhibition of the oxybate p450 12 enzyme system make it possible to predict that 13 drug-drug interactions should be minimal. 14 [Slide] 15 Although a serious withdrawal syndrome has 16 been described in the abuser population that 17 relates to escalation in both dose and frequency of 18 dosing, no evidence of withdrawal has been 19 demonstrated in patients maintained on long-term 20 therapeutic doses in narcolepsy. Following abrupt 21 discontinuation of long-term dosing in the blinded 22 study, only 2 patients reported anxiety but in the 23 presence of worsening cataplexy, with 1 patient 24 reporting mild dizziness and 1 report of insomnia. 25 [Slide] 83 1 We have not attempted in any way to 2 minimize the issue of abuse with GHB or its 3 precursors. We recognize that this is a serious 4 problem, but stress the fact that this has been 5 peripheral to the development program in 6 narcolepsy. We have detected no evidence of abuse, 7 diversion or self-escalation of dosing in patients 8 in clinical trials. Great efforts have been 9 applied to working with the appropriate expert 10 bodies to plan a restricted distribution system to 11 support in every way the unique bifurcated 12 scheduling legislated by Congress and to plan 13 physician and patient education to minimize the 14 possibility of diversion. This will be greatly 15 facilitated by the documentation centrally of 16 prescribing and patient use. This will be 17 described in detail to you later. 18 [Slide] 19 In conclusion, I would propose that we 20 have established statistically and clinically 21 significant evidence for the reduction in 22 cataplexy, and for improvement in daytime 23 sleepiness when used concomitantly with stimulant 24 medications. 25 Xyrem is generally well tolerated, with a 84 1 safety profile well characterized in this orphan 2 population by long-term exposure. The medical 3 benefits clearly outweigh the risks for a 4 therapeutic agent that may be the first single 5 agent to address the multiple symptoms of 6 narcolepsy. Thank you very much. 7 DR. REARDAN: I would just like to thank 8 the committee and FDA for your attention. I 9 believe Dr. Mani has some comments, or we are now 10 happy to take questions from the committee. 11 DR. KAWAS: The FDA will give us a 12 response to the presentation, and then we will 13 probably take a break before we have questions, 14 unless the committee has anything burning they need 15 to ask now. Dr. Ranjit Mani will present for the 16 FDA. 17 FDA Response to the Presentation 18 DR. MANI: What I propose to do in the 19 next few minutes is address two issues where our 20 views diverge somewhat from those of the sponsor. 21 [Slide] 22 The first is the effect of GHB on measures 23 of daytime sleepiness in narcolepsy. 24 [Slide] 25 This overhead illustrates how many 85 1 measures of daytime sleepiness there were in the 2 GHB efficacy trials. As you can see, GHB-2 had 3 3 measures of daytime sleepiness; the Scrima study 4 had 2, of which 1 was primary; and the Lammers 5 study had 2. I will draw your attention to the 6 fact that, with the exception of the Scrima study, 7 the remaining measures were all designated as being 8 secondary. 9 [Slide] 10 Because what is considered statistically 11 significant does depend or could depend on the 12 number of comparisons made, I think it is also 13 important to illustrate how many secondary efficacy 14 measures there were in each trial. In the GHB-2 15 trial I was able to count a total of 10; in the 16 Scrima study 17; and in the Lammers study 7. 17 [Slide] 18 This is based on data provided by Orphan. 19 As you can see, in the GHB-2 trial the Epworth 20 Sleepiness Scale measure did reveal a fairly 21 clear-but efficacy for GHB but only at the 9 g 22 dose. The p value of 0.001 probably remains 23 statistically significant even when adjustment is 24 made for multiple comparisons. 25 On the other hand, the frequency of 86 1 daytime sleep attacks and duration of daytime sleep 2 attacks should probably be considered negative 3 evidence of efficacy if adjustment is made for 4 multiple comparisons. 5 [Slide] 6 Again, in the Scrima study one primary 7 efficacy measure was sleepiness index of the 8 Multiple Sleep Latency Test. Here, the results 9 must be considered negative whether adjusted for 10 multiple comparisons or not. 11 [Slide] 12 The other measure was the frequency of 13 daytime sleep attacks, again negative whether 14 adjusted for multiple comparisons or not. 15 [Slide] 16 In the Lammers study the severity of 17 daytime sleepiness was 1 of 7 secondary efficacy 18 measures which is probably negative when adjusted 19 for multiple comparisons. On the other hand, the 20 frequency of daytime sleep attacks was positive, 21 but using an ANCOVA which was not a protocol 22 specified analysis. 23 [Slide] 24 So, here are the problems as we see them 25 with the proposed claim for excessive daytime 87 1 sleepiness. Most measures were secondary. The 2 only measure that was primary was negative. The 3 majority of measures were negative after adjustment 4 of the Type 1 error for multiple comparisons. The 5 effects were inconsistent across studies, and the 6 clearly positive results on the GHB-2 trial on the 7 Epworth Sleepiness Scale were not replicated. As 8 mentioned, the approval of modafinil for the 9 treatment of excessive daytime sleepiness was based 10 on replicated results in 2 efficacy studies. And a 11 minor point, the results on the GHB-2 study were, 12 to some extent, confounded by concurrent stimulant 13 use, raising the question, among other questions, 14 of whether Xyrem is effective as monotherapy for 15 the treatment of excessive daytime sleepiness. 16 [Slide] 17 The second issue that I want to address 18 briefly is that of sleepwalking. As you can see, I 19 have put it in quotes. As Bill Houghton has 20 already emphasized, we do not know what these 21 episodes represent. They have not been clinically 22 characterized. 23 [Slide] 24 The term sleepwalking does not correspond 25 to the medical entity of somnambulism. The term is 88 1 based entirely on patient diary entries, and there 2 has been no attempt to characterize the episodes 3 further and define what clinical entity they 4 correspond to. 5 The incidence of these episodes, whatever 6 they may represent, was approximately 32 percent. 7 The majority of patients did list as having more 8 than one episode. A single patient had a total of 9 346 episodes over a 5-year period. As already 10 said, an adequate clinical description is lacking, 11 and the episodes cannot be said to be completely 12 benign. 13 There was one patient who is reported to 14 have overdosed twice during two consecutive 15 episodes of sleepwalking. During one episode the 16 patient became comatose and needed to be 17 hospitalized, needed to be on a ventilator for some 18 hours but completely recovered. A second pat had 19 multiple episodes of sleepwalking. She was found 20 by her husband to be smoking, apparently 21 inadvertently. During one such episode her clothes 22 were set on fire. The fire was put out. She was 23 taken off GHB and did not have any further such 24 episodes. A third patient is reported to have 25 swallowed nail polish remover during an episode, 89 1 without any serious consequences. 2 I would also like to add one minor point 3 in response to Dr. Houghton's presentation. That 4 is, I believe that in the Scharf study there was 5 one patient who was withdrawn from the study 6 because he felt that he had benefitted from Xyrem 7 and decided that these benefits could be extended 8 to a circle of friends who also received part of 9 his own supply, again apparently without serious 10 consequences. Thank you. That is really all I 11 have to say. 12 DR. KAWAS: Thank you, Dr. Mani. Does the 13 committee have any questions they would like to ask 14 before the break? If not, we will reconvene this 15 meeting at 10:30 sharp. 16 [Brief recess] 17 Committee Discussion 18 DR. KAWAS: Will you please have a seat so 19 we can reconvene this session? This meeting of the 20 Peripheral and Central Nervous System Advisory 21 Committee is now reconvened. We appreciate the 22 presentations from the sponsor and the FDA, and the 23 floor is open for questions. The first question is 24 going to come from someone who has been patiently 25 sitting on the phone. Dr. Chervin, can you hear 90 1 me? 2 DR. CHERVIN: Yes, thank you. 3 DR. KAWAS: Dr. Chervin, we can't year you 4 yet, if you will give us a moment to do whatever it 5 is we have to do? 6 DR. CHERVIN: Can you hear me now? 7 DR. KAWAS: Give it a shot. 8 DR. CHERVIN: I have a question perhaps 9 for Dr. Houghton. In regard to the safety 10 experience with the 1328 patient years, were there 11 any reports that alcohol was taken in the evening 12 in combination with GHB? If so, what was the 13 outcome? 14 DR. HOUGHTON: It was certainly 15 recommended as a contraindication in our protocols. 16 The advice to the patient was that they not consume 17 alcohol during the studies. I can't vouch for the 18 fact that it was entirely complied with, but we 19 don't have protocol or database record of 20 consumption of alcohol during the trials. There 21 certainly is record of patients having imbibed 22 during the Scharf study and I am not in a position 23 to clarify that. 24 DR. GUILLEMINAULT: This is Dr. 25 Guilleminault. I have also a question, and it is 91 1 for Dr. Mani, about the sleepiness data. Was there 2 the slow wave sleep information looked at for 3 sleepiness? As you know, delta power greatly 4 improves alertness and there are many studies, 5 sleep deprivation studies and investigation into 6 sleep disorders such as obstructive sleep apnea, 7 where it is very clear that decrease in delta power 8 and in slow wave sleep has a big impact on the 9 alertness, and the more delta power you have and 10 the more slow wave sleep you have, the better 11 alertness the next day. 12 So, one of my understandings is that this 13 drug has an impact on slow wave sleep and delta 14 power. Was there any analysis of that in data 15 looking at alertness? 16 DR. MANI: To the best of my knowledge, it 17 was not listed as an efficacy measure in any of the 18 controlled studies that I looked at. 19 DR. GUILLEMINAULT: Okay. The second 20 question is maybe a question about my ignorance. I 21 did not understand exactly the statistic about the 22 ESS because in the investigation of the results of 23 the ESS there was an investigation with negative 24 studies. All the results, when you look at 25 everything there, was there a positive p value? 92 1 Was there a statistical difference? Because I 2 don't understand the manipulation which was done. 3 Maybe through poor knowledge, I have never seen 4 this type of manipulation. 5 DR. REARDAN: Dr. Guilleminault, which 6 study are you referring to when you ask about the 7 Epworth Sleepiness score? 8 DR. GUILLEMINAULT: I think OMS-2. 9 DR. REARDAN: Is that for Dr. Mani, or do 10 you want to pose that to the company? 11 DR. GUILLEMINAULT: No, I was asking that 12 because Dr. Mani reported that he looked at that 13 study and classified the results, and my 14 understanding, and it may be a wrong understanding, 15 is that he made a subdivision in looking at the 16 results and I did not see completely the 17 statistical rationale for that approach. 18 DR. MANI: Are you referring to the 19 statistical adjustments for multiple comparisons? 20 Is that what you mean? 21 DR. GUILLEMINAULT: No, the Epworth 22 Sleepiness Scale study in GHB-2, secondary efficacy 23 daytime sleepiness on your slide, and I did not 24 understand exactly how that was analyzed, the type 25 of analysis that was done or redone. 93 1 DR. MANI: Perhaps I should ask the Orphan 2 statisticians to explain that in greater detail, 3 but the analysis was an ANCOVA. 4 DR. GUILLEMINAULT: The microphone must be 5 poorly placed because we cannot hear the response. 6 DR. MANI: Can you hear me now? 7 DR. GUILLEMINAULT: Yes. 8 DR. MANI: The analysis was an ANCOVA. I 9 mean, perhaps I should get the Orphan study 10 statistician to explain the analysis to you in 11 greater detail. 12 DR. REARDAN: I am just asking Dr. Richard 13 Trout, the statistician, to comment on how the 14 Epworth Sleepiness score was statistically 15 analyzed. 16 DR. TROUT: Hi. My name is Dick Trout. 17 First of all, the analysis was just as you 18 described, that is to say it was an analysis of 19 covariance which was preplanned. I think the 20 concern that you expressed was the fact that it was 21 listed as a secondary efficacy measure -- 22 DR. GUILLEMINAULT: Right. 23 DR. TROUT: -- as compared to a primary, 24 and there was a number of secondary efficacy 25 measures, but even if one adjusted for the multiple 94 1 testing which I think you were concerned about, the 2 9 g separation from the placebo group would still 3 be significant. We already adjusted for the 4 multiple testing with regard to the dosing issue, 5 using Dunnett's test, but your concern was with 6 regard to the fact that there were a number of 7 secondary efficacy measures which would then 8 diminish the effect. 9 DR. GUILLEMINAULT: Okay, thank you. 10 DR. PENN: I can see that the claim for 11 helping daytime sleepiness is going to be one that 12 we will want to look into very carefully, and I 13 want to ask our FDA statistician a question about 14 that in a general sort of way. If you were a 15 gambling person, which I assume a statistician 16 would not be -- 17 [Laughter] 18 -- from the data that you have looked at 19 for 9 g, would you say that in a good controlled 20 trial you would bet on it working to decrease 21 daytime sleepiness? It looks like the strongest 22 data is at 9 g and that is what the company is 23 suggesting. I am going to ask you to bet on that, 24 and then I am going to make a point. 25 DR. MANI: You addressed the question to a 95 1 statistician; I am not a statistician. 2 DR. PENN: Oh, I am sorry. Anybody else 3 want to gamble with this? 4 DR. REARDAN: Coming up to the podium is 5 Dr. Sharon Yan, who is the FDA statistician that 6 has been working on the Xyrem program. 7 DR. YAN: Basically we rely on the results 8 that were prespecified, and a lot of results that 9 we looked at -- and you want me to bet -- after 10 looking at those results, most people would bet 11 that the data shown, for example, the 9 g it seems 12 that it is highly positive; it is highly 13 significant, but we rely on the analysis which is 14 prespecified. Without that, the data information 15 -- it is hard to bet on anything. 16 DR. PENN: But I am asking you how you 17 would bet on that if you had to make a bet now in 18 Las Vegas, and the point I am trying to make is 19 that it seems to me a reasonable bet that it does 20 help daytime sleepiness but that they haven't 21 presented two clean studies that show at 9 g that 22 that is the case. And, is there going to be some 23 middle ground to this where that claim can be put 24 in language that would be acceptable later on? So, 25 I wanted to see if you agree that that analysis 96 1 then presenting of the problem is the correct one, 2 that is, that there is very strong suggestive 3 evidence, not as strong as we often want for a 4 claim, that it helps daytime sleepiness. When you 5 sit back and you look at all the data, would you 6 bet on that helping daytime sleepiness? 7 DR. KAWAS: Perhaps Dr. Katz could help 8 with this response. 9 DR. KATZ: Yes, again, I will just sort of 10 reiterate something that Dr. Yan has already said, 11 which is that whether or not we personally believe 12 something is true or what we would bet on is not 13 really the standard. The standard which we apply 14 is what the law requires, which is substantial 15 evidence of effectiveness, ordinarily defined, 16 unless there is some compelling reason to do 17 otherwise, as data from at least two adequate and 18 well-controlled trials demonstrating effect. We 19 have adopted by tradition a usual sort of 20 statistical rule by which we decide whether or not 21 a study is "positive" for a particular indication. 22 So, I think that is the standard. Unless there is 23 some, as I say, very compelling reason to apply 24 some different standard, like what would I bet on 25 or what my personal belief is, that is the standard 97 1 we need to apply. Again, unless there is a view 2 that there is some compelling reason to apply some 3 different standard, we would ask you as a committee 4 whether you think that the evidence for that 5 particular claim meets that standard. 6 DR. PENN: So, once again the question 7 should go then to Orphan, whether or not they feel 8 they have met that standard on two separate 9 occasions using their 9 g amount, and I haven't 10 gotten a clear-cut idea in my mind whether they are 11 really claiming that or just showing us data that 12 would be for a good bet. 13 DR. YAN: May I clarify one thing? For 14 the analysis for daytime sleepiness for GHB-2 the 15 sponsor showed it was highly significant, with a p 16 value of 0.001, and I analyzed the data with the 17 original scale and, as I analyzed it, it shows that 18 the normal assumption was validated and then the 19 log transformation to then improve the data, and I 20 used nonparametric analysis to analyze the p value, 21 and it is not that small. As I remember, the p 22 value is 0.03 or something. 23 DR. REARDAN: I can comment on the trials. 24 We have GHB-2, obviously, where the trial was very 25 effective. I don't think there is a dispute with 98 1 FDA on that. The question is do we meet the 2 standard of two well-controlled trials for that 3 indication. The data in support of that comes from 4 the Lammers study. The sleepiness scale used there 5 was something he developed, not a validated scale 6 but it was statistically significant for daytime 7 sleepiness, albeit in a very small, 24-patient 8 crossover trial. 9 So, we have a small supportive study. We 10 have the large controlled study, GHB-2. That is 11 the evidence basically. Bill, do you want to 12 comment? 13 DR. HOUGHTON: Yes. We are not trying to 14 make this something that it is not in any way, and 15 if you apply the absolute, most rigorous standards 16 of normal drug development to our database, we have 17 a small database. We did have the two components 18 that were statistically significant. This was 19 supported by the reduction in daytime sleep attacks 20 which are very clinically significant to the 21 patient, and we had two components of statistical 22 significance there. 23 The other issue, and I know that this from 24 a pure mathematical sense is problematic, is the 25 evidence of long-term support in daytime sleepiness 99 1 claim with the GHB-3 protocol, which showed the 2 Epworth Sleepiness Scale and the daytime sleepiness 3 reduced and maintained over the long period of 4 time. The fact then that the objective data in 5 SXB-20 was so strongly supportive and the change in 6 Maintenance of Wakefulness Test is an objective 7 measure and was clearly positive was very 8 important. 9 The part that concerns me from a clinical 10 point of view is if you look at the patient 11 profiles as they enter the studies, they are on 12 stable doses of stimulants and, yet, their ratings 13 are very low. The real issue is that daytime 14 sleepiness with current medications isn't well 15 addressed. So, the question is not only have we 16 shown absolute irrevocable evidence of long-term 17 efficacy for daytime sleepiness with the existence 18 of the present treatments for long-term 19 effectiveness, what we didn't do is ask for a claim 20 in daytime sleepiness. 21 [Slide] 22 Our proposed indication was to improve the 23 symptom. We didn't attempt to do studies that 24 displaced the stimulant therapies. What we are 25 really looking at is a hand-in-glove approach that 100 1 actually makes patients better as an incremental 2 change, and all therapies up to now have been very 3 separate. The symptoms of daytime sleepiness and 4 those of the associated REM phenomena have been 5 treated by entirely separate medications. If there 6 is a component of Xyrem that assists in daytime 7 sleepiness as an incremental change, we think it is 8 very clinically important and that is what we 9 sought to present today. I want to stress very 10 clearly that we are not looking for the claim of 11 daytime sleepiness; we are looking at an 12 improvement in the symptom thereof. 13 DR. KAWAS: Dr. Houghton, can I ask you 14 then, to my reading, that indication is actually 15 two indications, I mean, cataplexy and sleepiness 16 being a separate one. When I was reading the 17 materials that you very carefully provided us, 18 obviously for cataplexy the GHB-2 and the SXB-21 19 study speak to that issue as pivotal trials. I was 20 going to ask you which were the two that speak to 21 the issue of daytime sleepiness. Now I understand 22 them to be the GHB-2 and the Lammers small trial 23 with the questionnaire that was developed there. 24 In both of those cases, however, we are talking 25 about subjective sleepiness from the Epworth scale 101 1 and the other question. Since there are factors 2 that can influence someone's subjective feelings of 3 sleepiness, do you have any objective measures that 4 support the indication of daytime sleepiness? 5 Specifically, the one trial that I am aware of that 6 had an MSLT and did daytime sleepiness as a primary 7 outcome measure, in fact, appears to be not 8 supportive of the indication. 9 DR. HOUGHTON: Yes, in the Scrima trial he 10 used the MSLT measure and that was not 11 statistically significant, as shown. The objective 12 data that we propose supports very strongly the 13 effect of adequate dosing of GHB was the SXB-20 14 trial that Dr. Black discussed. That is not only a 15 profound improvement in the MWT at the 9 g dose but 16 a defined dose response across all doses. That is 17 very positive data. 18 DR. KAWAS: In ten patients, it appears. 19 DR. HOUGHTON: Twenty-one. 20 DR. MANI: May I also add that that was an 21 open-label, non-randomized study? 22 DR. HOUGHTON: Sure, but using an 23 objective measure. 24 DR. RISTANOVIC: I am I am Ruzica 25 Ristanovic, medical director of Sleep Disorders 102 1 Center, in Evanston, Illinois. I would like to 2 comment on add-on Xyrem in the presence of other 3 stimulants. Other studies attempt to try to 4 document the effectiveness of other stimulants in 5 narcolepsy-related sleepiness documents, including 6 the most rigorous trial of modafinil in 7 double-blind, placebo-controlled studies. They 8 document that these drugs improve sleepiness but 9 very seldom outside of the range of pathological 10 sleepiness as measured by Multiple Sleep Latency 11 Test and Maintenance Wakefulness Test. So, the 12 patients remain sleepy. That is the message. 13 Add-on treatments are approved for other 14 indications in other neurological diseases, such as 15 epilepsy. So, I assume that this application for 16 that particular indication is not for monotherapy 17 but as an add-on to concurrent use of stimulants. 18 I would like to bring this to your attention. So, 19 patients do remain sleepy on stimulants and they 20 need additional treatments. 21 DR. KAWAS: Dr. Temple? 22 DR. TEMPLE: Dr. Houghton also seemed to 23 be distinguishing between monotherapy and add-on 24 therapy. That is not the problem. The problem is 25 whether there is adequate support for use as an 103 1 addition for whatever else the patient is on, and 2 whether there are well-controlled studies that 3 support that. So, add-on would be perfectly fine. 4 That is usually true in a lot of conditions, not 5 just neurological ones, where you continue to give 6 standard therapy and try to improve it. 7 I just want to make one observation about 8 the evidence. We do expect to see replicated or 9 reproduced findings. Some of the issues here are 10 whether the fact that the endpoints are secondary 11 and need some correction means that there isn't 12 adequate support. A lot of these things are 13 matters of judgment that the committee can weigh in 14 on. Not everything is, you know, a yes/no. Some 15 of the things are moderately subtle and that is why 16 this is being brought to you for judgment. There 17 is one study that is obviously stronger than the 18 rest but the others can be considered, and you sort 19 of have to think about how many real endpoints 20 there really are; how much of a correction is 21 needed. Those are difficult discussions but worth 22 considering. 23 DR. KAWAS: Dr. Katz? 24 DR. KATZ: I agree, but I think we would 25 still have to have the application meet the 104 1 standard of independent replication, in other words 2 two trials. You can decide that one of the other 3 trials actually does meet the usual standard, 4 again, taking into consideration the multiplicity 5 and that sort of thing. All I am saying is that I 6 don't think we can say we have one study that looks 7 good. If you believe that GHB looks good and the 8 others sort of contribute to a feeling that it 9 probably is okay, I mean, we really need two 10 independent sources that you believe demonstrate 11 the effectiveness. 12 The only other point I wanted to add is to 13 something, Claudia, you said which has to do with 14 Dr. Houghton's view that they are not going for a 15 claim of daytime sleepiness; they just want, I 16 guess, to have language in the labeling that says 17 that it improves that symptom. Most of the drugs 18 we approve are for symptomatic claims, so there is 19 no question that the inclusion of this language in 20 the indication is a claim as we always understand 21 that term. 22 DR. KAWAS: Dr. Guilleminault, followed by 23 Dr. Wolinsky, please. 24 DR. GUILLEMINAULT: If you look at all the 25 published data on modafinil, on amphetamine, on 105 1 methylphenidate, none of these drugs ever 2 normalized all the objective tests on alertness and 3 daytime sleepiness. None of them, including the 4 modafinil data which were approved by the FDA. The 5 MSLT and MWT for all these drugs are pitiful. The 6 only data which shows significance was the Epworth 7 Sleepiness Scale, which is a subjective scale, in 8 all these trials. So, we cannot expect to have any 9 positive result with subjective tests in any of 10 these drugs. We will always have to rely on 11 subjective tests even if the subjective test is not 12 great. Everybody in the field agrees that the 13 Epworth Sleepiness Scale is the most used scale 14 despite the fact that it has a lot of downfall, and 15 we have to remember that when we look at what has 16 been approved and what is being used. 17 DR. KAWAS: Thank you, Dr. Guilleminault. 18 I think that many people would agree with those 19 comments, but my question to you would be not 20 whether or not the Epworth Scale subjective 21 measurements are good but do we have two 22 randomized, controlled trials that show an 23 improvement in subjective sleepiness. 24 DR. GUILLEMINAULT: That was my initial 25 question because my understanding is, when the 106 1 statistician from the FDA responded, she said that 2 when she did a nonparametric analysis she found out 3 that she had a p value of 0.03. So, my 4 understanding is that she had a significant finding 5 even when she did the reanalysis. That was my 6 understanding of her response. 7 DR. KAWAS: Would you like to comment, Dr. 8 Yan? 9 DR. YAN: I am sorry, the previous number 10 is not right. I checked. The number for the 11 nonparametric analysis, the p value was 0.0109. 12 DR. WOLINSKY: I have a couple of 13 questions first for some information before I ask 14 the real question. For the informational questions 15 perhaps Dr. Mignot could help with. So, the first 16 question I have is if you could enlighten us or 17 re-enlighten us about how many patients that have 18 narcolepsy have had cataplexy as a component 19 symptom. What proportion? 20 DR. MIGNOT: In most case series it is 21 about 70 percent. 22 DR. WOLINSKY: The second question is that 23 at least for most of these studies which were done 24 and presented to us since cataplexy was being 25 measured, as is appropriate, the number of 107 1 cataplectic attacks was relatively high. I think 2 in these studies it was around 20 cataplectic 3 attacks per week. So, how many of the 70, 75 4 percent of patients with narcolepsy who have 5 cataplexy have cataplectic attacks at that level? 6 DR. MIGNOT: I would guess 20 percent. 7 DR. WOLINSKY: Thank you very much. 8 DR. MIGNOT: Yes, roughly. 9 DR. WOLINSKY: And then they would fall 10 down below that level for the remainder of the 55 11 percent of narcoleptics with cataplectic attacks. 12 DR. MIGNOT: If you analyze the spread of 13 the number of cataplexy episodes per week, but you 14 have to balance that also with the efficacy of 15 current treatments. A lot of people that currently 16 have cataplexy that is relatively mild just don't 17 want to take the antidepressants because they have 18 so many side effects, especially sexual side 19 effects, dry mouth, all these problems -- 20 DR. WOLINSKY: This is not the question 21 though. So, now the question to Orphan which has 22 really, truly become an orphan drug question, is 23 since all of the studies that have been done have 24 enriched for cataplexy, do we have any data that 25 would suggest that if cataplexy is adequately 108 1 controlled or if there is no cataplexy so we don't 2 have to worry about the control of cataplexy there 3 would be any effect of the drug on daytime 4 sleepiness in non-cataplectic narcoleptics? 5 DR. REARDAN: I think Jed Black wants to 6 make a comment on that. 7 DR. BLACK: Just a comment on the 8 prevalence of cataplexy in the 70-75 percent of 9 folks with narcolepsy that had cataplexy, the 10 frequency of events -- this is something that Dr. 11 Mignot is not aware of, the cataplexy was 12 subdivided into major events and minor events. 13 About 20 percent or so would have the major events 14 to that level, but when we look at the minor events 15 a far greater percentage of that 70 percent, which 16 may be up to 80, 90 percent of that 70 percent, 17 will have that number of minor effects. Those are 18 not complete attacks where they fall down. In 19 fact, with most narcoleptic patients, they 20 distinguish between the two and they will often 21 only report to the physician the major events. But 22 in the diaries that Orphan had set up all the 23 events are characterized. 24 DR. WOLINSKY: So, the second question -- 25 DR. BLACK: We have no idea. That is an 109 1 excellent question that I think needs to be 2 determined, but in the studies that have been 3 completed that question cannot be answered. 4 DR. REARDAN: Jed, the only study I can 5 think of maybe is SXB-20 where cataplexy was not an 6 entry criterion and I don't know what the cataplexy 7 incidence in that trial was. Bill is shaking his 8 head -- we didn't record it and we didn't 9 quantitate it. 10 DR. BLACK: We can't comment on that. 11 DR. REARDAN: It is true that in most of 12 our studies patients were selected because at entry 13 criteria they had to have a baseline cataplexy. 14 DR. KAWAS: Dr. Penix? 15 DR. PENIX: Before we address the two 16 separate indications issue -- and I guess, Dr. 17 Black, I could direct this question to you -- in 18 the GHB-2 study you did look at all cataplexy 19 events, I guess, and then total and partial 20 cataplexy. In the background material, in the 21 separation of the two it appeared that there was no 22 significant difference in any of the three doses of 23 GHB on total or complete cataplexy but your effect 24 was primarily in partial cataplexy. Is that 25 correct? 110 1 [No verbal response] 2 So, my question in that regard is what is 3 the clinical significance of partial cataplexy, and 4 you mentioned that patients frequently do not want 5 treatment for partial cataplexy. So, is this a big 6 problem? I presume that the patients that would 7 perceive a problem would be the ones with the 8 complete cataplexy but there we see no significant 9 difference. So, is there a problem there with 10 that? 11 DR. BLACK: I think this is a good point, 12 and the difficulty comes in trying to separate the 13 two because it is not sort of a box of partial and 14 a box of complete; it is a gradation, you know, 15 ranging from small partials to large partials and 16 the completes. So, I think this analysis is 17 difficult to perform. Clinically the degree of 18 improvement with traditional anticataplectic 19 medications that we use is similar. So, the 20 reduction in partial -- if that is all that is 21 being seen here and I am not convinced that 22 clinically that is the case -- while the 23 statistical analysis didn't demonstrate a 24 significant difference in the complete cataplexy 25 attacks, clinically there is an improvement in all 111 1 the different categories, and it is very 2 substantial in traditional anticataplectic 3 medications as well as with GHB. 4 DR. PENIX: Could Dr. Mignot comment on 5 the clinical significance of partial cataplexy? Is 6 it a big problem? 7 DR. MIGNOT: Yes, it is a big problem. In 8 fact, the problem is especially the social aspect 9 of cataplexy, when you have to realize that you are 10 just in the middle of a crowd and are meeting some 11 friends, and you can never tell when it is going to 12 happen. It may happen in very odd circumstances. 13 So, often even the doctors don't know what it is 14 and they just look at it and they wonder why this 15 person is kind of losing slight control and has to 16 sit down. There is also almost a social aspect 17 with fear of cataplexy that can occur at any time, 18 any moment and, yes, it is a very significant 19 problem. 20 Again, it is a balancing act because the 21 drugs that we use are somewhat effective but they 22 have all these side effects and you just have to 23 choose between two evils. I am pretty sure that, 24 for example, GHB, based on my relatively limited 25 experience, has less side effects than 112 1 anticataplectic classical tricyclic 2 antidepressants, and that a lot of patients would 3 prefer to take GHB even for partial cataplexy. 4 DR. PENIX: The case that you showed of 5 the nine-year child I assume is complete cataplexy 6 -- 7 DR. MIGNOT: Yes. 8 DR. PENIX: -- but you are also saying 9 that patients with partial cataplexy have a 10 significant impairment of their life. 11 DR. MIGNOT: Absolutely. But, as Dr. 12 Black mentioned, it is not an "all or none." I 13 mean, most patients, the ones that are complete, 14 have a lot of partial cataplexy. You never know 15 how bad it is going to be. Most of them are small, 16 little attacks, and sometimes they may even be 17 perceived only by the patient. Sometimes the face 18 may melt; the head drops. Sometimes they just have 19 to sit down; sometimes they don't have to sit down. 20 I showed a young kid because it is more dramatic, 21 but you would see the same thing in some of the 22 patients with partial cataplexy occasionally. 23 DR. BLACK: I am realizing that a 24 definition may be useful here. In general when we 25 were describing patients who documented the partial 113 1 versus complete, we told them to think about 2 complete as an episode where they fall to the 3 ground with complete paralysis or where, if they 4 weren't sitting, they would have fallen to the 5 ground with complete paralysis. Otherwise, 6 anything else is partial -- so, slurred speech, 7 head drops, dropping things are the partials, and 8 those become very important for quality of life and 9 daytime performance. Driving, those kinds of 10 things can become a very significant event for 11 partials. 12 DR. MIGNOT: Yes, one thing I should also 13 emphasize is that in a very large number of series 14 that, for example, have analyzed several hundred 15 patients with narcolepsy and cataplexy, as a mean 16 the large majority of patients have several attacks 17 per day, several attacks per week. Between several 18 attacks per day and several attacks per week, that 19 is generally partial or complete attacks and it is 20 not something that appears just once, you know, 21 every ten years. It is really something that 22 occurs regularly and sometimes totally 23 unexpectedly. 24 DR. KAWAS: Dr. Falkowski? 25 DR. FALKOWSKI: That leads me to a 114 1 question just for clarification. For the purposes 2 of these clinical trials, were the cataplectic 3 events something that was just perceived by the 4 patient and recorded in a diary, or were they 5 verified by some third party? 6 DR. REARDAN: These were taken from 7 patient diaries. So, it is patient recorded 8 episodes. 9 DR. HAGAMAN: I am Dr. Hagaman and I just 10 wanted to address the partial versus the complete 11 cataplectic events. I think that you have to take 12 it on an individual basis. We have patients that 13 come in that are teenagers that have tests in front 14 of them and they have a partial cataplectic event 15 and they drop their pencil; people that cut hair 16 that have scissors in their hands and they drop 17 their scissors. So, even though they have not had 18 a complete event, this has been a very debilitating 19 event in their lives. So, it is a continuum and I 20 think you just have to really look at each person 21 as an individual and what they are doing. 22 DR. KAWAS: Dr. Dyer? 23 DR. DYER: How variable in the same 24 patients are the number of cataplectic attacks per 25 week? What is the variance in that? 115 1 DR. MIGNOT: We have looked at that quite 2 a bit. 3 Actually, I did some diaries in a large number of 4 patients with cataplexy. It is really totally 5 unpredictable and that is one of the most scary 6 parts about cataplexy when you have narcolepsy. Of 7 course, if something emotional is going to happen, 8 say a patient is going to go to a wedding, often 9 they will kind of fear that event much more because 10 they think it is very likely that they are going to 11 have cataplexy in front of everyone and, indeed, 12 they may actually have a lot more cataplexy because 13 it is an emotional event. 14 Still, I have followed, for example, 15 patients and sometimes they may have like 80 for 16 one week and then the following week they may have 17 only three or four. I mean, it can really vary 18 quite a bit. And, one of the main reasons is 19 really that emotion is something that is very 20 variable. In fact, someone mentioned how easy it 21 is to observe cataplexy. It is very difficult to 22 get it on tape because typically the patient come 23 to your office; he really wants to show you what it 24 is but, you know, he is tense and it just will not 25 occur but as soon as he leaves the office and 116 1 something happens -- boom, he is going to collapse. 2 So, it is very difficult to predict and it is quite 3 variable. 4 DR. ROMAN: For Dr. Mignot also, you 5 mentioned that cataplexy probably is the result of 6 what you called dissociated REM. However, if I 7 recall correctly, the polysomnographic analysis has 8 shown that Xyrem actually decreases the amount of 9 REM sleep and increases delta sleep. Would you 10 like to speculate on what could be the mechanism of 11 action to improve the cataleptic component? 12 DR. MIGNOT: That is a very, very 13 difficult question. One of the difficult 14 questions, of course, is the mode of action of GHB. 15 I have looked into it myself for quite a while 16 because I was trained as a pharmacologist, and it 17 is not clear. There are two camps. Some people 18 think it acts on GHB receptors, specific receptors; 19 others think that it acts through the GABA-B 20 receptors. We know that it has some strong effect 21 on dopamine transmission. If you inject GHB in 22 animals the rate of activity of dopaminergic cells 23 shuts down and dopamine can increase in the brain 24 proportionally to the dose. We have done quite a 25 bit of studies that have shown that the 117 1 dopaminergic system is very important to regulate 2 both wakefulness and also cataplexy and the 3 regulation of emotion. I believe it is by changing 4 the balance of the dopaminergic system, that 5 improves cataplexy the following day maybe by 6 increasing dopamine in the brain during the night, 7 but this is highly speculative and a lot more 8 research needs to be done. 9 The fact that it does not increase REM -- 10 first, it is quite variable because some studies 11 have shown that it does increase REM and this 12 contrasts dramatically with what all hypnotics do. 13 If you take MVN or all the other 14 benzodiazepine-like hypnotics, what they do is 15 actually, rather, reduce slow wave sleep and reduce 16 REM sleep. Xyrem doesn't do that. It actually 17 promotes slow wave sleep and, if anything, would 18 promote REM sleep or doesn't change it. That is 19 still, you know, much more in the right direction 20 of promoting normal sleep, including REM sleep. 21 The last comment I want to mention is that 22 it is not sufficient -- if you know a lot about 23 narcolepsy, it is not sufficient to just explain 24 narcolepsy as a disorder of REM sleep. Indeed, 25 they have all this transition to REM sleep but they 118 1 also have impaired wakefulness per se. For 2 example, if you do MSLTs they don't always go into 3 REM. They will often just fall asleep into normal 4 sleep. So, it is not only REM sleep that is 5 disregulated in narcolepsy, it is also wakefulness 6 and by improving slow wave sleep you presumably 7 also can improve the wake aspect of narcolepsy. My 8 answer may be a little complicated but I would be 9 happy to discuss it in more detail. 10 DR. KAWAS: Dr. Van Belle? 11 DR. BLACK: Just another comment on that, 12 the Broughton study showed an increase in REM at a 13 lower dose. The first dose of the SXB-20 that I 14 participated in showed at 4.5 g the first night an 15 increase in REM, which was then followed by a 16 dose-related decrease in REM over time, which is 17 very different from REM suppressant agents where 18 there is a robust, or in fact the largest effect 19 that can often be seen on the first night of 20 administration. 21 So, we don't know exactly why it is that 22 over time the REM with higher doses is reduced, and 23 why with the first dose, and with the lower doses, 24 as has been demonstrated here with Roger 25 Broughton's work, why the REM is increased. There 119 1 has been established sort of a competitive reaction 2 between slow wave sleep and REM sleep. It appears 3 that there may be factors that regulate slow wave 4 sleep that also are important in regulating the 5 appearance, or lack thereof, of REM sleep. It may 6 be that gama hydroxybutyrate is sort of normalizing 7 slow wave activity which then results in a more 8 normal control or regulation of the REM or 9 REM-related events. 10 DR. KAWAS: Can I ask for my 11 clarification, what dose the company is proposing? 12 DR. REARDAN: Bill, can you take that 13 question? 14 DR. HOUGHTON: Yes, the dosage regimen 15 that we are proposing is that patients be started 16 at 4.5 g and then titrated between the range of 3-9 17 g to clinical efficacy. Although in the strictest 18 mathematical sense the only statistical efficacy in 19 the GHB-2 study was clearly defined at 9 g, that 20 may well represent that the study was too short 21 because in the open-label study that followed, as I 22 showed, the maximum nadir occurred at 8 weeks, and 23 when those patients were followed over the course 24 of 12 months they maintained efficacy across the 25 dose range. Certainly, there is an advantage in 120 1 terms of the important side effects to dose 2 titration. In all of the treatment IND protocols 3 and the safety studies the data was generated at 4 between 3-9 g. Now, 80 percent of the patients 5 were maintained between 6 g and 9 g, but there was 6 certainly facility for down-titration from the 4.5 7 or maintenance there as well. 8 DR. KAWAS: Thank you. Dr. Van Belle? 9 DR. VAN BELLE: It seems to me that there 10 is reasonable agreement with respect to efficacy 11 for cataplexy at least between the FDA and the 12 sponsor. So, I would like to get back to the 13 secondary endpoints. I would like to ask a 14 question to the sponsor's statistician, Dr. Trout, 15 as to whether he thinks that multiple comparisons 16 is a problem. Secondly, if multiple comparisons 17 are a problem, how he would adjust. 18 DR. REARDAN: Do you want to put this in 19 relation to a specific trial or all the trials in 20 general? 21 DR. VAN BELLE: Well, I bring it up in 22 connection with the analysis of Dr. Mani where he 23 clearly comes to conclusions that differ from yours 24 with respect to the efficacy of some of these 25 secondary endpoints. 121 1 DR. TROUT: You know, it is hard to answer 2 that question. I think the way I would answer that 3 is as follows: The GHB-2 analysis, the results 4 that we found and also that were expressed earlier 5 were very strong. So, even with the fact that 6 there is some multiplicity, we also have, remember, 7 some other outcome measures which were related to 8 this particular general area in terms of daytime 9 sleep attacks. So, there were at least two 10 measures that suggested improvement with respect to 11 that particular outcome. 12 The other second study that has been 13 discussed is the Lammers study, and that study is 14 obviously much smaller. It is obviously a weaker 15 study, and there is some issue with regard to 16 whether the appropriate method of analysis was 17 there. So, I think that is a harder one to 18 address. 19 Now, there are two kinds of multiplicity 20 going on here, which you are well aware of. One is 21 the multiplicity with regard to the multiple dosing 22 levels and that was accounted for in our analyses. 23 The question that was brought up by Dr. Mani with 24 regard to the multiplicity of secondary endpoints, 25 and I am not a betting man but I think there is 122 1 certainly evidence to suggest that daytime 2 sleepiness is being affected possibly. But I don't 3 go to Las Vegas nor Atlantic City. 4 DR. KAWAS: Actually, while we have Dr. 5 Trout up, I would ask him with regard to excessive 6 sleepiness on the Epworth Scale in the GHB-2 study, 7 while there certainly was a difference in the two 8 groups, there were also major baseline differences 9 in sleepiness for the responders and the 10 non-responders. In fact, those that appeared to 11 respond had a baseline that was better than the 12 improvement in the other group. There was a 13 significant difference. Are you concerned about 14 these and how these might affect the results? 15 DR. TROUT: There is always concern about 16 baseline differences, and that was attempted to be 17 accounted for in two mechanisms, one, we looked at 18 change from baseline and we also did a covariate 19 adjustment to try to account for that. 20 DR. KAWAS: Dr. Katz? 21 DR. KATZ: I would like to ask Dr. Trout a 22 question also. Dr. Yan mentioned that we didn't 23 believe that the data were normally distributed, 24 and when you transformed the data it didn't really 25 help very much. I don't want to get bogged down in 123 1 a hyper-arcane discussion about normally 2 distributed data, but when we did that we got a p 3 value for that comparison -- I guess it was the 4 Epworth, of about 0.01 -- 5 DR. MANI: I am sorry, it wasn't the 6 Epworth. You are talking about the Lammers study 7 where you are talking about the frequency -- 8 DR. KATZ: I thought we were talking about 9 GHB-2. 10 DR. MANI: Oh, sorry, fine. 11 DR. KATZ: So, if we are right, it takes 12 the p value which was 0.0001 or something like that 13 to 0.01, and then when you get to the multiple 14 comparisons issue it makes it less weak. I agree if 15 you take a p value of 0.001 or 0.0001, no matter 16 what you do to it as far as a multiple comparison, 17 it is still going to be significant. But if it is 18 0.01 it is a little different story. So, I am just 19 wondering, again without getting into excruciating 20 details, what about this question of the data being 21 normally distributed and not necessarily being 22 improved very much by transforming it? Is there 23 common agreement about that or not? 24 DR. TROUT: My recollection, and it has 25 been sometime since I have seen the results of the 124 1 analysis, is that it suggested that we didn't see a 2 particular problem with the normal distribution as, 3 for example, was the case with cataplexy which was 4 clear. I am not sure if Dr. Yan did a 5 nonparametric covariance analysis or not. I 6 haven't seen those analyses. And, I think the 7 point was made earlier that that would be, I think, 8 an appropriate thing to do in order to account for 9 some potential baseline differences. If she did, 10 then whether it is a reflection of a decreased 11 sensitivity of a nonparametric analysis or whether 12 it is a normal distribution -- I can't answer that 13 without seeing the data. Maybe it was just a 14 standard, nonparametric analysis which might help 15 account for the difference. 16 [Comment away from microphone; inaudible] 17 DR. TROUT: No, I know that but Dr. Yan 18 did a nonparametric analysis because she was 19 concerned about the normality, and did look at the 20 log transformation and it didn't have any impact on 21 that, which doesn't surprise me at all. 22 DR. KAWAS: I would like to ask the 23 sponsor, I mean, there clearly was a dose 24 relationship in terms of the adverse events. Were 25 any other factors looked at that may be related to 125 1 the adverse event profile, things like age, even 2 previous psychiatric history, other medications? 3 Whether or not they drank alcohol? Anything? 4 DR. HOUGHTON: No, we didn't go as far as 5 an alcohol history. Certainly for the major 6 psychiatric, a preexisting history of major 7 psychiatric disease emerged. Major psychiatric 8 disease was actually a protocol exclusionary 9 criterion, but in those that, for instance 10 attempted suicide, post-study it was discovered 11 that they had a previous psychiatric history and in 12 actual fact in one of the patients a previous 13 suicide attempt had been made. There was major 14 depressive disease reported in those, but for those 15 who developed psychosis there was definite recorded 16 preexisting psychiatric history. 17 In terms of age, we haven't done a 18 breakdown of the database, and in most instances 19 there was not a dose relationship. There were just 20 instances that were mentioned in the presentation. 21 Confusion and sleepwalking suggested a dose 22 relationship. In the GHB-2 protocol which was 23 obviously blinded, there was the association with 24 nausea, vomiting, confusion and enuresis that was 25 definite, but that didn't extend across the whole 126 1 study database. So, the relationship with dose is 2 not well defined. 3 DR. KAWAS: But how about relationship 4 with anything else? For example, were the patients 5 who had confusion more likely to be the elder 6 patients? You might be able to tell I am in aging. 7 DR. HOUGHTON: I can identify well. Do we 8 have a breakdown of confusion by age? A range 9 would be still useful. 10 [Slide] 11 Here is a slide that shows that the 12 distribution of age was between 25 and 73 years, 13 with 67 percent over 50 years of age, but the range 14 is still wide. There is the distribution across 15 doses. Four events at 3 g, 10 at 4.5, 12 at 6 g, 8 16 events at 7.5, and 13 events at 9 g. 17 DR. KAWAS: Thank you. Do we have any 18 other questions from the committee? If not, we 19 will move on. Dr. Katz? 20 DR. KATZ: A quick question, if I heard 21 you correctly, there were 14 events reported as 22 convulsions, but when you went back and looked at 23 that, 13 of them were actually cataplexy. So, 24 presumably cataplexy was a verbatim term. How is 25 it that cataplexy got coded as convulsions? 127 1 DR. REARDAN: The COSTART dictionary puts 2 cataplexy in as a convulsion. It is a definition. 3 Convulsion has ten different terminologies, 4 verbatim events, and they all code up to 5 convulsion. 6 DR. WOLINSKY: Along those lines, how come 7 there were only that few number of convulsions when 8 we were studying cataplexy in the trial? I mean, I 9 don't know that it is easy to explain this in both 10 sides of one's mouth. 11 DR. HOUGHTON: No, and we are not trying 12 to. If there was a cataplexy event that occurred 13 of a severity to be seen as unusual for that 14 patient, and the patient volunteered it as an 15 event, then it was recorded as an adverse event. 16 Or, there may have been injury related to the 17 cataplexy events. We do have representation in the 18 database. I can recall absolutely a fractured 19 ankle in the washout study. So, there were 20 traumatic events associated with a major cataplexy 21 event that would have been of sufficient impression 22 on the patient to report as a separate event. 23 DR. WOLINSKY: But then the event would 24 not have been withdrawal from the primary measure 25 of efficacy even though it was also registered as 128 1 an adverse event? 2 DR. HOUGHTON: I am sorry? 3 DR. WOLINSKY: Was it still counted as an 4 event in the measure of efficacy if it was also 5 shifted to be counted as an adverse event? 6 DR. REARDAN: Yes, the patient diaries 7 recorded cataplexy. If they record cataplexy as an 8 event itself, that was part of the efficacy 9 outcome. It wasn't necessarily an adverse event. 10 If they had an adverse event -- fall and break an 11 ankle, cataplexy is coded as part of that adverse 12 event. It is the cause of the adverse event and so 13 it shows up in the database. 14 DR. KAWAS: Dr. Simpson? 15 DR. SIMPSON: I have two questions. One 16 really was just a clarification of this business 17 about the sleepiness. I think we have all agreed 18 that there has to be some adjustment for multiple 19 comparisons on the sleepiness index, and the GHB-2 20 study, even if you make an adjustment, there are 21 certainly some of the indices about sleepiness 22 which seem to be significant. But coming back to 23 the Lammers study, have we established whether or 24 not, once we have made an adjustment, we have any 25 significance there or not? Because that is the 129 1 pivotal trial, isn't it, because we need two? 2 DR. REARDAN: Remember that the Lammers 3 study was a very small trial, 24 patients. Daytime 4 sleepiness was a secondary endpoint in that study, 5 and I forget the p value. Maybe Dr. Yan or Dr. 6 Katz could comment. I don't think any formal study 7 of multiple analysis was done, except maybe by Dr. 8 Yan -- 9 DR. YAN: No. 10 DR. REARDAN: -- and I think she needs to 11 comment on that. 12 DR. YAN: For Lammers study there was no 13 prespecified analysis, except the Wilcoxon assigned 14 rank test. It was across the study and we 15 considered it not very appropriate, and for a 16 secondary analysis none of the statistical analyses 17 were specified. The problem with this Lammers 18 study is that if you use different statistical 19 analyses which are considered appropriate, you get 20 a very different result. Some could be less than 21 0.05 and some ranged to something like 0.2. So, 22 the results are not consistent and we don't have a 23 reliable method to see which one we could consider. 24 DR. REARDAN: We don't disagree with that. 25 I mean, the problem with Lammers is that it was a 130 1 one-sentence statement about how he was going to 2 analyze it, and it was an inappropriate statistical 3 analysis for a crossover study. So, that creates 4 issues about not having a prospective statistical 5 plan appropriate for the study. But even in that 6 initial Wilcoxon analysis the daytime sleepiness 7 was statistically significant. It was not 8 corrected for multiple analyses. 9 DR. KAWAS: Dr. Simpson? 10 DR. SIMPSON: I just have another question 11 that I wondered if you could clarify. In a lot of 12 these studies you talk about an intent-to-treat 13 analysis, but when I read it I wasn't clear whether 14 or not that meant the patients that were randomized 15 were actually included always in the analysis or 16 not. 17 DR. REARDAN: Yes, the intent-to-treat 18 would include every patient who received drug. Is 19 that correct? 20 DR. TROUT: Yes, every patient who 21 received at least one dose. 22 DR. SIMPSON: So, how did you then deal 23 with the patients who dropped out? 24 DR. TROUT: In the GHB-2 analysis we 25 selected an endpoint. So, in order for the patient 131 1 to be included in that analysis there had to be at 2 least one post-baseline measure of cataplexy or 3 sleepiness, or whichever outcome you want. So, it 4 was an endpoint analysis that was done in order to 5 accommodate that. 6 DR. KAWAS: It looks like we are 7 completely behind schedule and we will have a very 8 late lunch, I will warn everyone. The FDA's 9 invited speakers on risk management issues is the 10 next component of this discussion. The first 11 speaker is going to be Dr. Carol Falkowski, of the 12 Hazelden Foundation, in Minnesota, who will be 13 speaking on the epidemiology of GHB abuse issues. 14 FDA Invited Speakers on Risk Management Issues 15 Epidemiology of GHB Abuse Issues 16 DR. FALKOWSKI: Hello. Good morning, 17 almost afternoon. 18 [Slide] 19 This is the title of my talk, GHB Abuse in 20 the United States. I am Director of Research 21 Communications at the Hazelden Foundation. I have 22 been a member of the National Institute on Drug 23 Abuse's Community Epidemiology Work Group since 24 1986. I am author of a book, called, "Dangerous 25 Drugs: An Easy-to-Use Reference for Parents and 132 1 Professionals." What is missing from this overhead 2 is that I served on the Drug Abuse Advisory 3 Committee for the FDA from 1995 through 1999. 4 [Slide] 5 In the very short time that I have, I am 6 going to try and just hit the big points about what 7 we know about the abuse of GHB in the United 8 States, starting off with measuring drug abuse. 9 There are a number of things that are thought to 10 bear when we talk about measuring something as 11 complex and multi-dimensional as drug abuse. This 12 includes population surveys. It includes hospital 13 emergency room episodes; medical examiner data; 14 addiction treatment data; law enforcement data, as 15 well as ethnographic studies that look at specific 16 populations of users that are more anthropological 17 and ethnographic in nature. 18 [Slide] 19 I also want to make the point that all 20 data systems have limitations, and this is 21 particularly true in the case of new drugs of 22 abuse. For example, if we are talking about GHB 23 and trying to measure the number of patients who 24 have presented to addiction treatment centers 25 across the country with GHB as their primary drug 133 1 of abuse, it is now the case that it is often 2 grouped in a category of drugs called sedative 3 hypnotics. It is not its own line item. So, in 4 preparation for a meeting like this it is very hard 5 to get an accurate count of the extent to which GHB 6 itself is the presenting drug of abuse. 7 Similarly, surveys that are conducted -- 8 we have not added GHB to the National Household 9 Survey or the Monitoring the Future Survey, 10 although to the Monitoring the Future Survey that 11 looks at drug use among 8th, 10th and 12th graders 12 ecstasy, another club drug, has been added. 13 Also, in terms of law enforcement 14 indicators, there is no field test for GHB so it is 15 hard to also get that indication of it as well. 16 In addition, new methods of abuse are hard 17 to track. I recall, in 1986, when we started at 18 the national level wanting to track crack cocaine, 19 we knew about how to track cocaine but, all of a 20 sudden, we were looking at it by a different route 21 of administration. So, it was a challenge to all 22 of us to start switching our data systems just to 23 measure crack instead of cocaine, to make that 24 distinction. 25 Existing data systems are slow to respond, 134 1 and there is a system-wide learning curve when a 2 new drug of abuse appears on the scene. That means 3 it is a learning curve in terms of emergency room 4 personnel, treatment providers, law enforcement, as 5 well as prevention agencies, and that is why we 6 rely on a lot of the scientific literature put out, 7 particularly in emergency medicine, to inform the 8 field about emerging drugs of abuse and how people 9 present with those problems. 10 [Slide] 11 My background in this has been as part of 12 the Community Epidemiology Work Group. This is a 13 group of drug abuse researchers from twenty cities 14 in the country that has been convened by the 15 National Institute on Drug Abuse since 1976. This 16 model of drug abuse epidemiology has also been 17 adapted in different parts of the world. There is 18 a similar group in Europe, in Canada, Mexico and 19 Asian cities. 20 [Slide] 21 The Community Epidemiology Work Group is 22 an early warning epidemiological surveillance 23 network that detects new drugs of abuse, patterns 24 of use and populations at risk. 25 [Slide] 135 1 It involves researchers looking at the 2 same data from different geographic areas and in 3 this case, as I mentioned, there are people like me 4 in twenty cities in the country who write 5 quantitative reports on drug abuse twice annually, 6 and we are convened by the National Institute on 7 Drug Abuse twice a year. 8 [Slide] 9 Having done this and written over twenty 10 reports on drug abuse trends in my city and met 11 with my colleagues, it has given me a sort of 12 broad-based perspective on how emerging drugs are 13 measured and how we get a handle on them. But 14 everyone looks at medical examiner data. We look 15 at the data from the Drug Abuse Warning Network, 16 which is data from a representative sample of nine 17 federal short-stay hospitals with 24-hour emergency 18 rooms, and that is conducted in 21 cities, as well 19 as some other areas of the country. 20 We also look at treatment data, law 21 enforcement data and price, purity, trafficking and 22 the sale of drugs, as well as supplemental research 23 data and information from multiple sources. 24 [Slide] 25 I want to start my introduction to GHB by 136 1 telling you about the abuse of a group of drugs 2 that are called club drugs. That is really the 3 first time in a long time we have had a name like 4 club drugs applied to drugs because they are used 5 in a particular setting. That is why they came to 6 be called club drugs. It is a mixed category of 7 drugs. It includes stimulant drugs as well as 8 depressant drugs that are used in nightclub 9 settings. GHB is also known in these settings as 10 liquid X, gamma, G, easy lay, Georgia Home Boy or 11 great hormones at bedtime. MDMA or 3,4 methylene 12 dioxide methamphetamine is ecstasy, e or x. 13 Ketamine is known as special K. It is a veterinary 14 anesthetic, a dissociative drug similar in effects 15 to PCP. Flunitrazepam, Rohypnol is a long-acting 16 benzodiazepine, which was dubbed the original date 17 rape drug which is a drug not approved for medical 18 use in this country; methamphetamine and LSD. 19 If there is one point to make about club 20 drugs as a term, one thing that has emerged is the 21 fact that clearly these drugs are not limited to 22 club settings and I will be talking to that in a 23 moment. It is not just clubs where they are used. 24 [Slide] 25 To give you a little slice of the 137 1 progression of GHB and how it came on the CEWG 2 radar screen, it was first mentioned in 1990 3 through a poison information center from my 4 colleague in Miami. Then, from 1990 to 1994 it 5 appeared in the Miami and the New York city 6 reports. In 1996 it appeared in 6 other cities, 7 and by the year 2000 most cities in this 21-city 8 work group were reporting GHB. It reports 23 9 deaths in the 20 CEWG cities, and I refer you to a 10 handout that I prepared that sort of gives the 11 chronology of how my colleagues describe the 12 growing abuse of GHB in their cities. 13 [Slide] 14 Now, in terms of user typologies, they 15 tend to be young adolescents through adulthood. 16 There is really no age group but when we look at 17 population surveys in this country of who are drug 18 abusers, by and large the biggest bulk of drug 19 abusers are people who are under the age of 35. 20 The motive for use is multiple. It 21 includes not only intoxication, but also people 22 seeking intoxication effects in the absence of 23 alcohol. I have had people describe it to me as it 24 gives them the effects of alcohol without having to 25 waste that time drinking alcohol. This is by young 138 1 people who haven't developed the taste. 2 It is also used by weight lifters and body 3 builders for its alleged anabolic effects. It is 4 also marketed in nutritional supplements to promote 5 better sex, better sleep and some people take it to 6 counter the effects of other club drugs. One of 7 the characteristics of drug abuse in nightclubs 8 that has come up over the past year is the fact 9 that people seem to have the impression that if you 10 take just a little bit of this and a little bit of 11 that nothing can really hurt you in a club setting. 12 So, you might take a little bit of ecstasy to get 13 you going, with a little bit of cocaine to keep you 14 there, and maybe a little bit of heroin to take the 15 edge off. This sort of mixing and matching is also 16 part of the user typology. 17 The settings it is used in are nightclubs, 18 raves, parties, but also in homes, in health clubs, 19 gyms and other settings. The sources of it come 20 from health food stores, mail order kits, the 21 Internet or at these clubs where it is being used 22 by the capful. Sometimes at these clubs, because 23 ecstasy dehydrates you, people have a lot of water 24 bottles and it is not unusual to have a water 25 bottle that may have GHB mixed in it, and for ten 139 1 bucks someone can get a swig of it. This makes it 2 very imprecise dosing, as you can imagine. 3 [Slide] 4 In terms of deaths, in terms of the 5 consequences of use -- there is a huge bullet 6 missing from this slide, which I will get to. So, 7 if everybody wants to find their slides and write a 8 bullet in it, I would appreciate it. Deaths -- 9 there have been 71 documented deaths, according to 10 the Drug Enforcement Administration, through 11 November of last year. Again, the problem is that 12 because it is a new drug of abuse people don't 13 know. You know, you have to know what you are 14 looking for to be able to find something and this 15 has clearly been the case in trying to document GHB 16 deaths. It is a huge issue and I hope we get 17 enlightened on that this afternoon. 18 Also, there have been adverse medical 19 reactions, not only people who come into emergency 20 rooms, but the countless people, which is quite 21 hard to quantify, who have episodes but never get 22 emergency room treatment for it. But there have 23 been medical reactions, adverse ones. 24 Dependence -- there has been a reported 25 increase in people presenting to addiction 140 1 treatment centers with GHB as their primary 2 substance of abuse, and an increase in the reported 3 addiction to GHB by those who may not make it to 4 treatment programs. 5 I work at the Hazelden Foundation. We are 6 based in Center City, Minnesota, with campuses in 7 Chicago, New York City and West Palm Beach. There 8 were 5 patients in 1999 who had a history of GHB 9 abuse, and that had grown to 39 in the year 2000 10 and we are just one treatment center. 11 Finally, the missing bullet on here is 12 drug rape. One thing we have seen in this country 13 since the early 1990's is the use of drugs, this 14 predatory use of drugs where you administer drugs 15 to people without their knowledge for the purpose 16 of disabling them to commit crime on them. The 17 first drug that came to this sort of notoriety was 18 Rohypnol, but now we are in a situation where GHB 19 is often used in drug-induced rape. In fact, 20 several years ago when President Clinton signed the 21 federal date-rape law, the Samantha Reid and Hilary 22 Farris Date Rape Act, that was in response to two 23 cases of drug rape that were not related to 24 Rohypnol but to GHB. So, that bullet should be up 25 there, drug rape. 141 1 Also, another bullet would include the 2 trafficking, sale and manufacture, the law 3 enforcement consequences. 4 [Slide] 5 Let's look at hospital emergency room 6 episodes of GHB. This looks at them from 1994 7 through 1999. You can see the increase in hospital 8 emergency department mentions of GHB. Mentions is 9 sort of unusual term for people who aren't familiar 10 with the Drug Abuse Warning Network, and it quite 11 literally means, in a retrospective review of 12 patient records, that they find a mention of GHB. 13 Sometimes it is the sole drug that precipitated the 14 medical emergency and sometimes it is used in 15 combination with other drugs. For every drug abuse 16 episode in the Drug Abuse Warning Network there can 17 be the mention of 4 drugs and alcohol, but when 18 alcohol is used in combination with other drugs; it 19 is not an alcohol tracking system. 20 [Slide] 21 So, this is what it looks like through 22 1999. This looks at it by half year increments. 23 You can see this takes us into the year 2000 and we 24 have the first half of the year 2000. 25 I want to go back to just my opening 142 1 remarks about club drug abuse. I think in the 2 general population when we think of club drugs, you 3 know, what we hear about, what everybody is talking 4 about, what seems to be in U.S. News and World 5 Report, in Newsweek and Time Magazine is ecstasy. 6 [Slide] 7 This is from exactly one year ago. This 8 is Time Magazine from June 5, 2000. It talks about 9 ecstasy. For many folks, club drugs -- you think 10 ecstasy. 11 [Slide] 12 This was, I believe, from Time magazine as 13 well. You see the water bottle there. If you 14 didn't see Time magazine, you may have seen The New 15 York Times Sunday magazine insert. This is from 16 January of this year, talking again about ecstasy. 17 This is from January 2001. 18 So, since it is in the same category of 19 drug, I think it is relevant to look at how GHB 20 emergency room episodes compare with those of 21 ecstasy. 22 [Slide] 23 Ecstasy, or MDMA, is in the pink and GHB 24 is in blue. You can see in the first half of the 25 year 2000 that GHB hospital emergency episodes have 143 1 surpassed those of ecstasy. 2 [Slide] 3 Efforts to control GHB -- a number of 4 states have done things to try to control GHB abuse 5 in their states. This is sort of a listing of the 6 scheduling of it in various different states. It 7 was added, as you know from the materials the 8 committee received, to the Federal Control 9 Substance Act. 10 [Slide] 11 Finally in conclusion, GHB is a 12 significant, growing drug of abuse. We have seen 13 rapid growth in the adverse medical consequences 14 related to GHB since 1999 and, in fact, hospital 15 emergency mentions of GHB now surpass those of 16 ecstasy or MDMA. We have seen rapid growth in 17 adverse medical reactions despite not only federal 18 scheduling but the scheduling in numerous states. 19 We have multiple user typologies. This is not a 20 substance that is sought after simply by people at 21 parties and raves. These products that contain GHB 22 as well as its precursor drugs, GBL and 1,4-BD, are 23 sought after by people who believe the claims on 24 these nutritional supplements and take them for 25 promoting muscle growth, for sleep; and take them 144 1 for better sex, as well, and as I said, use it in 2 sort of predatory way. Dependence is clearly 3 possible. 4 So in closing, here we have a drug with an 5 established widespread abuse record. With GHB we 6 needn't talk about abuse potential. With GHB we 7 have abuse reality. We have a decade of GHB abuse 8 in this country; a decade of deaths and hospital 9 emergency room episodes and dependence. We have 10 escalating abuse of GHB in spite of recent efforts 11 to control it and, yes, people acquire this drug 12 and its precursors in many ways. But make no 13 mistake, the effects being sought are the GHB 14 effects. The chemical agent in the body that is 15 producing these effects is GHB, and this 16 undisputable fact is entirely relevant to our 17 discussions today. 18 I have to take issue with the statement 19 from the sponsor that says Xyrem is not the 20 problem. If Xyrem equals GHB, then I believe it is 21 a problem. This drug, if approved, will exist 22 outside the confines of this room. Patients will 23 use it outside the confines of clinical trials. In 24 America, in 2001 we have a serious, significant and 25 growing problem with GHB abuse in this country, and 145 1 it just so happens that this coincides with Orphan 2 Medical seeking approval for this drug. 3 This drug already has avid followers, and 4 there is no reason to assume that another source of 5 GHB would not be sought after by these folks, and I 6 think we need to bear that in mind throughout our 7 discussions. Thank you. 8 DR. KAWAS: Dr. Falkowski, can I ask you 9 one question? With regards to the emergency 10 department data for GHB, I recognize the 11 difficulties of all of this kind of data but, for 12 example, MDMA is not infrequently the only drug and 13 when they go to the emergency room that is clearly 14 because of the MDMA. Can you give us any kind of 15 quantification or semi-quantification? You 16 mentioned that sometimes GHB is the only drug. 17 DR. FALKOWSKI: The question was how often 18 is GHB used in combination, and let me find that. 19 DR. KAWAS: For the emergency room data. 20 DR. FALKOWSKI: Yes, that is what I am 21 looking for. I have it right here. It is 70 22 percent of the time. Like many other drugs, GHB 23 episodes involve drugs other than GHB as well. 24 I would also like to add that I believe 25 these hospital emergency room episodes 146 1 underestimate GHB because drugs that are used in a 2 predatory way, that are administered to people 3 without their knowledge are not DAWN reportable. 4 So, if someone comes to the emergency room and says 5 I believe somebody gave me something and it is 6 making me sick, that is not a DAWN reportable 7 thing. That is being addressed by the Substance 8 Abuse and Mental Health Services Administration. 9 But what that means is that people who are drugged 10 with any sort of drug are not picked up by this 11 particular reporting system. 12 DR. KAWAS: And, what are the most common 13 drugs or classes of drugs that go along with GHB 14 when people take them in combination? What are the 15 favorites? 16 DR. FALKOWSKI: It is probably ecstasy, 17 MDMA, and to a lesser extent ketamine and also 18 alcohol. 19 DR. SANNERUD: I have some data on the 20 DAWN statistics too. When drugs are used in 21 combination, 50 percent alcohol, 11 percent 22 stimulants, 8 percent marijuana, poly drugs, 23 hallucinogens and sedatives and all these are at 24 least at 3 and 2 percent each. 25 DR. KAWAS: Dr. Dyer, I believe you are 147 1 our next speaker. 2 DR. KATZ: Claudia, if I could just ask a 3 question, and I don't know who best to direct it, 4 but you said 70 percent of the time the reports are 5 of GHB in association with something else. So, 6 presumably 30 percent of the time it is the sole 7 drug. I have a sort of methadologic question. How 8 reliable would you say that information is, just in 9 general? What is sort of the nature of the 10 information that is recorded and from whom that 11 allows us to conclude that, in fact, GHB is the 12 only drug that was taken? Who reports that, and 13 how reliable are those reports, just as a general 14 rule? Number one. 15 Number two, how many of the deaths and 16 very serious adverse events were associated with 17 GHB use alone? 18 DR. FALKOWSKI: I believe you could 19 address the reliability of DAWN. You are a DAWN 20 reporter. Again, regarding the deaths, you know, 21 the Drug Abuse Warning Network also collects data 22 from medical examiners, but the people in the 23 20-city work group of mine rely more often on 24 getting data directly from the medical examiners, 25 first because it is more timely and also because it 148 1 casts a better net. It captures situations that 2 are not only due to drug-related toxicity but also 3 ones where the use of drugs were considered by the 4 medical examiner to be significant contributing 5 factors to the death. So, that is what I can say 6 about deaths. 7 Also, I have a table, if you are 8 interested, that I could make available that shows 9 exactly DAWN emergency room data for 1999 and what 10 were the co-ingestants. 11 DR. KAWAS: Our next speaker is Dr. Jo 12 Ellen Dyer, from the California Poison Control 13 System at UCSF, speaking on adverse medical effects 14 with GHB. 15 Adverse Medical Effects with GHB 16 DR. DYER: Thank you and good afternoon. 17 [Slide] 18 In 1990 I identified and made the first 19 reports on GHB abuse from over-the-counter sales of 20 GHB. Over the next 11 years I have been following 21 GHB. I have an interest in it and I have been 22 reporting on the progress, the adverse effects and 23 the trends in use. 24 [Slide] 25 This is a description of the California 149 1 Poison Control System data of GHB reports to our 2 center. We logged these reports over 10 years. 3 The first years are when the San Francisco center 4 stood alone so it is a population base of 7 or 8 5 million. We became a system in '97 so we have 4 6 years of data for the entire state. 7 We are a medical toxicology consult 8 service, so we are not a required or mandatory 9 reporting center. So, this reflects just the tip 10 of the iceberg of use and abuse and adverse effects 11 that are out there. 12 [Slide] 13 In our experience GHB produces a profound 14 coma. This has been known for over 40 years, 15 starting out in surgical anesthetic studies where 16 it was evaluated as an anesthetic and now through 17 numerous occurrences of coma in users through this 18 widespread public use, where accidental overdoses 19 are occurring because of the narrow and variable 20 therapeutic index for this drug. 21 [Slide] 22 Looking at 5 studies, anesthetic studies 23 that cover over 700 patients -- there are many 24 other studies; I just picked a small set of them -- 25 you see the effects of GHB in a controlled 150 1 situation. GHB causes unconsciousness and a 2 profound coma. This is what is intended with an 3 anesthetic. The respiratory effects that are seen 4 are Cheyne-stokes respiration. There were 5 aspirations. There was a case of unexplained 6 pulmonary edema. In many of these cases the 7 patients are intubated and the airway is attended 8 to. If their airway was left to chance in these 9 situations, it would be compromised. They lose 10 their airway protective reflexes. They have no 11 gag. So, with the high incidence of vomiting, 12 about 30 percent in these studies, combined with 13 the loss of gag, it is not difficult to see how 14 aspiration is going to occur. 15 There are cardiovascular effects, like 16 bradycardia, and then there are isolated incidences 17 where blood pressure rose up to 30-60 mmHg for 18 unexplained reasons really. There is myoclonus 19 that we see. There is an emergence delirium, 20 confusion. There are also secretions like 21 salivation, vomiting, incontinence and diaphoresis. 22 [Slide] 23 If I look at 16 reports that cover 175 24 cases of adverse events where GHB was in public 25 use, you see these same physiologic responses to 151 1 GHB. You have profound coma. They develop a mild 2 respiratory acidosis; bradycardia; myoclonus; 3 confusion; emergence delirium; and then the 4 secretions. This raises doubts for safety of use 5 among a generalized public population. 6 [Slide] 7 If we look at a closer group where we did 8 a study in our emergency department, and this is 9 the San Francisco County emergency room that sees 10 over 200 patients a day -- we looked at GHB 11 overdoses that we had over 3 years. This is just a 12 retrospective descriptive study where we were 13 trying to get a handle on what is going on. We 14 found that of those cases, about 33 percent had no 15 co-ingestion. This was documented by either 16 toxicology or patient report. Those patients came 17 in, a quarter of them, with Glasgow Coma Score of 18 3. So, they were profoundly comatose and 33 19 percent of them had coma scores between 4-8. The 20 coma lasted 15 minutes to 6 hours. 21 Again, a third of the patients had these 22 same symptoms, bradycardia, respiratory acidosis, 23 hypothermia, vomiting. We saw hypotension in about 24 11 percent. Those cases were primarily cases where 25 alcohol was co-ingested. Then, on emergence these 152 1 patients are difficult to manage. They can have an 2 emergence delirium which includes combative, 3 agitated behavior. 4 [Slide] 5 Because of that evidence and wanting to 6 focus in closer and get some GHB levels to find out 7 if that is truly what we were looking at, we did a 8 prospective study over 6 months, looking at 15 9 cases of GHB overdose, and 73 percent of those came 10 in with a Glasgow Coma Score of 3. Our intent was 11 to document the presence of GHB, to detect the 12 co-ingestants and what they were or if there were 13 none, and then to verify that our ability to 14 predict an overdose is truly GHB by the toxidrome 15 that we are using, whether or not that was 16 effective. 17 So, all of these 15 cases did have GHB 18 that was measurable. They were young, ages 20-39; 19 73 percent were male. The study inclusion criteria 20 were patients presenting with Glasgow Coma Scores 21 less than 8 and 73 percent of these patients had a 22 Glasgow Coma Score less than 3. 23 In 5 of the cases there were no other 24 drugs or alcohol detected. The GCS was 3 in 80 25 percent of those cases. So, profound coma from 153 1 accidental overdose; no other obvious cause. 2 [Slide] 3 It is clear to us that there is really 4 substantial evidence that GHB causes coma. Coma is 5 life-threatening, and these deaths are occurring 6 from accident or injury and from respiratory 7 compromise. We are seeing that through aspiration; 8 through apnea; through positional asphyxia -- these 9 are profoundly comatose people, they can't even 10 move to open their airway -- and through pulmonary 11 edema. 12 [Slide] 13 So, I have reviewed 20 GHB related 14 fatalities where I had autopsy reports. I just 15 sent letters to medical examiners asking for their 16 reports. In these cases, the ages ranged from 15 17 to 46 years. Three-quarters of them were male; 20 18 percent of them had no concurrent ingestions. If 19 we look at those that had co-ingestants, the 80 20 percent. We will see that many of these substances 21 are legal commonly ingested things. Tylenol was 22 one of them; caffeine; alcohol. The levels of 23 alcohol went up to 0.17 percent. The legal limit 24 for driving ranges from 0.08 to 0.1. So, most of 25 these cases were in the lower range, right around 154 1 the legal limit of driving, saying that they had 2 maybe one or two drinks and none of these would 3 reach an alcohol level that would cause coma. 4 [Slide] 5 The societal costs that were seen from GHB 6 abuse, there are many driving under the influence 7 arrests that have occurred with GHB. There were a 8 whole lot that were not recognized until GHB 9 testing became available and now they are being 10 recognized. I don't go out really and collect this 11 data but there are two vehicular manslaughter, I 12 guess they would call it, cases where a person 13 driving under the influence of GHB has hit and 14 killed another individual. One of those was in '96 15 and one was in 2000. 16 Another societal cost is the assaults 17 where the victim is under the influence of GHB 18 given to them or slipped to them by the assailant. 19 It is common enough that they have a term for it. 20 It is called being "scooped" by GHB. The assailant 21 then attacks the victim while they are unconscious 22 or amnestic to the effects of the drug, making 23 prosecution and even reporting of these very, very 24 difficult. 25 These are 4 cases. There are others. But 155 1 in these GHB was clearly documented as the cause. 2 The first was a woman who was drugged and assaulted 3 by her boss as they went out with a group of 4 colleagues after work. She had GHB in her urine. 5 There were 10 victims of some DJs in Los Angeles 6 that were slipping GHB into drinks and then 7 assaulting them. There was a 24-year old that was 8 eventually prosecuted more for trafficking drugs 9 after a woman had reported an assault to them and, 10 in kind of the bargaining, he admitted, yes, he had 11 drugged her twice with GHB and she has no memory of 12 the first event at all. Nothing. The last is two 13 15-year old females who were unconscious at a 14 party. One was hospitalized and one of these girls 15 died. 16 [Slide] 17 We also see addiction as another burden 18 from GHB abuse. We are currently seeing one to two 19 cases a month at our poison center, and this is 20 eight cases that I collected. The age range is 21 young, 22-38, again three-quarters male. The 22 pattern just continues through all these of the 23 demographics of who is using. Of these, 63 percent 24 started taking GHB for body building. They had 25 what they thought was kind of a legitimate use of 156 1 this dietary supplement. In this group, 88 percent 2 of them were employed or students. These were 3 functional members of society that have had trouble 4 now because of this drug. These are not people 5 that really had drug-seeking behavior. The onset 6 of symptoms we see within 1-6 hours. It progresses 7 over a couple of days. The duration is 5-15 days. 8 Now, these are often unrecognized by 9 healthcare professionals when they present for 10 treatment. GHB abuse addiction is not really very 11 well known out there. These are severe 12 neuropsychiatric symptoms with autonomic 13 instability that we see. I have had physicians who 14 have treated many, many cases of severe alcohol 15 withdrawal that have called me up and said, my 16 gosh, I am impressed; I am so impressed by this 17 withdrawal symptom. The patients become agitated, 18 combative, delirious. They are hallucinating. 19 They require sedation, a milligram a minute of IV 20 Ativan has been used over a few hours to gain 21 control. They require four-point leather 22 restraints and intensive care. One of the 23 patients in this series died while being 24 hospitalized for GHB withdrawal. 25 [Slide] 157 1 Substantial and compelling evidence from 2 case reports of accidental poisoning and from 3 toxicology supported adverse events really shows us 4 that these effects are due to GHB. It is not some 5 contaminant or something else that is causing 6 these. And, there is an insufficient or no safety 7 margin between the effective level of the 8 therapeutic dose of these drugs that these people 9 are taking and the dose that causes these effects. 10 As you can see from the sponsor's study, the 11 adverse effects that they are reporting are very 12 similar. The confusion, the nausea, the vomiting 13 are very similar to the things that we are seeing. 14 One physician, Dr. Gallamberti from Italy, 15 who is doing therapeutic use of GHB withdrawal 16 states talks about a 15 percent problematic GHB use 17 among his population. This can be dose escalation. 18 This can be GHB overdoses up to 10 times a year, or 19 GHB dependence. 20 [Slide] 21 This slide just looks at the kinetics to 22 illustrate that there is really a very narrow 23 therapeutic index with this drug and there is a lot 24 of variability. The pharmacokinetics of GHB are 25 capacity-limited absorption, capacity-limited 158 1 elimination. The coefficient of variation of some 2 of these parameters is 50 percent. There is a lot 3 of variation and we don't really know what the 4 consequence in different populations and different 5 people of these really variable kinetics is going 6 to be, or why they are so variable. You are used 7 to using phenytoin. It has capacity-limited 8 elimination. We know that when you are bumping the 9 dose of a patient on phenytoin you have to be 10 really careful because they can exponentially 11 increase their level. Well, the same thing happens 12 with GHB and we don't know where that is yet. 13 There is not enough experience. And, with 14 phenytoin the absorption is pretty good. We know 15 the bioavailability of IV phenytoin and oral 16 phenytoin. Here, I don't think it is so constant. 17 It really changes with food and there is a 18 capacity-limited absorption that is going to vary 19 between patients. So, this is a really difficult 20 drug to control, particularly orally on an 21 outpatient basis. 22 [Slide] 23 So, what is the current level of GHB abuse 24 that is out there? We really don't know. If we 25 wanted to project from one survey that was done, 159 1 Dr. Miotto, a UCLA physician that works addiction 2 medicine did a 45-minute structured interview with 3 42 GHB users. Among that group, 69 percent had 4 admitted that they had lost consciousness, had 5 periods of consciousness laps from minutes to 6 hours. There was variability in the amnesia 7 dependent upon how often people used. Twenty-eight 8 percent admitted having an overdose; 9 percent had 9 been to the emergency department for an overdose. 10 Now, there is an interesting misconception 11 here where they don't consider the loss of 12 consciousness to be an overdose, and people 13 overdose and when they are in a profound coma are 14 not taken to the emergency department. So, there 15 are really some problems there, and this gives us 16 an example of the kind of under-reporting that is 17 out there. 18 If we try and extrapolate from the amount 19 of drug that we are seeing marketed illicitly, this 20 is just one arrest in Marin County, a small county 21 north of San Francisco, where they had 207 L of 22 butanediol. The average street dose varies around 23 2 g. If you look at that, that is 103,500 doses in 24 one capture at one house, and there are many, many 25 of these. There are lists of different amounts 160 1 that have been busted all over. 2 Then there is the problem that Carol has 3 already talked about, surveying and policing the 4 issues of this type of new drug abuse. There is no 5 systematic method in place for data collection on 6 this. 7 There is rapid metabolism of the drug. It 8 clears from the blood in within about 6 hours; it 9 clears from the urine within about 12 hours. We 10 can't test these people and find it. When we are 11 trying to get evidence in a drug assault case, it 12 is gone. It is really difficult to detect. And, 13 should we increase our level of detection to the 14 very, very minute nanogram kind of range, then we 15 are going to start running into the biological 16 background so we aren't even going to be able to do 17 that if we increase our ability to detect. There 18 are also very poor assays currently out there. 19 None of the hospitals have an assay for this, and 20 none of the law enforcement has a field kit for it. 21 So, it has to be taken into a lab and specifically 22 run through a complicated GC mass spec procedure to 23 get a level out, which is expensive. 24 The current documentation clearly grossly 25 underestimates the amount of use that is out there. 161 1 And, it is very clear that there is a little, if 2 any, safety margin with GHB use in the therapeutic 3 doses that are proposed. GHB is a very potent new 4 drug of abuse. It has been around 10 years. We 5 thought it was going to come and go as a fad, it 6 hasn't and it is not going to. The use is still 7 increasing. 8 There is a very high acute toxicity in 9 accidental overdose -- coma, bradycardia, 10 myoclonus, vomiting, aspiration -- we are seeing a 11 lot of it, and it has very high abuse and addiction 12 potential. So, I think that we have to be very 13 careful and it is very difficult to try and 14 minimize these potential risks, the risks of having 15 it get out into the drug abusing population but 16 also among patients that we are going to be giving 17 this drug to take at home. At the poison center, 18 every night at bedtime, 9 to 11 o'clock I am called 19 by people that say, oh, I'm sorry, I accidentally 20 took a double dose of my medication. What should I 21 do? In this case, they are all going to go to the 22 emergency room. There is really not a margin of 23 safety with this drug. Thanks. 24 DR. KAWAS: Thank you, Dr. Dyer. The next 25 presentation is from the sponsor, presentation on 162 1 risk management and abuse liability, Dr. Bob 2 Balster, from the Medical College of Virginia. 3 DR. REARDAN: Yes, I would like to now 4 introduce Dr. Balster who will present his views 5 with respect to abuse liability of Xyrem and GHB. 6 Dr. Balster is a previous chair of the FDA Drug 7 Abuse Advisory Committee and a widely published 8 abuse pharmacologist from the Medical College of 9 Virginia. He is editor and chief of a leading 10 addiction journal, Drug and Alcohol Dependence, and 11 a past president of the College on Problems of Drug 12 Abuse. 13 Sponsor Presentation on Risk Management 14 and Abuse Liability 15 DR. BALSTER: Thank you very much, Dayton. 16 Good morning or good afternoon, I guess it is now. 17 [Slide] 18 Well, as you have just heard, the 19 development of Xyrem as a medication has taken 20 place in a context of a national epidemic of the 21 abuse of its constituent GHB, and also the abuse of 22 a number of GHB-related drugs that I will tell you 23 about. 24 As Dr. Houghton told you, Orphan is very 25 well aware of this problem and has consulted many 163 1 drug abuse experts to try to understand the problem 2 better. My own analysis of this situation is that 3 Xyrem has certainly not contributed to the problem 4 that exists today with the abuse of this class of 5 compounds. I guess where I may disagree a bit is 6 that I am pretty convinced that Xyrem is not going 7 to be a player in this over the long term. 8 I think in order to understand and make an 9 appropriate public health response to this 10 situation, you need to know a little bit about what 11 some of the causes are of this GHB abuse problem. 12 [Slide] 13 So, I hope to make two points in this 14 presentation. The first point is that I believe 15 that the recent abuse of GHB-like substances 16 probably reflects a ready availability more than 17 their inherent pharmacological propensity for 18 abuse. 19 I think I will make this point by first 20 off reviewing for you the incredible availability 21 of these compounds, and then also review very 22 quickly scientific studies that have been done on 23 the abuse liability of GHB as it is compared to 24 other drugs of abuse you might be familiar with. 25 Secondly, I believe that Xyrem, if approved for 164 1 medical use, will not contribute to the public 2 health problem of the abuse of these GHB-like 3 substances in any significant way. 4 [Slide] 5 Before we continue, it is very important 6 to know the cast of characters here. I think next 7 to the federal government, the next worst developer 8 of abbreviations is a drug abuse research 9 community, with MDMA, and PCP, and GHB, and BD -- 10 it must be hard to kind of keep track of the 11 players but, of course, the drug we are talking 12 about here is GHB, gamma hydroxybutyrate. But 13 there are a bunch of other drugs that are basically 14 part of this national drug abuse problem. 15 You have heard a little bit about them, 16 but these precursors, gamma butyrolactone or GBL, 17 1,4 butanediol or 1,4-BD are precursor compounds 18 that, if obtained, can be easily and readily 19 converted into GHB. They also can be consumed 20 directly because they are metabolized by the body 21 into GHB. So, they themselves are drugs of abuse 22 like GHB. Then there are others that are also 23 available. 24 Now, of all these chemicals only GHB is 25 actually a scheduled drug. It is Schedule I under 165 1 the Controlled Substances Act for the abusable 2 versions, GHB; Schedule III for an approved medical 3 product. So, only GHB is scheduled. Now, GBL is 4 what is called listed so its availability is 5 diminished. These others are still freely 6 available without any drug abuse controls. 7 [Slide] 8 You have heard a lot about GHB abuse but I 9 am pretty convinced that what we are seeing here is 10 something that has resulted from an amazing 11 situation of the availability of these compounds. 12 To remind you, GHB was available legally and 13 legitimately through health food stores up through 14 1990 when you could buy it anywhere, and the abuse 15 problem with this drug began during that period of 16 time. 17 Then through that time and afterwards GHB 18 could be obtained through the Internet. There was 19 an amazing number of sites set up to sell GHB. 20 Then, as GHB became less easy to get because 21 Internet sources dried up, the Internet sources 22 were selling the precursors, etc., etc. I will 23 show you some data a little bit more, but these 24 precursors are not going to disappear any time soon 25 from public availability. Now that the 166 1 availability of GHB has been restricted by the 2 federal scheduling actions and actions by the FDA, 3 people can now purchase the precursors and make 4 their own GHB. Essentially anyone can do that. It 5 is a very simple thing and the recipes are right 6 there on the web. As I said before, they 7 themselves are widely abused. So, we have a class 8 of chemicals here that are really basically part of 9 what has been referred to as a GHB abuse problem, 10 but it is really an abuse of a class of drugs, and 11 you saw some evidence on that. 12 [Slide] 13 At this point I want to review the 14 scientific literature on the laboratory studies of 15 the abuse potential of GHB. You may wonder why I 16 would want to do that, I mean, why would I want to 17 review literature on abuse potential when the 18 reality of GHB abuse is clear to us from 19 epidemiological data that Dr. Falkowski mentioned 20 and clinical data. The reason to do this is to try 21 to understand what the basis for this is, and to 22 know whether or not this wide abuse is due to some 23 features of this incredible availability, or 24 whether the drug has sort of the inherent 25 pharmacological desirability that you would 167 1 associate with a really dangerous drug like cocaine 2 or heroin where, no matter how many billions of 3 dollar we throw at the problem, we are getting 4 nowhere with it, or does GHB represent a drug which 5 is less desirable or has less propensity for use. 6 [Slide] 7 Just to remind you, there is a 8 well-established science of abuse liability 9 evaluation, and it is used in evaluating new 10 compounds that are under development. It is useful 11 in making decisions about drug abuse control, and 12 data such as these are used widely by the FDA for 13 making regulatory decisions. All of these data are 14 reviewed in your packages, but just to quickly tell 15 you, first off, GHB is a unique drug. It is not 16 just another depressant drug like barbiturates or 17 even benzodiazepines that have its own receptor and 18 its own characteristics. 19 In studies which are called drug 20 discrimination studies, which allow you in a way to 21 compare unknown drugs to known drugs of abuse, 22 again, GHB lacks equivalence to these classical 23 depressants like barbiturates or any other classes 24 of drugs to which it has been directly compared. 25 In self-administration studies -- these 168 1 are laboratory studies where you can actually 2 measure what we call the reinforcing effects of the 3 euphorigenic potential of these drugs -- actually 4 in this particular class of studies GHB has very 5 weak reinforcing effects. It is difficult to 6 obtain them in laboratory studies and there have 7 been a number of those. We did one of these 8 ourselves in our laboratory and we essentially 9 found no evidence of GHB self-administration under 10 conditions where we reliably get 11 self-administration of cocaine, heroin, 12 barbiturates, etc., etc. 13 The case of physical dependence is a 14 little bit more complicated. You heard from Dr. 15 Dyer about the fact that abusers can develop 16 dependence and show withdrawal signs, and there is 17 no question about that. These people are taking 18 maybe 10 or more times the therapeutic dose. We 19 are talking about 70, 80, 100 grams a day, and they 20 take them every 3 hours or so because they have to 21 maintain the blood level. Yes, in those cases you 22 get dependence, but in patients receiving Xyrem, 23 where they are getting it in lower doses and they 24 are taking it only in the evening, as you have 25 heard from the reports, there have not been 169 1 significant problems of dependence. So, yes, it 2 can occur in abusers but it isn't really an issue 3 in patients. Importantly, animal studies, for 4 example, where you try to show the dependence of 5 GHB and compare it, for example, to barbiturates, 6 it is not easy to develop a model for GHB 7 dependence in animal studies because it has less 8 inherent dependence producing properties than these 9 other drugs. 10 [Slide] 11 So, my conclusion when I reviewed the 12 literature on the scientific studies of GHB, when I 13 was asked to do that, I basically thought it looked 14 a lot like what I would say is a Schedule IV drug. 15 Schedule IV drugs, you remember, are 16 benzodiazepines and chloral hydrate and drugs of 17 this type, and that is sort where it fit. It isn't 18 like cocaine. It isn't like heroin. In fact, that 19 analysis of looking at the data has been made by 20 others with very much the same recommendation as 21 mine, that is, it sort of fits pharmacologically 22 with Schedule IV. 23 For example, the WHO expert committee 24 which met not too long ago to make a recommendation 25 to the UN Commission, the WHO expert committee 170 1 recommended Schedule IV and, in fact, the UN 2 Commission ultimately placed GHB in Schedule IV. 3 Schedule IV, under the Psychotropic Convention is 4 very analogous really to our Schedule IV that you 5 are familiar with under the Controlled Substances 6 Act. 7 [Slide] 8 We are not here to talk about GHB abuse 9 which we know is a significant problem. We are 10 here to talk about Xyrem and what its role may be 11 in the drug abuse problem in the United States. 12 There are two issues we are really worried about 13 here. Number one, we are worried about the 14 possibility that patients legitimately prescribed 15 Xyrem will abuse it in some way, or misuse it or 16 escalate and then, secondly, we are worried about 17 whether or not it might be diverted into sort of 18 illicit channels and become part of a problem in 19 that way. 20 [Slide] 21 Turning first to the issue of patients, 22 first off, I think most of you know, and it is 23 important to always know this, that the development 24 of abuse among patients receiving therapeutic doses 25 of abuse drugs is a much smaller problem than some 171 1 people realize. It is actually fairly unlikely to 2 occur in a general sense. Of course, in the trials 3 with Xyrem there weren't problems of abuse; there 4 wasn't evidence that people were escalating their 5 dose or complaining and asking for more, and that 6 sort of thing. 7 It is important also to recognize that 8 narcolepsy patients are patients that are receiving 9 controlled substances all the time. The stimulant 10 class of drugs, all those drugs that Dr. Mignot 11 spoke about are all scheduled compounds. In fact, 12 many of them are Schedule II where they can't even 13 get them half the time because the production 14 controls on Schedule II reduce their availability. 15 Then the issue about their dependence, if 16 you understand, for example, that with 17 benzodiazepines, when you discontinue 18 benzodiazepine administration you will often see a 19 withdrawal syndrome, well, that is because 20 benzodiazepines have this incredibly long duration 21 of action with active metabolites that accumulate 22 so that the body continually maintains levels of 23 benzodiazepines. So, when you quit using them 24 there is a withdrawal syndrome. With GHB, as you 25 saw from Dr. Houghton's presentation, the duration 172 1 of action is just a couple of hours. It would take 2 many, many, many multiple daily uses, way more than 3 the patients are going to get, to maintain the kind 4 of levels of GHB that would be expected to produce 5 dependence. So, yes, in abuse cases where people 6 are just going all day and all night but not with 7 patients. 8 [Slide] 9 Turning now to illicit diversion of Xyrem, 10 first off, that hasn't happened yet. So, we are 11 not aware of any diversion of any Xyrem through any 12 of the products. This is, of course, only in 13 clinical development but I think it is important to 14 know. Most importantly, the company has been very 15 much worried about this issue and has developed a 16 distribution system that you are going to hear 17 about, called the Success Program, which I 18 personally believe is going to substantially 19 prevent any opportunities for diversion. Lastly, 20 Xyrem, whether you approve it or not -- it is going 21 to make very little difference in the overall 22 availability of this whole class of chemicals in 23 the national scene. 24 [Slide] 25 To illustrate that, this slide shows you 173 1 the product amounts anticipated, annual production 2 amounts for this class of chemicals I mentioned to 3 you. So, if Xyrem is approved the anticipated 4 first year production amounts of gamma 5 hydroxybutyrate are about 82,000 kg. GBL, gamma 6 butyrolactone, the precursor that can be made into 7 GHB easily and consumed itself, is 83 million kg, a 8 thousand times more. 1,4-BD which is not a 9 controlled substance and has no drug abuse control 10 under it whatsoever right now, is widely available 11 through all sources in large amounts, and is made 12 in the neighborhood of 377 million kg. For those 13 of you who don't do the metric system, that is 14 400,000 tons of 1,4-BD. And, all of these drugs 15 are basically substituting for one another. So, 16 whether you take Xyrem in or out of that graph, it 17 is not going to make a difference. 18 [Slide] 19 In conclusion, I believe that the epidemic 20 of abuse of GHB-like drugs has resulted really 21 primarily from its extraordinary availability. In 22 fact, when GHB was controlled -- it is hard now to 23 get GHB. It is hard even for me to get GHB as a 24 research scientist. So, the problem has now 25 switched to these precursors that are available. 174 1 Secondly, the scientific studies of GHB 2 show that you are not talking here about cocaine or 3 heroin. It is a weak depressant of maybe the 4 benzodiazepine, chloral hydrate type. Thirdly, I 5 believe that Xyrem abuse is very unlikely among 6 patients for the reasons I said. Lastly, the 7 contribution of Xyrem to the public health problem, 8 which is the matter of concern, is essentially not 9 significant. So, you know, have your way with the 10 drug in terms of efficacy and safety but I don't 11 think you need to be worried that this drug is 12 going to be a major factor in the drug abuse 13 problem with this class of drugs. Thank you. 14 DR. KAWAS: Yes, a quick question, Dr. 15 Leiderman. 16 DR. LEIDERMAN: Yes, I would like to ask 17 Dr. Balster two questions. I would like you to 18 comment on the species of animal that you are 19 addressing when you talk about self-administration 20 in drug discrimination studies. Two, I would like 21 you to comment on the data that those models show 22 with other classes of drugs. 23 DR. BALSTER: All the studies are reviewed 24 on that slide on abuse potential with laboratory 25 animal studies, using fairly well developed 175 1 methodologies. The self-administration studies 2 that Dr. Leiderman referred to were studies that 3 were done in monkeys in sort of a standardized 4 method that is done through a program directed by 5 the College on Drug Dependence. Those are the 6 models, and I can show you data if you give me the 7 time to do it. Maybe later, if the committee is 8 interested, I can show you data. But these are 9 models in which basically it is extremely easy to 10 get animals to actually literally self-inject most 11 of the drugs of abuse -- cocaine, amphetamines, 12 opiates of all types, barbiturates, depressants, 13 benzodiazepines -- benzodiazepines are a little 14 harder but in the model that was used that I showed 15 the negative results from, benzodiazepines were the 16 positive control. So, basically the only area 17 where that model has been not very successful and 18 underestimates abuse potential is with 19 hallucinogenic drugs and marijuana type drugs. 20 DR. LEIDERMAN: Yes, many of the Schedule 21 I drugs. DR. REARDAN: We just 22 have about another ten minutes. If we can prevail 23 on the committee, we have one last speaker, and 24 that will be Patti Engel, who is going to describe 25 for you the risk management system that the company 176 1 has developed to help control diversion. Patti? 2 Risk Management 3 MS. ENGEL: Good afternoon. My name is 4 Patti Engel, and I am here today to talk to you 5 about the risk management program for Xyrem, which 6 we call the Xyrem Success Program. 7 [Slide] 8 This program will ensure the responsible 9 distribution of Xyrem, namely, to meet two goals. 10 First, to ensure that patients who desperately need 11 the medicine can get it. Secondly, to keep this 12 out of the hands of those people who might abuse 13 it. 14 [Slide] 15 To develop this program we consulted 16 broadly with a number of people interested in the 17 issues not only germane to patients but also that 18 of drug abuse. As you can see, we spoke with drug 19 diversion investigators, field law enforcement, 20 forensics experts, toxicologists, pharmaceutical 21 distribution experts, drug abuse trend experts. 22 [Slide] 23 Through those discussions we followed 24 FDA's proposed risk management guideline, which is 25 risk management through risk confrontation, in 177 1 essence egging the partners and the shareholders to 2 not only identify the issues but also assess the 3 risks, identify the options and select a strategy. 4 The program that I am going to be sharing with you 5 today is certainly a draft program that the company 6 has designed after discussions with these numerous 7 stakeholders. 8 [Slide] 9 This slide I show to you really to point 10 out the standard route of distribution of a 11 pharmaceutical product in our country today. This 12 includes not only commonly used medications like 13 products for blood pressure control or products for 14 arthritis, but also products under Schedule II, 15 including such agents as amphetamines. Typically, 16 a product is manufactured and goes to a number of 17 national, regional and local wholesalers, 18 eventually getting to 63,000 retail drugstores 19 around the country. One can only imagine the 20 number of loading docks, transport vehicles and 21 hands that touch a pharmaceutical product in this 22 traditional distribution system. 23 [Slide] 24 As we contemplated the distribution of 25 Xyrem and how to do this responsibly to meet the 178 1 prior stated goals, we determined that a closed 2 distribution system would best fit everyone's needs 3 for this product. The product is manufactured at 4 one single manufacturing facility. It is sent to 5 one single national specialty pharmacy. Eventually 6 it goes by courier to patients with narcolepsy. 7 [Slide] 8 The benefits of this program are that not 9 only is the product distributed from a central 10 location, but all the controls and all the records 11 are in one place. 12 [Slide] 13 So, how will this work? Because a number 14 of doctors prescribe medicines for narcolepsy, we 15 will focus our promotional effects on those 16 physicians. They include such specialists as 17 neurologists, pulmonologists, psychiatrists, 18 internal medicine physicians and several primary 19 specialties who practice sleep medicine. 20 [Slide] 21 Our small sales force will call on these 22 physicians, communicating the clinical benefits of 23 Xyrem in narcolepsy. At those calls, the sales 24 representatives will also review with each 25 physician something that we call the physician 179 1 Success Program. I will go into the details of 2 this program in a more in depth fashion in just a 3 moment. But it is important to know that each 4 physician will sign that they have reviewed this 5 program with the sales representative and 6 understand the program. I should also note that at 7 no time will we embark upon physician sampling. 8 [Slide] 9 I promised to come back to the components 10 of the physician Success Program. I know that many 11 of you received copies of this but I would like to 12 highlight some of the main points. First, because 13 we know individuals all learn differently -- some 14 by hearing, some by reading, other methods -- we 15 have made this a multi-faceted program which 16 includes videos, brochures, pamphlets that describe 17 four main areas. 18 The first is to highlight to physicians 19 that the distribution process for Xyrem is 20 different, that their patients won't be able to get 21 this at the corner drugstore. The second important 22 issue that this binder points out to physicians is 23 the dosing and administration of Xyrem. The next 24 important issue is that of home storage and secure 25 handling. The fourth is an important module that 180 1 we call "doctor be wary" which is an educational 2 module that educates doctors about the ways that 3 drugs are commonly diverted in this country so they 4 can be aware of patients who are attempting to 5 illegitimately get a prescription from them for 6 this product. Each of the kits will also contain a 7 number of unique prescribing forms for Xyrem which 8 will be necessary in order for the prescription to 9 be filled. This is, in essence, a special 10 prescription form. As well, contact information 11 will be provided should the doctor have any 12 questions at all about the program. 13 [Slide] 14 Once the physician decides to prescribe 15 Xyrem the physician faxes this special prescription 16 to the specialty pharmacy. Now, I am going to come 17 back to how this prescription is verified. So, I 18 will ask you to hold on that point for just one 19 moment. But, based on that prescription and based 20 on the patient's geographic location, the pharmacy 21 assigns that patient to a dedicated pharmacy team. 22 So, each time that the patient deals with the 23 system they are talking with the same pharmacist 24 and the same reimbursement specialist. 25 [Slide] 181 1 I mentioned that as the prescription comes 2 to the specialty pharmacy there will be a number of 3 checks to determine if the physician is, in fact, 4 eligible to prescribe Xyrem. We will be utilizing 5 DEA's NTIS or National Technical Information 6 Services database to ensure that each physician has 7 an active valid medical license, and also to ensure 8 that that physician has current prescribing 9 privileges which allow him or her to prescribe 10 Schedule III medications in this country. As a 11 backup check, the specialty pharmacy will also be 12 checking with the appropriate state medical board 13 to determine that there are no pending actions on 14 the behalf of the state for that given physician. 15 [Slide] 16 As a secondary step, the specialty 17 pharmacy will also do a check on the patient in 18 essence. What they will do is when that 19 prescription comes in they will call the 20 prescribing physician's office to determine that, 21 in fact, that patient is real and a prescription 22 has, in fact, been written for that patient. 23 [Slide] 24 Once insurance reimbursement is obtained, 25 the specialty pharmacy contacts the patient, first, 182 1 to determine the patient or the patient designee's 2 location and availability for shipment, and also to 3 describe to them the contents of the shipment. I 4 will come back to the details of this in just a 5 moment, but it is important that you know that each 6 patient, when they get their first prescription of 7 Xyrem will receive a multi-faceted educational 8 program called the Xyrem patient Success Program, 9 and I will come back to the details of that in just 10 a moment. 11 In that same shipment they will also 12 receive their Xyrem, and that will look something 13 like this, with child resistant closure not only on 14 the primary container but also on the dosing cups 15 which are provided by the company. 16 [Slide] 17 The shipment that goes to the patient is 18 sent by a special system that has a special, unique 19 tracking system called the Rapid Trac System. this 20 system will allow detailed real-time tracking of 21 that package which is delivered only by the 22 authorized signature. If the patient or their 23 designee is not available for receipt of the 24 package at the time agreed upon with the specialty 25 pharmacy, the package will be returned to the 183 1 specialty pharmacy after one delivery reattempt. 2 So, a package will not sit on a delivery truck or 3 in a hub for weeks at a time or anything like that. 4 If the package is lost the system will allow an 5 investigation to begin regarding the shipment's 6 whereabouts at that point of loss. 7 [Slide] 8 I spoke a moment ago about the patient 9 Success Program. Again, this is a multi-faceted 10 program which includes video, brochures and 11 pamphlets which deal with a number of important 12 issues for patients. First addressed, of course, 13 is the distribution process since it is so 14 important that the patients understand that the 15 only way they will get Xyrem is through the special 16 pharmacy and not at their corner drugstore. 17 There is information about Xyrem's dosing 18 and administration because we feel that that is an 19 important message to be delivered in an 20 understandable and a very consistent manner. 21 There is information on home storage and 22 secure handling, and we also are very clear with 23 patients about the criminal and civil penalties 24 that the public law assigns to any illicit use of 25 Xyrem. So, if I were, as a valid narcolepsy 184 1 patient, to take my Xyrem prescription and use it 2 to conduct a rape or in an assault situation, or if 3 I were to sell it to someone for illicit use I 4 would be penalized, I would be subject to C-I 5 penalties. The patient Success Program also 6 includes contact information for the specialty 7 pharmacy should the patient have any questions at 8 all, and also reimbursement information. 9 [Slide] 10 After the Rapid Trac System shows that the 11 package has been received by the patient, the 12 specialty pharmacist will call the patient within 13 24 hours not only to confirm receipt of that 14 package but also to again reiterate certain 15 important points with the patient. Those include 16 the penalties for illicit use of Xyrem; Xyrem's 17 dosing and administration; home storage and secure 18 handling. The pharmacist will also take the 19 opportunity to discuss with the patient the 20 child-resistant features on the primary container 21 as well as the child-resistant features on the 22 dosing cups that are provided. 23 [Slide] 24 The central data repository designed for 25 Xyrem really allows for identification of a number 185 1 of unusual types of behavior, including any 2 duplicate prescriptions, any attempts of 3 over-prescribing, or any attempts at over-use by 4 patients. The benefit here is that that 5 information is available prior to filling the 6 prescription so appropriate pharmacist intervention 7 can occur. 8 [Slide] 9 As you can see, the Xyrem Success Program 10 is a comprehensive program which is designed to 11 responsibly distribute this important medication in 12 order that patients who need it have it available, 13 and it is inaccessible for those who might abuse 14 it. Thank you. 15 DR. REARDAN: Dr. Kawas, that completes 16 our presentation and we will turn this back over to 17 you. 18 DR. KAWAS: Thank you very much. I want 19 to thank all of you for all of your nice 20 presentations but, rest assured, you will have more 21 questions in the afternoon. We are running quite 22 late so we are going to cut lunch a little short 23 and we will plan on reconvening at 1:30, at which 24 time the public hearing component of this meeting 25 will happen. 186 1 [Whereupon, at 12:50 p.m., the proceedings 2 were recessed for lunch, to resume at 1:30 p.m.] 187 1 A F T E R N O O N P R O C E E D I N G S 2 DR. KAWAS: We will reconvene the meeting 3 of the Peripheral and Central Nervous System 4 Advisory Committee discussing Xyrem. We are now in 5 the open public hearing portion of this meeting, 6 and we have quite a few people that we will be 7 hearing from and additional people who want to add 8 to the list. I would like to ask all of the 9 speakers to please do their best -- not their best, 10 you must stay to five minutes. You will have a 11 one-minute warning sign with your timer. If you go 12 over, please don't take it personally but you might 13 hear my voice ending your part for the meeting. 14 This is in order to allow us to hear from everybody 15 who wants to speak, as well as to get onto the 16 deliberations of this committee. The first speaker 17 in the public forum is Sharon Fitzgerald of 18 Littleton, Colorado. 19 Open Public Hearing 20 MS. FITZGERALD: Good afternoon. I am 21 Sharon Fitzgerald from Littleton, Colorado, and I 22 am a narcoleptic. I am a volunteer member for the 23 Orphan Medical Patient Council and the Narcolepsy 24 Network is paying for my travel and my hotel to 25 allow me the privilege of speaking with you today. 188 1 Five minutes isn't long enough. I have provided a 2 longer version for you to read. Please, please 3 read it. It explains my experiences with the five 4 major symptoms of narcolepsy and how Xyrem gave 5 back my American dream, the ability to pursue 6 happiness without stumbling on the way when it gets 7 tough, and without literally falling on my face 8 when the goal of happiness is reached. 9 I have had daytime sleepiness since 1969. 10 It threatened my ability to be a good mother and 11 protect my children, and it trapped me in a series 12 of entry level jobs. Not knowing it had a name, I 13 tried to hide my problem from employers and I hid 14 in restrooms for many years for 15-minute naps at 15 unpredictable times lots of the time. 16 My symptoms dramatically increased and 17 worsened in 1977 when I was in law school. I was 18 raising school age kids on my own, being widowed at 19 the age of 32. In daytime, against my will, I took 20 naps in my classes, going instantly from 21 consciousness to dream state sleep, the switch 22 being so quick that I actually wrote words from my 23 dreams in my class notes about things like my 24 mother and helicopters, and wondered where they 25 came from when I read them. These were usually 189 1 followed by a mark where I dropped my pen as I 2 stopped writing, and that would startle me into 3 wakefulness and I would stay awake for a while and 4 take more notes. 5 Going to sleep nearly every night, my mind 6 created vivid illusions of my very worst fears, 7 often a murderous rapist breaking into my house 8 from behind wherever I was sitting or lying. My 9 knowledge of where I was, was accurate. I could 10 not scream. I was paralyzed and I couldn't turn 11 around to defend myself. These are called, as you 12 know, hypnagogic hallucinations. I didn't know 13 that at the time. 14 At the same time, the symptoms of 15 nighttime wakefulness became more severe. I 16 experienced long hours of anxiously lying awake, 17 punctuated by times of intense dreaming so real and 18 so vivid that in the daytime I couldn't remember 19 whether events I remembered were real or dreamed. 20 You may understand that I feared for my sanity, and 21 this is when I sought medical help. 22 I was my doctor's first experience with 23 narcolepsy. It took a very long time for him to 24 find a diagnosis. When he did, it was because of 25 my mild cataplexy and he found the diagnosis an 190 1 announced that was the good news because the bad 2 news was there was no treatment. I self-medicated 3 for years with Sudafed and coffee. 4 With determination -- if you knew me you 5 would know about it -- and special accommodations 6 from the university I actually finally managed to 7 graduate from law school, but I turned down the 8 dream job that was offered, clerking for a district 9 court judge, because I feared I would fall asleep 10 in front of the courtroom. He told me our first 11 case would be about two nuns who had been brutally 12 murdered and I feared I might experience cataplexy. 13 By this time my cataplexy had increased to 14 the point that all my facial muscles would relax 15 and my speech would become momentarily slurred. It 16 passed so quickly that I couldn't hide it. I was a 17 sole practitioner. I couldn't bill enough hours to 18 earn a living. I took Ritalin; I took 19 antidepressants unsuccessfully. I found a job with 20 the State of Colorado. It didn't require my legal 21 expertise but I got lucky, I found out about the 22 trials. I had rebound cataplexy, like what they 23 showed you in the pictures, and it was horrendous 24 for several weeks, waiting to be on Xyrem and my 25 secret was brought out at work. But they didn't 191 1 fire me because I told them I was going on Xyrem. 2 Its effects were immediate and dramatic. 3 I have experienced no side effects. I get good 4 restful sleep. I awaken refreshed. I stay 5 reliably awake at work with fewer stimulants and I 6 don't fall down. My supervisors noticed my 7 increased wakefulness and rewarded it with 8 committee chairmanships and memberships and, in 9 1999, a promotion. In 2000, January 1, I became an 10 administrative law judge for the Division of 11 Workers Compensation in the Colorado Department of 12 Labor and Employment. It is responsible; it is 13 emotional. I can do it. My colleagues know I have 14 narcolepsy and they know that with Xyrem it doesn't 15 interfere with my job performance. For years I was 16 unable to safely carry my children or 17 grandchildren. I carried my newborn to his first 18 examination and that is just the beginning of my 19 story. 20 DR. KAWAS: Thank you, Ms. Fitzgerald. 21 Next is Richard Gelula, the executive director of 22 the National Sleep Foundation. 23 MR. GELULA: Thank you. The National 24 Sleep Foundation is an eleven-year old organization 25 that was developed by the American Academy of Sleep 192 1 Medicine to educate the public about sleep and 2 sleep disorders, and our leadership has always been 3 drawn from the top tier of sleep experts, sleep 4 scientists and sleep physicians. We are 5 independent. We raise our money in a variety of 6 ways including government grants, corporate grants, 7 and many memberships, individual contributions that 8 have played a major part, particularly from people 9 and families affected by sleep disorders. Our 10 funding from Orphan Medical for the last two years 11 has been a total of 160,000 out of a two-year total 12 of about 5 million. Our budget is about 2.5 13 million a year. And, their support has gone to 14 broad activities -- sponsorship for National Sleep 15 Awareness Week where they join in with other 16 sponsors, and there is no name or brand specific 17 recognition or benefit for them. So, I wanted to 18 point that out. 19 The Foundation is qualified to address 20 this and our interest is due to the fact that we 21 have invested about a million dollars in narcolepsy 22 research, including center grants for the genetic 23 research done at Stanford. We presently have one 24 of our postgraduate fellowships at UCLA studying 25 the neurophysiology of cataplexy. We also have 193 1 established the National Narcolepsy Registry which 2 has registered to serum DNA registry with about 700 3 patients and family members registered. That is 4 managed at Montefiore Hospital in the Bronx, and it 5 has been a resource for seven scientific 6 investigations. 7 To summarize the position of the National 8 Sleep Foundation on sodium oxybate, the National 9 Sleep Foundation calls upon this panel to fully 10 consider the safety and efficacy of sodium oxybate 11 for the treatment of narcolepsy and cataplexy, and 12 to do so in a comprehensive context that fully 13 recognizes the extreme psychological, emotional, 14 economic, social and health toll that this 15 affliction exacts from people who suffer from it. 16 NSF does not presume to second-guess the 17 evidence that has been submitted about the safety 18 and efficacy of this drug, but it goes on record to 19 say that such considerations should only pertain to 20 affected patients and not other societal 21 considerations. It is safe and effective for 22 people with narcolepsy, like the speaker before me. 23 Sodium oxybate should be made readily available to 24 them. Any concern for illicit use should be 25 addressed strongly through other channels, such as 194 1 law enforcement and professional licensing. The 2 fact that narcolepsy is an orphan disease, for 3 which only one medication is currently indicated, 4 would be weighed as a factor in favor of approval 5 of sodium oxybate because it is likely that 6 availability of an approved drug will foster faster 7 diagnosis and more appropriate treatment, and will 8 also -- and we think this is very important -- 9 stabilize patients who usually first experience the 10 dreadful effects of narcolepsy and cataplexy during 11 their developmental years, before the completion of 12 their educations and initiations of a career. 13 I would like to summarize a few key 14 background points. Narcolepsy and all of its 15 primary characteristics, including cataplexy, are 16 truly life-altering afflictions, a term that best 17 connotes the life-diminishing and debilitating 18 aspects of this disabling disease. Untreated, 19 narcolepsy not only causes vivid nightmares and 20 undermines the safe and secure feeling that most 21 people get when they go to sleep, but it makes 22 daily existence, both objectively and subjectively, 23 frightening and strange, even alienating to the 24 self and others. It makes the well-controlled 25 process that routinely governs the existence for 195 1 almost all other humans, the alternating cycle of 2 sleep and alertness, into something entirely 3 different, an uncontrollable process where the 4 maintenance of conscious attention becomes random 5 and cannot be sustained or relied upon. Both the 6 phenomenon of overwhelming sleep attacks and the 7 muscular weakness and collapse that occur with 8 cataplectic attacks undermine the sense of 9 predictability and confidence required to fully 10 develop and function in our contemporary world. 11 But a true understanding of narcolepsy 12 goes beyond physiology. The cumulative effects of 13 the distinctive daytime and nighttime 14 characteristics of this disease are truly 15 traumatic. They not only disrupt; they undermine 16 and frighten and change the core experience of the 17 individual, exacting a toll that ranges from 18 difficulty coping and functioning to total 19 disability. 20 I think some key characteristics that 21 should be taken into consideration are that 22 narcolepsy is not well understood or accepted. 23 People who suffer from this suffer alone. They 24 don't have generally the benefit of support groups, 25 even though there is a fine support organization 196 1 out there, but the people are just spread out. 2 There isn't enough concentration. Most people with 3 narcolepsy do not have a relative with the disease 4 such that it is even strange to them. People 5 suffer a double blow because it is thought their 6 sleepiness is volitional and a sign of laziness. 7 Thus, I think it should come as no 8 surprise that people with narcolepsy suffer from a 9 high rate of depression. It has been estimated 10 from 30-70 percent in various studies. The good 11 news is that in one study health quality of life 12 was improved through effective administration and 13 medical treatment, and I think that would pertain 14 as well to sodium oxybate. 15 In sum, the National Sleep Foundation 16 believes that narcolepsy exacts an unusual and 17 cruel toll. We really call upon this panel to 18 continue to do the professional job that brought 19 you here today and fully consider the personal, 20 psychological, emotional and human aspects of this 21 disease and the great need for an effective 22 medication. Thank you. 23 DR. KAWAS: Thank you, Mr. Gelula. The 24 next speaker is Ms. Abbey Meyers, who is president 25 of the National Organization for Rare Disorders, 197 1 Inc. 2 MS. MEYERS: The National Organization for 3 Rare Disorders, which is known as NORD, came 4 together initially because voluntary agencies for 5 many rare diseases worked together to pass the 6 Orphan Drug Act. So, we are the orphan drug folks 7 who work to monitor the development of these drugs. 8 I have several conflicts of interest with 9 this drug because for 20 years I begged practically 10 every company I ever met to pick up this drug and 11 to adopt it. It is a 20-year saga. And, I wrote 12 something for you that you will be able to read 13 about the history of development of the drug. 14 Also, about a year ago I bought some stock 15 in this company. If I wanted to make money I would 16 have put it in Merck, but the idea with the drugs 17 that they are developing is I feel I have to make 18 my own personal investment in the survival of the 19 company. 20 For this drug FDA, rightfully, has asked 21 for a risk management program, and there are 22 several really good models to look at, most 23 notably, I would like you to remember when you are 24 discussing the risk management what happened with 25 Clozaril because when Clozaril first got on the 198 1 market with the drug for schizophrenia, they had a 2 very stringent distribution program, and they were 3 sued by 30 states, attorneys general, because the 4 laws in those states said that you could not 5 restrict the distribution. In the settlement of 6 that case, the federal court assigned us, NORD, 7 with the task of distributing the drug to the 8 people in this class action settlement. 9 So, I am very sensitive to what happens. 10 FDA approved Clozaril's distribution program but 11 then the law said that they couldn't do it. So, 12 people really want the freedom to be able to get 13 the drug when they want it, when their doctor 14 prescribes it. 15 The other program you should look at is 16 thalidomide because it is an extraordinarily 17 important drug, again very orphan. Nobody wanted 18 to go near it because of the liability problem. 19 But they have a wonderful distribution program and 20 I think it should be a good model for the field 21 when there are drugs with specific dangers 22 involved. 23 I also want to give you several cautions. 24 Don't make the distribution too restrictive. For 25 example, don't allow just certain specialists to 199 1 prescribe it because people with narcolepsy have a 2 great deal of travel problems. Many of them don't 3 have driver's licenses in many states. They may 4 hold on to their driver's license but actually if 5 it was ever reported to the state that they had 6 narcolepsy they would lose it. It is just like 7 epilepsy. So, you have to be sensitive to that. 8 There are many current problems with 9 Ritalin and Dexedrine and the amphetamines that 10 they are using because the government limits the 11 amount of manufacture every year. So, at the end 12 of the year they run out of drug and there are 13 times when they just aren't able to fill their 14 prescriptions and they can't order it by mail order 15 because it is a controlled substance. So, these 16 people have suffered so tremendously because of 17 these distribution problems. With those drugs, 18 pharmacies don't stock a sufficient amount and they 19 will only dispense one month at a time. 20 Don't require a distribution program that 21 is going to cause legal problems. So, ask yourself 22 that, whether the program that has been designed by 23 Orphan Medical could be loosened up a bit. 24 The other thing goes back to what you were 25 talking about this morning, labeling. You know, 200 1 does this drug help with daytime sleepiness, etc.? 2 I want to caution you that if you label this drug 3 just for cataplexy with no effect on daytime 4 sleepiness, there are a lot of insurance companies 5 that are not going to reimburse for it. So, 6 labeling on a drug is extraordinarily important to 7 patients because of the managed care insurance 8 system. So, try to be as liberal as you can on 9 that, thinking about whether insurance companies 10 are going to say no, except to just people with a 11 particular type of narcolepsy. 12 Also, recognize that it is a unique 13 disorder that is just as crippling as epilepsy, and 14 that these people are already paying a very heavy 15 price because of the problems they have with their 16 current drugs. 17 Illegal use has to be handled, which I 18 know that you are going to do, but you must pay 19 attention to the valid use of this drug. Thank 20 you. 21 DR. KAWAS: Thank you, Ms. Meyers. You 22 are the first one who hasn't used all of your time 23 and that is greatly appreciated. The next one is 24 Robert L. Cloud, from the Narcolepsy Network. 25 MR. CLOUD: Good afternoon, and I wish to 201 1 thank the committee for the opportunity to address 2 you on this issue. My name is Bob Cloud, and I 3 would like to briefly talk to you, first about my 4 own long, personal use of Xyrem, and I will call it 5 Xyrem not GHB or sodium oxybate and, secondly, my 6 very serious concerns as director of Narcolepsy 7 Network, which is a national non-profit, primarily 8 patient organization. In that capacity we have 9 received funds, a minor amount of funds, perhaps 10 ten percent of our revenues, from Orphan Medical 11 over the last several years. 12 First, my personal experience with Xyrem 13 as a narcolepsy patient with cataplexy. I am 57 14 years old, married, have two adult children, and I 15 am an attorney in private practice, primarily 16 family, probate and criminal law which sometimes 17 can be intense and have a few emotions attached to 18 it. 19 I believe I am the first American to have 20 used Xyrem for narcolepsy, and I am probably the 21 longest continuing user of Xyrem which now 22 approaches 19 years every night without fail. My 23 narcolepsy/cataplexy symptoms began in the mid-30's 24 and by age 39 included severe and recurring 25 cataplexy together with excessive daytime 202 1 sleepiness and sudden sleep attacks. My cataplexy 2 caused numerous daily episodes of complete body 3 collapse, such that I couldn't leave my office or 4 home without risk of harm to myself or others. 5 Feeling any emotion, humor, anger or mere 6 enthusiasm, would result in sudden immediate 7 collapse. I guess we are all ignorant of what 8 diseases feel like that we don't have them, but my 9 best description of the sudden collapse of 10 cataplexy would be to imagine a puppet on strings 11 and suddenly the strings, which are your muscle 12 tone, are immediately let go and so you fall to the 13 ground immediately, and your head comes down last 14 and whips against whatever -- sidewalk or table 15 corner or escalator or whatever might be there. I 16 have been rescued by police and emergency squads 17 and life guards and well-meaning strangers and 18 friends. 19 Obviously no injury for me has been fatal 20 because I am here, but unfortunately I do know 21 others whose fall has occurred at the top of the 22 stairs and they fell down backwards and killed 23 themselves, and there are others that I don't know 24 about. 25 In 1982 my treating physician sent me to 203 1 Sunnybrook Medical Center in Toronto, Canada to 2 begin prescriptive use of Xyrem under the research 3 being conducted by Dr. Mortimer Mamelak. After 4 three weeks I returned home and continued using 5 Xyrem, always prescribed by my local physician 6 under his own individual investigational new drug 7 application. My severe cataplexy symptoms 8 disappeared almost overnight. I was immediately 9 able to return to my full-time law practice and I 10 have continued to this day to use Xyrem under that 11 individual application and subsequently in the 12 clinical trials under the Orphan Medical 13 application. During these 19 years, I have never 14 changed the dose. I have never experienced 15 tolerance. I have never noted side effects. 16 Simply stated, the drug is as safe and effective as 17 it was on day one. It is hard to imagine a 18 pharmaceutical product having such a quick, 19 complete, safe and enduring benefit. 20 As director of Narcolepsy Network, I have 21 said on a number of occasions that I think the 22 greatest tragedy in the treatment of people with 23 narcolepsy is that Xyrem or GHB has not been 24 available so that other patients could benefit from 25 it as I have. Hopefully, this committee will 204 1 remedy that. 2 We are sensitive to the reports of 3 injuries and deaths and other victimizations from 4 the abuse of GHB and, as an organization, we work 5 with law enforcement and community drug agencies to 6 partake in their activities to limit that and 7 correct that. I think it is obvious that Orphan 8 Medical is going above and beyond the call of duty 9 to also contribute to restricting the unlawful use 10 of GHB. 11 In closing, I submit that our concern for 12 patients with narcolepsy should receive your 13 highest considerations so that people can 14 rediscover their economic and particularly their 15 family lives and avoid disability. Thank you. 16 DR. KAWAS: Thank you, Mr. Cloud. The 17 next speaker is Cindy Pekarick from Pennsylvania. 18 MS. PEKARICK: Hi. My name is Cindy 19 Pekarick, and I am here today to tell you how GHB 20 killed my daughter. In October of 1998, my 21 daughter Nicole, a college student and gym 22 enthusiast met a new boyfriend who introduced her 23 to a product called Renewtrient. In November she 24 researched the product over the Internet and 25 received only positive information. She could take 205 1 it before bedtime and wake up in only four hours 2 feeling refreshed, well-rested, and all her muscles 3 would be completely recovered and ready for another 4 workout. 5 In December I found out she was taking 6 this supplement. I didn't believe the promises 7 made by the advertisers. Arguments ensued and she 8 promised she wouldn't drink it anymore. She was 9 away at school from mid-January until April. 10 In April she returned home. She was 11 behind in all her bills. She was black and blue on 12 her arms and legs. She stopped attending classes, 13 and she kept losing things. In May I discovered 14 she had essentially dropped out of school. 15 In June I could see mild changes in her 16 behavior. She began taking power naps, as she 17 called them. She would sleep three hours in the 18 middle of the day and get up at four o'clock and go 19 to work. She continued losing things and having 20 difficulty paying her bills. I searched her room 21 and car but found no evidence of substance abuse. 22 By July, my younger daughter, Noelle, 23 informed me that Nicole was having problems. She 24 said, "mom, she isn't taking anything bad or 25 illegal. She takes a muscle supplement that 206 1 doesn't agree with her. Sometimes she has bad 2 reactions and she doesn't even know it. She 3 embarrasses herself and me when she acts weird and 4 then goes to sleep. When she awakes she never 5 remembers anything that she did. She started 6 taking it once in a while so she could go to sleep 7 right away after work when she got home. Then she 8 started using it more often. It disgusts me to see 9 her out of control." 10 It was at this time I discovered Nicole 11 had been taking GHB since November. I then began 12 my own search over the Internet for more accurate 13 information. In August, Nicole was found having a 14 seizure in a public bathroom. She had urinated and 15 defecated on herself while pulling at her clothes 16 and hair and flailing her arms. She was rushed to 17 the hospital where we arrived to find her 18 unconscious, intubated, with her arms, legs and 19 waist strapped to the bed. They claimed her 20 seizure was violent. She barely had a pulse when 21 they found her. 22 It was at this time I knew my daughter was 23 addicted to whatever she was taking. There is 24 absolutely no other reason why a young, bright, 25 healthy woman would take a supplement that was 207 1 harmful. I begged the doctors to transfer her to a 2 treatment center for chemical dependency, but they 3 said they wouldn't do it without the patient's 4 permission. She was clueless as to why she was 5 hospitalized and she had no recall of anything that 6 happened to her. She was discharged. 7 In September, Nicole, sweating profusely, 8 with a red face and shaking hands while crying 9 said, "mom, I have to talk to you. I'm really 10 scared. I have a problem. I can't stop drinking 11 it." I stood up, wrapped my arms around her and 12 hugged her as hard as I could. I told her that she 13 was on her way to getting better, that 14 acknowledging that "g" had a hold on her was a step 15 in healing. 16 On Monday morning, on her way to the 17 treatment center, Nicole refused to go in. She 18 claimed that "g" wasn't addictive; that she did the 19 research and she was just having reactions to it. 20 She said she was now in control of her life and 21 future. She stayed in counseling and, by the end 22 of September, Nicole had applied, transferred, and 23 was accepted at the university. She was excited. 24 Things seemed okay on the surface but she was 25 hiding tremors, hallucinations and insomnia. She 208 1 went days without sleeping but never told me. 2 On October 3, 1999 at 2:00 p.m. she said 3 she needed to take a nap before she went to work 4 since she hadn't slept the night before. She set 5 the alarm for 4:00 p.m. but she never heard it. 6 She was in her final sleep. My firstborn child was 7 found in bed, blue, at 6:00 p.m. We found a bottle 8 of GHB in the trunk of her car. The autopsy 9 revealed she had GHB and GBL in her system at the 10 time of her death. No other chemicals were found. 11 Nicole was an honor student, captain of 12 two varsity teams and graduated third in her class. 13 For her undergraduate studies she majored in 14 biology, with a plan to major in engineering for 15 her master's degree. Her ultimate goal was to 16 become a biomedical engineer. She wanted to be 17 able to design body parts to help extend people's 18 lives. She understood that to function well, one 19 had to be healthy. She was a loving, sensitive, 20 caring and intelligent woman. Her only fault was 21 that she was naive. Thank you. 22 DR. KAWAS: Thank you, Mrs. Pekarick. The 23 next speaker is Eric Strain. Doctor Strain is from 24 the College on Problems of Drug Dependence. 25 DR. STRAIN: Thank you. I would like to 209 1 thank the FDA and the members of the Peripheral and 2 Central Nervous System Drug Advisory Committee for 3 providing me the opportunity to speak. My name is 4 Eric Strain. I am a professor in the Department of 5 Psychiatry at Johns Hopkins University School of 6 Medicine. I am a board-certified psychiatrist with 7 qualifications in addiction psychiatry, and I am 8 here today representing the College on Problems of 9 Drug Dependence, CPDD. 10 The College is the leading organization of 11 drug abuse scientists in the United States. I am 12 also the former chairman of the FDA's Drug Abuse 13 Advisory Committee. I have sponsored my own travel 14 to today's meeting, and I have no relationship with 15 Orphan or other pharmaceutical companies that make 16 narcolepsy products. 17 There are two point that I would like to 18 make during these brief comments. The first is 19 that the College on Problems of Drug Dependence 20 would like to emphasize the importance of 21 science-based assessments of new medications, 22 especially as they relate to issues such as abuse 23 liability evaluation and safety of abused products. 24 The College wishes to stress the long history that 25 has led to the establishment of reliable and valid 210 1 methods for determining abuse potential. This work 2 includes both preclinical as well as clinical 3 studies. Several academic medical centers contain 4 rich experience in this area of research. Methods 5 have been well tested, and outcomes from previous 6 studies have helped inform and guide agencies such 7 as the FDA in making determinations regarding abuse 8 potential, therapeutic efficacy, and safety of new 9 medications. 10 CPDD has played a key role in such 11 matters, as its members are the primary group that 12 have conducted such studies. The College wishes to 13 strongly and forcefully advocate that decisions 14 made by the FDA grow out of and be based upon 15 well-conducted research, and whenever possible 16 decisions should be derived from well-controlled 17 studies and data driven. In order to achieve such 18 goals, advice on substance abuse related matters 19 should be solicited from experts in the field. 20 The second point I would like to make has 21 to do with the Drug Abuse Advisory Committee. As 22 the former, and the last chairman of this advisory 23 committee of the FDA, I believe it is important for 24 me to comment upon its termination. The Drug Abuse 25 Advisory Committee has been dissolved by the FDA, 211 1 and in the process the FDA has lost an important 2 resource that can inform decisions regarding 3 substance abuse. To my knowledge, today's meeting 4 is the first FDA advisory committee meeting since 5 this termination where issues of drug abuse are an 6 important element in your discussions. 7 I am pleased to see that there are several 8 drug abuse experts represented here today, however, 9 I am concerned that the numbers do not allow the 10 breadth of expertise that would have been found on 11 the DAAC. Such breadth is essential to fully 12 consider all of the issues involved in advising the 13 FDA on the abuse potential of new medications, the 14 extent of the public health consequences of such 15 abuse, additional data that the FDA should require 16 companies provide, and recommendations regarding 17 post-marketing surveillance. 18 The College is particularly concerned that 19 comparable experience and knowledge brought to the 20 Drug Abuse Advisory Committee by experts in the 21 drug abuse field is no longer readily available to 22 the FDA. In my experience as chairman of the 23 committee, I was able to witness firsthand on 24 repeated occasions the value of having a group of 25 scientists and clinicians who could provide 212 1 informed knowledge and experience to the FDA on 2 matters such as those that appear to be on today's 3 agenda. 4 The loss of the DACC to the FDA is 5 significant and substantial, and adequate 6 representation of drug abuse issues on other 7 advisory committees needs to be clearly 8 demonstrated by the FDA. I speak on behalf of the 9 College in expressing the College's continued 10 concern regarding the dissolving of this advisory 11 committee. Given the tragic consequences of drug 12 abuse to our society, as exemplified by the 13 previous speaker, its prevalence and the growing 14 body of medications for the treatment of substance 15 abuse disorders, it is particularly concerning that 16 the FDA has decided to terminate this particular 17 advisory committee. 18 Again, I wish to thank the FDA and this 19 advisory committee for allowing me to make these 20 comments today. The hope of the College is that 21 these companies will spur tangible demonstration of 22 FDA's commitment to having adequate outside input 23 by experts in the drug abuse field in the advisory 24 committee process either through the renewal of the 25 Drug Abuse Advisory Committee or through adequate 213 1 and substantial representation by drug abuse 2 experts on other advisory committees where issues 3 of drug abuse may be of substantial importance. 4 Thank you. 5 DR. KAWAS: Thank you, Dr. Strain. The 6 next speaker is Deborah Zvorsec. Dr. Zvorsec is 7 from Hennepin County Medical Center in Minnesota. 8 DR. ZVORSEC: Thank you very much. My 9 research is in the area of gamma hydroxybutyrate 10 abuse toxicity, addition and withdrawal. Dr. Steve 11 Smith and I, with others, published a case series 12 in Morbidity and Mortality Weekly Report in 13 February of '99 that described adverse events due 14 to ingestion of dietary supplements containing GBL, 15 GHB precursor. I was the lead author of a case 16 series of 1,4 butanediol toxicity that was 17 published in The New England Journal of Medicine in 18 January 2001. These toxicity episodes included two 19 deaths that occurred with no co-intoxicants and no 20 evidence of aspiration or asphyxiation or 21 adulterants. 22 I will focus today on GHB addiction. In 23 the course of our work, Dr. Smith's and my name 24 were listed on the project GHB help site. We 25 received calls from over 40 addicted patients from 214 1 25 states, and have treated an additional 5 cases 2 of inpatient withdrawal at HCMC in Minneapolis. 3 The vast majority of these addicted people 4 began using GHB to treat insomnia, anxiety, 5 depression, chemical dependence or for 6 body-building purposes, as recommended by 7 marketers, websites and fringe pro-GHB physicians. 8 Many, indeed, began with GHB, continued with GHB 9 and never used any of the dietary supplement 10 analogs. Our patients began with small doses, 11 often only at night, and discovered that it made 12 them feel good; increased dosing frequency and, as 13 tolerance developed, needed more GHB in order to 14 feel good. Within months, they were taking GHB 15 every one to three hours around the clock to avoid 16 withdrawal symptoms. By the time they realized 17 that they might be physically dependent, attempts 18 to abstain resulted in severe anxiety, insomnia, 19 panic attacks and hallucinations. 20 Their addiction destroyed their lives. 21 They lost their spouses. They lost access to their 22 children, their jobs. They acquired tremendous 23 debt to support their habit. They became comatose 24 while driving and crashed their cars, frequently on 25 multiple occasions. They called us in absolute 215 1 desperation. Their detox was frequently similar to 2 the worst cases of delirium tremens, requiring 3 large and often massive doses of sedatives, often 4 with intubation. 5 Almost all patients suffered weeks or 6 months of profound depression and anxiety after 7 detox, and some also experienced muscle twitching 8 and tremors. Of the over 40 patients we have 9 worked with, only a scant handful have remained 10 GHB-free, frequently despite CD treatment. Many 11 have detox'd numerous times but continue to 12 relapse, sometimes within hours of discharge from 13 treatment. Unfortunately, many never lost faith in 14 GHB and continued to be convinced that they could 15 get back on it and use it responsibly. They 16 continue to argue its health benefits. 17 One of our patients was a 50-year old 18 businessman with his own business who began using 19 GHB, not an analog, five years ago, initially for 20 body-building purposes. Within months he had 21 increased his dosing to around the clock. His life 22 was entirely controlled by the need to have GHB 23 with him at all times. He tried numerous times to 24 quit. His wife was unaware of his addiction. She 25 described witnessing frequent frightening hypnotic 216 1 states, punctuated with clonic movements. She 2 believed that his frequent states of apparent 3 somnambulism were due to a sleep disorder but 4 despaired when a sleep specialist could not cure 5 him. This woman is a very bright professional who 6 was totally unaware of GHB, as is the case with 7 many family members. It was only on the morning of 8 his admission that she learned the truth. After 9 six days of detox he was through the worse and 10 appeared to be on the road to recovery. 11 Psychiatrists treated him with sleeping meds and 12 antidepressants, but within three days he was using 13 GHB again to control anxiety attacks and 14 depression. 15 GHB is perhaps the most addictive drug 16 ever abused. Experienced drug users describe a 17 euphoria that surpasses that of any other drug. 18 Availability of off-label prescription presents 19 profound personal and public health risks. The 20 fringe physicians who now promote GHB will be 21 joined by thousands of mainstream physicians with 22 the approval of the FDA. The majority of 23 physicians are ignorant of the diverse health risks 24 of GHB, as are toxicologists and law enforcement 25 officials. Users will seek Xyrem from physicians 217 1 who don't recognize sodium oxybate as GHB and are 2 unfamiliar with the health risks. Patients will 3 obtain prescriptions for sleep disorders, also for 4 insomnia, depression, anxiety, treatment of alcohol 5 and drug dependence and other conditions for which 6 it has been touted. 7 We know that addicts often use GHB and its 8 analogs interchangeably. Their compound of choice 9 is dependent on access, determined by cost, 10 perceived quality, ease of procurement. Clinical 11 literature reports one user who spent $200 per day. 12 That comes to $70,000 per year. Our patients 13 report ingestion of up to a bottle every one to two 14 days, coming to $11,000 to $36,000 per year. A 15 Xyrem prescription will be a bargain for such 16 users, who will then avoid the high prices, erratic 17 availability and risks of supplement and solvent 18 purchase. We know that many people are afraid to 19 buy or make their own GHB due to risks of 20 contamination or errors of production. Xyrem, a 21 pharmaceutical product of controlled quality, 22 available by legal prescription, and with very 23 little risk if found in their possession, will be 24 very attractive. We know that users are watching 25 for the release of Xyrem. Recreational drug sites 218 1 post links to narcolepsy sites and publications 2 about Xyrem. One hotyellow98.com, for example, 3 instructs users "click here to find out when GHB 4 will be released under the trade name of Xyrem." 5 DR. KAWAS: Your time is up, Dr. Zvorsec. 6 Please finish. Thank you very much, Dr. Zvorsec. 7 Our nest speaker is Trinka Porrata of California. 8 MS. PORRATA: I wish I had time to tell 9 you the stories of 200 dead people that I know of, 10 hundreds of rape victims and thousands of GHB 11 overdoses, and About Caleb Shortridge, to whom our 12 website www.projectghb.org is dedicated, about 13 Matthew Coda and Joshua Parks to whom our GHB 14 addiction hotline is dedicated. I wish I could 15 tell you about Ben Croman, Mike Fox, Tyler Johnson 16 and other young men from New Zealand to Sweden who 17 either have or are right now considering suicide 18 because of the withdrawal from this drug; about 19 more than 300 people I personally know about who 20 are horribly addicted to GHB, and who could each 21 name at least one dozen people more just like them. 22 I have lived and breathed GHB since June 23 of 1996 when I was first assigned to handle it for 24 the LAPD. Four young men collapsed. Two literally 25 died and were brought back to life by the 219 1 paramedics. One thing was clear, people were dying 2 from GHB and it was being missed. It has been a 3 heartbreaking five years, mixed with the privilege 4 of learning more and teaching others to recognize 5 the rape, overdose and deaths and getting rape 6 victims into treatment and addicts help. It has 7 been very lonely at times when the agencies who 8 should care haven't. 9 DEA has reviewed and documented 71 deaths 10 related to GHB but, basically, stopped counting 11 once the drug was controlled, for obvious reasons. 12 No one at FDA has ever expressed interest in these 13 cases. My database now includes over 200 14 GHB-related deaths. In fact, Robert McCormick, of 15 the FDA's Orphan Drug Unit, told me emphatically he 16 did not care how many people had died nor were 17 addicted to it because he intended to approve it 18 anyway. Something is wrong with this picture. 19 This is the most horrid drug I have encountered in 20 25 years as a police officer. 21 Much new has come to light during the past 22 two years, none of it good. Around the world 23 countries are just now awakening to their problems 24 with GHB. Schedule IV by WHO is simply an 25 awakening to the problem. As we speak, countries 220 1 are restricting it. France is backing away. 2 England is struggling with it. Sweden has an 3 unrecognized addiction and suicide problem. New 4 Zealand tried it as a prescription drug and now 5 realizes they screwed up royally. NIDA is 6 currently releasing $2 million in research on this 7 drug. This is not a time to be pushing it forward 8 on an unsuspecting American citizenry. 9 You are here today to approve GHB, 10 disguised as sodium oxybate, for use with 11 narcolepsy/cataplexy. Orphan's investors have been 12 assured that you will do so. When the last meeting 13 was cancelled the stock dropped 30 percent in 14 frustration over it. You have not seen my 15 videotapes of the day-to-day struggle of GHB 16 addicts showing that GHB clearly gives previously 17 healthy people symptoms that can only be described 18 as temporary narcolepsy/cataplexy, just like the 19 nine-year old you saw in the tape. Their heads 20 ricochet off board room tables around this country. 21 They break mirrors. They are cut up. They crash 22 cars. They die and kill others. It is destroying 23 them. Their wives are terrified of their husbands 24 and have no idea what is happening. They are 25 locked in psychiatric wards because doctors and 221 1 emergency rooms do not recognize GHB psychotic 2 episodes. 3 There are no answers for them. So, how 4 can you approve this drug for use? My addicts 5 suffer alone, much as narcoleptic/cataplectic 6 patients do. Many do not have insurance or their 7 insurance will not pay for this drug that is not 8 recognized as an addictive drug. 9 I am deeply concerned about the off-label 10 use policy, enabling any doctor ultimately to 11 prescribe it for any condition as I have no faith 12 that its use will be limited to 13 narcolepsy/cataplexy. Look at the chatter around 14 Orphan about fibromylagia, a condition with vague 15 symptoms for which a drug seeker could easily get a 16 prescription. I know the vast majority of doctors 17 do not realize that sodium oxybate, Xyrem, is GHB. 18 I see no significant talk on the legitimate 19 narcolepsy websites about it, but the message 20 boards where GHB addicts hand out are buzzing. In 21 fact, the key figures in illegal GHB Internet sales 22 were posting on the website www.xyrem.com. 23 There is very little drug diversion 24 enforcement in the United States. Only a handful 25 of agencies devote any time to this. It is a small 222 1 portion of DEA effort. States are not prepared. 2 They are not able to handle it. Therefore, 3 Orphan's proposed voluntary -- key word, voluntary 4 -- promises of distribution are frightening. 5 More importantly, the issue goes beyond 6 diversion of Orphan's product to use of Orphan as a 7 shield for possession of GHB in general. It would 8 be unrecognized by law enforcement. Once in 9 possession of that prescription and a bottle of 10 Xyrem, the addict will be home free. There is no 11 field test kit. All investigations of GHB are 12 difficult. Encountering a prescription, real or 13 counterfeit, and a bottle of Xyrem, real or 14 counterfeit, the officer would have zero ability to 15 identify it -- none; zero; nada. 16 To those who claim real GHB is safe and 17 only street stuff is dangerous, poppycock. My 18 addicts have used everything from European 19 pharmaceutical grade to bad stuff. The 20 unprecedented split scheduling of GHB was unwise 21 and unenforceable. We were forced to accept it. 22 It was political, not science. The people in the 23 clinical trials have reason to obey; people in the 24 streets do not. 25 If I were to convey to you but one 223 1 thought, it would be that not enough information is 2 known about GHB to approve it for any purpose at 3 this time, and certainly not appropriate for 4 off-label use. Any approval at this point will 5 trigger an absolute further epidemic of general 6 abuse because you will create an aura that it is 7 safe. I ask you please table this issue until the 8 NIDA research comes in. Please do not make this 9 mistake. 10 DR. KAWAS: Thank you, Ms. Porrata. Our 11 next speaker is Matt Speakman from West Virginia. 12 While Mr. Speakman is coming up, I just want to 13 remind everybody I am not trying to be mean; I am 14 not trying to be difficult, but we are trying to 15 keep the public hearing section of this meeting 16 down to under two hours and that will only happen 17 if everyone sticks to their five minutes. We would 18 like to let the committee get a chance to have some 19 more discussions for everyone. So, we greatly 20 appreciate honoring the time constraints. Mr. 21 Speakman? 22 MR. SPEAKMAN: Thanks. I just wanted to 23 say thanks. This is kind of a unique experience 24 addressing doctors. It is really cool. 25 My name is Matt Speakman and I have 224 1 narcolepsy. I will describe very briefly my 2 experience. I have cataplexy also. My first 3 experience was in chemistry class my junior year in 4 high school. The professor pulled out the liquid 5 nitrogen experiment and was freezing flowers and 6 flicking them, making them shatter. I got very 7 excited and he called us to the front of the room 8 and, on my way up to the front of the room, I felt 9 my legs start to buckle. This was the first time 10 anything like this had happened. I had had trouble 11 laughing a little bit because cataplexy sometimes 12 has onset with laughter and emotion, but it wasn't 13 very serious. 14 I eventually just realized that I was 15 going to fall. So, I went back to my desk and 16 collapsed on the desk with my face down in my arms, 17 kind of draped over the thing. It was humiliating. 18 I couldn't move. I was awake and aware and I could 19 still hear the class kind of looking around and 20 what-not. 21 This started to happen more regularly and 22 I started to fall asleep during class. My grades 23 started slipping. I had to stop swimming. I was 24 on the swim team. Falling asleep in the pool is 25 kind of dangerous. So, I quit doing that. Most of 225 1 my teachers suspected drug use and I don't blame 2 them. 3 But I managed to get into the University 4 of Kentucky and I went there for a year. I was 5 unable to meet friends and my grades weren't very 6 good because I spent most of my time in my dorm 7 room. I didn't make it to class very often; very 8 hard to wake up. It is very hard to keep 9 consistent notes when you are falling asleep all 10 the time. 11 My parents weren't happy so they found, 12 you know, I needed some other treatment. So, I 13 went to a doctor in Cincinnati who was part of the 14 study for what is now Xyrem. That was four years 15 ago, and I am taking it nightly unless I pull an 16 all-night study session or something like that. I 17 don't have any withdrawal symptoms when I don't 18 take it. I don't have any side effects when I do 19 take it. I sleep well. I have no cataplexy. I am 20 here speaking to you right now and I certainly 21 wouldn't be doing this without this treatment. I 22 used to take stimulants and antidepressants to 23 control the cataplexy, none of which worked; they 24 just had nasty side effects. It wasn't very good. 25 Two weeks ago I graduated from West 226 1 Virginia University with honors. I am looking for 2 a job -- 3 [Laughter] 4 -- and I am thinking about going to grad 5 school. That is definitely on the bill, but I am 6 going to need some money first. So, first things 7 first. Right? 8 I understand all the concerns about the 9 illicit use and that definitely needs to be 10 addressed, but this drug is working for 11 narcoleptics and, you know, I have a girlfriend and 12 I have a life, and I live normally. A couple of 13 years ago I got a job as a full-time camp counselor 14 in Maine; drove there myself; had no problems. I 15 read the review they gave me after the summer was 16 up and it said, this guy has the energy of a small 17 power plant, which was nice to hear after suffering 18 from narcolepsy for a couple of years. So, I am 19 happy. I am working on success, and I just wanted 20 to thank you for giving me the time to speak with 21 you and I hope you can work all this thing out, but 22 my main point was that the drug is working for 23 narcoleptics and I want to thank the Narcolepsy 24 Network for paying for my travel arrangements and 25 my hotel. I am not in any way tied to Orphan 227 1 Medical. I don't care who makes it. I just want 2 to let you guys know it is working. Thank you. 3 DR. KAWAS: Thank you, Mr. Speakman. The 4 next speaker is Charles Cichon, president of the 5 National Association of Drug Diversion 6 Investigators. 7 MR. CICHON: Good afternoon and thank you. 8 My name is Charlie Cichon. 9 DR. KAWAS: My apologies. 10 MR. CICHON: No apology. The nuns never 11 got it in grade school; nobody has ever got it 12 right. I go everywhere from Ceechon to Chicken. 13 [Laughter] 14 I have a 16-year background in law 15 enforcement, but for the last 12 years I have 16 worked in the health regulatory field with the 17 Maryland Board of Physician Quality Assurance, the 18 state medical board licensing and regulatory agency 19 for Maryland. But I am here today as the president 20 of the National Association of Drug Diversion 21 Investigators. 22 Established in 1987, the National 23 Association of Drug Diversion Investigators, NADDI, 24 was formed in Maryland, in Annapolis by a sergeant 25 in the Ann Arundel County police department. Our 228 1 organization is a unique organization whose members 2 are responsible for investigating, prosecuting and 3 preventing pharmaceutical drug diversion. 4 NADDI has proven to be a valuable asset to 5 law enforcement, the pharmaceutical industry and 6 health regulatory professionals. NADDI principal 7 activities comprise cooperative education and 8 training in the specifics of pharmaceutical drug 9 diversion, investigation and prosecution; the 10 sharing of investigated information and 11 communication with a wide variety of interested 12 parties with regard to the nature, scope and impact 13 of pharmaceutical drug diversion; and the 14 development of stronger effective measures to 15 combat the problem of pharmaceutical drug 16 diversion. 17 NADDI supports the safety and efficacy of 18 the new drug application, NDA 21-196, Xyrem, 19 proposed to reduce the incidence of cataplexy and 20 to improve the symptoms of daytime sleepiness for 21 persons with narcolepsy. 22 NADDI is aware that in many reported cases 23 the use of GHB has changed from homemade GHB to 24 ingesting of industrial chemicals that convert to 25 GHB in the body. (My car got towed away yesterday; 229 1 I lost my other glasses. I noticed that when I was 2 sitting in the back and I couldn't read my paper. 3 So, I apologize.) 4 We are also aware that there are no known 5 cases which involved Xyrem. Rather than consider 6 the above issues as tangential, Orphan Medical has 7 gotten involved, helping to educate and uncover 8 solutions in conjunction with stakeholders such as 9 NADDI. In fact, since November of 2000, an Orphan 10 representative appeared at our national conference 11 in Columbus, Ohio, and for the last several months 12 has been involved in several states in 13 multi-regional training with over 600 NADDI 14 members. 15 Input has been sought regarding 16 distribution systems that will minimize and 17 identify potential diversion situations, allowing 18 diversion investigators to more easily perform 19 their jobs. It is the job of the pharmaceutical 20 diversion professionals to investigate potential 21 diversion, however, Orphan is willing to cooperate 22 with the appropriate local, state and federal 23 agencies. Thank you. 24 DR. KAWAS: Thank you. The next one is 25 Debbie Alumbaugh from Florida. 230 1 MS. ALUMBAUGH: Good afternoon. My name 2 is Debbie Alumbaugh, from Florida, and I am the 3 surviving mother of Michael Tiedemann. He was 15 4 years old when he died. That was just over two 5 years ago. The cause of Michael's death was 6 aspiration vomitus and GHB toxicity. 7 Michael was a sophomore at a high school 8 in Florida. He was a black belt in karate, and he 9 was also an instructor. He had won several 10 academic awards for reading, spelling, mathematics 11 and music. 12 On October 1, 1998, Michael came home from 13 school and asked if he could go to the show with 14 his friends. It was unusual for a school night but 15 we decided to let him go. We required Michael to 16 bring home a progress report every week from school 17 and he had brought one home and he was making A's 18 and B's in all of his subjects. Before they left, 19 one of Michael's best friends came into our home 20 and they shot into Michael's bedroom. This boy was 21 only in there for five minutes and when he left 22 Michael was passing out within ten minutes of this 23 young man leaving our home. 24 We found out 18 months after Michael died 25 that when they left our home they drove three 231 1 blocks and started to play a game of basketball on 2 the way to the show. Michael had the ball and was 3 going for a lay-up, and when he came down he was 4 unconscious. He lay there several minutes. His 5 friends, not knowing what to do or recognizing the 6 red flags, giggled and laughed. They scooped my 7 son up and took him on to the movies. We 8 understand Michael never saw the first five minutes 9 of the movie. He passed out again. 10 When they brought our son home, my husband 11 looked at him and he asked him, Michael, are you on 12 something? Did you take something, son? He said, 13 no, dad, nothing. Brad decided not to lecture 14 Michael this late at night; he would talk to him 15 tomorrow. Brad never got that chance. Michael 16 died that night, alone in his bed. 17 The next morning, when Brad went to wake 18 Michael for school he could hear Michael's alarm 19 blaring. Michael had full intentions of getting 20 up. When he opened our son's door he knew he was 21 dead. The first thought that ran through his mind 22 was to run, run out of the house and not look back. 23 My son was on his bed, his eyes wide open, his 24 mouth hanging open, his tongue swollen so much that 25 my husband couldn't shut his mouth. He had dry 232 1 vomit running down his chin into a puddle on his 2 collarbone. His hands were in a clawed position 3 where he had tried to roll himself over but 4 couldn't. GHB takes away the gag reflexes and it 5 paralyzes you. 6 We didn't know why Michael had died. None 7 of his friends would speak up. It took 12 weeks 8 for us to find out that Michael had ingested GHB 9 that evening. It was the first and only time that 10 this had happened. 11 In the last three years, in Florida alone, 12 we have lost 207 young lives to these drugs. From 13 1999 to 2000 our numbers have more than doubled in 14 Florida alone. 15 After several months after Michael died, 16 he came to his father in a dream and said, dad it 17 is wrong to destroy the body the way I have done. 18 I need you and mom to go out and tell my friends 19 and my generation of people my story, our tragedy. 20 This put a burden on our hearts and we seemed to 21 stop healing until one day Michael's father 22 gathered up enough courage and strength and he made 23 the first phone call. 24 We now go to schools all over and share 25 our story with students about GHB, and the tragedy 233 1 of our family. Friday, June 1 our son would have 2 been 18 and he would have graduated on that day. 3 When we went to his grave one Friday, his 4 graduating class had left white roses and the 5 mascot for the graduation cap. We missed prom; we 6 missed graduation because of this drug. Our voices 7 have to be heard. Please investigate this drug. 8 It is not safe. It is killing our children and it 9 is not the pushers that are dying; it is our good 10 kids that we are losing. Thank you. 11 DR. KAWAS: Thank you, Ms. Alumbaugh. The 12 next speaker is Brian Hunter, of the Young Adults 13 with Narcolepsy. 14 MR. HUNTER: Good afternoon. My name is 15 Brian Hunter. I am the founder of Young Adults 16 with Narcolepsy or YAWN. I am also a medical 17 student at the University of Minnesota and a person 18 with narcolepsy and cataplexy. 19 I would like to preface my comments today 20 by disclosing that Orphan Medical has provided my 21 organization with a minor grant and it provided a 22 general grant to the Narcolepsy Network who has 23 paid for my travel and accommodations to attend 24 this meeting. 25 YAWN is the first youth-focused online 234 1 narcolepsy support and advocacy organization. We 2 work at the grass roots level to advance public 3 awareness of narcolepsy on behalf of young adults 4 and others whose lives are affected by this often 5 debilitating sleep disorder. 6 As founder of YAWN, I believe I am in a 7 unique position to comment on the issue currently 8 under consideration by this committee. I do not, 9 and have not used Xyrem for treatment of my 10 cataplexy but as the representative of many young 11 adults in need of an effective treatment for their 12 narcolepsy, I am compelled to present my views on 13 the risk management issues pertaining to the safety 14 and efficacy of Xyrem. 15 Narcolepsy is most commonly diagnosed by 16 the middle of the third decade of life, often 5-15 17 years after the onset of symptoms, the most 18 dramatic of which is cataplexy. Excessive daytime 19 sleepiness, combined with the impact of sudden 20 attacks of cataplexy that may last from a few 21 seconds to hours can be profoundly damaging to the 22 interpersonal, educational and professional 23 development of these young adults at an extremely 24 critical point in their development. Although I am 25 fortunate only to experience rare and mild attacks 235 1 of cataplexy, I know others who are completely 2 incapacitated by cataplexy and have not, or would 3 not been able to achieve their personal 4 professional goals without a medication like Xyrem. 5 I submit that the risk for experiencing 6 the negative impact of untreated cataplexy on the 7 potential of so many young adults with narcolepsy 8 is a serious issue that must be included in any 9 discussion of risk management of Xyrem. 10 Xyrem offers a singularly important 11 therapy for the 65-70 percent of young adults with 12 narcolepsy who suffer with cataplexy. We must 13 recognize the consequences of failing to approve 14 Xyrem to treat the 1/1000 people suffering with 15 narcolepsy. For example, after forming YAWN, I was 16 contacted by the parents of a 16-year old boy, 17 living in a small town not three hours away from 18 the nearest city. This young man was bright. He 19 did well in school, and was active in his community 20 until his 12th birthday when he began experiencing 21 severe episodes of cataplexy that lasted for hours. 22 When I first spoke to him on the phone he 23 told me that his condition was so severe that he 24 was forced to spend five days a week in a nursing 25 home, and he is still there. What are the costs of 236 1 providing nursing home care in a public institution 2 for a 16-year old boy for the next 60 to 70 years? 3 By not adequately controlling his cataplexy, what 4 are his chances for becoming a contributing member 5 of our society? Unfortunately, this man's story is 6 all too common. Unless something is done about the 7 current environment of limited access to inadequate 8 pharmaceutical therapies, the future of young 9 adults suffering with cataplexy will remain bleak. 10 This, however, does not have to be the 11 case. In fact, a brighter future has been achieved 12 by the lucky few who have participated in Xyrem 13 clinical trials. They have become success stories. 14 To these young adults with narcolepsy Xyrem has 15 meant the difference between a life within an 16 institution and having the opportunity to achieve 17 their goals, free from the physical constraints of 18 their disease. Xyrem has enabled many young 19 adults, my friends, to earn their Ph.D.'s or become 20 lawyers, doctors or to simply be good parents. 21 These are people who took Xyrem and 22 couldn't have succeeded otherwise. Yet, there 23 continue to remain thousands of other talented and 24 capable young adults who have not yet had a chance 25 to fulfill their dreams. They are the reason I 237 1 formed YAWN and why I am here testifying before you 2 today. We can no longer afford to neglect the 3 potential of so many young adults by failing to 4 provide them with the only medication known to be 5 safe and effective. It is our responsibility to 6 protect their right to pursue a happy and 7 productive life by having access to medications 8 like Xyrem that will effectively treat their 9 disease. 10 Thank you for allowing me to present these 11 remarks to you today. I urge you to approve the 12 NDA for Xyrem. There really are lives at stake. 13 DR. KAWAS: Thank you, Mr. Hunter. The 14 next one is Joe Spillane. 15 DR. SPILLANE: I would like to also say 16 thank you for an opportunity to speak to the FDA 17 and to this committee on this important issue. 18 I work at Broward General Medical Center 19 which is a community hospital in south Florida. My 20 experience with GHB is as a pharmacist and in 21 clinical toxicology. I also teach as an associate 22 professor at the College of Pharmacy at NOVA 23 Southeastern University. 24 Our experience in the emergency department 25 is very similar to what Dr. Dyer mentioned. We 238 1 have a lot of GHB overdoses. We had 48 overdoses 2 associated with GHB in 1999. That number increased 3 by 60 percent to 77 in 2000. We have more GHB 4 overdoses than ecstasy. We have more GHB overdoses 5 than oxicondon. I think it is important that I 6 just underscore the immensity of the problem 7 associated with GHB abuse. Most of our overdoses 8 come in with people who have altered mental status 9 and, basically, they just need a short period of 10 supportive care, airway management. Most wake up. 11 Many of them -- and I think this is important to 12 point out, many of them mention that somebody had 13 given them GHB, put it into their drinks, and so 14 forth. As such, the media an many people have 15 advised don't accept a drink from anybody but the 16 bartender. We had a bartender up in our ICU about 17 a month ago, and when he did recover I spoke with 18 him and he said, yes, I chronically use GHB. A lot 19 of my friends in the beverage industry also do. 20 And, I think we can understand what the potential 21 problems could be with that. 22 We have also treated five withdrawal cases 23 and, again, the numbers might not be that big but 24 this is just one hospital and, since it is a 25 difficult thing to identify, we are probably 239 1 missing cases and I am sure there are cases missed 2 throughout the country. 3 There have been nine deaths where, in the 4 estimation of the medical examiner in Broward 5 County, a county of 1.6 million people -- nine 6 deaths were caused by GHB and I think it is 7 important to point out that at least one of those 8 deaths was with GHB alone, with no co-intoxicants 9 and no alcohol level. 10 I guess my major concerns are with the 11 scheduling and some of the off-label prescribing 12 issues, and the voluntary nature of this 13 distribution system. I kind of just want to 14 summarize briefly by saying I think there are four 15 questions that are major concerns of mine and I 16 hope this committee addresses those concerns. 17 The first one is, is it really wise to 18 rely upon an essentially voluntary, supposedly 19 closed-loop distribution system, designed by the 20 manufacturer, to prevent diversion of an 21 increasingly popular, highly lethal, addictive and 22 abused substance? 23 My second question is, is it prudent to 24 require very little governmental regulatory 25 oversight of such a system when the strict 240 1 adherence to that system may not be in the best 2 financial interest of the entity responsible for 3 that strict adherence? 4 My third question is, is it responsible to 5 rely solely on those with a vested interest in 6 demonstrating little or no diversion to verify that 7 little or no diversion is occurring? It is my 8 understanding that that is essentially what we may 9 be doing here. I think there was an example of how 10 this could be problematic just in today's 11 proceedings. I certainly was under the impression 12 by several people who spoke today that there was no 13 diversion in the clinical trials. I think Dr. 14 Mani, from the FDA, said that, indeed, there were 15 some cases of diversion. So, I just think that is 16 a potential concern. 17 My fourth question is does it demonstrate 18 judicious foresight to establish a precedent for 19 sort of circumventing existing scheduling and 20 distribution processes, and couldn't such a 21 precedent be used in the future to the financial 22 benefit of pharmaceutical manufacturers and to the 23 detriment of drug diversion prevention? 24 I would like to commend Orphan for their 25 work and bringing a medication that they feel is 241 1 effective to those who could benefit from it. I 2 think a mandatory, not voluntary, system of 3 distribution, with adequate governmental regulatory 4 controls and any restrictions on off-label 5 prescribing would advance another one of their 6 stated goals, which is reducing abuse and 7 diversion. Thank you very much for having me. 8 DR. KAWAS: Thank you, Mr. Spillane. The 9 next one is Ms. Mali Einen. 10 MS. EINEN: Hello, and thank you for the 11 opportunity to speak before you today. I could 12 tell you my story of my scars and bumps and bruises 13 from my many falls from cataplexy, or I could tell 14 you about my disappointment from having had to give 15 up my career that I was dedicated to and loved, not 16 to mention the loss of income and security. 17 Instead, the part of my story I share with you 18 today is the loss of the normal, everyday things 19 that most parents take for granted. 20 My name is Mali Einen. I am a single 21 mother from California with narcolepsy and what is 22 considered severe cataplexy -- and a lot of 23 nervousness. As a person with narcolepsy, I was 24 fortunate to be diagnosed fairly quickly after the 25 onset of my symptoms. I was diagnosed at the age 242 1 of 22 after first noticeable systems of narcolepsy, 2 appearing at about age 22. 3 In the early years my cataplexy was 4 triggered mostly by strong emotions -- a truly 5 funny joke or my young daughter saying something 6 adorable. As the years progressed, my cataplexy 7 worsened, requiring less and less of an emotional 8 trigger to cause a complete collapse -- unable to 9 move or talk for seconds, sometimes even minutes at 10 a time despite my daily medications. 11 As my daughter grew and my cataplexy 12 worsened, I was unable to attend her performances, 13 school programs or sports activities without 14 several full collapses. My young, then seven or 15 eight year old daughter would complain, why do you 16 bother to come? You spend most of your time passed 17 out. That is what she called cataplexy. I 18 wondered would she ever understand that it was my 19 joy for her success and my love for her that 20 prevented me from participating in these 21 milestones. 22 Several years later my daughter's simply 23 relaying a story to me, excitedly, about her latest 24 crush or her experiences with her friends would 25 cause me to crumble, much like the film that Dr. 243 1 Mignot showed earlier today. It dawned on me that 2 I had not only given up my experiencing anything 3 that might involve positive emotion, it had become 4 difficult for me to even participate as a spectator 5 in my daughter's life. 6 During the years, I had been able to 7 maintain success in my developing career as a money 8 manager. My workaholic, nose to the grindstone 9 withdraw kept me away from the usual office fun and 10 water cooler moments, while allowing me to avoid 11 embarrassing cataplexy. But this too had begun to 12 erode. Although the various medications allowed me 13 to keep my cataplexy partially in check, it seemed 14 that my nighttime sleep became more and more 15 disrupted, sleepy during the day, yet never able to 16 sleep more than an hour or two at a time at night. 17 By 1996, my spotty nights of a few hours 18 of sleep, my sneaking naps during the work day, and 19 collapsing in exhaustion any time I sat still had 20 affected my ability to continue to perform my job 21 adequately. Long ago my daughter had given up on 22 my being able to read her a story or to help her 23 with her homework. My life had become dragging 24 myself to and from work, attending to the basic 25 needs of my daughter, while constantly working to 244 1 keep my emotions in check. There was little room 2 for fun and interaction. Sole provider for my 3 daughter and myself, I finally voluntarily left my 4 job. 5 By this time I had become a complete slave 6 to my next dose of medication to either control my 7 cataplexy or to help keep me awake. The 8 medications didn't make me feel well; they made me 9 feel horrible, yet, I was their slave. I had never 10 taken a back seat to finding better or best 11 treatment options. I tried no less than five to 12 seven different antidepressants over the years with 13 varying degrees of success, but each with such a 14 cost. 15 Within a year after I had left work, I 16 became aware of a new medical study through 17 Stanford, an experimental treatment for narcolepsy 18 and cataplexy. I started Xyrem. My life changed! 19 After a horrific washout period when, unmedicated, 20 I was faced with my inability to care for myself, 21 let alone my daughter, with mere thought causing 22 collapse after collapse, I found that Xyrem 23 controlled most of my cataplexy and I was thrilled 24 how the better quality nighttime sleep allowed me 25 to feel normal, almost good upon waking. 245 1 Although not required by the medical 2 study, I began to voluntarily decrease my daily 3 doses of amphetamines. The better, less disrupted 4 nighttime sleep allowed me not to be a slave to my 5 next dose of stimulants in order to make it through 6 the next several hours. I now go many days without 7 stimulants at all, and other days take 5 mg or less 8 of Dexedrine. 9 I not only began to be able to listen to 10 my daughter's glee-filled stories of her day, I 11 started to volunteer at her school. I could joke 12 with the kids; I could even watch Kelsey smash a 13 winning serve across the volley ball court. I must 14 admit, occasionally a funny story or my evening 15 interaction with my daughter still causes my facial 16 muscles to slacken with a bob of the head, but my 17 daughter now uses these opportunities to give me a 18 hard time, knowing that I will recover in a second 19 or two and we will have fun and enjoy our life 20 together. 21 I asked my now 17-year old, upon 22 contemplating being here today, would you say my 23 taking Xyrem has made a difference in your life? I 24 had expected the usual teenage disinterested reply. 25 Instead, Kelsey responded, as tears welled in her 246 1 eyes, as much as I hate it sometimes, you are 2 really a part of my life now; you know everything 3 that's going on with me. 4 It is for this that I am truly grateful to 5 Orphan Medical and Xyrem -- and I think I forgot to 6 say my conflicts of interest. 7 DR. KAWAS: That is the only reason we are 8 going to let you go more over time. 9 MS. EINEN: I am a shareholder of Orphan 10 Medical and a number of other stocks of products 11 that I believe in. Narcolepsy Network has 12 generously paid for my air fare and accommodations, 13 but they have not compensated me for my time, nor 14 am I paid for the time away from my brand-new job 15 back in the career which I had to leave five years 16 ago. 17 DR. KAWAS: Thank you, Ms. Einen. Next is 18 Ms. Sandra Jones from California. 19 MS. JONES: Good afternoon, ladies and 20 gentlemen. My name is Sandra Jones, and I am from 21 Los Angeles, California. My travel expenses are 22 being reimbursed by the Narcolepsy Network. I am 23 50 years old. It was only 19 years ago that my 24 mother truly became a mother to me, my brother and 25 sister. Nineteen years ago my mother began taking 247 1 what we now call Xyrem. Within a week after she 2 started taking this medicine we noticed the 3 incredible change in her. She could cook dinner 4 without collapsing to the floor. She could sit 5 down and eat dinner with us without falling asleep. 6 She could make a sound that we hadn't heard in a 7 very, very long time -- laughter, and more laughter 8 without falling to the floor. 9 She became a totally new person to our 10 family. That was not the case nearly thirty years 11 ago. She quit her career as a nurse for fear of 12 how the disease might affect her care of her 13 patients. She became sort of a recluse in her home 14 and we grew used to seeing her sleeping throughout 15 the day and staying up all night. She was afraid 16 she would fall and bring embarrassment to herself 17 and especially to her family. People just did not 18 understand her disease. She once collapsed at a 19 party and people dismissed her as being a drunk. 20 My mother didn't drink. It was what the narcolepsy 21 had done to her. 22 This is an evil, evil disease and unless 23 you have witnessed it firsthand you cannot 24 understand the horrible ways it affects a person's 25 live. Imagine having a newborn child, my sister, 248 1 and not being able to hold her for fear of dropping 2 her. Imagine not being able to go to the grocery 3 store for fear of falling in the aisle. Imagine 4 not being able to read stories to her children 5 because she would fall asleep, not us. Imagine not 6 being able to drive a car for fear of collapsing 7 behind the wheel. This was my mother. 8 But Xyrem changed all that. It was a 9 difference between night and day and mother quickly 10 rediscovered the joys that she had missed for 11 decades -- playing games with us, going dancing, 12 going to the movies, celebrating family birthdays 13 and holidays. The day-to-day tasks that you and I 14 take for granted, she could finally do as a normal 15 person. This was the mother that we had never 16 known until Xyrem gave us her life back and her 17 family back. I have seen the difference. I have 18 lived the difference. Please make this valuable 19 medication available to people who have narcolepsy. 20 They and their children will see the change in 21 their lives. Thank you. 22 DR. KAWAS: Thank you, Ms. Jones. That 23 concludes the section of open public hearing, and I 24 want to thank everybody who expressed their views, 25 information and helped the committee keep sight of 249 1 all the issues here. 2 We will now reopen the questions from the 3 committee to the invited speakers, sponsor and the 4 FDA. In particular, I would like to focus on the 5 presentations that we had right before lunch 6 involving the epidemiology, adverse medical events 7 and the sponsor presentations on risk management 8 and abuse liability. So, who wants to start the 9 questions from the committee with regard to some of 10 those presentations? 11 Continued Committee Discussion and Deliberations 12 DR. SIMPSON: I put up my hand under false 13 pretenses because I had just one question really -- 14 DR. KAWAS: We don't like false pretenses 15 around here! 16 DR. SIMPSON: It was really relating to 17 the efficacy. I mean, a lot of the presentations 18 we have just heard give the impression that the 19 cataplexy was, if not completely controlled, almost 20 completely. Yet, when we look at the data we see 21 that the median number of events at the end of some 22 of the studies is about eight or so on drug. So, 23 do we have any data about how many people actually 24 had no cataplectic events? 25 DR. REARDAN: I think that this question 250 1 was discussed to some extent this morning. It 2 dealt with complete cataplexy -- 3 DR. SIMPSON: No, no, I am saying do we 4 have data on the people who were, quote, cured? 5 Were there any? 6 DR. REARDAN: We have a slide on that, I 7 understand. 8 [Slide] 9 DR. HOUGHTON: This is an example of the 10 long-term data, and one of the problems with the 11 controlled GHB-2 trial is that it may be too short. 12 The reason that the time was restricted is because 13 of the imposition of patients on placebo for longer 14 periods of time. But that represents a picture of 15 the long-term response in terms of percentage 16 change. So, we have a control across all doses, 17 demonstrated here for a 12-month period, around the 18 90 percent or better mark. Now, that doesn't mean 19 to say people don't have any cataplexy, but it is 20 certainly very significantly reduced. 21 DR. KAWAS: Dr. Katz? 22 DR. KATZ: Yes, we have seen this slide a 23 number of times. I just want to remind the 24 committee that this is open, uncontrolled, 25 non-randomized data, not the sort of data that we 251 1 would ordinarily rely on to draw any sort of 2 conclusion about effectiveness of any sort. 3 DR. KAWAS: Maybe the sponsor could show 4 us some of this data from one of the randomized 5 trials? 6 DR. HOUGHTON: We could show you the 7 change in the GHB-2 study again, which is the 8 four-week study. 9 [Slide] 10 The data is median change from baseline. 11 We had a median incidence of 23.5 in the 9 g group, 12 a change from baseline of 16.1. If we present that 13 again as percentage change -- because, once again, 14 it is complicated by the spread in the data. 15 DR. SIMPSON: I guess my question is if 16 the median at the endpoint is 8.7, it means 50 17 percent of the people were above it and 50 percent 18 were below. Now, how many were below, say, 1 or 2? 19 DR. HOUGHTON: Well, it depends on what 20 their starting level was, and the conditions of 21 entry were 3 cataplexy or more attacks per week. 22 We did have patients with very high incidence. So, 23 in terms of absolute numbers, that is a very 24 difficult response. I am not trying to be evasive. 25 DR. WOLINSKY: The other piece of that 252 1 data though that you presented and might be worth 2 looking at quickly is the randomized stop component 3 of the trial. 4 DR. HOUGHTON: Sorry? 5 DR. WOLINSKY: When patients were 6 randomized to be taken off -- 7 DR. KAWAS: The 21 study. 8 DR. REARDAN: Right. The question is on 9 a-patient-by-patient basis, how many patients went 10 from X amount of cataplexy to zero cataplexy. Is 11 that what you are trying to get at? 12 DR. SIMPSON: Zero or close to zero. 13 DR. REARDAN: That is in the data listings 14 for the trial. We didn't bring individual breakout 15 of the data. We brought summary information for 16 the committee. I don't know if Dr. Mani has a 17 recollection or Dr. Katz. 18 DR. KATZ: You don't have a distribution 19 of how many events patients had? In other words, 20 you know, X percent had two or fewer events; Y 21 percent had between two and five events. 22 DR. HOUGHTON: No, we didn't break it down 23 like that. I think the slide that you were 24 referring to was the one that I showed with 25 individual patient plots, and I can show you that 253 1 quickly. 2 [Slide] 3 That is just an example of absolute 4 numbers. These were individual patients plotted. 5 That was their incidence at the baseline, and that 6 was some two years after this was conducted. That 7 is the sort of response they got when their active 8 treatment was withdrawn. That is the group in 9 active treatment. So, in terms of just absolute 10 numbers, that is just a snapshot. That is not a 11 statistical presentation. It happens to be every 12 patient that came from that original trial through 13 into this trial, and I show it as individual plots. 14 It is the best impression of individual patient 15 data I can give you to answer your question. 16 DR. BLACK: Just a comment on that. In 17 this section we do have placebo-controlled data and 18 we have the number of cataplexy attacks on placebo 19 versus active medications after patients have been 20 on treatment for a long period. Dr. Katz' comment 21 is very good. The data that has been generated 22 over the open label, though it does suggest there 23 is a time course till optimal effect of at least 24 two months, is open label. But this is 25 placebo-controlled data, suggesting that the 254 1 average there of cataplexy attacks per day -- I 2 don't know if you have the numbers of that, Dr. 3 Houghton, but it is very low during the time of 4 treatment unless they are taken off and then on the 5 placebo-controlled portion. 6 DR. KAWAS: I have a question for the 7 company as well as probably Dr. Dyer. I want to 8 hear both sides of why we heard such very different 9 descriptions of the potential for withdrawal 10 syndromes with this disorder. I recognize fully 11 that the company has studied individuals with 12 narcolepsy and it is possible that alone could 13 comprise the difference, but we do have a very nice 14 withdrawal study in study 21, which is not 15 typically the case, and the findings that were 16 collected from that are in fairly sharp contrast to 17 the stories that we have heard from Dr. Dyer with 18 regard to withdrawal syndromes, and I wondered if 19 both sides could tell me what the difference was. 20 Is it dose? What is the difference here? 21 DR. REARDAN; I will ask Dr. Balster, but 22 I believe it is dose and frequency. Bob, do you 23 want to comment? 24 DR. DYER: I doubt that we disagree. 25 Clearly, in my set of patients and what we use 255 1 nearly as a diagnostic parameter and which patients 2 we should admit, even though their early symptoms 3 are mild, is the frequency with which they are 4 using it. So, the kinetics of the drug show us a 5 duration of activity around three or four hours. 6 When these patients increase their frequency so 7 that their body constantly is exposed to GHB, those 8 are the ones that we feel become severely 9 physically dependent and then go through this 10 withdrawal syndrome that can have an onset within 11 hours of discontinuing the drug. 12 DR. KAWAS: So, in your opinion it is 13 frequency of dosing, not even the number of grams 14 per day. 15 DR. DYER: As far as I can tell, it is 16 frequency because if I take the sponsor's 17 information, and for years I have spoken to the 18 investigators that are doing this and they have 19 said they have had no trouble. Their patients have 20 a 12-hour drug holiday daily, which is two to maybe 21 three times what they are calling a half-life for 22 this drug. So, the drug is completely eliminated 23 from the body for a time period, and the patients 24 have that become severely addicted, all of them -- 25 I mean, that is kind of diagnostic for the severe 256 1 withdrawal, somebody who is taking it every three 2 hours around the clock. 3 DR. BALSTER: Yes, I agree completely with 4 that, and maybe the analogy that would help you 5 understand it would be the analogy, for example, 6 with alcohol where really alcohol can produce a 7 very significant physical dependence but you can 8 drink it every evening with your meal and you won't 9 become dependent because between that evening use 10 and the next day it has cleared the body. So, 11 whatever physiological adjustments are necessary 12 have corrected themselves. So, we are in complete 13 agreement. 14 DR. KAWAS: Thank you. Dr. Katz? 15 DR. KATZ: Just as an extension of that, 16 there was also the implication or the explicit 17 statement that in some of those people who took it 18 very frequently and ultimately, presumably, became 19 addicted, they were compelled to take it more 20 frequently. In other words, there was a tolerance 21 that developed and they had to increase their 22 frequency to get the same sort of pharmacologic 23 effect. 24 So, I will just ask the same question that 25 Dr. Kawas asked about withdrawal. We have heard 257 1 from the company that patients who have taken the 2 drug for years and years and years don't develop 3 tolerance; they don't have to increase their dose; 4 they don't increase the frequency of 5 self-administration. But, we are hearing that on 6 the outside there are people in whom this 7 phenomenon apparently does occur. So, I will ask 8 the same question. Why the disparity? 9 DR. DYER: Again, there haven't been 10 really good studies or anything scientific. It is 11 kind of my thoughts or opinions but, again, it is 12 accommodation because you are taking it around the 13 clock. So you are accommodating. Also, in the 14 patients that are taking it -- well, I don't know, 15 they are not really patients -- in the people who 16 are abusing it there is a lot of the feeling that 17 if a little is good, a lot is better. They are 18 taking it initially, these body builders, for this 19 growth hormone burst. So, they really feel like 20 they are doing the right thing. So, there is 21 nothing to have them diminish their dose or hold 22 their dose as it is. Then, once they start taking 23 it more frequently, the duration of the drug as it 24 wears off in three or four hours, we think, gives 25 them kind of a dopamine surge for which then they 258 1 are going to feel a little depleted and want to 2 take that next dose. Then there is also physical 3 craving for that kind of high. They are awake and 4 feeling that kind of high as opposed to the 5 patients that are taking it immediately upon going 6 to bed and then sleeping through this euphoric -- 7 whatever the kids are trying to get that are 8 abusing it -- if you can roll that into an answer. 9 DR. BALSTER: That is exactly the way I 10 would see it too. Just to add one further thing to 11 that, the way to look at tolerance, you have to 12 understand that it occurs through different effects 13 at different rates and in different ways. So, the 14 therapeutic effect is one effect. The intoxicating 15 effect is a different effect. And, commonly in 16 abuse situations where persons are trying to 17 maintain an intoxication, they have to escalate 18 dose and frequency in order to do that, whereas the 19 data obtained in these clinical trials, of course, 20 is on the therapeutic effect. 21 DR. DYER: One other comment, in the 22 alcohol abuse trials they did escalate their dose 23 in more of a craving kind of manner. That was 24 about 15 percent. 25 DR. KAWAS: Dr. Roman? 259 1 DR. ENGEL: I would like to add something, 2 if I may, to this point that is based on the risk 3 management system proposed by the sponsor. As you 4 saw, the data collected by the specialty pharmacy 5 will include dose by patient. And, because of 6 that, the specialty pharmacy will be able to 7 predict when is the appropriate timing for a given 8 patient to have their prescription refilled. So, 9 for example, there are patients attempting to 10 refill too soon, so to speak, that will be 11 identified and it will be an opportunity for the 12 pharmacist to interact with the physician very 13 quickly, before a patient might get into a 14 situation like which Dr. Dyer is describing with an 15 overuse syndrome. 16 DR. ROMAN: A question perhaps again for 17 Dr. Balster. Is the pharmacology of GBL and 1,4-BD 18 similar in animal experience to GHB? Number two, 19 if there is a difference, did I understand 20 correctly that GBL and 1,4-BD are not currently 21 drugs of abuse? 22 DR. BALSTER: Well, the first question, 23 pharmacological comparisons of GBL, GHB and 1,4-BD, 24 these haven't been very extensively done. So, 25 hopefully some of those NIDA grants that someone 260 1 was talking about will really take that question 2 on. But let me say that in a number of those 3 studies that were done to describe the pharmacology 4 of GHB, in some of these studies actually GBL was 5 administered to the animal with the view that it 6 was a prodrug for GHB. I forgot who said it but 7 someone said that so far as we know, all of the 8 effects of GBL and 1,4-BD are really as a 9 consequence of their conversion to GHB. I believe 10 that would be the current state of knowledge about 11 that although it is imperfect. 12 Now, the question about control, in a 13 sense, yes, all of these drugs are potential drugs 14 of abuse because they can be taken and basically 15 are active in the case of precursors with 16 metabolites. So, yes, all of these are potentially 17 drugs of abuse. Only one of them is a controlled 18 substance and one of them, by congressional action 19 of last year, became what is called a listed drug, 20 and I could explain that to you or, actually, Dr. 21 Sannerud would know better than I what exactly that 22 means. But it essentially means that there is 23 limited distribution. 24 DR. ROMAN: So, with GBL and 1,4-BD there 25 is no control. 261 1 DR. BALSTER: Well, as I say, for 1,4-BD, 2 to my knowledge, there is no control. I need to 3 step back a little bit from that because we could 4 get into too long of a discussion about what 5 constitutes an analog under the specific language 6 of the legislation. So, it is possible for 7 prosecuting attorneys to claim that one or another 8 of these drugs are analogs of a controlled 9 substance. The Controlled Substances Act, in a 10 sense, regulates analogs. Now, 1,4-butanediol is 11 questionably an analog, but that would be something 12 that would be worked out in court. So, I am not 13 trying to tell you that someone could absolutely, 14 with impunity, sell 1,4-BD to children and say that 15 it wasn't a drug of abuse because I am sure that 16 there would be authorities and prosecutors who 17 would try to do something about that. But in terms 18 of the actual language of regulation, only GHB is a 19 controlled substance. 20 DR. SANNERUD: GHB is a Schedule I 21 controlled substance. Butanediol and GBL are 22 considered controlled substance analogs under 23 federal law, which means they can be prosecuted, as 24 GHB, with penalties and other things would apply if 25 someone is caught trafficking, distributing or 262 1 clandestinely manufacturing or selling these 2 compounds as well. GBL is listed as a List I 3 chemical, which means that there is record-keeping 4 and registration required. There are no retail 5 sales of butanediol, and there is a graph in here 6 with the product. These are used in industrial 7 uses. So, this comparison is really a little bit 8 misleading. I don't know the numbers but GHB is 9 not even marketed yet, so this number on production 10 is only for clinical trials I assume. 11 As far as the GHB and Xyrem they are both 12 GHB. There is no forensic analysis that is going 13 to differentiate between the two. So, when samples 14 are submitted to labs there is no way to tell if it 15 is the product or if it is something that is made 16 at home. So, for someone to say that there has 17 never been any diversion of the product, there is 18 no way to tell that because there is no way to 19 differentiate between the two under forensic 20 laboratory conditions. 21 Another question I wanted to address is 22 the quota issue. Ms. Meyers brought up quotas for 23 Schedule II compounds, the stimulants. DEA sets 24 the quota, as it will with GHB as well. It has 25 never been the case that drug has run out at the 263 1 end of the year because the quotas are set too low. 2 If there is a problem with the drug manufacture the 3 quotas can always be increased throughout the year, 4 and they are done so on a regular basis. So, there 5 has never been the case where a drug has run out. 6 DR. KAWAS: Dr. Mani? 7 DR. MANI: I would just like to touch upon 8 the issue of drug diversion during the clinical 9 trials once again briefly. Many speakers have 10 asserted that there has been no evidence that Xyrem 11 or GHB used in the clinical trials included in the 12 database was diverted. That may very well be true, 13 barring the one exception that I cited earlier, and 14 I have no firm evidence to the contrary. However, 15 I have gone through the NDA, reviewed the whole 16 NDA, and I would be a little more hesitant in 17 making that assertion, and I will tell you why, and 18 that has to do with the way the drug was dispensed 19 in the Scharf study which, as you know, occupied 20 about 30 percent of the database in terms of 21 patient numbers and about 70 percent of the 22 database when you are talking about patient years 23 of exposure. 24 What happened here was that patients saw 25 Dr. Scharf in Cincinnati, at least for an initial 264 1 visit, and had an appropriate diagnosis made and 2 were then enrolled in the trial and then went back 3 to whatever part of the country they came from. 4 Prescriptions for medication were filled based on 5 their returning completed diaries. In some 6 instances it appears, at least from my looking at 7 the case report forms, that prescriptions were 8 sometimes filled in advance or the diaries being 9 returned, obviously to prevent the patient from 10 running out of the drug. But the important thing 11 is that patients were not required to return unused 12 supplies of medication prior to getting a fresh 13 prescription, or to provide any formal accounting 14 of how much medication they used or did not use. 15 In the absence of any active surveillance of that 16 kind, as I said, I would be quite hesitant in 17 making the assertion that no medication was 18 diverted. 19 DR. REARDAN: I need to make a qualifying 20 statement here. We do not disagree with Dr. Mani. 21 However, under the company's clinical IND, our 22 patients under IND didn't begin entering trials 23 until 1996. Patients were required to document 24 their dose; to return their bottles. The bottles 25 were all qualified by volume in terms of what was 265 1 returned. The incident that Dr. Mani refers to, I 2 believe, occurred in 1986, when GHB was available 3 as a nutritional supplement and Dr. Scharf's trial, 4 again, was clinical practice. There were a lot of 5 issues on GCP compliance in that trial. We do not 6 take responsibility for accountability of drug 7 under Dr. Scharf's trial. So, I will just qualify 8 that. Okay? 9 DR. MANI: I agree. 10 DR. FALKOWSKI: I have a question and it 11 has to do with the fact that we are talking about a 12 method of taking this drug where you take half the 13 amount at bedtime and then you wake up several 14 hours later, but don't really wake up, and take the 15 rest of it. And, I am just wondering what would 16 happen if you were confused. It also involves 17 mixing it ahead of time to the right strength. I 18 am asking this both to Dr. Dyer and the sponsor, 19 what would happen if someone took 9 mg at once? 20 You know, if someone got confused and took it all 21 at once, what would be the expected outcome? 22 DR. REARDAN: I had a number of questions 23 about this at the break from a couple of members of 24 the committee -- how do they make it up, and so on. 25 It might be worthwhile to ask Patti Engel to go 266 1 through that. The other point about narcoleptic 2 patients waking up, maybe Dr. Black, you could 3 comment on how they wake up and take their second 4 dose. 5 DR. FALKOWSKI: Right, but my bottom line 6 question is what would happen to a person who 7 inadvertently took all of their dose at once, and I 8 really insist on an answer to that. Thank you. 9 DR. BLACK: That question has been 10 answered by patients who have taken inadvertently 11 larger doses. As far as the waking up at night, 12 the patients that are here could probably respond 13 to that, but the overwhelming majority are awake 14 actually before the four hours later on their own 15 and they are fully awake. The medication is 16 premixed so there is no mixing that needs to occur 17 at that point. There are folks who have taken 18 extra doses and there is more sedation that occurs 19 with the extra duration and the period of sleep is 20 longer with the higher dose. 21 DR. FALKOWSKI: Is the answer then 22 increased sedation? Is that the answer to my 23 direct question? 24 DR. BLACK: Yes, if the dose is increased 25 there is increasing sedation and a longer sleep 267 1 period. 2 DR. FALKOWSKI: Okay. Dr. Dyer, could you 3 respond to that? 4 DR. DYER: It is my opinion that the dose 5 would be around 100 mg/k and at that point you are 6 going to have coma and some of the other side 7 effects that we see in our club goers are very 8 likely to be what you would see. So, vomiting and 9 aspiration is a possibility. You know, the ability 10 to hear and react to fire alarms, children, 11 whatever, that is all going to be blunted. 12 DR. FALKOWSKI: Is it a possibility then 13 that some of these people who may have double dosed 14 would be in a coma but who would know, you know? 15 Is that a possibility, sponsor? I mean, who is to 16 know? 17 DR. BLACK: I think that the question is a 18 good one, and what I might call deep sleep someone 19 else might call a coma. But when we look at the 20 brain wave activity of the folks with the higher 21 doses, they have nothing in the EEG that would be 22 consistent with straightforward coma. 23 DR. FALKOWSKI: But you didn't take EEGs 24 on these people when they were sleeping in 25 situations like this. 268 1 DR. BLACK: Well, we have done EEGs on the 2 folks when they have been sleeping at the 9 g dose 3 but not on double the 9 g dose. 4 DR. FALKOWSKI: Okay. 5 DR. KAWAS: Dr. Katz, please. 6 DR. KATZ: Yes, a couple of things. Maybe 7 the best way to get at this if it is possible is to 8 ask the company to show us any data that they have 9 about what happened to patients who took, let's 10 say, a single 9 g dose. I don't know how many 11 patients did that, but if there is data on that it 12 would be nice to see. 13 So, I don't know, maybe you could look for 14 that while I get to the second part which is, 15 again, just another variant about the question we 16 were talking about before, this perceived disparity 17 between patients and non-patients who take the drug 18 recreationally. We have heard again, not just in 19 terms of withdrawal and addiction and tolerance but 20 just in terms of serious adverse events, a number 21 of the serious adverse events that we have heard 22 about in the emergency room situation seem to have 23 occurred at doses, presumably -- I don't know how 24 reliable the dose information is in that setting, I 25 am not sure, but presumably at doses that patients 269 1 routinely get and which they tolerate extremely 2 well. So, I will ask the same disparity question 3 again there. 4 DR. MIGNOT: I think you have to realize 5 also that you are talking about narcoleptic 6 patients who also experience daytime episodes of 7 overwhelming sleepiness that sometimes lead to 8 confusion, and there are a lot of horror stories 9 about narcoleptic patients, independently of GHB, 10 at any moment of their life where they can 11 sometimes be in a risky situation just because they 12 have what we call automatic behavior, this 13 overwhelming sleep attack where they really don't 14 know what they are doing, where they may be driving 15 or doing something dangerous. I think that is also 16 important to keep in mind. The danger of taking 17 two doses at a time, if it is relatively well 18 dispensed, for narcolepsy patients I think needs to 19 be put in perspective for their other symptoms. 20 DR. REARDAN: I am only aware of one case 21 in our database. It was a patient who 22 inadvertently took 18 g and I think, Dr. Mani, you 23 are well aware of that. He did fall on his head. 24 So, it is confusing as to whether it was a result 25 of his 18 g dose -- you know, that was the best 270 1 estimate we had -- or in the fall he hit his head, 2 but he did end up being taken to the emergency 3 department and did need supportive care. Oh, Bill 4 is saying that was a normal dose. I am sorry, let 5 me get him to clarify. 6 DR. HOUGHTON: Yes, I am sorry. That is 7 one of the cases that we know very little about. 8 It was a patient who was in the kitchen. There was 9 a loud bang. His wife heard the noise and came in, 10 and her husband was on the floor. So, we got no 11 dose relationship to that event. We know nothing 12 as to whether it is related to Xyrem. 13 The 18 g overdose was the patient who was 14 supposedly sleepwalking, in the Scharf database, 15 who supposedly then took 18 g on top of his normal 16 dose and was taken to hospital and ended up on a 17 ventilator. 18 Really, the best prevention we have of 9 g 19 being taken together is the fact that the dose has 20 to be made up into separate doses. The 21 instructions to the patient are very clear. They 22 make two doses up together, dilute it in the water; 23 drink one when they get into bed and the other, in 24 a sealed cup, put away. Now, if they took the 25 second dose in ten minutes or two hours, we have 271 1 not done that study and it is very dangerous to 2 extrapolate that sort of dosing. On one hand, I 3 can quote the patient who took 180 g and was taken 4 to hospital unconscious and walked out of hospital 5 four hours later to be admitted to the psychiatric 6 unit. 7 I certainly don't want to propose that as 8 the normal pharmacodynamic response. We have not 9 done a study that has escalated beyond the 4.5 g 10 dose twice a night, and I think it is very 11 dangerous to extrapolate. It is also very 12 dangerous to extrapolate the anesthesia data or 13 some of the data that Dr. Dyer talked about this 14 morning. Doses were given up to 100 mg/kg 15 intravenously. If we believe the bioequivalence 16 data, the absolute bioavailability data, that is 17 equivalent to at least 300 mg/kg as an anesthetic 18 dose, and that would be the best dose relationship 19 we could give to dose escalation. Again, without 20 true data I am not prepared to extrapolate from 21 that. 22 DR. KAWAS: Dr. Mani, do you still want 23 the floor? 24 DR. MANI: Yes, very briefly, just as 25 further evidence of how much individual variability 272 1 there is in response to this drug. There is a 2 subject who Dr. Houghton had referred to in his 3 presentation this morning, a healthy subject 4 participating in a pharmacokinetic trial, a healthy 5 young subject who received a single dose of 4.5 g 6 and afterwards became obtunded, developed 7 obstructed respiration perhaps because of his jaw 8 falling back, became incontinent or urine and 9 stool, and took a number of hours to recover but 10 did not need any special supportive care. So, even 11 a 4.5 g dose may not be entirely safe for 12 everybody. 13 DR. HOUGHTON: That story is somewhat true 14 but not quite accurate. The patient was easily 15 arousable, walked to the bathroom after the event 16 of passed urine, after resting back in bed had a 17 normal sleep and, two hours later was awake and ate 18 a normal lunch. So, again, I can't account for the 19 degree of obtundation but that still represented 20 the maximum single dose in our database. It was a 21 single dose of 4.5 g after a 10-hour fast. 22 DR. MANI: Although those details about 23 the patient being able to get up and go to the 24 bathroom and eat her lunch, and so on, wasn't in 25 the narrative that we have available. 273 1 DR. HOUGHTON: We were collecting urine 2 samples every two hours and I can assure you the 3 patient was walked to the bathroom. She certainly 4 vomited at the time. 5 DR. KAWAS: Dr. Leiderman? 6 DR. LEIDERMAN: Very briefly because Dr. 7 Mani raised one of the points that I wanted to, but 8 the other question I had for the sponsor and the 9 sleep neurophysiologists here, do you think that in 10 some of the differential response that we are 11 seeing in the narcolepsy patients as compared to 12 the subjects who become dependent, addicted, have 13 overdose problems that there may be a role not only 14 of the basic neurophysiology of the narcoleptic 15 brain but, of course these patients tended to be 16 co-medicated with stimulants, and what role do you 17 think that might be playing in the narcolepsy 18 population? 19 DR. REARDAN: Is the concern that 20 stimulants would still be present on board when 21 they take their nightly dose of Xyrem? Is that 22 what you are after, or what? 23 DR. LEIDERMAN: Well, I am asking for your 24 thoughts on, shall we say, the differential effects 25 of GHB on the two populations, and one of the sort 274 1 of clear differences, taking sort of the first cut, 2 is that narcolepsy patients are co-medicated with 3 stimulants generally, whereas the abusing drug 4 population, if anything, is self co-medicating with 5 other CNS depressants or using GHB at high doses 6 alone. 7 DR. BLACK: I think there are a number of 8 questions that surface. We have patients in 9 protocols where they are wanting to remain on the 10 protocols or wanting to be drug compliant. There 11 are reasons that they wouldn't abuse in addition or 12 outside of the fact of co-pharmacy with stimulants 13 and so forth. So, it is hard to compare those two 14 groups clearly. 15 I think the best we can do is speculate. 16 We have a number of patients that were not 17 co-treated with stimulants as well, that were on 18 just Xyrem, and they didn't self-escalate the dose 19 or abuse the agent either. I think the only way to 20 do it would be to give high dose frequently to the 21 narcolepsy patient population and see if they are 22 similarly addictable, and then it would be also 23 interesting to find out what percentage of the 24 normal population is addictable as well. 25 Obviously, those studies couldn't be done. But I 275 1 think we can't compare the two and it is real hard 2 to try to extrapolate the information we have to do 3 a comparison. 4 DR. KAWAS: Dr. Dyer, followed by Dr. Van 5 Belle, followed by Ella Lacey, followed by the 6 questions that the FDA has asked us to consider. 7 In between, we will get a quick demonstration of 8 the mixing. 9 DR. HOUGHTON: Could I just add one point 10 of clarification to Dr. Leiderman's question? 11 There were patients in all of the studies that were 12 not on stimulants. In the GHB-2 study I think it 13 was about 15 percent when we did a recent look at 14 the database for Dr. Mani. So, there was at least 15 a proportion of patients represented in the 16 database that weren't on stimulants as concomitant 17 medication. 18 DR. DYER: There was one study, I believe 19 it was done in rats where amphetamines and then a 20 second with caffeine, where those were shown to 21 kind of be antidotal to GHB poisoning, where it 22 prevents the rats' loss of riding reflex. So, 23 there may be some of that issue if they are taking 24 it concurrently. One of the other things about the 25 disparity, where I don't see the disparity as being 276 1 so much is that the narcoleptics are taking their 2 dose at night. We know pretty commonly from the 3 surgical studies from what we see coming into the 4 emergency room and from the adverse effects of the 5 study, that GHB causes vomiting and incontinence. 6 So, we are seeing that in both populations of 7 patients. 8 DR. CHERVIN: Is anybody there? 9 DR. KAWAS: Yes, is that one of our phone 10 consultants, Dr. Chervin or Dr. Guilleminault? 11 DR. CHERVIN: Sorry, it seems like we were 12 completely cut off. 13 DR. KAWAS: Can you hear us now? 14 DR. CHERVIN: Just barely. If there is 15 any way you can make this signal more than barely 16 audible, it would be helpful? 17 DR. KAWAS: We can barely hear you but it 18 sounds like we are going to have to get the AV 19 people on it, if you give us a moment. 20 DR. CHERVIN: I do have questions if I 21 have time to ask them. 22 DR. KAWAS: I know that you are on a 23 timetable, so we will put you in the middle of the 24 six-person pileup, if we could let the speaker that 25 is going now finish though. 277 1 DR. DYER: So, there was another study 2 where they took the patients and the patients that 3 they gave the dose to and then forced or tried to 4 maintain themselves awake, those were the patients 5 that became confused. 6 The other thing is that in our emergency 7 department study where we were trying to verify our 8 ability to predict GHB by toxidrome, we looked at 9 patients that came in with a GCS score less than 8 10 that were spontaneously breathing. So, unlike most 11 CNS depressants that cause profound coma, generally 12 the breathing is still spontaneous and maintained. 13 You see mild respiratory acidosis but it is not 14 very common that these patients need to be 15 intubated. So, it is not contrary to be thinking 16 that a patient might be comatose and survive the 17 night. 18 DR. KAWAS: Dr. Van Belle, while we are 19 still working on the audio, do you want to go ahead 20 and ask your question? 21 DR. VAN BELLE: I just have a brief 22 question with respect to age eligibility. Will 23 this medication be available to people under 18 24 years old? 25 DR. REARDAN: The company has not 278 1 specifically developed data for pediatrics, and I 2 think this would have to be something we work out 3 with the agency but, typically, a medication 4 approved for adults is not denied children. FDA 5 and Congress have tried to put incentives in to get 6 sponsors to develop pediatric information. In 7 addition, narcolepsy is not generally a pediatric 8 disease. I don't know if either Dr. Mignot or Dr. 9 Black want to comment further. Dr. Katz? 10 DR. KATZ: Well, generally speaking, 11 unless there is a good reason not to, we would 12 limit the age that would be at least included in 13 the indications or in labeling or dosage 14 administration to the age of the lower limit of the 15 age studied in the trials. I don't know exactly 16 what the youngest patient was in these trials. 17 DR. REARDAN: Bill Houghton is saying 12. 18 DR. KATZ: Okay, 12. Again, if there was 19 one patient who was 12 and everybody else was 18 20 and above, we would say adults or 18 and above, 21 that kind of thing. It is true that there is no 22 prohibition, obviously, from a physician writing a 23 prescription for a drug for a child if it is only 24 explicitly approved for an adult. It happens 25 obviously all the time. But one of the questions 279 1 when we get to it with regard to risk management 2 and that sort of thing is if there were no children 3 studied, or children studied below a certain age, 4 do you think attempts should be made to restrict it 5 in this case? So, you know, it is open for 6 discussion. 7 DR. MIGNOT: To answer the question, onset 8 of the disease is roughly between 15 and 25. That 9 is really when the bulk of the patients are coming 10 in, especially for cataplexy, and I think it is 11 very important to treat them early. As there is 12 more and more knowledge about narcolepsy being an 13 important disease and being recognized early -- I 14 think you have heard a lot of testimony about how 15 important it is to treat them early so that they 16 can go through normal schooling. I think it will 17 be very important to not be too restrictive towards 18 the lower age. 19 DR. KAWAS: Dr. Lacey? 20 DR. LACEY: Two questions, one regarding 21 the packaging. With the packaging being in a 22 bottle and it is child-resistant dosing, and all, 23 but hearing about adolescents and their involvement 24 with GHB, I wondered if you considered other 25 packaging. In deciding on this packaging, did you 280 1 consider individual dosage packaging at all, and 2 what happened with that? 3 DR. REARDAN: We considered individual 4 dosing packaging for sure. We thought that was a 5 greater potential for diversion as it is easy to 6 take those individual doses. I think maybe you 7 would get some reassurance if Patti Engel can go 8 through how we instruct the patients to dose and 9 what the controls are for that. Patti? 10 MS. ENGEL: Thank you. To the point of 11 individual dosing, we did speak quite extensively 12 about that with law enforcement. 13 DR. LACEY: Yes, I am pretty convinced 14 about the patient. I am more concerned about 15 others in the household who are exposed to a 16 bottle. 17 MS. ENGEL: Right. I will address that as 18 well. On the individual dosing, law enforcement 19 was concerned about small containers that could be 20 stuck in a pocket or purse, or slipped in someone's 21 drink more easily. One of the things I shared with 22 you earlier is that the bottle itself comes with a 23 child-resistant closure. What is difficult to see 24 from this distance, but it is something called a 25 press-in bottle adaptor. When the patient gets 281 1 this, there is a little well, if you will, in 2 there. Even if a child can get this lid off, you 3 can't drink it down. What has to happen is there 4 is a metered syringe provided. It gets stuck in 5 here and the patient removes a metered dose. Okay? 6 They then have two child-resistant dosing cups and 7 these aren't fancy. We took them because they are 8 CPIS tested for child resistance, of course, and 9 they put it in, preparing both doses by their 10 bedside. 11 Now, the dose itself is metered. This 12 Xyrem, to be frank, is not good tasting stuff. It 13 is sodium oxybate. It is very salty. Many people 14 will dilute it. How much they dilute it really is 15 to their taste. We did not want to cherry flavor 16 it or anything like that that may make it more 17 attractive to children. Okay? Does that answer 18 your question? 19 DR. LACEY: It really wasn't the small 20 children that I was concerned about as I was about 21 the older, the adolescents in the household who can 22 open it the same as I could. So, I guess your 23 answer was that law enforcement was concerned about 24 the small dosages just being put in a pocket. 25 MS. ENGEL: That is right. Remember, 282 1 illicit use of Xyrem also falls under C-I 2 penalties, like heroin or LSD. So, we will never 3 be able to find a package that a 19- or a 21-year 4 old will not be able to get into. What we do, 5 however, is to educate the Xyrem patient on a 6 number of occasions of the penalties should that 7 occur. So, there is an element of patient 8 responsibility with this. 9 DR. LACEY: Thank you. The second 10 question I have is about the suicide attempts that 11 were presented by Dr. Houghton this morning. That 12 was in that list of adverse events I believe, and 13 it has continued to bother me that we talk about it 14 as a suicide attempt as though nothing else 15 happened and I am just curious, I guess, in those 16 attempts were some of the other adverse events also 17 experienced by those persons who were suicide 18 attempters? 19 DR. REARDAN: As you heard from Dr. 20 Mignot, depression is very common in narcoleptics, 21 but I will ask Bill to comment on that. 22 DR. HOUGHTON: In all the patients who 23 attempted suicide there was preexisting disease. 24 In terms of response to the dose taken, only one of 25 the suicide attempts involved Xyrem, and that was 283 1 the patient who took a very large dose, about 300 2 ml of the drug which is equivalent to at least 150 3 g, and he became comatose, incontinent of feces and 4 urine, continued to breathe spontaneously, was 5 found by his wife in the bathroom, transported to 6 the emergency medical care, did not require 7 intubation or ventilation, and walked out of 8 hospital four hours later to be admitted to the 9 psychiatric unit. I certainly don't propose that 10 as the norm. There will be certainly unconscious 11 patients at much lower doses. So, please don't 12 think I am proposing that as the pharmacodynamic 13 profile of the drug. But you asked me what the 14 side effects of the suicide event were and that is 15 the only data that I can give you. 16 The second suicide event that was not 17 fatal did not involve Xyrem. One of the fatal 18 attempts did not involve Xyrem at all. The last 19 suicide attack in the bipolar disorder patient was 20 a real pharmacologic cocktail involving 21 benzodiazepines, opiates, a number of drugs and 22 some Xyrem. 23 DR. LACEY: But for those individuals who 24 did have the suicide attempts, they did not have 25 other -- not with the attempt directly but other 284 1 adverse events also in their report? 2 DR. HOUGHTON: No. One of those was a 3 lady who had a group of people to her home. She 4 asked them all to leave early, and when attempted 5 to be contacted the next morning didn't respond, 6 and when her attentions were sought she was found 7 dead in the home. 8 The second attempt was a young lady who 9 took an overdose of buspirone and told her father 10 immediately. Her behavior was normal to that 11 point. So, that is an example. 12 DR. KAWAS: Dr. Chervin or Dr. 13 Guilleminault, can you hear us now? You guys are 14 next in the line up. 15 DR. CHERVIN: Thank you. I have two 16 questions. Please tell me if it has been covered 17 and I just was not able to hear it, but I read in 18 some of the material that was distributed prior to 19 the meeting about comparisons of the therapeutic 20 index or the therapeutic window for GHB to that of 21 other drugs that are currently approved and used. 22 I was wondering if perhaps Dr. Dyer or Dr. 23 Falkowski or Dr. Balster could address that 24 comparison. 25 DR. DYER: Is that the comparison of LD-50 285 1 in rats? 2 DR. CHERVIN: I guess it was rats, and it 3 was LD-50 and effective dose, and they looked at 4 the ratio. 5 DR. DYER: The problem I have with some of 6 the rat data, lethal dose data, is the deaths we 7 see are often secondary to coma. It takes high 8 doses to cause pure respiratory depression. We 9 have some patients that idiosyncratically have a 10 pulmonary edema, but most of the deaths are 11 secondary to unprotected coma and loss of airway. 12 So, I don't know that that would extrapolate or 13 come from rat data at all. I don't think you would 14 see that. 15 DR. CHERVIN: Is there any other way to 16 get at the issue of is Xyrem going to be more 17 dangerous than other drugs that are used carefully 18 when indicated? 19 DR. REARDAN: Dr. Chervin, I have some 20 data on LD-50 that will help. Oral GHB has an 21 LD-50 on the order of 9000 mg/kg in rats, and about 22 3500 mg/kg in mice. The IV LD-50 is about a third 23 of that for GBL and for butanediol it is on the 24 order of 2000 mg/kg. If you look at the effective 25 dose, we are in the range, I believe, of about 286 1 50-120 mg/kg recommended for the narcoleptic 2 patients. Now, that is just on an LD-50 basis. I 3 don't know if Dr. Mani wants to comment on the 4 therapeutic range, or Dr. Katz. 5 DR. KATZ: I don't think we really know. 6 I am not sure if the animal data is relevant at 7 all. And, I don't think we have data that, in a 8 systematic, adequate way, explores the full dose 9 response both with efficacy or tolerability. As 10 you have said, you have done a trial where the 11 maximum dose, fixed dose, was 9 g per night and, 12 you know, we either decide that that was a 13 tolerable dose or it wasn't. And, you have the 14 dose response for the effectiveness, and that is 15 all you have. As you acknowledge, you haven't 16 explored higher doses so I don't think we really 17 know, and I don't know how you would really get at 18 the question of how the therapeutic window, if 19 there is one, compares to other drugs that are in 20 common use. Some drugs that are used, there is a 21 belief that they have a very narrow therapeutic 22 windows, and some are wide. I don't think you can 23 say more than that. 24 DR. REARDAN: I don't disagree. 25 DR. GUILLEMINAULT: I have a question. 287 1 Narcoleptic patients have hypnagogic 2 hallucinations. They may even shoot -- if a gun is 3 available they may hurt their bed partner because 4 they are keeping their hallucination. How much 5 does Xyrem decrease hypnagogic hallucinations, 6 which is a very significant side effect which may 7 kill neighbors and may kill even bed partners? 8 DR. REARDAN: If I understand the 9 question, Dr. Guilleminault, it is how much did 10 Xyrem reduce hypnagogic hallucinations in our 11 trials, and I guess my first response is the 12 incidence was very low and we did not see a 13 statistical significance in GHB-2. I don't know if 14 Dr. Houghton wants to comment further on hypnagogic 15 hallucinations. 16 Just while they are finding the data, it 17 is fair to say that the incidence of hypnagogic 18 hallucinations recorded in the four-week trial was 19 very low. There was a trend towards improvement 20 that certainly didn't reach statistical 21 significance. There was a better representation in 22 the long-term open-label study and we could show 23 that but I am loathe to do so because I certainly 24 don't want to claim it as efficacy. I think we 25 will be able to find the GHB-2 data. 288 1 [Slide] 2 DR. HOUGHTON: In the Lammers study there 3 was a reduction from 0.87 hypnagogic hallucinations 4 per night over the 4-week treatment period to 0.28 5 incidence per night, with a p value of 0.008. That 6 is one set of figures. 7 DR. MIGNOT: Just to sort of expand on 8 what you said, if only about 40-60 percent of 9 patients we narcolepsy/cataplexy have hypnagogic 10 hallucinations as their symptoms or sleep 11 paralysis, then obviously that must reduce the 12 power for the trial because they have only about 13 half of the patients they included who even had 14 that symptom. 15 [Slide] 16 DR. REARDAN: This is a slide from GHB-3. 17 I guess that is open label, I don't know if we want 18 to go into that. What it shows is median change 19 from baseline to visit number and out through 12 20 months. You see a median change in hypnagogic 21 hallucinations, a reduction of 0.35 per day. Is 22 that right? 23 DR. KAWAS: Dr. Penis and then Dr. 24 Falkowski and then this committee will be looking 25 at the questions that the FDA has asked us to vote 289 1 on. 2 DR. PENIX: I think we have to anticipate 3 several different possibilities in the treatment of 4 patients with any drug, and I am somewhat concerned 5 about the fact that the effective dose of Xyrem 6 appears to be the maximum dose available, number 7 one. Secondly, in regards to the possible 8 protective effects of stimulants on the side effect 9 of sedation, and whether we should consider Xyrem 10 as a monotherapy drug or as an adjunctive 11 treatment, and the question I would like to ask -- 12 I think Dr. Houghton may have presented this data 13 of talked about it, of the 15 percent of patients 14 who did not receive stimulants while on Xyrem 15 whether there was a difference in the maximum dose 16 escalation in those patients compared to the ones 17 who were on stimulants. I am not sure if we can 18 answer the question, but if there is data on that, 19 if there is a difference. 20 DR. HOUGHTON: No, we don't have data 21 separate for those on stimulants and those not on 22 stimulants. There was only about 15 percent in 23 that controlled trial that were not on stimulants. 24 So, we hadn't plotted that at all. Remember that 25 stimulants are taken in the morning and usually the 290 1 last dose at lunch because narcoleptics are really 2 trying to sleep at night and stimulants really 3 complicate that, and the half-life of the gama 4 hydroxybutyrate is about an hour. 5 So, even after their second dose their 6 plasma levels on awakening in the morning are 7 extraordinarily low. So, a contribution of 8 stimulants to change that is quite unlikely. We 9 certainly didn't see an abnormal sleep response in 10 the normal volunteers in any of the pharmacokinetic 11 studies, except the one patient who became 12 obtunded, and she was awake four hours later and 13 ate lunch, and then went home that day. So, the 14 only real suggestion of data I could give you in 15 the absence or stimulants is the single dose 16 response or the repeat dose response in the 17 pharmacokinetic studies, and that certainly didn't 18 appear to be different at all. 19 DR. BLACK: I would just comment on the 20 notion of a potential protective effect with 21 stimulants. With the traditional stimulants, they 22 are relatively short acting and there is a 23 phenomenon called rebound hypersomnia as the 24 medication wears off -- well demonstrated in 25 animals and humans -- where the individual becomes 291 1 more sleep than they would have been had they not 2 taken a medication; often a problem for those with 3 narcolepsy who are using those medications. 4 Rather than those stimulants keeping 5 people more awake and less affected by the Xyrem 6 dose, there is the potential for even greater 7 sleepiness with that rebound hypersomnia. That has 8 not been well explored, but I think it would be 9 erroneous to assume that there is any protective 10 effect from the traditional stimulants. From the 11 longer acting stimulant, modafinil, sleep studies 12 have been done to suggest that there is no impact 13 one way or the other on sleep in terms of depth of 14 sleep and so forth. 15 DR. KAWAS: Dr. Falkowski? 16 DR. FALKOWSKI: I have to take issue -- 17 well, I already did with the statement that Xyrem 18 will not contribute to the public health problem of 19 abuse of GHB-like substances because I think it 20 will and I want to take just a few minutes to 21 elaborate on why that might be something I couldn't 22 cover in the confines of my 15 minutes as well as 23 covering those other points. 24 I had occasion last week, in Philadelphia, 25 to present at a conference on drug abuse addiction 292 1 professionals from around the country, and since I 2 speak about drugs of abuse, when I got to GHB I 3 said, so, tell me about GHB in your community. 4 Having heard from 15 people from 15 distinct parts 5 of the country on this, a common theme emerged and 6 that had to do with the fact that people who were 7 abusing it couldn't quite get the dosing right 8 because they kept passing out. Passing out became 9 sort of a way of life. I think in Dr. Dyer's data 10 we even saw that as well. 11 This is a drug that causes people to lose 12 consciousness and in some cases respiratory arrest. 13 Well, I think this is particularly relevant because 14 if dosing is the problem I believe that this will 15 only make more attractive a predictable dose as a 16 known entity in a prescription product. "Gee, I can 17 get around all these dosing problems by getting the 18 prescription." 19 I am also concerned that none of the 20 sponsor's packaging that I looked at even mentions 21 the word gamma hydroxybutyrate, or did I miss that? 22 I looked for it; I didn't see that. That concerns 23 me because, as we have seen with oxicodon, we know, 24 for example -- and I think it is a good case, we 25 know that narcotic addicts will seek out 293 1 prescription narcotics for predictable dosing and 2 for predictable purity. And, we have seen an 3 increase once long-acting oxicodon was developed -- 4 we have seen an expansion in its prescribing not 5 just for chronic pain but for the treatment of even 6 acute pain. That plays out to the tune of 300,000 7 oxicodon prescriptions in 1998 and over 5 million 8 oxicodon prescriptions in the year 2000. 9 What people have to do, what drug seekers 10 have to do to acquire it is go to a doctor and 11 feign pain. This happens with unsuspecting doctors 12 and it is happening in all parts of the country. 13 Now, diversion of drugs does not occur by 14 people storming with machine guns the one central 15 manufacturing. It occurs at the patient-doctor 16 level. And, I am very concerned about the 17 possibility of folks who are having trouble. 18 Again, this is a diverse population; it is not just 19 kids using drugs. This is weight-lifters, these 20 are people seeking effects, going to a doctor and 21 saying, gee, you can get around all that; just go 22 to a doctor and tell him you are sleepy. Just go 23 to a doctor and tell him you collapsed. This is 24 really seriously my concern about this, and I don't 25 think that these two issues are separate. This 294 1 drug has a huge following. 2 DR. KAWAS: I would now like to focus on 3 the questions that the FDA has asked us to vote on. 4 Do you feel very strongly that your comments are 5 necessary before that? 6 DR. RISTANOVIC: I am going to make a 7 comment extremely brief. The comment is very brief 8 because in today's time we know how to diagnose 9 narcolepsy. So, there is no way, even if someone 10 is trying to malinger, to be given a diagnosis 11 without appropriate testing in the sleep lab. That 12 is a prerequisite. 13 DR. KAWAS: Thank you. 14 DR. RISTANOVIC: That is all. 15 DR. KAWAS: The FDA has given us three 16 questions that they want this panel to vote on, and 17 a whole page and a half of other items that they 18 would like this committee to discuss. 19 So, I would first like to ask them if it 20 is acceptable to facilitate the discussion, can I 21 make the decision to split the first question into 22 two? 23 DR. KATZ: Absolutely. 24 DR. KAWAS: Thank you. It might be the 25 only thing that gets done quickly today. The first 295 1 question is going to be has the sponsor 2 demonstrated efficacy of Xyrem for the proposed 3 indication to treat cataplexy? I am opening the 4 floor for discussion on that. Yes, Dr. Katz? 5 DR. KATZ: Again, I think it is very 6 important for us to hear a discussion about dose 7 and which dose. I mean, I mentioned that earlier 8 in my comments this morning, but if you could 9 address that it would be very helpful. 10 DR. KAWAS: Absolutely. In fact, maybe I 11 would like to facilitate this part because I think 12 this is the easiest thing that is going to happen 13 in the next hour. To my mind, there have been two 14 pivotal studies that have suggested efficacy for 15 this drug in relationship to cataplexy at the 9 g 16 level. Maybe by making that not overly provocative 17 comment we can stimulate discussion. Does anyone 18 want to comment on the dose or the effect on 19 cataplexy before we vote? 20 DR. FALKOWSKI: Is that the recommended 21 dose? It is not. That is why I am sincerely 22 confused because the study seemed to show efficacy 23 at 9 g, yet, the recommended dose is something 24 other than that and that needs explanation. I 25 don't understand that. 296 1 DR. KAWAS: Any other comments? Richard? 2 DR. PENN: I was going to make it a motion 3 so we would save some steps. I think it is very 4 clear that what you said is a good summary of the 5 case that, in fact, they haven't set the dose at 9. 6 They have suggested a different dose regimen and 7 that has to be looked into very carefully. But the 8 one thing I think we all we agree on is your 9 statement. I would, therefore, put it as a motion, 10 since we are supposed to do a motion so that that 11 has been shown. 12 DR. KAWAS: Would you like to make a 13 comment, Gerald, before we pick the motion that is 14 about to be on the floor? 15 DR. VAN BELLE: Sure. Well, I think it is 16 the issue of dose response that I am struggling 17 with in this case in terms of the pharmacokinetic 18 model. If you assume that there is a 19 pharmacokinetic model that is dose related, I would 20 say if evidence has been shown for an effect at 9 21 there is probably an effect at 8.5 as well. Well, 22 where do you draw the line at that time, and I 23 don't quite know where to do that. I think there 24 is ambiguous evidence for an effect at 6 and one 25 study showed that. So, if you want the technical 297 1 answer, I think there is only evidence for clinical 2 effectiveness at 9 but that ignores, to my mind, 3 the pharmacokinetic aspects of the data so I am 4 struggling with this. 5 DR. KAWAS: Could we restate Dr. Penn's 6 motion that this committee vote on whether or not 7 there has been efficacy demonstrated of this drug 8 for the treatment of cataplexy and, specifically at 9 the dosage of 9? 10 DR. SIMPSON: This may be my ignorance, 11 but when something is labeled, for example, that it 12 is efficacious at a dose of 9, does that mean that 13 a doctor would necessarily prescribe it at 9? He 14 could prescribe it quite a lot higher, couldn't he? 15 DR. PENN: That is going to get us into 16 the next thing, which is how this is going to be 17 monitored. Because it sounds like we want to put 18 an absolute dose limit and we don't want to allow 19 variability in the population. By the technical 20 way we are going to allow this out, if they are 21 going to be watching how much a patient can take, 22 then is a doctor going to be allowed the latitude a 23 patient more, and you are asking can they be given 24 less? I think the answer is usually the doctor 25 makes that decision. Everybody understands that is 298 1 the mean does that you have to use but that doesn't 2 mean your patient will respond to it. So, there is 3 the latitude unless we put into force this 4 voluntary program. 5 DR. KAWAS: I would like to focus this 6 committee back on the questions or we will never -- 7 well, we will have everyone on a plane without a 8 quorum in order to vote on these issues. 9 The first question really isn't so much 10 about safety and what a doctor will do, the FDA has 11 just asked us have they demonstrated efficacy for 12 this drug in either of the two indications. 13 DR. FALKOWSKI: I believe they have 14 demonstrated efficacy for reducing cataplexy in 15 cataplectic narcoleptics on stimulant drugs. I 16 think that is what their studies have shown us 17 today. 18 DR. KAWAS: Okay. We will be taking a 19 vote and everyone's vote is going to count. Are 20 there any other comments people want to make before 21 we put Dr. Penn's motion on the floor? 22 DR. SIMPSON: I really agree that they 23 haven't necessarily demonstrated efficacy in 24 treating cataplexy but really in reducing 25 cataplexy. 299 1 DR. KAWAS: Do you want to put your motion 2 on the floor again? 3 DR. PENN: The company has shown efficacy 4 at 9 g per day using a 4.5 divided dose for 5 treating cataplexy in narcoleptic patients. 6 DR. KAWAS: These votes are going to have 7 to be recorded individually I think. So, can we 8 start with everyone who agrees that the sponsor has 9 demonstrated efficacy of Xyrem for the proposed 10 indication to treat cataplexy? Please raise your 11 hands now. 12 I just want to remind everybody that the 13 voting members of the committee actually are sort 14 of in the central part of the table, beginning with 15 Dr. Simpson and then going around to Dr. Penix. 16 All who agree the company has demonstrated efficacy 17 for cataplexy, raise your hand. 18 [Show of hands] 19 How about if we go around and identify, 20 and start with Dr. Penix for the record? 21 DR. PENIX: I agree. 22 DR. KAWAS: Just your name. 23 DR. PENIX: Dr. Penix. 24 DR. VAN BELLE: Van Belle. 25 DR. PENN: Penn. 300 1 DR. KAWAS: Kawas. 2 DR. WOLINSKY: Wolinsky. 3 DR. ROMAN: Roman. 4 DR. KAWAS: All the people who do not feel 5 the company has shown efficacy for the treatment of 6 cataplexy, please raise your hand and start 7 identifying. 8 [Show of hands] 9 DR. SIMPSON: Simpson. 10 DR. FALKOWSKI: Falkowski. 11 DR. LACEY: Lacey. 12 DR. KAWAS: I think that was everyone, so 13 no abstentions in that case. 14 Moving on to the next hard one, has the 15 sponsor demonstrated -- 16 DR. KATZ: Dr. Simpson and Falkowski, I 17 believe in your comments you said you thought there 18 was an effect demonstrated, or something, but the 19 vote went the other way. I just want to 20 understand. 21 DR. FALKOWSKI: Right, I believe that they 22 have demonstrated that there is some evidence of 23 efficacy for reducing cataplexy in cataplectic 24 narcoleptics on stimulant drugs. These studies 25 have been conducted on people who were already on 301 1 stimulant drugs. We don't know about the 2 cataplectic narcoleptics who weren't. So, I wanted 3 to reflect what we actually looked at, the 4 scientific evidence. 5 DR. KATZ: And, would that be the basis 6 for your no vote as well? 7 DR. SIMPSON: Well, mine is really that 8 they reduced cataplectic events. I guess my 9 understanding of treating it is that they couldn't 10 sort of cure it. 11 DR. PENN: May I just clarify? I didn't 12 mean cure. My motion was not cure, nor did I say 13 monotherapy. 14 DR. KATZ: Right. From the point of view 15 of an effect, you know, that sort of language only 16 being applied to a cure, the vast majority of 17 things we treat and give claims for in indications 18 are for symptomatic, non-curative treatment. So, 19 it is perfectly acceptable for us -- and I think it 20 was implied in Dr. Penn's motion that to vote yes 21 you wouldn't necessarily have to conclude that the 22 drug cures it or wipes these attacks out, but just 23 that there is a decrease in these attacks compared 24 to the control. 25 DR. FALKOWSKI: And you can call it 302 1 monotherapy but what the subjects were in these 2 studies were subjects with the condition that were 3 already under medication for this condition. So, 4 to take that leap to say, well, therefore, if you 5 have people with this condition who are not on 6 stimulant drugs, does that follow? I don't believe 7 it does. 8 DR. KATZ: We will take that under 9 advisement. 10 DR. KAWAS: The next question, has the 11 sponsor demonstrated efficacy of Xyrem for the 12 proposed indication to reduce excessive daytime 13 sleepiness in patients with narcolepsy? The floor 14 is open for discussion on this point. 15 At the risk of putting myself back in the 16 same place as last time, I would summarize what we 17 have seen today with regards to excessive daytime 18 sleepiness that there was one study, in a 19 double-blind fashion, that showed subjective 20 changes in sleepiness with the Epworth Scale, and 21 that would be the GHB-2 study. The other study 22 which is being held up as a pivotal study with 23 regards to daytime sleepiness was the Lammers 24 study, which is a small study. Otherwise, I feel 25 that the evidence with regards to daytime 303 1 sleepiness was very weak at best, in particular, 2 the only study that proactively made daytime 3 sleepiness the primary outcome measure as well as 4 using objective measures with the MSLT was, in 5 fact, negative. All the other studies were open 6 label. So, here I have a little more -- 7 considerably more difficulty actually seeing that 8 the sponsor has demonstrated efficacy for daytime 9 sleepiness. So, what are the committee's thoughts 10 on this? What are the committee's comments on 11 this? Jerry? 12 DR. WOLINSKY: As I tried to point out 13 before, I think this is such an enriched patient 14 population for purposes of the endpoints that were 15 studied, it is hard to know that one could 16 generalize daytime sleepiness effects in a full 17 population of narcoleptics. So, I agree that the 18 data is weak and it is also in a very enriched 19 population. 20 DR. KAWAS: I am not sure I understand. 21 For clarification, enriched with what? You mean 22 enriched for cataplexy? 23 DR. WOLINSKY: Enriched for cataplexy 24 which is not present in all narcoleptics and is not 25 always present at this frequency. So, I don't 304 1 think that we would know. I would not know as a 2 clinical that if I had a narcoleptic with sleep 3 attacks or daytime sleepiness but no cataplectic 4 attacks whether I could expect the drug to work or 5 not, and I saw no data to tell me that I could. 6 DR. KAWAS: Any other comments? Any other 7 thoughts before we call the vote on this question? 8 DR. PENN: I move that the company has not 9 provided information to prove that daytime 10 sleepiness is affected by Xyrem, and I would make a 11 comment on my motion, that if the company sees this 12 as an important thing they can do a post-approval 13 study on that specific item and that would be 14 appropriate. I was leaning at the beginning of 15 this to think that there was too much need for full 16 proof on an orphan drug that this might be the case 17 and I was going to give them the benefit of the 18 doubt, but considering the potential for abuse in 19 patients who will say they are just sleepy and the 20 regulatory problems with that, I think we had 21 better be quite strict on this. 22 DR. KAWAS: Can you make that motion 23 without the addendum? 24 DR. PENN: No, no, the addendum is just my 25 comment. 305 1 DR. KAWAS: Good. Give me the short 2 motion. 3 DR. PENN: They didn't prove their point. 4 DR. KAWAS: The language is has the 5 sponsor demonstrated efficacy of Xyrem for the 6 proposed indication to treat excessive daytime 7 sleepiness in patients with narcolepsy? So, a vote 8 of yes the way I just worded it would suggest that 9 the company has shown efficacy, similar to the last 10 vote. A vote of no would suggest that the company 11 has not shown efficacy for that particular 12 indication. So, all in favor of yes, the company 13 has shown efficacy for the indication of daytime 14 sleepiness, please raise your hand. 15 [No show of hands] 16 All if favor of no? 17 [Show of hands] 18 Let the record show that it was unanimous. 19 It might be the only time today. 20 DR. TITUS: And enter nine names please 21 into the record. 22 [Drs. Penix, Van Belle, Penn, Kawas, 23 Wolinsky, Roman, Falkowski, Simpson and Lacey voted 24 against the motion] 25 DR. KAWAS: Now, the second question that 306 1 the FDA has asked us to vote on is has the sponsor 2 established the safety of Xyrem when used for the 3 proposed indication for which substantial evidence 4 of effectiveness has been submitted? 5 Now, given our previous vote, we are 6 talking about substantial evidence for the 7 effectiveness to treat cataplexy, and I want to go 8 ahead and put in here that I think most of the 9 committee members have been of the opinion that the 10 substantial evidence is almost exclusively in the 9 11 g dose range. So, I think we are talking about has 12 the sponsor established safety of Xyrem when used 13 for cataplexy at a dose of 9 g per day, for the 14 most part. The floor is open for discussion on 15 this question. 16 DR. SIMPSON: Could one of the physicians 17 put the adverse events that one can see in the 9 g 18 in perspective? 19 DR. KAWAS: Let me let Dr. Katz and Dr. 20 Mani answer the question. Dr. Katz? 21 DR. KATZ: Yes, this is why the dose which 22 you think is effective is important. It might be 23 useful, before you decide whether or not the safety 24 has been established at 9 g, to have a look at what 25 the total exposure at the 9 g dose is and whether 307 1 or not you think that is acceptable, as a first 2 step, independent of whether or not it seemed to 3 have been tolerated, with enough people at 9 g with 4 sufficient duration. So, I don't know if the firm 5 could put up a slide. I think Ranjit has an 6 overhead. 7 DR. KAWAS: Slide 67 from the company, 8 updated ISS database, summary patient exposure by 9 dose. By my calculations we are talking about 60 10 years, person years of exposure on the 9 g dose 11 from the integrated data set. 12 DR. MANI: I am sorry, I don't believe it 13 is patient years, is it? It is the number of 14 patients. 15 DR. KAWAS: Well, I calculated it because 16 there were 13 patients who had been on it for 2 17 years or more and 34 patients who had been on it 12 18 months or more. So, it was just 2 times 13 plus 19 34. That is the way I cam to the 60 person year 20 estimate. I actually didn't give them any credit 21 for the 6-month exposure. 22 Actually, I have a question to ask of the 23 company, do each years subsume the others? So, the 24 13 individuals who were in the 2-year category, are 25 they also included in the 62 who are in the 6-month 308 1 category and the 34? 2 DR. REARDAN: Yes, I believe that is 3 correct, Dr. Kawas, the 13 patients would be 4 included in the 34, and the 34 would be included in 5 the 62. 6 DR. KAWAS: So, the math is more 7 complicated than I made it out to be, actually. It 8 still comes to about 47 patient years of exposure 9 by my calculation. I believe that the standard 10 generally if it is considered acceptable is 11 considerably higher than that. Perhaps Dr. Katz 12 would like to comment on that, particularly in the 13 case of an orphan drug with a relatively small 14 patient population. 15 DR. KATZ: Yes, the typical minimum 16 requirements for an application for a standard drug 17 that is not an orphan -- we will start there 18 because we have such standards written, is at least 19 1500 patients total or subjects total, with at 20 least 300-600 for 6 months for a chronic disease 21 and at least 100 for a year. That is the standard 22 ICH minimum data package for safety. 23 As you point out, this is an orphan 24 condition. I guess the company estimates the 25 prevalence of narcolepsy patients with cataplexy is 309 1 about 25,000 or 24,000, something like that. And, 2 we had agreed prior to the submission of the NDA 3 with the company that, because it is an orphan with 4 a fairly small prevalence, that they wouldn't 5 really have to have the full data set that a 6 typical NDA would have, and we agreed that a total 7 of about 500 would be in the ball park. It is 8 understood that at least some significant 9 percentage of those patients should be at a 10 therapeutic dose because the safety accrued at the 11 dose that is less than therapeutic isn't 12 particularly contributory. 13 So, while I don't believe -- the company 14 can correct me if I am wrong, but I don't believe 15 we set in stone what would the minimum numbers be 16 that would be sufficient for either 6 months or a 17 year or total active therapeutic dose. I don't 18 believe we signed a contract about that, but I 19 think the implication is that a big chunk of the 20 data ought to be at therapeutic dose. So, I can't 21 give you an absolute answer but I will throw it 22 back to you and ask would you think that the 23 exposure at the therapeutic dose that you have seen 24 is sufficient to characterize the safety profile 25 reasonably and that we could write labeling that 310 1 would adequately inform prescribers about what the 2 panoply of risks is at 9 g? 3 DR. ROMAN: Could that be solved with a 4 post-release very strict follow-up on these 5 patients, Dr. Katz? 6 DR. KATZ: We really have to be assured 7 that the drug is safe in use at the time of 8 marketing. We cannot rely on post-marketing data 9 to say, well, we will find out if it is safe in 10 use. We have to make a decision about whether it 11 is safe in use as described in labeling, whatever 12 that is going to look like, at the time of 13 approval. There may be additional information we 14 would like to have in Phase IV but the fundamental 15 finding of whether or not it is safe in use must be 16 made prior to approval. 17 DR. ROMAN: A second point that I would 18 like to make is that probably you can say that up 19 to 9 g per day, not that there is sort of the 20 middle of the road -- probably it would be 21 recommended to start with a lower amount and 22 increase according to tolerance and effects, but it 23 is up to 9 g per day. That is sort of the upper 24 limit. It happens to be the most effective one and 25 sort of therapeutic dose but probably you would 311 1 like to start with the lowest possible amount. 2 DR. KAWAS: I think the company shares 3 your interest, but my take on this is we don't want 4 to put out there that a drug is efficacious at one 5 dose and safe at another. I mean, I think it is 6 incumbent on us to feel confident that both of 7 those characteristics go with whatever dose we 8 think is appropriate. 9 In response to your question, Dr. Simpson, 10 and I don't know if I understood it correctly but 11 you said what is the clinical significance, is that 12 from the perspective of a clinical? 13 DR. SIMPSON: Well, that is part of it. 14 Just speaking as a statistician though, the safety 15 evidence isn't there with those kind of numbers, 16 obviously. I mean, I think everybody knows that. 17 DR. KAWAS: I think that is really more 18 the question that is on hand here -- 19 DR. SIMPSON: Yes. 20 DR. KAWAS: -- because from the 21 perspective of a clinical, this drug actually -- 22 you know, if you didn't tell me what the drug was 23 and just showed me ten safety profiles that have 24 gone by this committee in the last decade, or 25 whatever, I suspect this would look like one of the 312 1 best ones. Nobody died from it. No major 2 laboratory abnormalities were detected. But it is 3 very, very, very few subjects that we are talking 4 about, and I think that is considerable concern to 5 us. 6 DR. SIMPSON: There actually was one 7 suicide which could be attributed to this. 8 DR. KAWAS: It still puts it in probably 9 the best of the ten. Dr. Katz? 10 DR. KATZ: Dr. Racusin, on our safety 11 team, just reminded me of sort of a simple rule 12 that we use to decide what sort of size of a risk 13 you can cap with a given exposure, it is called the 14 rule of thirds, but basically with a cohort of 60 15 patient years you could be comfortable with ruling 16 out a risk of no greater than 1/20, which is 17 --what? -- 5 percent. So, in other words, there 18 could be a rate of 5 percent of something bad with 19 a cohort of 60 that you would not have even seen in 20 that cohort. So, just to sort of give you an idea 21 of what sorts of potential risks are there that we 22 might not have seen yet with this cohort size. 23 DR. VAN BELLE: Just a small correction, 24 Dr. Katz. I believe that it should be 3/60, which 25 is 15 percent rather than 20 percent. 313 1 DR. KAWAS: Do we have any other comments 2 before we give a shot at trying to vote on the 3 safety? 4 DR. WOLINSKY: I very much share your 5 concern about approving the drug at one effective 6 dose and then saying the safety is really at a 7 lower dose than what is effective. On the other 8 hand, I do think that we have some reasonable data 9 on the efficacy side that says that the dose ranged 10 somewhere between 6-9 g is effective for a 11 substantial proportion of patients, which we then 12 give us not roughly 50 years of patient exposure 13 but closer to 200 years of patient exposure. 14 DR. KAWAS: I agree with that comment, Dr. 15 Wolinsky, but I really would want to point out that 16 almost all of the SEs appear at the 9, not at the 6 17 range. So, you know, you are stacking the deck a 18 little. 19 DR. WOLINSKY: I thought actually, as I 20 saw the listing of the adverse reactions, they 21 clustered in two modal distributions. One was at 22 the high range and one was, surprisingly, below 6. 23 DR. KAWAS: Actually, maybe we will take a 24 look at that. Could Xyrem put up slide number 70 25 for us, updated ISS database does distribution of 314 1 adverse events? 2 [Slide] 3 I think that is what you are talking 4 about. It is not a perfect dose response. I mean, 5 something pops up in the middle, the 6 range 6 actually in terms of SAEs at 12 percent for the 6 g 7 dose. 8 DR. WOLINSKY: And if I heard correctly, 9 and I don't know how they were distributed, at 10 least some of those serious adverse events were 11 cataplectic episodes. 12 DR. KAWAS: But even then, I mean, I would 13 point out that we are talking about a 3-fold 14 increase in discontinuations due to AEs in the 9 15 versus the 6. I mean, it is a 3-fold difference. 16 DR. WOLINSKY: I take your point. 17 DR. PENN: On the other hand, once again, 18 that looks like a pretty safe drug to me when you 19 are only talking about 15 percent of people 20 dropping out for AEs, and the real-life situation 21 is that these patients are going to be titrated up 22 to the 9 and, as we saw from that graph of the 23 unacceptable information from the standpoint of the 24 study results, in experience over a number of years 25 you can run patients certainly at lower doses than 315 1 9. So, I think that should be influencing our 2 opinion of the safety data. 3 DR. KAWAS: Thanks. Dr. Katz? 4 DR. KATZ: Yes, I think the critical 5 question here is not whether those numbers at 9 g 6 are acceptable or not, although that is an 7 important question, but to me the question is -- 8 and you have certainly been talking about that, do 9 you have enough experience to be comfortable at the 10 dose you think is effective. I think, I mean my 11 sense of what people are saying -- you didn't vote 12 on it yet, but my sense is that you felt that at 9 13 g there just isn't really that much data. I don't 14 want to preempt your vote, but it sounds like the 15 general consensus was there wasn't enough data 16 there -- forget about what the data actually 17 showed, but there just wasn't enough to be able to 18 be comfortable that we have adequately 19 characterized the safety at 9, which is what we 20 have to do. The only vote you took on 21 effectiveness was effectiveness at 9 g. So, if you 22 think it is useful to reopen a discussion about 23 whether or not you think there is effectiveness at 24 6 g, and if you do, then you have considerably more 25 exposure to think about. So, that is your call. I 316 1 mean, Dr. Wolinsky suggested that he thought there 2 might be some evidence of effectiveness at 6. I 3 don't know how the others feel, and I leave it up 4 to you as to whether or not you want to reopen that 5 question because if you do think there is 6 effectiveness at a lower dose, it increases your N 7 from the point of view of safety. So, I just throw 8 that out. 9 DR. KAWAS: I actually think that is 10 probably worth our doing. With regards to 11 effectiveness at 6 g, what are the thoughts of the 12 committee? I will start by saying that I suspect 13 that there is effectiveness for at least many 14 patients at 6 g, partly for all the reasons that 15 other members of the committee have said, but also 16 because there appears to be a fairly prominent 17 dose-response curve not only in terms of AEs but 18 also in terms of efficacy. And, what isn't 19 factored into a total dose is the levels of 20 particular patients, the weights of particular 21 patients or whatever, but the data shows me that at 22 least a subset of patients appear to be responding 23 at least in some of the trials to 6 g. Dr. Katz? 24 DR. KATZ: Study 21, the withdrawal study. 25 DR. HOUGHTON: That is the slide that I 317 1 would really like to show if I could. 2 DR. KATZ: The dose there was 50 mg/kg, is 3 that correct? What was the distribution of doses 4 in that study? 5 [Slide] 6 DR. HOUGHTON: This is shown here. There 7 was an equal distribution of patients at the 6, 7.5 8 and 9 g and if you look at that paradigm of acute 9 withdrawal, the response to placebo randomization 10 is obviously very robust at 6 and 7.5 g, as it is 11 at the 9 g. The problem with the GHB-2 study is 12 that it is only a 4-week study and the slope of the 13 line hadn't plateau'd at the end of 4 weeks. When 14 we did apply that to open label, even though it was 15 open label we still saw the maximum nadir at 8 16 weeks. So, if you then take a group of patients 17 who have been on active treatment for a very long 18 time and are then randomized to placebo, if you 19 believe that is a support for long-term efficacy 20 then efficacy is supported at 6 g and 7.5 g. 21 DR. KAWAS: Would members of the committee 22 like to comment on this data or any other data 23 showing efficacy or non-efficacy at 6 g? Yes? 24 DR. SIMPSON: I do think that this trial, 25 in fact, is very impressive. I just want to remind 318 1 everybody of the caveat of this, that the people 2 that you were looking at long-term exclude all 3 those people who have dropped out for adverse 4 events. 5 DR. KAWAS: I think that is a very good 6 point. I mean, this was a study done in responders 7 rather than just random narcoleptics. Individuals 8 in this group represented probably are individuals 9 who felt they were getting benefit or saw benefit. 10 DR. SIMPSON: And provided the drug is 11 safe, then in fact this might be a fair rule to 12 look at to say, yes, the drug is effective. 13 DR. MANI: I would just like to point out 14 that these comparisons are not of randomized 15 groups. 16 DR. KATZ: They are not randomized to 17 dose. 18 DR. MANI: They are not randomized to 19 dose. 20 DR. KATZ: It is obviously a randomized 21 study. So, they are not randomized to dose in the 22 sense of typical dose response. These are doses 23 that presumably they had been responding to in open 24 experience, and there is not as balanced across the 25 doses, that is true. And, the numbers are quite 319 1 small on each dose. On the other hand, you have 2 already decided that in toto it is a study that 3 demonstrates effectiveness. 4 DR. KAWAS: I mean, I think even though we 5 all recognize these are responders, the fact that a 6 group of individuals on 6 g who, when withdrawn, 7 showed this effect at least told me that there was 8 a subgroup that did respond, as I said before, to 9 6. The question is how big is that subgroup, and 10 when we are talking about indications and efficacy 11 do we feel that on the whole 6 is a dose to which 12 people respond based on all the evidence that we 13 have seen so far? 14 DR. FALKOWSKI: And I would also like to 15 say I am a little uncomfortable with the idea of 16 saying that we have so many patient hours for most 17 drugs but, because this is orphan status, we have 18 it but we don't have -- Dr. Katz' remarks -- but we 19 don't have any numbers. Well, that, to me, puts 20 the sponsor in a difficult situation about, you 21 know, what is adequate in trying to develop a new 22 drug and it makes it very difficult for us here to 23 try to reach a conclusion. Enlighten me, here. 24 DR. GUILLEMINAULT: Can we make a comment, 25 as a sleep expert, on the issue? 320 1 DR. KAWAS: I am sorry, who is speaking? 2 DR. GUILLEMINAULT: Yes, can we make a 3 comment on that issue as sleep experts? 4 DR. KAWAS: Please. Yes, you are on the 5 air. 6 DR. GUILLEMINAULT: Okay. The comment 7 that I want to make is that currently there is no 8 drug for cataplexy which is at a fixed dosage. 9 None. Because there is a certain amount of 10 variability from patient to patient, and a patient, 11 for example, can respond at 20 mg of fluoxetine or 12 60 mg of fluoxetine. In general terms, it is 13 unrealistic to believe that there will be a single 14 dose which will control all cataplectic attacks for 15 all narcoleptic patients. So, you have dose 16 ranges, and I think that that is what these studies 17 are showing. Looking at the data that you have, 18 efficacy for some patients is at 6 or for some 19 patients at 9. And, that is the clinical 20 experience, 20 years of clinical experience. That 21 is the best that you are going to get. So, your 22 efficacy for some is 6 and for some is 9. All 23 drugs used for cataplexy are like that. All 24 patients respond following that scheme. 25 DR. KAWAS: Thank you. Dr. Katz, would 321 1 you like to comment on Dr. Falkowski's concerns 2 about the orphan status? 3 DR. KATZ: The only written rules that I 4 am aware of which talk about numbers that are 5 adequate, or are potentially adequate, for an NDR, 6 or for a typical NDR, there are no numbers written 7 down anywhere as policy or guidance. 8 So, as I say, had agreed that a total of 9 500 was appropriate -- we, the company and the 10 division. 11 DR. FALKOWSKI: So they came up short. 12 DR. KATZ: Well, that is the question we 13 are asking. There was, on our part, that at least 14 a big chunk of that would be at a therapeutic dose. 15 So that is why we are asking you whether or not you 16 think it is adequately chararacterized. 17 I just want to make one other comment with 18 regard to the 6-gram effectiveness and to ask the 19 company just -- should make this explicit, although 20 I think Dr. Trout said it a couple of times. 21 In Study 2, the p-value for the 6-gram 22 versus placebo contrast was 0.0529, or 0.053, I 23 believe. That was including a correction for 24 multiple comparisons given the three doses. 25 So you have one study which, basically, 322 1 has a p-value of 0.05 at the 6-gram dose; right? 2 And then you have what you have seen. So I just 3 remind the committee of that. 4 DR. FALKOWSKI: And that was the four-week 5 study, the GHB-2 study; right? Okay.DR. KATZ: i 6 7 DR. KAWAS: Any final comments before we 8 take a vote on the sponsor establishing the safety 9 of Xyrem when used for the proposed -- well, 10 actually -- 11 DR. SIMPSON: Would it be appropriate to 12 do a revote on the efficacy? 13 DR. KAWAS: Not revote, but we can do 14 another vote on whether or not the panel thinks 15 that there was efficacy demonstrated at -- 16 DR. SIMPSON: A dose between 6 and 9. 17 DR. KAWAS: Well, I think we will have to 18 say either a dose of 6 or a dose of 7.5 or 19 something like that. 20 DR. KATZ: Well, if you conclude it is 21 effective at 6 and you have already concluded it is 22 effective at 9, it would be sort of odd if it 23 wasn't effective at 7.5. So, if you just want to 24 vote it at 6, we will take it from there. 25 DR. KAWAS: Okay. We are voting on 6. 323 1 Has the sponsor demonstrated efficacy of Xyrem for 2 the proposed indication to treat cataplexy at the 3 dose of 6 grams per day? All in favor? All who 4 agree that the efficacy has been demonstrated, 5 raise your hand. 6 [Show of hands.] 7 DR. KAWAS: Let's start and identify 8 yourself as we are going around. 9 DR. SIMPSON: Simpson. 10 DR. ROMAN: Roman. 11 DR. WOLINSKY: Wolinsky. 12 DR. LACEY: Lacey. 13 DR. KAWAS: All who do not feel that the 14 company has demonstrated efficacy at 6 to treat 15 cataplexy, raise your hand. Start identifying at 16 that end. 17 DR. PENIX: Penix. 18 DR. VAN BELLE: Van Belle. 19 DR. PENN: Penn. 20 DR. KAWAS: And I am the lone abstention, 21 I think. 22 DR. FALKOWSKI: Over here. 23 DR. KAWAS: Oh; and Falkowski. So we have 24 a split committee for you on 6. If I vote, I break 25 it. Actually, I am fairly convinced that there is 324 1 efficacy at 6. So Kawas. 2 Now, safety. We are now talking safety 3 between 6 to 9. We are now talking about a lot 4 more patient hours, patient years. The floor is 5 open for discussion for safety between 6 and 9 6 grams a day. 7 DR. PENN: Can the company give us the 8 number of patient years exposure 6, 7, 9, total 9 because we can't do it from your data that we have 10 seen here. How close to the magic 500 are you? 11 Patient years; excuse me. 12 DR. KATZ: Not patient years. 250 13 patients greater than six months, if I added that 14 up correctly. That is without Dr. Scharf. This is 15 now with, so the numbers are bigger. Without Dr. 16 Scharf, I calculate about 250 patients for at least 17 six months. Is that about right? 18 DR. VAN BELLE: I got 399. 19 DR. KATZ: Greater than six months? 20 DR. VAN BELLE: Yes. 21 DR. KATZ: At 6 and above? We can just 22 split the difference. 23 DR. VAN BELLE: How many Ph.D.s does it 24 take to add nine numbers? 25 DR. KATZ: I am not a Ph.D. I can't be 325 1 expected to. Can you put the slide back without 2 Dr. Scharf? 3 DR. KAWAS: I come to about 150 patient 4 years of exposure just looking at the individuals 5 who were on at 12 months or more. 6 DR. REARDON: This is the data without Dr. 7 Scharf included from the ISS. 8 DR. KAWAS: I think it is important that 9 we know exactly what we are looking at so thank you 10 for pointing that out to us. On the other hand, I 11 will say that it is to -- my personal impression 12 was that Dr. Scharf's data, although it was the 13 most extensive and the longest term, was collected 14 the least systematically. Given some of the other 15 issues that were brought up about it, it is 16 probably to your advantage to stick with this 17 dataset in terms of AEs. 18 Okay; then the vote is about to be called 19 for. If the sponsor has established the safety of 20 Xyrem when used for the proposed indication at the 21 dose of 6 to 9 grams per day. All who think yes, 22 raise your hands. 23 [Show of hands.] 24 DR. KAWAS: Wait a minute. Something very 25 funny just happened here. It seemed like more 326 1 people were willing to say it was safe at 9 than 2 are willing to say it is safe at 6 to 9? Let me 3 try again. Who thinks it is safe, raise your hands 4 now. 5 [Show of hands.] 6 DR. KAWAS: Identify yourself from that 7 end. 8 DR. ROMAN: Roman. 9 DR. WOLINSKY: Wolinsky. 10 DR. PENN: Penn. 11 DR. KAWAS: Kawas in there. Anyone else? 12 Who does not think it is safe, raise your hands, 13 that safety has been demonstrated, established 14 safety at the dose from 6 to 9 raise your hand now? 15 [Show of hands.] 16 DR. KAWAS: Has not been demonstrated to 17 your satisfaction. Falkowski, Simpson, Lacey, 18 Penix? Anyone else? 19 DR. VAN BELLE: Van Belle abstains. 20 DR. KAWAS: And one abstention. We are 21 really helping a lot. 22 DR. KATZ: I didn't count. Was that a 23 split? 24 DR. KAWAS: Right down the middle. Really 25 helping. 327 1 The third question that the FDA has asked 2 us to consider is the adoption of a risk management 3 plan necessary for the safe use of Xyrem. I would 4 like to focus us on that question. First, in a 5 yes/no way rather than the details of whether or 6 not, of what belongs in a management program if we 7 think yes, or what doesn't belong if we think yes. 8 DR. FALKOWSKI: I thought part of our 9 discussion was going to be different elements of 10 that. 11 DR. KAWAS: That is the next part. First, 12 let's decide do we need a risk-management program, 13 yes or no. And then, if we do, what should be the 14 elements. Jerry? 15 DR. WOLINSKY: I think there are really 16 two issues here. I wish there weren't, but there 17 are two. One is the risk-management program and 18 whether it is critical for the patient population 19 in which the drug seems to be indicated. I 20 actually don't think that is important. 21 Then the question is is there a risk-management 22 program that is necessary for the 23 concerns about the societal risk at large. There, 24 I think the answer is absolutely yes. Because of 25 that conflict, we may be in an unusual position if 328 1 we favor this drug, favoring, potentially, making a 2 precedent step in which we put unusual controls on 3 physicians and patients, more so than we have had 4 in the past. 5 I am not sure there is anything wrong with 6 that, but I am not sure that this is a large enough 7 forum in which this question should be addressed. 8 DR. KATZ: There certainly are precedents 9 for risk-management programs being necessary for 10 the safe marketing of the drug. I don't know that 11 there are many, but there are certainly -- and I 12 think you heard about some. So there is this 13 precedence for a risk-management program. 14 Now, the details--I don't know 15 specifically which details you are thinking about--may make 16 this more of a precedent. But, certainly, 17 risk-management programs of this type or similar 18 type have been used and have been approved. 19 DR. WOLINSKY: I don't disagree with that, 20 but I think we are talking about whether or not 21 there is an inherent problem with the drug in terms 22 of the efficacy, safety level that we are seeing. 23 Most of the risk-management programs that I am 24 aware of that have been put in place have been put 25 in place for the protection of the patient not the 329 1 protection of society. 2 DR. KATZ: Again, you have made a 3 distinction which we have not yet explicitly made. 4 It is a fair distinction. I am not sure everyone 5 agrees that there would be no need for a risk-management 6 program if it was just--if you weren't 7 worried about the societal questions. But it is a 8 fair point for sure. 9 DR. PENIX: Also, isn't it the difference 10 in the fact that this is a controlled substance and 11 the other drugs are not that the safety measures 12 that are put in place for the protection of the 13 patients are usually not controlled substances. So 14 that may be a difference in this particular case. 15 DR. WOLINSKY: This is controlled, but I 16 am not sure that the controlled substances have 17 this much potential control on them is what we are 18 suggesting here. 19 DR. FALKOWSKI: I have a question which is 20 has the FDA ever been in a position where they have 21 a drug coming before them that has already been 22 scheduled? This seems to be unique. 23 DR. LEIDERMAN: Could I just answer a 24 couple of these questions? 25 DR. KAWAS: Please, Dr. Leiderman. 330 1 DR. LEIDERMAN: Let me refer you to a 2 table. It is actually the last page in your blue 3 FDA briefing package book. It actually lists 4 several examples of risk-management plans for 5 different drugs that come from different classes 6 and for different therapeutic indications that are 7 all in place for various safety reasons within the 8 FDA, and they range from other controlled 9 substances, potent opiates in the case of Actiq and 10 fentanyl, to mifeprex and thalidomide. The risks 11 and the intended protected individuals may be 12 different in each case. Obviously, in thalidomide, 13 the risk isn't to the patient but to the accidental 14 fetus. Similarly, much of the consideration in 15 Actiq, which is a potent opiate, was concern for 16 other individuals within the household and, again, 17 not for an opiate-tolerant severely debilitated 18 pain patient. 19 So, to answer Dr. Penix' question, in 20 fact, or Dr. Falkowski's, some of these have been 21 already scheduled drugs. I think what is unusual 22 but not absolutely unique is to start out with a 23 drug that is basically in Schedule I and then to be 24 bringing it into the therapeutic arena but, again, 25 it is not entirely unprecedented either. 331 1 DR. KAWAS: Thank you. I can't help but 2 point out that it is probably unprecedented, but 3 this drug has gone from over the counter, a 4 completely unregulated food supplement that could 5 be bought by anybody ten years ago to Schedule I, 6 which seems to me even more unusual. 7 So we are back to the question about the 8 adaption of a risk-management plan necessary for 9 the safe use of Xyrem. I think the comments that 10 have been made, that Dr. Wolinsky made, was it may 11 not be necessary for the safe use but it is 12 necessary for other reasons. 13 Can we amend what we vote on, whether or 14 not it is necessary, period, for whatever reasons 15 and vote on it in that regard? 16 DR. KATZ: Yes; I would prefer you did, 17 actually. 18 DR. KAWAS: Okay. The real question is is 19 a risk-management program necessary. I have a 20 feeling we are ready to vote on that. So I will 21 call the question. All in favor say aye. 22 [Chorus of ayes.] 23 DR. KAWAS: No? 24 DR. PENN: No. 25 DR. KAWAS: Let the record show that Dr. 332 1 Penn voted no. Any abstentions? 2 [No response.] 3 DR. KAWAS: Dr. Penn, do you want to give 4 your comments, since you were the descending 5 opinion. 6 DR. PENN: I think this is a very 7 complicated issue and I don't think we can resolve, 8 at the end of a committee meeting, the 9 responsibilities toward the general population of 10 controlling the drug and the FDA controlling it for 11 a group of patients. 12 I see that the whole issue is being 13 distorted in the same way that drugs for treating 14 pain have been a problem and that is if we limit 15 the drug with all these regulations, that the 16 patient population, which is quite small, will not 17 be served. 18 That certainly has been true with narcotic 19 drugs over the years, that many, many physicians 20 have underprescribed narcotics for a long period of 21 time. I think we will see the same here except 22 there won't be the same push to get it accepted by 23 cancer patients. The narcolepsy group is much too 24 small. 25 So it is going to be a very hard balance. 333 1 I also worry about the idea of "voluntary" ways of 2 doing this. They are not voluntary on the company. 3 The company wants to get the drug out and they 4 realize that they can't do it unless there are 5 societal controls on the drug and they are willing 6 to do it. 7 But I don't like the precedent of the drug 8 company deciding for a physician whether, for 9 example, somebody 17-years old will get the 10 medication or whether somebody, because of 11 different metabolism of the drug, might not be used 12 on a slightly higher dose than 9. 13 Those are things that we have 14 traditionally let the treating physician do and we 15 have also not let the company choose who are the 16 treating physicians. So I think this is something 17 that needs a large amount of debate and that is why 18 I was being obstinate and voting no on this without 19 qualification. 20 DR. KAWAS: Thank you. Rusty? 21 DR. KATZ: Just as far as the dose and the 22 limitations, that is something that can be 23 discussed in the context of what type of risk-management 24 program you think needs to be in place. 25 You could have a risk-management program that 334 1 doesn't say you cannot ever give a dose greater 2 than 9 grams. 3 In a typical drug, when we have labeling, 4 we have information that the drug is effective or 5 safe only up to dose X, we don't usually say, "You 6 can't possibly give any more." We just say, "Here 7 is the data. There is no data above dose X." 8 So it isn't part and parcel of any risk-management 9 program that you would automatically 10 limit the dose. I supposed you could, but it is 11 not presupposed that that must be the case. 12 DR. PENN: But you might limit age. The 13 other thing is who is going to make these 14 decisions. We were given this in the context of a 15 very particular type of risk management. I think 16 the devil is in the details in these types of 17 situations and to vote yes or no is very difficult 18 without knowing exactly what details we are talking 19 about. They make major substantive differences. 20 DR. KAWAS: Let's go on. 21 DR. KATZ: That is why I wouldn't ask you 22 to vote on the details. 23 DR. KAWAS: That is what I was going to 24 say. Let's go on to the details. I want to remind 25 the committee, particularly because of the lateness 335 1 of the hour, if there is a detail that is not 2 important to you, please don't fill up too many of 3 the airwaves with it so we can get to the ones that 4 are important to you. 5 So the first one is should there be a 6 requirement for additional safeguards; i.e., 7 keeping drugs in a locked storage space in the 8 patient's home. Just for a straw vote to begin 9 with. How many people think that there should be 10 the requirement for a locked cabinet in the 11 patient's home? Anyone who thinks yes? Straw 12 vote. Anyone who thinks no? Straw vote. 13 I think we have got a clear preponderance 14 here. I think I will at least express my thinking 15 is that we don't require patients to keep Demerol 16 or Valium or Halcion or anything else in a closed 17 cabinet, many of the drugs that are potentially at 18 least as abusable as this. 19 Having said that, I think that almost all 20 drugs belong in a locked cabinet. That is the real 21 issue here and I am not sure to what extent 22 requiring it would make one difference or another. 23 So, should there be a requirement for 24 additional safeguards? Can I say, in general, that 25 the committee felt that that was not essential, necessary. 336 1 Should there be additional warnings on the 2 labeling of the dose cups and/or bottle? Any 3 comments? 4 DR. WOLINSKY: I heard something that I 5 thought was very insightful from one of the people 6 who talked to us in the public session and that it 7 would be useful if there was some distinguishing 8 feature about the bottles that could not easily be 9 counterfeited and this was be in everyone's best 10 interest. 11 DR. KAWAS: Thanks. I assume that would 12 be something that the company would do to the 13 bottle rather than something the patient-- 14 DR. WOLINSKY: I assume so. 15 DR. DYER: Are the dose cups to be labeled 16 because those are not? So additional would be 17 additional to that or additional to what is 18 required by law, because they should definitely be 19 labeled. 20 DR. KATZ: If I can just interject. I 21 don't think there is anything required by law. 22 This is what the patient keeps at home. Right now, 23 I think they are just as you see them. There is 24 nothing on them. There is no labeling of any sort; 25 is that right? They are just blank? 337 1 DR. KAWAS: Would the company like to 2 comment? Is any additional labeling planned for 3 the dose cups? Or maybe it is about to be planned 4 for the dose cups? 5 MS. ENGEL: Actually, no. As you know, 6 the poison-control system nationwide is going to a 7 central 800 number as well as having a logo that is 8 "Mr. Yuck" like but better tested for kids. That 9 we expect to be ready in October. At that point, 10 the central pharmacy will put into each of the 11 packages three stickers, one for the bottle and one 12 for each dose computer that will include that "Mr. 13 Yuck" type symbol plus the central 800 number for 14 the entire poison-control system nationwide. 15 DR. DYER: My concern is that if the 16 bottle ever leaves the little dose caps--if you go 17 away for a night, I am going to take my two doses 18 with me. If they are separated from that bottle, 19 no one is ever going to know what it is. 20 MS. ENGEL: As I said, there are three of 21 those labels that will go, so one for each--no; it 22 does not. 23 DR. DYER: It needs to say what it is. If 24 you go stay at a friend's for the night and you 25 have narcolepsy and you take those two bottles with 338 1 you, child-resistant caps are designed to keep 2 children out for one to two minutes. That is it. 3 Somebody will get into that and, if they do, there 4 is no way to know what it is. 5 When they call that number to the poison 6 center, they say, "I have a bottle with a "Mr. 7 Yuck" sticker on it." It needs to say Xyrem and 8 now many milligrams. 9 DR. KAWAS: I would like to call the 10 question. Should there be additional warnings on 11 the labeling of the dose cups and the bottle of 12 GHB? Do I need to separate those two out or can I 13 put the dose cups together with the bottle. 14 Let's start with should there be labelings 15 on the bottles. All in favor raise their hands? 16 [Show of hands.] 17 DR. KAWAS: Is that almost unanimous? No? 18 Labels on the dose cups saying that it is Xyrem or 19 GHB or something. That is unanimous, please note 20 on the record. 21 How about should there be additional 22 warnings on the dose cups and/or bottle of GHB? I 23 am not sure, maybe I should ask, what is the 24 definition of additional? What is supposed to be 25 on there already? Dr. Katz? 339 1 DR. KATZ: I think we are probably mostly 2 thinking of the cups. There was supposed to be 3 nothing on cups. So anything you put on is 4 additional. I don't know about the bottle. I 5 don't know if we were thinking specifically about 6 the bottle. I assume that has all the usual 7 required statements, whatever they are. 8 DR. KAWAS: Are you satisfied by our vote 9 that there needs to be labeling on the dose cups? 10 I think, though, I am starting to feel from the 11 committee that there is some expression of wanting 12 certain kinds of warnings added? No? 13 DR. DYER: If I could just add in, by law, 14 you have to have "Keep out of reach of children," 15 "Don't take with depressant drugs," "Avoid 16 hazardous machinery." So those kinds of standard 17 things would be on there and I don't know that 18 anything else would be required. 19 DR. KAWAS: Dr. Lacey? 20 DR. LACEY: If this is a scheduled 21 substance with implications for--legal 22 implications, why wouldn't we put that type of 23 warning in as few words as possible there. Maybe 24 it would deter someone. 25 DR. DYER: There is already a requirement 340 1 for "Federal law prohibits dispensing of this drug 2 to other than who it is prescribed." There is 3 already a label like that required on 4 prescriptions. 5 DR. PENIX: It could also attract certain 6 people as well, I think. 7 DR. KAWAS: Yes; these warning labels have 8 a mixed response. Can we move on to special 9 concern or advice regarding limitations on the 10 quantity supplied at any one time. Perhaps the 11 sponsor can correct me but my recall is that it is 12 going to be dispensed at one month and then--a 13 maximum of one-month supply at a time? Is that 14 correct? 15 DR. REARDON: We had proposed to the 16 agency initially to start at one month with each 17 patient. As the patients and pharmacists get 18 experience, that might be extended to three months 19 or could be kept to one month. 20 I think the FDA is asking should there be 21 a regulatory or legal description on the length of 22 period that a Schedule III drug should be 23 prescribed. 24 DR. KAWAS: Rusty? 25 DR. KATZ: I am not sure we meant that 341 1 question to be generic with regard to any Schedule 2 III. We want to know whether or not, in this 3 particular risk-management program, there ought to 4 be a provision that says you only get one month at 5 a time, or you only get three months at a time. We 6 just wanted to know what you felt about that. 7 DR. KAWAS: The floor is open for 8 discussion. First, do people think there should be 9 any restrictions on the amount, period, and then we 10 can discuss the timing. So straw vote. All people 11 who think that we should be talking restriction of 12 some sort or another raise their hand. And people 13 who don't think we need to be talking restriction 14 on length of time, raise your hands. 15 We have got a roughly split straw vote 16 with the probable preponderance on the no time 17 limit. Does that help enough? 18 DR. KATZ: Sure. If that is what you 19 think, it is helpful. I can't guarantee we will 20 agree. 21 DR. KAWAS: Having worked in sleep 22 laboratories as well as doing other physician 23 things where certain drugs--I mean, my personal 24 rule has been that drugs that have the kind of 25 potential for trouble, of which there are many, 342 1 many, many of them already in our armamentarium, I 2 never give out more than one month's supply with 3 three refills. 4 DR. FALKOWSKI: That is why I think that, 5 particularly with this, we need to be cognizant of 6 that and that there should be a limitation on that. 7 That is all I wanted to say. And I also don't know 8 where it comes in, or where this discussion 9 happens, but I really believe that a drug, if you 10 look at the third page from the back of the 11 materials the FDA provided about just the 12 scheduling criteria for drugs, that this drug, 13 although it is efficacious for people with 14 cataplexy, with narcolepsy or else on stimulant 15 drugs, that it clearly-- 16 DR. KAWAS: Your point it getting lost. 17 DR. FALKOWSKI: It should be in Schedule 18 II. I believe it should have the dispensing 19 restrictions that are more consistent with a 20 Schedule II drug and I don't believe that would put 21 undue burden on the patients because most of them 22 are already on Schedule II drugs because they are 23 on methamphetamines or other drugs. 24 Somehow, I wanted to say that today. 25 Thank you. 343 1 DR. KAWAS: Do you feel satisfied with 2 what you have heard on that question, Rusty? 3 DR. ROMAN: Claudia, one more point is how 4 are the patients going to be selected. I think 5 would should at least mention that the patient 6 should have a clear diagnosis of narcolepsy with 7 polysomnogram and MSLT 8 DR. KAWAS: You are jumping to Question 6, 9 but why don't we go ahead and do that since I agree 10 that is an important point and I am worried we 11 won't get to it. 12 So what are your thoughts? 13 DR. ROMAN: That patients should have a 14 recent polysomnogram followed by MSLT in order to 15 confirm the diagnosis of narcolepsy. 16 DR. PENN: Who is going to decide whether 17 it really is narcolepsy or not? The government? 18 The company? The person who reads the test? The 19 doctor that is taking care of the patient? That is 20 why I mean the details are very important. You can 21 say that it sounds good that we should have a 22 diagnosis, but these are important points. 23 DR. KATZ: Can I just clarify what we 24 meant? 25 DR. KAWAS: Thank you. 344 1 DR. KATZ: We meant the treating 2 physician, in other words, would make the 3 diagnosis. We certainly, obviously, are not going 4 to get involved in the diagnosis of a patient from 5 where we sit. The company didn't anticipate that 6 they would either if I can speak for them. 7 No; we just meant do you think that the 8 patients have to have a bona fide diagnosis, does 9 the physician who is writing the prescription have 10 to assert, in writing, before the prescription will 11 be filled that, yes, this patient has narcolepsy. 12 Then you can throw this apart and say do 13 they have to assert that the patient has cataplexy 14 and that is what you have decided the effectiveness 15 data supports. So that is a subtlety or nuance of 16 the question you can get to. But specifically with 17 regard to who is going to make the diagnosis, if 18 you meant that question seriously, we meant the 19 prescribing physician. 20 DR. KAWAS: Response to that? Dr. Roman, 21 do you want to give your opinion and then Dr. 22 Wolinsky has a question or comments. 23 DR. ROMAN: I think that there are 24 diagnostic criteria that are sort of fairly well 25 accepted, at least here in the USA. The question 345 1 of should it be a certified polysomographer or 2 should it be one of the certified centers in the 3 nation, we will start getting into the problem of 4 what happened with the patient who lives in the in 5 the middle of nowhere and has no way to get to the 6 next sleep center at 500 miles. 7 DR. KAWAS: Excuse me, but that is not 8 what Dr. Katz asked you. He wants to know do you 9 think the physician needs to certify, however they 10 come to this decision, that the person has 11 narcolepsy, that they need to certify up front, 12 this person definitely has narcolepsy. 13 DR. ROMAN: One of the speakers mentioned 14 that it is relatively simple to get a sleep attack 15 and narcoleptic episodes that are real enough to 16 fool the best unsuspecting doctor. So, since we 17 have objective ways of making a diagnosis of 18 narcolepsy, I think we need to use that for the 19 protection of the public at large. 20 DR. KAWAS: Thanks. Jerry? 21 DR. WOLINSKY: I think this actually 22 frames what is my concern from before about 23 protecting, or treating patients and protecting 24 society. Now I want to get back more to protecting 25 people who are treated. That really gets to an 346 1 issue that we run away from in this country and 2 that is, if we want to be able to push the envelope 3 to be able to provide drugs that may be helpful for 4 patients with true orphan diseases, we probably 5 also have to say that we are willing to make sure 6 that those people have what they say they have and 7 that the drugs are being used in the context of the 8 set of patients in whom they were originally 9 tested. 10 It is one thing to talk about hemorrhoid 11 cream but it is another thing to talk about a drug 12 with a narrow therapeutic window and a diagnosis 13 which can be made with accuracy by experts most of 14 the time and could be misapplied by others a lot of 15 the time. 16 This becomes a critical issue so that if 17 someone is not willing to monitor this, all that we 18 do, in looking at the hard science of what is 19 presented to us, flies out the window as soon as 20 the drug gets approval. 21 DR. HAGAMAN: Can I make one quick 22 comment? I think, as a physician treating these 23 patients, if they have had a PSG and MSLT in the 24 past, there is really no need to bring them back in 25 for another one. At that point, you have to trust 347 1 the physician's judgment that yes, they do have a 2 diagnosis of narcolepsy, they have had the PSG MSLT 3 done. 4 DR. WOLINSKY: I don't think the panel was 5 questioning that at all. 6 DR. MIGNOT: Especially because, in such 7 cases, you will have to stop medications which is 8 another problem. 9 DR. KAWAS: I don't think that was being 10 suggested. So let's move on if we could, please. 11 DR. SIMPSON: I don't know if this fits 12 under it, but the way the question is worded, 13 should there be restricted prescribing for the 14 product. I just want to put in a plea for 15 prescribing for children. As far as I can see, 16 there have been no pharmacokinetic studies in 17 children and children's pharmacodynamic and 18 pharmacokinetic profile can be very different from 19 adults. 20 So, given its complex pharmacokinetic 21 profile, as it is, I would be very concerned if it 22 was prescribed in children based, as is usual, on a 23 way to a BMI. 24 DR. KAWAS: I am not sure that we have 25 answered your question. Actually, I still have a 348 1 question that I want the committee to focus on 2 unless Dr. Katz feels otherwise. Is it important 3 that we decide whether or not it needs to be 4 restricted to people with cataplexy as a component 5 of their illness? 6 DR. KATZ: I am not sure whether or not 7 you think you have made some sort of recommendation 8 about whether or not it needs to be restricted to 9 patients with narcolepsy globally yet. Do you 10 think you have, because I didn't hear it if you-- 11 DR. KAWAS: No; I don't think we have. 12 You are talking now about certifying that the 13 person has narcolepsy, at least on some signature 14 level. 15 DR. KATZ: We did not put in how we you 16 would know that the patient has narcolepsy. We 17 anticipated that the physician would make the 18 diagnosis appropriately. We didn't ask--I don't 19 think we did anyway--about whether or not there 20 should be specific diagnostic criteria that they 21 have checked off or they have had a recent, or ever 22 had a polysomnogram. 23 We anticipate, for purposes of this 24 question, that the diagnosis would be up to the 25 physician to make appropriately without any 349 1 additional specific requirements, but I suppose you 2 could say patients must have a history of 3 polysomnography and other tests, a multiple sleep 4 latency test or an MPT before they can be 5 prescribed this. 6 You could decide that you think that that 7 is appropriate. We left it open intentionally. 8 DR. KAWAS: I think the committee needs to 9 discuss that particular point. I want to make the 10 comment, though, before we get too far, I would 11 tend to leave it open and I recognize all of the 12 things of modern medicine that all of the people in 13 this committee are familiar with because we sit at 14 major medical centers. 15 But there are people with narcolepsy and 16 cataplexy at places that do not have access to 17 sleep-disorder centers and polysomnography. I 18 think that needs to be kept in mind or discussed on 19 some level as we are cogitating about this. 20 DR. ROMAN: The problem is that you need 21 to go through the differential diagnosis of 22 excessive daytime sleepiness and the differential 23 diagnosis of cataplexy. In most cases, that is 24 going to require at least a polysomnogram, a sleep 25 test, to rule out obstructive sleep apnea, 350 1 restlessness, and what have you. 2 So, in most patients, at least those who 3 present for the first time to get this medication, 4 I don't see how you can avoid doing these tests. 5 DR. BLACK: I hate to interrupt, but a 6 point that I think is worth bringing up is that the 7 condition indication here is cataplexy. Cataplexy 8 is a clinical diagnosis not confirmed by any 9 testing or MSLT. If you are going to limit it to 10 cataplexy, I think it is important to recognize 11 that you can't make any verification on the 12 diagnosis with MSLT as far as the cataplexy goes. 13 DR. KAWAS: Since we have you up there, 14 what percentage of people have isolated cataplexy 15 without narcolepsy and sleep attacks? 16 DR. BLACK: It is incredibly rare. 17 DR. KAWAS: Thanks. 18 DR. BLACK: Incredibly so. But, on the 19 other hand, the incidence of cataplexy and 20 sleepiness without an MSLT that confirms it is a 21 modest subset. In other words, if you have 22 cataplexy, you won't necessarily have two sleep-onset REM 23 periods on your MSLT, so we need to keep 24 that in mind so that we don't potentially limit 25 folks with true sleepiness and cataplexy and 351 1 narcolepsy that don't show the MSLT findings. 2 It is not 100 percent specific or 3 sensitive. 4 DR. KAWAS: We have some people over on 5 this side who wanted to-- 6 DR. LEIDERMAN: I just wanted to be clear 7 about the question that I think we were asking. 8 What was discussed internally within the agency was 9 the concern about off-label use. We all know that 10 drugs are used often more frequently for other than 11 their labeled indications. The question we wanted 12 to pose for this specific drug, does the committee 13 recommend restricting its prescription to the 14 labeled indication. 15 DR. KAWAS: So, actually, I think maybe, 16 put in that context, we could call the question and 17 try a vote here. In the opinion of this committee, 18 are we recommending that this drug needs to be 19 restricted in some fashion to on-label use? All in 20 favor? 21 [Show of hands.] 22 DR. KAWAS: Almost unanimously. Negative? 23 [One hand raised.] 24 DR. KAWAS: One negative vote from Dr. 25 Penn. 352 1 DR. VAN BELLE: I am going to abstain 2 because I was out of the room. 3 DR. KAWAS: Dr. Van Belle is abstaining. 4 Everyone else voted yes; am I correct? So, did we 5 give you a better answer this time? 6 DR. KATZ: Yes. All your answers are 7 good. 8 DR. PENN: Isn't this the first time 9 anybody has ever suggested that the FDA should be 10 restricting off-label use of drugs? 11 DR. KATZ: I doubt. I don't know. 12 DR. PENN: Isn't it stated in the FDA, all 13 of your regs, that you do not regulate medicine and 14 off-label use is up to the physician? 15 DR. KATZ: I don't know if it says we 16 don't regulate medicine but, certainly, I think we 17 have the authority to do, I think, plenty of things 18 that some people might consider practice of 19 medicine. So I don't think, as far as I know, 20 there is any--as far as I know, there is no legal 21 bar to this if that is the question you are asking. 22 I think we have done it in the past. 23 DR. KAWAS: I think that I want to make 24 the comment that even if it was the first time that 25 the FDA was doing this, it certainly is not new to 353 1 medicine. Now, insurance companies routinely make 2 us do this. 3 DR. FALKOWSKI: I have one question, I 4 guess, or one concern, and I just want 5 clarification. Did I not read this correctly? I 6 tried to read it all, but nowhere does it says 7 gammahydroxybuterate. Is this correct, sponsors, 8 that there is not the word gammahydroxybuterate in 9 any of these doctor or patient things. 10 In terms of issues here, I think it is 11 very important that the doctor information says 12 what this is. 13 MS. ENGEL: As we worked with our 14 colleagues in law enforcement, they urged us not to 15 put gammahydroxybuterate as the generic name of the 16 materials, et cetera, because they felt, for 17 example, if you are a patient, and you have 18 something in your home that says 19 gammahydroxybuterate, that might actually be an 20 attractant to a babysitter or someone else. 21 So the attempt, based on the advice of law 22 enforcement, was to separate that out. 23 DR. FALKOWSKI: I am not talking about 24 patient materials--to the doctors. Will the 25 doctors get to know? They don't have their 354 1 materials sitting around their home. 2 DR. KAWAS: Excuse me. Dr. Katz, is this 3 a question you would like the committee to discuss? 4 DR. KATZ: I think it is an interesting 5 question. I think we can work it out. The point 6 is well taken and, as the company says, they have 7 gotten conflicting advice for good reasons as well. 8 I think we can work it out. 9 DR. KAWAS: Great. Thanks. 10 DR. LEIDERMAN: I just wanted to respond 11 to Dr. Penn's comment about restrictions on 12 prescribing. Actually, there is some very recent 13 precedence in the non-CNS drug arena. The drug, 14 mifepristone, in fact, was approved under very 15 restricted distribution. It requires signed 16 documents by both physician and patient to be 17 returned to the distributor before--and only a 18 restricted group of physicians who certify to a 19 certain ability to handle the complications are, in 20 fact, allowed to prescribe the drug. 21 So that is a precedent in the non-CNS 22 arena. 23 DR. KAWAS: I am told that somebody on one 24 of our phone lines would like to make a comment? 25 Can you hear us? 355 1 DR. CHERWIN: Yes; I had wanted to make a 2 comment several comments ago, just to briefly 3 reiterate. I agree with Dr. Black said which may 4 be important that not all patients with cataplexy 5 have positive sleep studies. So, in addition to, 6 perhaps, in some cases, sleep studies not being 7 available, this is another concern. 8 DR. KAWAS: Thank you. 9 DR. CHERWIN: Another thing is that 10 cataplexy is not always a crystal-clear diagnosis. 11 Not too many people have talked about that, but 12 there can be cataplexy in the eye of one physician 13 that does not exist in the eyes of another 14 physician. That is a potential problem. 15 Finally, the International Classification 16 of Sleep Disorders, which is to the sleep field 17 similar to what the DSM is to psychiatrists, does 18 not specifically require a sleep study diagnose 19 narcolepsy. 20 I thought those three things might be 21 salient to the discussion especially--since we sort 22 of jumped to the appropriate prescribing section, 23 maybe we can run through the questions there and 24 see how many of them we can quickly comment on for 25 Dr. Katz and the agency. 356 1 Should physicians document that they read 2 the material sent to them before the pharmacy fills 3 the initial prescription? If we took a straw vote 4 right now, how many people would say yes? How many 5 people would say no? Since we have got a split 6 here, of the people who are on the yes side right 7 now, would some of you like to comment on what kind 8 of documentation you want? 9 I mean, are we talking a signature saying, 10 "I have read the materials that were sent to me," 11 or are we talking about something more than that?. 12 Jerry? 13 DR. WOLINSKY: Again, it sort of depends 14 what we require or what might be expected for a 15 diagnosis rather than what would be required. I 16 think if a sleep specialist is comfortable with the 17 diagnosis in that patient, and refers the patient 18 back to treatment to that physician who is back in 19 North Dakota that you keep mentioning that can't 20 possibly have all of the diagnostic tests around, 21 then I think it is important that that physician in 22 North Dakota knows what they have signed on to. 23 If it is the sleep specialist who has got 24 150 patients on treatment because they are very 25 expert at this, if they have signed the document 357 1 once, that is probably enough for me. 2 But I think these are details that I am 3 not sure that we need to work out today. There are 4 plenty of things that can be worked out by Russ and 5 his people. 6 DR. KAWAS: Russ and his people gave us 7 this question. 8 DR. KATZ: And we didn't anticipate, 9 necessarily, a vote. But right now, as I 10 understand the program, the initial prescription is 11 filled and then the physician and the patient have 12 to send back a card that says, "Yes; I read this 13 stuff." It was just some sentiment internally for 14 all of that documentation that, "Yes; I have read 15 it. Yes; I understand it," that is to happen even 16 before the first prescription was filled. 17 We are going to get into major problems if 18 we try and apply a different standard to different 19 types of treating physicians, the expert versus the 20 non-expert. Actually, this was one of the issues 21 that I actually did want. A lot of them are not 22 necessarily that critical but this was one of the 23 few that I really wanted some discussion on. There 24 are a lot of other details I think we can take care 25 of. 358 1 DR. WOLINSKY: But I guess I was saying 2 that, that even the expert would sign it. He just 3 wouldn't have to sign it every time he gives out a 4 new dose. 5 DR. KATZ: No, no, no, no. We don't 6 anticipate that. 7 DR. KAWAS: Once. 8 DR. KATZ: I just meant the first time you 9 give a dose to a particular patient, you would sign 10 a card before the initial prescription was filled 11 for that patient. That is what I think we 12 anticipate. 13 DR. FALKOWSKI: On a patient by patient? 14 DR. KAWAS: I want to make the comment 15 that I am comfortable with the notion of physicians 16 having to sign for this potentially, but I am not 17 comfortable with what was suggested as a mechanism 18 to have it happen by the sponsor and that is 19 sending a drug representative to the physician's 20 office. I really feel very strongly that is not 21 the way this should be done. 22 Dr. Penix? 23 DR. PENIX: This is a question for Dr. 24 Katz. What is the purpose of the physician signing 25 such a document? 359 1 DR. KATZ: It is just to acknowledge that 2 they have read the material and that they are 3 familiar with its safe use and that they have 4 spoken to the patient about its safe use. 5 Actually, that is a separate question, but it is 6 all combined--that they know how the drug should be 7 used, what its risks are, what the penalties are 8 for inappropriate use. 9 DR. KAWAS: Doesn't it also sort of 10 acknowledge that this is a somewhat unusual drug in 11 some sense because every drug has all these risks 12 in prescribing and we don't ask any physician to 13 sign for all those drugs. 14 I sense on the committee a growing concern 15 that the more drugs we have to sign for, the more 16 uncomfortable they are becoming. But I think, 17 really, it points out to the physician who is 18 signing it that there is something different here. 19 DR. PENIX: I think, also, in that sense, 20 it is important for the physician-information 21 packet that they are aware that this drug is GHB 22 and so, therefore, they may understand why it is 23 required for them to sign this information. 24 I think that is really the bottom line. 25 So I think it would be useful for a treating 360 1 physician to know what type of drug this is. 2 DR. FALKOWSKI: I would say yes only if it 3 says it is GHB. 4 DR. DYER: Wouldn't CII make that implicit 5 to know that this is a drug that has illegal 6 implications and would be dangerous? 7 DR. KATZ: It is Schedule III. 8 DR. DYER: I am saying it belongs in 9 Schedule II. 10 DR. KATZ: I think that question has been 11 dealt with definitively. It has been legislated as 12 Schedule III by Congress. 13 DR. FALKOWSKI: Right. That was 14 legislated at another time. 15 DR. PENIX: Not to belabor this, but I 16 agree with that drug company's position not to let 17 the patient information--or not include GHB in the 18 patient information. But I think the treating 19 physician should be aware of that. 20 DR. KAWAS: I think that is a very 21 important point because physicians do have a 22 knowledge base of GHB even if it is from the 23 newspaper or whatever to insure that they 24 understand what it is. 25 DR. ROMAN: It also has the legal 361 1 implications of a physician somewhere who has been 2 prescribing this at a higher rate than expected for 3 that population. He may find his licensing--and a 4 problem if they find that he is prescribing more of 5 these, let's say more than a couple of patients in 6 a year, or whatever it is that delimits. 7 So we need to look into that because there 8 is potentially a risk for medical licensing. 9 DR. KAWAS: Can we see if we have shifted 10 the straw vote from about a 50:50 split to 11 something that is more consensuslike for the 12 agency? On the question, should physicians 13 document that they read the material sent to them 14 before the pharmacy fills the initial prescription, 15 presumably, some of those materials would 16 incorporate the fact that what this drug really is 17 is GHB whether or not it is on the bottle. 18 All in favor? 19 [Show of hands.] 20 DR. KAWAS: Nos? 21 [Show of hands.] 22 DR. KAWAS: And no abstentions. So let 23 the record show that nos were Dr. Richard Penn and 24 Dr. Gerald Van Belle. The remainder of the 25 committee voted yes. No abstentions. 362 1 Should physicians be required to 2 demonstrate safe use and appropriate dosage 3 preparation to patients before the first 4 prescription and be required to document that it 5 has been accomplished? Do we want to try a straw 6 vote and see if we can keep on going? 7 I think I will make the comment that 8 patient education is too important and sorely 9 underdone in this medical world that that is true 10 for everything. I think, personally, that it would 11 be the hope that, with all drugs, that the 12 healthcare team will insure these demonstrations. 13 I am going to suggest that we do not need to 14 require any specific demonstration or any specific 15 certification of this process. 16 I see some heads going in different 17 directions. Let me get a straw sense on this one. 18 Should physicians be required to demonstrate safe 19 use and dosage? How many people are going to say 20 yes? Straw vote. 21 DR. FALKOWSKI: Is the intent here that it 22 just be demonstrated regardless of who does it, 23 whether it is a nurse or a physician? What is your 24 intent? 25 DR. KATZ: The intent was that--I don't 363 1 think we necessarily meant the physician but 2 someone responsible in the physician's employ. It 3 shows them how to draw it up and how much your dose 4 is. 5 DR. FALKOWSKI: Should somebody 6 demonstrate how you administer this drug before the 7 patient takes it. So I think that is a good 8 question. Can we take a vote on that? 9 DR. KAWAS: You mean someone in the 10 physician's office should be required to 11 demonstrate it and, in some way, ascertain it. The 12 question is called on that. Who votes yes? 13 DR. VAN BELLE: Before we vote, there is a 14 further addition to that statement here, and it 15 says, "And be required to document that it has been 16 accomplished." Are you intending to have that 17 included as well? 18 DR. KAWAS: I think everything that 19 happens in a physician's office needs to be 20 documented. So, yes. That is why we are writing 21 twenty-seven page H&Ps right now. 22 So we have got one vote yes? Is that all? 23 Dr. Falkowski. No votes? 24 [Show of hands.] 25 DR. KAWAS: Abstentions. 364 1 [One hand raised.] 2 DR. KAWAS: We have got one abstention 3 with Dr. Simpson and the remainder of the committee 4 voted no. 5 DR. WOLINSKY: Having voted no on that in 6 terms of the office personnel and the physician, it 7 seems to me that it would be advantageous to the 8 company to have first doses shown in the home when 9 medication arrives. This is actually the effective 10 education. 11 What goes on in the physician's office, my 12 bias is, may not be as effective as with home nurse 13 agents. 14 DR. KAWAS: I think we are not going to 15 repeat the restricted prescribing for the drug 16 question. We have gone over that adequately, I 17 hope. 18 But the next one, does the risk-management 19 program assure appropriate prescribing or 20 sufficiently reduce the risks of misuse or 21 overdose. I am not quite sure where to start with 22 this one. Actually, Dr. Katz, which components of 23 the risk-management program are you asking us to 24 comment on? 25 DR. KATZ: That is a fair question. This 365 1 is sort of a global question, I think. To the 2 extent that you have seen the details of the 3 proposal, is there anything that leaps out at you 4 as being absolutely inappropriate, or is there 5 something that is not there that is a glaring 6 omission that you all believe absolutely should be 7 there? 8 I think that is sort of the sense of the 9 question. 10 DR. PENN: Yes. I don't think the 11 potential problems of the drug are explained to the 12 patient adequately. That is, the narcoleptic 13 patient won't necessarily know that this is an 14 abused drug or if they take it in the wrong way 15 that they can get into a lot of trouble and that 16 the real education has to be to the patient in some 17 manner. 18 I usually think that is the responsibility 19 of the physician to do that, but I don't see that--I mean, 20 we are protecting the patient from knowing 21 what the name of the drug is. We are protecting 22 them from knowing what the real side effects might 23 be. 24 It doesn't say that if you take double the 25 dose, it may have more than double the effect and 366 1 that you may go into coma and become incontinent 2 and have seizure--well, probably not seizure but 3 stop breathing or something unpleasant like that. 4 I think the emphasis should be on the 5 patient understanding the medication and how to use 6 it. The narcoleptic community suffers enough and 7 has pretty good ways of letting each other know 8 about the disease. Maybe you should use their 9 ability to instruct patients on the proper way to 10 do it and combine it in some way. 11 But that is where I think the glaring 12 error is. This is a drug with very little leeway 13 for dosing and people have to understand they 14 shouldn't use it during the day, for example, 15 because they won't have this period of time off. 16 So I think there is a huge amount to be 17 done. I just don't like to see it done in this 18 mandatory fashion because I don't think it will 19 work. You will get a lot of signed papers, but you 20 won't get the education you need done. 21 DR. KATZ: But I just want to clarify. I 22 understand your reservations about the entire 23 process but, given that there is a document that 24 goes to the patient that ostensibly tells them what 25 they need to know about using the drug safely, you 367 1 believe that that document that is currently 2 written really needs to be beefed up as far as 3 communicating to the patient what the risks are and 4 how to use it? 5 DR. PENN: Yes; I think that the patient 6 has to know what it is, that it is an abused 7 substance that potentially can be abused. It would 8 be like our not telling patients who use oxicodon 9 not to chop it in two and take it. That gets them 10 into trouble and they ought to know about that. 11 So there is a lot of education that has to 12 be done with this medication. 13 DR. FALKOWSKI: I think I already 14 addressed this question by saying I think the word 15 gammahydroxybuterate should appear for patients and 16 particularly for the physicians, the prescribing 17 physicians. What is the secret? The way to have a 18 drug come into the market when it is already a 19 substance of abuse is not to pretend it doesn't 20 exist and not even call it what it is. 21 I don't think that is an informed approach 22 for physicians to know what it is. 23 DR. LACEY: Just as one presenter, and I 24 don't remember who, today gave us the common names, 25 the club names and everything. I think the patient 368 1 actually should be provided with as much of that 2 information as possible. To not want to put it on 3 the printed book or something because it is exposed 4 to someone else is one thing. But the patient 5 should be provided as much information as possible 6 to know what they are dealing with. 7 DR. KAWAS: Any other comments before we 8 move on to the next question? Jerry? 9 DR. VAN BELLE: Let me just make a 10 comment. I agree with that and, also, from the 11 practical point of view, we have already heard this 12 afternoon that the narcolepsy website network is 13 just far flung. If this is going to be approved by 14 the FDA, the word will be out in the next fifteen 15 minutes. 16 So to play coy and not put it on one set 17 of labels is just not going to work. 18 DR. ROMAN: I completely agree. The USA 19 Today had the title, "Company wants date-rape drug 20 approved for a sleep-disorder treatment." If that 21 is in the newspapers-- 22 DR. FALKOWSKI: This question is--it is my 23 understanding, and I asked for clarification for 24 this prior to the beginning of this meeting today--that we 25 are voting here on specific questions. Is 369 1 the determination of approval made upon FDA's 2 consideration of what we talked about today? 3 DR. KATZ: Well, sure. 4 DR. FALKOWSKI: Is it made today? 5 DR. KATZ: Is the decision about what to 6 do with the application made today? Absolutely 7 not, no. Your opinions are all advisory. We take 8 them very seriously and then we go back and we 9 discuss it internally and we come to a decision, by 10 the PDUFA due date. 11 DR. KAWAS: Going to the next question, 12 can I ask, Dr. Katz--tell us what do you mean by 13 certification and certification of physicians for 14 prescribing? 15 DR. KATZ: There was some sense, 16 internally, on the part of some people that 17 physicians should--first of all, that it might be 18 restricted to use only by sleep experts or 19 physicians would have to somehow take a test to 20 show that they know about narcolepsy, that sort of 21 thing, that they are appropriate prescribers in 22 some sense. 23 DR. KAWAS: So we are not talking about 24 the same thing that we were talking about 25 previously, documenting that they have read 370 1 whatever materials with the first prescription that 2 they write? 3 DR. KATZ: It is something more than that. 4 DR. KAWAS: Okay. Let's take a straw vote 5 on that. I think we can get past that one 6 potentially fast, then. We are talking about more 7 than just documenting that you have seen materials. 8 Should certification of physicians, or some other 9 restrictions, for prescribing Xyrem be required? 10 Straw vote. How many people think yes? How many 11 people think no? How many people are abstaining? 12 Let the record show that Dr. Wolinsky 13 abstained. I am not sure, but I need to know why. 14 DR. WOLINSKY: Well, I am internally 15 conflicted on this. When I say conflicted, I don't 16 mean that I have some stockholdings anywhere but 17 that I am-- 18 DR. KAWAS: Anyone knows when they use 19 that word they have time on the floor. 20 DR. WOLINSKY: I haven't come to a final 21 decision in my own mind, but I would lean towards, 22 I guess, certification of physicians when the 23 circumstances are special. That doesn't actually 24 keep patients from assessing care. It may mean 25 that they have to be diagnosed in an appropriate 371 1 situation and then can be cared for by a physician 2 who is willing to educate themselves about how to 3 best use the drug. 4 I know that most of my colleagues won't 5 like this but I think that this is where we have to 6 go if medicine is to maintain credibility with an 7 increasingly complex medical world that we live in. 8 DR. KAWAS: Now to go backwards to No. 5, 9 which the questions deal with safe use by the 10 patient. Should the patient sign an informed 11 consent form before receiving the initial shipment 12 of the drug? Straw vote. How many people think 13 yes? How many people think no? 14 I won't ask Dr. Penn. 15 DR. PENN: I am worried about the medical-legal 16 implications of informed consent in this 17 situation. What does informed consent mean? Who 18 signs it? All the things we get to in the 19 controlled trials and that we deal with daily in 20 the university setting. 21 It seems to me that, unless we work out 22 the details, I can't feel comfortable voting for 23 it. 24 DR. KAWAS: Actually, I abstained on the 25 straw vote. My concern, and maybe my question is, 372 1 informed consent about what? Presumably, we are 2 talking about some version of the education that we 3 have said they need to have. So is this just an 4 acknowledgment of that education? What is it we 5 want to make sure that they are informed about and 6 get a signature to verify that? 7 DR. KATZ: Usually, informed consent is--it mostly 8 emphasizes the potential risks. There 9 are drugs, of course, that have informed consent as 10 part of their approval. So that was the question. 11 Given the potential risks of this particular 12 treatment, do people think that patients need to 13 sign an informed consent. 14 It is unusual, but there certainly are 15 precedents for it. 16 DR. PENIX: I think informed consent does 17 imply a certain medical-legal situation but, 18 perhaps, a contract like they use in many pain-management 19 centers so that the patients acknowledge 20 the problems with the dispensing of the drug and 21 that type of thing. So maybe a contract would be a 22 better idea than an informed consent. 23 DR. KATZ: Again, we put it on the list 24 because it was raised internally at several 25 discussions that we had. It doesn't mean that we 373 1 necessarily, as a group, endorse it or most of us 2 think it is a good idea. It was an option. We 3 wanted to see what you thought about it. 4 DR. WOLINSKY: Call that question again. 5 DR. KAWAS: Does that mean you want to 6 change your vote? 7 DR. WOLINSKY: I would like to withdraw my 8 yes because this is much more complicated than 9 immediately meets the eye and goes beyond what we 10 really need, given all the other things that are 11 already in this package. 12 DR. KAWAS: Okay. Do we need any more 13 discussion before we call the question the second 14 time? Any other comments people want to make? 15 Should patients sign an informed-consent form 16 before receiving the initial shipment of the drug. 17 All who think yes, raise their hand. 18 [Show of hands.] 19 DR. KAWAS: Let's go around the table and 20 identify the yes votes. 21 DR. SIMPSON: Simpson. 22 DR. FALKOWSKI: Falkowski. 23 DR. ROMAN: Roman. 24 DR. LACEY: Lacey. 25 DR. VAN BELLE: Van Belle. 374 1 DR. KAWAS: All who think no. 2 DR. WOLINSKY: Wolinsky. 3 DR. KAWAS: Kawas. 4 DR. PENN: Penn. 5 DR. PENIX: Penix. 6 DR. KAWAS: Okay; we are set there. 7 Furthermore, should the patients be 8 required to return a registry form before receiving 9 the first shipment? Now, I assume that a registry 10 form that we are talking about is kept by the 11 sponsor? 12 DR. KATZ: Again, this analogous to what 13 we talked about with the physician. The idea here 14 was right now, the plan calls for such a form to be 15 submitted after the first prescription is filled, 16 that they have read the materials, they have 17 received them and they have read them. 18 The question here was just whether or not 19 you think that all has to happen before they even 20 get the first dose. 21 DR. KAWAS: To my mind, that simplifies it 22 considerably, then. Straw vote. How many people 23 think yes, it should be done before not after or 24 with the first dose. 25 DR. SIMPSON: Is this in addition to the 375 1 consent form? 2 DR. KAWAS: This is different than the 3 consent form; yes. 4 DR. SIMPSON: So, would it be in addition? 5 I mean, if they did the consent form, would they 6 need to fill out another form and send it in? 7 DR. KAWAS: I am not sure I am the right 8 person to answer that because I don't know whether 9 or not there is going to be a consent form. But 10 maybe Dr. Katz could-- 11 DR. KATZ: We asked it separately. They 12 are two different things, although they are very 13 closely related, I suppose. If you sign a informed 14 consent that says, "I know what the risks are. 15 "The card--what do we call it--a registry card. 16 That presumably could be something that says, "I 17 have read the material. I assert that I know how 18 to draw the appropriate dose up. I know how to mix 19 it. I know that I have to mix both doses first." 20 They have a sense of how it is supposed to 21 be taken. So you would imagine it would have 22 different information, could have different 23 information, than an informed-consent form. 24 DR. KAWAS: So the registry, actually, 25 has--it is not just a name, address, serial number 376 1 of a person who is getting the drug. That is not 2 what we are talking about in the registry form? We 3 are talking about-- 4 DR. KATZ: I think the idea here was, as I 5 said before, whether or not, analogous to the 6 question with regard to the physicians, that they 7 have read the materials, what I intended, anyway, 8 for this question was the exactly analogous 9 situation for the patient. 10 Should the patient have to send the form 11 back. It would be a registry form, I suppose, in 12 terms of who they are, but the pharmacist already 13 knows who they are so they get into the registry 14 that way, I suppose. 15 But whether or not they have read the 16 material and they understand what the risks are and 17 they understand how to take the appropriate dose, 18 just before the first dose. 19 DR. KAWAS: Okay. Now I think we can 20 better take a straw vote. 21 DR. SIMPSON: I just wanted to say I 22 thought the consent form was that. 23 DR. KAWAS: But, having rephrased it for 24 us, I think essentially what we are saying is now 25 we have said that we want the physicians to certify 377 1 that they have read, know and understand some of 2 the issues, the question is, should we ask the 3 patients to do the same thing. 4 All who think yes, raise your hand. 5 [Show of hands.] 6 DR. KAWAS: And nos? 7 [Show of hands.] 8 DR. KAWAS: I think we have got a bunch of 9 abstentions, mostly. Would you like to comment no 10 your thinking? 11 DR. PENIX: I think it is just pretty 12 complicated. I am not sure what a registry is 13 going to do, what the drug company is going to do, 14 with the information, who should keep the 15 information. There are a lot of different issues, 16 so I guess, in the late hour, I am going to 17 abstain. 18 DR. LACEY: I would think these two things 19 could be combined into one some way or the other. 20 If they can't, it is just getting to be too 21 complicated in terms of all the forms and whatever, 22 so they are losing interest in it. 23 DR. KAWAS: Are you talking about the 24 patient or the committee? No; I think that 25 something really important was just said here, 378 1 actually. I think that if we put too many layers 2 that nobody is going to pay attention to any single 3 layer here. The whole idea is to do exactly the 4 opposite, to have both the patients and the 5 physicians taking this seriously. 6 Anybody can write in a patient's chart, "I 7 have demonstrated how to do a safe dosage through 8 the patient," and signed their initials. That only 9 takes a few seconds. Getting them to spend the 10 time to do it in the office is quite a different 11 thing. 12 Obviously, what is more important is what 13 is actually done and not what is certified. But 14 let me see if I am getting the flavor from this 15 committee that, in general, they think there should 16 be one certification, registration, informed-consent process 17 or whatever for both physician and 18 for patient. Is that the gist of what we have been 19 saying? 20 All who agree with that statement, straw 21 vote, yes. All who think no. 22 DR. PENN: I abstain. 23 DR. KAWAS: Oh, gosh. And Dr. Penn 24 abstains and we are not going to even bother 25 finding out why. 379 1 Dr. Katz? 2 DR. KATZ: Given the late hour and the 3 list that still remains, I don't think we really 4 need much in the way of discussion or even a vote, 5 or a straw vote, on any of the other remaining 6 issues. 7 I would ask, though, the committee members 8 to just sort of quickly glance at it, or not, as 9 you wish. But, again, if there is anything that 10 strikes you as being a glaring omission in the 11 program as proposed and as amended by your previous 12 votes, just sing out. But I don't think we need 13 any detailed discussion of the rest. I think we 14 can sort of work it out. 15 DR. KAWAS: I would like to make the 16 comment that, at least on the postmarket 17 surveillance, I think there should be required 18 postmarketing reporting, surveillance, monitoring. 19 DR. PENIX: In addition to the usual 20 adverse effects, of course. 21 DR. KAWAS: Are there any other comments 22 or thoughts from the committee particularly on the 23 items we didn't specifically discuss like central 24 pharmacy, postmarketing surveillance or other 25 recommendations on protecting-- 380 1 DR. SIMPSON: I guess there was just one 2 issue brought up about who would police the 3 policemen. 4 DR. KAWAS: You want to more specific in 5 which policemen we are talking about? 6 DR. SIMPSON: The issue was whether the 7 drug companies should be policing the correct usage 8 of the drug and then, if that were the case, who 9 would be policing that the drug company were doing 10 it right. And, if the physicians are supposed to 11 be making sure that the patients are doing it 12 right, and so on. That is what I mean. There is 13 layer on layer here. 14 DR. KAWAS: Let's start with the first 15 layer about if there is a surveillance or whatever 16 from the company. 17 DR. KATZ: Again, in some sense, we are 18 always in a position to oversee what the companies 19 do in terms of meeting their appropriate reporting 20 requirements and this sort of thing. 21 I think there is an understanding that 22 what comes out of this registry and the experience 23 will be reported to us. It will have to be 24 reported to us. We will be working in close 25 cooperation with the company to make sure that this 381 1 happens. 2 We won't be down at the first line making 3 sure that the pharmacist is calling the patients 4 within 24 hours. But, like many other things, 5 there is an understanding that the company is 6 responsible for making sure any given system of 7 surveillance is working appropriately and we have 8 interactions with them periodically. 9 So that is as far as we have gotten. 10 DR. LEIDERMAN: There are also precedents, 11 at least for independent monitoring committees. 12 And that has certainly been in approval agreements 13 in the past. So that is the kind of thing that I 14 think we need to work out. 15 DR. KAWAS: Unless there are any more 16 burning comments or thoughts or theories, I would 17 really like to thank the company, the agency, the 18 members of the panel and all the invited speakers 19 as well as the speakers from the public forum for 20 this interesting and challenging day 21 This meeting is now adjourned. 22 [Whereupon, at 6:00 p.m., the meeting was 23 adjourned.] 24 - - -