The AIDS Malignancies Working Group (AMWG) was convened in response to emerging opportunities in HIV/AIDS-related malignancies research and the rapid development of tools enabling the pursuit of crucial research questions in this area. By design, the AMWG represents a spectrum of disciplines currently working in or with an interest in AIDS malignancies, and those conducting research that may have relevance for these cancers. At this fourth meeting of the AMWG, participants received an update on progress of the NCI's AIDS-related malignancy initiatives. Several speakers provided comments on their experiences with these initiatives and presented preliminary findings of research funded by supplements to Cancer Center Support Grants. Presentations were also provided on the serology of Kaposi's sarcoma-associated herpesvirus (KSHV) or human herpesvirus 8 (HHV-8) and on the performance of existing KSHV/HHV-8 assays. Thereafter, the group considered two major issues: (1) the impact of highly active antiretroviral therapies (HAART) on the incidence of AIDS-related malignancies and the data needed to adequately monitor and understand the impact of these therapies on malignancies; and, (2) whether enough is known about the serology and virology of KSHV/HHV-8 to initiate prevention studies exploring strategies for preventing Kaposi's sarcoma in high risk populations. After discussing these issues in subcommittees, the group reconvened to share conclusions and formulate recommendations for proceeding in these research areas.

WELCOME, MEETING OBJECTIVES, AND UPDATE ON NCI PROGRESS ON WORKING GROUP RECOMMENDATIONS

In opening the fourth meeting of the AIDS Malignancies Working Group (AMWG), Dr. Ellen Feigal indicated that:

• The creative input of the Working Group has been extremely helpful in helping the NCI organize and advance its AIDS malignancy research program.

• Since early 1994, the NCI Clinical Trials Cooperative Groups have had supplemental funds for the purpose of forming working groups within the clinical trials group structure to facilitate the transition of research activities initiated under the AIDS Malignancy Consortium (AMC) to larger clinical trials through the cooperative groups.

• The Cancer Centers Support Grant supplements were intended to be seed money aimed at forging alliances between investigators working in cancer centers and AIDS investigators working at centers for AIDS research or at AIDS clinical trials units. The goal was to facilitate work in the area of AIDS malignancies; the results of this first-year effort will be summarized by Dr. Bhorjee later in the morning.

• The NCI Handbook on Resources in AIDS and AIDS Malignancies, available since January 1997, is in the process of being reconfigured and reissued. The Handbook currently consists principally of a directory of AIDS and AIDS malignancy investigators within NCI and highlights of the types of research underway in this area. Dr. Feigal invited the Working Group to provide input on additional information that would be of use given the changing AIDS malignancies research environment.

• The first National AIDS Malignancies Conference, attended by 400-500 academicians, laboratory and clinical scientists, industry-based investigators, and regulators from around the world, was held in April 1997. The conference was developed in response to observations by the AMWG that no fully appropriate forum existed at which to discuss AIDS malignancies. Based on the success of the first conference, NCI has organized a second national conference, to be held April 6-8, 1998. The program committee, again consisting of many members of the Working Group, is planning a roundtable on highly active antiretroviral therapies (HAART) and their impact on the incidence of AIDS malignancies. Plenary presentations are planned on the human papillomavirus (HPV)-related tumors, as is a KSHV workshop. Findings from the first national conference are available on the Internet, and the Journal of the National Cancer Institute will soon publish a monograph based on the conference plenary sessions. The Working Group was invited to provide input on the topics proposed for discussion at the upcoming April conference. Dr. Feigal indicated that the cost of the first national conference was $123,000, plus $15,000-$20,000 in travel- and lodging- related costs.

• Based on discussion at the first AMWG meeting suggesting a lack of formalized training systems to enable clinicians to gain a better understanding of current directions in AIDS malignancies research, key disease management issues, and possibilities for innovative treatment, NCI issued in March 1997 a Request for Applications (RFA) to develop such training opportunities. The many applications received were reviewed in November 1997, and results of the study section evaluations will soon be available. NCI was pleased by both the number and quality of the applications received.

• Between 1995 and 1996, NCI shifted a significant portion of the AIDS malignancy portfolio from intramural to extramural grants. NCI has endeavored to maintain this new balance in 1997.

• The NCI is interested in the Working Group's input as to the most important research opportunities and scientific goals in AIDS malignancies research for the next five to ten years.

Key Points

• NCI decided to increase cross-disciplinary research activity in AIDS-related malignancies in the NCI-designated Cancer Centers through supplements to their existing CCSGs. Since the funding mechanism was already in place, this was judged to be an expeditious way of deploying funds for developing new AIDS-related malignancies research programs. This initiative included support for pilot research projects, shared resources and for pilot clinical training programs in AIDS oncology. Applications were received from 38 institutions for a total of 161 pilot research projects and 12 applications for core resources support. The applications were reviewed in a two-step peer review process, with participation by intramural and extramural scientists from NCI, the National Institute for Allergy and Infectious Diseases (NIAID), the Office of AIDS Research (OAR), and the Food and Drug Administration (FDA). In FY 1996, NCI funded 11 applications for developing full programs in AIDS malignancies research that included support for 45 pilot research projects and three core resources. In addition, individual projects/resources were funded in 11 other applications for a total of 12 pilot research projects and one core resource support effort. Eighteen applications were received for clinical training programs in AIDS-oncology of which five were funded. Total costs for the initiative were $5,490,380, including $602,545 for AIDS-oncology training.

• The pilot research projects funded focused predominantly on AIDS-associated Kaposi's (KS) and lymphomas, HHV-8-associated KS, and Epstein-Barr virus (EBV) in AIDS-associated lymphomas. These studies were directed toward increasing understanding in the areas of basic biology, immunobiology, molecular biology, and molecular genetics of AIDS-malignancies, and the processes of signal transduction and oncogenesis, including approaches to gene therapy, and the development of relevant animal models for AIDS-related malignancies.

• Although some of the funded research projects had been in progress for less than a year, awardees were surveyed to assess their progress and the utility of the funding mechanism for its intended purposes. Specifically, five areas were queried:
  • Interesting preliminary or completed results from research projects. The results submitted by the respondents are compiled in a separate volume as part of the proceedings of the AMWG.

  • Competitive research projects resulting from pilot research projects. A major intent of the supplement was to help initiate AIDS malignancies research activities at the cancer centers, with the hope that successful pilot efforts would result in competing for continuing support through mainline RO1, PO1, or other grant mechanisms. To date, 16 such applications have been funded, some from industrial sources, and another 15 applications are planned. Moreover, in addition to helping centers establish NCI-funded AIDS malignancy research programs, funding through the supplement mechanism is also having the effect of helping the institutions use the data from the pilot projects in applications for their Centers for AIDS Research (CFAR) programs.

  • The extent to which new alliances were formed or existing alliances reinforced between cancer and AIDS investigators, and barriers to successful alliances. Respondents indicated that funding through the supplement has: served uniquely to stimulate intra-institutional linkages among cancer center/CFAR/AIDS Clinical Trials Units (ACTU) investigators; stimulated new research alliances between cancer and AIDS investigators; and attracted new investigators into AIDS-associated malignancies research. Respondents noted, however, that while the supplement funds have helped initiate activities, the one-year support was inadequate for follow-up efforts.

  • Success of and barriers to developing cross-disciplinary clinical training programs in AIDS-oncology. According to the survey, funding for these efforts has successfully attracted new people into the field of AIDS-oncology and has helped significantly in submission of proposals for the AIDS-Oncology Clinical Scientist Development Program (K12). Because of one time, one-year funding of this initiative, some respondents indicated problems in recruiting trainees.

  • Usefulness of the supplement mechanism for achieving the NCI's goals for this initiative for the AIDS malignancy community. Respondents indicated that this special initiative was a success and would be worth repeating from the standpoint that it created new research opportunities in AIDS malignancies and helped in submission of competitive grant applications. The initiative was also deemed successful in bringing together a diverse body of basic science and clinical investigators to focus on AIDS-associated malignancies, and according to some respondents, it had an extraordinary impact on the coordination of AIDS-associated malignancies research at the cancer center. At some centers, the supplement award raised the visibility of AIDS malignancies research sufficiently that these centers now plan to include this research program area in their CCSG competing renewal application. The supplement was also seen as a good vehicle for pursuing certain high-risk research unlikely to be funded through the RO1 or PO1 mechanisms. Moreover, support of pilot projects through supplemental mechanisms was seen as a cost effective way of rapidly funding, testing, and evaluating novel exploratory studies.

• NCI believes the overall sentiment of the respondents was that the supplement was a positive step. The supplement appeared to meet the Institute's objectives of increasing centers' focus on AIDS malignancies, fostering collaboration between cancer and AIDS investigators, in particular at institutions that have CFARs and ACTUs, and of attracting new investigators to the field. NCI is interested in AMWG's suggestions and recommendations as to how such supplements (or other mechanisms) might again be used to continue progress toward these objectives.

• The one-year, $5.5 million CCSG supplement included total direct and indirect costs. The initiative allowed for the possibility of extending the supplement award second year, but with no additional funding. This supplement represents approximately 1.5 percent of the current $225 million NCI extramural AIDS malignancies budget. Though a relatively small amount, the CCSG supplement provided an excellent opportunity to leverage institutional resources for developing new AIDS-related malignancies research programs with this kind of initiative.

Key Points

• The supplemental grant to CWRU/UH Ireland Cancer Center, funded in the fall of 1996, has served as a springboard for bringing together in an extraordinary fashion investigators, clinicians, and scientists from a variety of NIH-funded centers at the university. The grant supports five exploratory laboratory projects and a small administrative component. The pilot laboratory projects focus on:
  • Expression of HHV-8 mRNA and viral antigens in African KS tumor lesions. This project has been completed and has been published in the Journal of Infectious Diseases (JID); it is the basis of several other ongoing projects. The investigators have identified the KSHV antigen in KS tumor lesions in African (Ugandan) KS patients, irrespective of HIV serostatus. EBV also was detected in a great number of these lesions, and there was some discordance in the appearance of CMV. KSHV was not detected in peripheral blood mononuclear cells (PBMCs) in HIV- positive and HIV-negative patients who were KS patients have an increased risk of non-Hodgkin's lymphoma (NHL); EBV and HHV-8 have been detected in the skin lesions of patients with African KS, and both viruses produce factors capable of stimulating B lymphocyte proliferation. Preliminary results indicate that KS lesions contain an oligoclonal array of B cells not seen in reactive tonsil controls. The next phase of the study will be to assess the presence of this array in circulating lymphocytes of the KS patients. If not detected in peripheral blood, the investigators will test the biopsy specimens for the presence of B-cell lymphoma-associated somatic mutations.

  • Tyrosine kinase signals in African KS. The investigators are assessing tyrosine kinase profiles in KS lesions using an improved reverse transcriptase polymerase chain reaction (PCR) method.

  • Angiogenic signals in African KS cells. It is known that tat activates stress pathway SAPK/JNK. The study is exploring whether tat may stimulate more than one pathway; it has been shown that tat-stimulated cells die by apoptosis.

  • Interactions among EBV, HIV, and HHV-8. This study is exploring the role of viral interleukin-6 (vIL-6) expression in a patient with multicentric Castleman's disease with the hope of identifying a monoclonal antibody for vIL-6 that may form the basis of a series of investigations.

• As a result of the supplemental grant, an AIDS-related Malignancies Working Group was established at the institution. The Working Group meets monthly and is attended by investigators working in a wide variety of disciplines.

• The cancer center's Developmental Therapeutics Program is now working on the clinical development of a second generation photosensitizer silicon phthalocyanine (Pc4), which is synthesized at CWRU's Department of Biochemistry. In addition to other cutaneous malignancies, a primary target for this modality will be KS.

• In addition, members of the cancer center and CFAR have joined with the Skin Diseases Research Center to sponsor a joint KS symposium in March 1998.

• Two grants have been submitted stemming from the supplemental grant: (1) a CFAR competitive renewal proposal that would establish a new core facility, the International Clinical Coordination Center, linking six NIH-funded research centers at CWRU, the CFAR at CWRU, and the Joint Clinical Research Center in Uganda, and (2) a proposal to develop an AIDS-oncology clinical scientist development program at CWRU.

• The CWRU/Uganda collaboration dates to 1986 and initially focused on relationships between HIV infection and tuberculosis. Around 1994, the cancer center became involved in the work. The new coordination center will be heavily invested in the area of AIDS-related malignancies. The major themes planned for the AIDS-related malignancies program in Uganda are to extend the spectrum of studies of viral polymorphism and expression in KS tumor lesions and perhaps other malignancies; to continue the studies on TILs in KS; and to begin a clinical trials program in AIDS malignancies (particularly NHL) for dose modified oral combination chemotherapy.

• As part of the CFAR supplement, it is planned to establish a video conferencing and telecommunications center that will facilitate research interactions between the investigators in Uganda and Cleveland and also make possible a monthly AIDS Malignancy Tumor Board via teleconferencing including investigators in Cleveland and clinicians in Uganda.

• Although the total dollar amount of the supplemental grant and its addition to the already vigorous AIDS-related programs at CWRU was not large, the supplemental grant has had a significant impact in a very short time at CWRU. It is likely that without the grant, the recent collaborations and interactions would not have occurred in so timely a manner.

Key Points

• UCLA received a supplemental award of $101,755 of the $335,570 requested; the grant funds three pilot projects and provides a small amount of administrative support.

• One of the pilot projects focuses on the role of galectin-1 in T cell apoptosis. Contrary to the original hypothesis, it has been found that galectin-1 expression is inversely correlated with CD45 expression (i.e., cells with the highest level of expression lack expression of CD45). In addition, galectin-1 expression has been found to be modulated by O-glycans rather than by N-glycans, as originally thought. This latter finding is of interest because branch chain O-glycans are preferentially expressed in the T cells of individuals with HIV and may be related to the apoptotic process. Thus far, high-level expression of galectin-1 has not been identified in AIDS-related malignancy tissue samples; this work is ongoing.

• The second of the pilot projects explores two aspects of the interaction between HPV and HIV in viral expression in cervical cancer patients. The investigator has been able to demonstrate that HPV-infected keratinocytes secrete soluble factors that increase HIV expression in monocyte macrophage cultures. A panel of neutralizing antibodies is now being used to identify which factor is secreted; several candidates have been identified to date. The part of the study assessing the effects of HIV on HPV infection of keratinocytes in regulation of viral expression is ongoing.

• The third pilot study focuses on KSHV in primary effusion lymphomas (PEL) and related disorders.

• The supplemental grant funds have enhanced already established interactions between the UCLA CFAR, Cancer Center, and AIDS Clinical Trials Group (ACTG). The CFAR provides the vehicle for funding seed grants, using CFAR and state set-aside funding. Think tanks have been developed to explore single subjects in intensive day-long sessions aimed at generating ideas for studies; these sessions typically involve faculty from the UCLA campus, affiliated hospitals, and private industry. In addition, conferences are used to bring in individuals from other fields. A semi-annual peer review process is in place to review grant proposals generated by these sessions. The process is aimed at involving senior fellows and junior faculty who have an existing interest in this area, with the overall goal of developing faculty with a long-term interest in AIDS research.

• Two of the three pilot project investigators have secured RO1 funding for continuation of their work; the third investigator has an RO1 planned for submission in January 1998. From this perspective, the supplemental grant has been quite useful.

• As implemented, however, the supplemental grants did not allow local expertise to decide how best to utilize the funds. Pilot projects were selected for funding as they would under a PO1 mechanism and thereby did not make use of the well-established review process at UCLA. In addition, important activities (e.g., think tanks, fellows research support, malignancy conference) for which support was requested were deleted from the final award.

• Though the supplemental funding has been useful, its magnitude is insufficient to generate long-term interest in AIDS malignancy. The faculty who were funded at UCLA were relatively well-established individuals who were likely to secure RO1 funding even without the pilot project funding under the supplement. Although the cancer center was given to understand that work in AIDS malignancies was of high importance, the size of the award undercut that message. In the future, if awards are to be small, UCLA recommends allowing the centers to manage the distribution of the funds, with accountability for results.

• At UCLA, there is a crucial need for bridge funding to help senior fellows make the transition to junior faculty positions. It must also be recognized that long-term, sustainable research programs with support are needed to attract junior faculty to AIDS malignancies research careers.

Key Points

• Potential difficulties exist in proposing that funding of specific pilot projects with the supplemental grant be left to the awarded institution; investigators' incentive to write a proposal for inclusion in the overall grant proposal may be diminished if the project will then also have to undergo local peer review; at some institutions, an old boys network makes outside review by NCI an attractive feature of the supplemental grant; it may be difficult to specify how grantees are to be held accountable for results. Possible criteria for assessing success of the supplemental grant might include the number of RO1s submitted, number of successful RO1s, and number of junior faculty who are appointed as a result (in whole or in part) of the funding. It also was suggested that if supplemental funding was available in years after the first, another way of holding investigators accountable would be to make funding for subsequent years dependent upon results in year 1.

Key Points

• The supplemental grant represents the only NIH funding at the University of Michigan provided to investigate properties of KSHV. This was the only study funded under the supplemental grant at the institution, so that unlike some of the other awardee institutions, opportunities for local collaboration were relatively limited. There was, however, interface with the cancer center and with other AIDS clinical groups nationwide.

• The work on HHV-8/KSHV grew out of an interest in HIV and also in relation to work on gene transfer into vascular cells. In collaboration with dermatologist Brian Nickoloff, Dr. Nabel and colleagues cultured HHV-8 from primary KS lesions. Initial studies used primary cell lines derived from KS lesions and grown in a medium containing a constituent known as "scatter factor," which is used instead of HTLV-1 condition media. Cells cultured in this way were found to be negative for all of the usual contaminant herpes viruses and positive, with a few exceptions, for HHV-8/KSHV.

• The cytopathic effects of the virus in cell cultures have been of particular interest and their study has been the work primarily supported by the supplemental grant. Following infection of primary human umbilical vein epithelial cells (293 cells) with HHV-8/KSHV, progressive zones of lysis develop rapidly; by day 5, the culture is essentially nonviable. This is unusual for 293 cells, which normally are hardy. When virus particles are isolated from the supernatant of the infected cultures, viral DNA can be detected in the presence of pronase and DNase. If, however, the particles are treated with NP40, susceptibility to DNase and protease treatment occurs, suggesting that some structure is released in the infected cultures having a membrane associated with the viral DNA.

• As a result of these studies, a replication-competent herpes virus, detected at a fairly low copy number per cell, was isolated from primary KS lesions, with a DNA sequence nearly identical to the putative HHV-8/KSHV. This virus was then used to further study its role in KS (particularly replication-competent versus latent virus) and in other diseases, such as prostate-related diseases. However, the virus is difficult to propagate in vitro since it has low titers. To identify alternative cell lines that could be used to propagate the virus, a number of other cell lines were exposed to the virus; several prostate and colon carcinoma cell lines were found to support viral replication. Some epithelial tissue does not support propagation of the virus; notably HeLa cervical epithelial cells. In addition, it has been possible to propagate the virus in CT-26 mouse colon carcinoma cells.

• Cytopathicity of the virus decreased in a manner consistent with the amount of viral DNA that can be detected in the cell culture. However, more cytopathicity in culture was observed than would be expected given the numbers of infected cells, suggesting that either a released factor was present in the cell culture that was transmitting the cytopathic effect, or that the virus itself was highly cytopathic. It also seemed possible that cytopathicity was occurring at levels so low that virus replication or assembly could not be detected.

• The HHV-8/KSHV virus being grown by the University of Michigan team from primary KS lesions appears to be different in character from that derived from EBV lymphomas, which does not appear to be cytopathic. To enable a meaningful comparison with lymphoma-derived virus, the Michigan researchers have adapted their HHV-8/KSHV virus to grow in a B-cell lymphoma from a nasopharyngeal carcinoma. Noteworthy differences in biological properties have been observed; the viability of 293 cells infected with the B-cell lymphoma-derived virus was largely unaffected (virus invades the cell, but does not replicate), while the viability of KS lesion-derived virus dropped dramatically. It is important to understand the mechanisms at play both to understand differences in the virus and to provide a cell line in which the virus can be propagated most effectively. The replication-induced cytopathicity of the KS lesion-derived virus has been a barrier to growing the high titer levels of virus desired.

• The virus particle itself does not appear to be causing cytopathicity. Subsequent tests indicated that viral DNA and RNA replication must be intact for cytopathic effects to occur. A potential mechanism for improving viral replication while suppressing cytopathic effects is to inhibit apoptosis in the host cells. Previous attempts to do this by a variety of mechanisms had been unsuccessful. More recently, the research team induced overexpression of viral BCL-2 (made in both the lymphoma-derived and primary KS lesion-derived virus) in 293 cells; viability of the cells was increased in a dose-dependent fashion.

• The current goal is to improve the robustness of the virus propagation method, and to distinguish between lytic replication of the virus and dissemination to other cell types, particularly to lesions and to B-cells in which transformation may occur. The research team's current working model is that lytic infection may be important for transmission; that cells of the urogenital tract are more permissive of viral replication; and that when these cells are transferred by secretions to the colonic epithelium, B-cells also support viral replication and dissemination to other tissues. The challenge now is to understand to what extent latent infection of these other cell types causes lesions, and to what extent it may be possible to interfere with this process by interfering with viral replication (as opposed to viral gene expression of latently-infected virus).

Key Points

• The Michigan research team has been able to serially propagate approximately 20-25 passages of cell-free HHV-8/KSHV supernatant. At that point, titers begin to drop and it is difficult to maintain viability. It has also been possible to propagate the virus from the Louckes cell line ; persistent infection can be maintained for approximately four to eight weeks. A wave of propagation tends to occur in the first week or two, followed by decline.

• Dr. Nabel emphasized that differences between lesion-derived virus and long-term cultured lymphoma-derived virus are more important than differences between the B- lymphoma cell lines. Primary lesion-derived virus, whether in a lymphoma or a KS lesion, is highly cytopathic. Dr. Nabel and colleagues have sequenced reading frames from the HHV-8/KSHV virus they have isolated, and have observed that whenever polymorphisms are found, that the sequences are identical to the primary lesion virus.

• Epidemiologic evidence suggests that fecal-oral exposure may be the root of transmission of HHV-8/KSHV in developing countries and among gay men in the U.S. This appears to be consistent with Dr. Nabel's hypothesis that genitourinary tract cells are relatively permissive of viral replication. Animal models will be useful in exploring this idea further; this should be possible since the virus has been demonstrated to grow in murine colon carcinoma cells.

• Referring to the first slide presented by Dr. Nabel, it was clarified that in the early stage of growth, endothelial cells can be supported simply by production of endogenous growth factors in the mixed culture; with time, smooth muscle cells grow out preferentially under the growth conditions used by Dr. Nabel's group. It is not, however, known if the epithelial cells are infected. In other work, Dr. Nabel's group has studied other endothelial cell lines. They also have studied microvascular and human umbilical vein epithelium, neither of which could be infected with virus and propagated serially. They have also observed that DNA sequences disappear even in spindle cell cultures as they grow; as a result, these cells are not good hosts for viral replication.

• It was suggested that the HHV-8/KSHV virus could infect endothelial cells latently and escape detection by in situ staining; this was identified as an area for further research.

• With the support of the supplemental grant, Dr. Nabel's group has screened a bank of prostate tissues from the University of Michigan cancer center for the presence of viral DNA using an automated polymerase chain reaction (PCR) sequencing strategy (Tacman ABI 7700 methodology). Thus far, no viral DNA has been detected in lesional areas from uninfected normal patients who have had prostate carcinoma. Limited sensitivity of the assay may in part explain these findings.

• Dr. Nabel clarified that there does not appear to be a differential growth characteristic of the smooth muscle cells from the KS lesions compared with normal cells in the presence of scatter factor; scatter factor may simply promote the outgrowth of the KS lesion smooth muscle cell population.

ADDITIONAL DISCUSSION AND PRESENTATIONS

Prior to breaking into subcommittees, there was additional discussion and two presentations relevant to all of the subcommittee deliberations.

Key Points

• An estimated 900,000 people in the United States are infected with HIV; approximately two thirds are aware that they are HIV-positive. Of the 600,000 who know they are infected, it is estimated that 250,000-333,000 have received antiretroviral chemotherapy. Of these, at least half are on HAART (well over 100,000 have been put on protease inhibitors). Thus, the fraction of the HIV population on HAART is less than 20 percent.

• Of those who have been put on protease inhibitors, at least half have been put on inappropriate regimens. As a result, there are a significant number of treatment failures. The most important cause of failure has been inappropriate use (using protease inhibitors as sequential therapy rather than in appropriate combinations); adherence problems have also contributed to treatment failure. However, many people who are appropriately using protease inhibitors develop resistance to the drugs, preventing them from achieving effective viral suppression.

• It was suggested that the proper criterion of effective use of HAART is having virus below the level of detection of the most sensitive assay (currently about 50 copies/cell). This can be achieved in approximately 80-90 percent of an adherent, well-educated, population not previously treated with antiretrovirals. Factors that mitigate against successful viral suppression are prior therapy, and multiple sequential use of the available drugs. The proportion of people in these categories who can succeed is reduced, although prior experience with nucleosides has a relatively insignificant impact on successful therapy if alternatives are well-designed. Low CD4 cell counts, even in antiretroviral-naive patients, also reduce the likelihood of treatment success. Early use of drugs, use in people with limited prior therapy, and proper use of drug combinations all improve the probability of success.

• Available data (through three years of treatment) indicate that once a person's viral load is successfully suppressed (still negative at 9-12 months), suppression will be maintained, perhaps indefinitely, as long as the patient adheres to the treatment regimen.

• Soon to be published papers in Science suggest that CD45 RO memory cells are a source of latent infection and are maintained for very long periods of time, despite sustained suppression.

Key Points

• The first assays for KSHV were based on the BC-1 cell line, which is both EBV-infected and KSHV-infected. The first of these, developed at Columbia University by Dr. Moore and colleagues, was able to identify a latent nuclear antigen in the BC-1 cells. A second assay developed at Yale University was able to induce with butyrate a 40 kilodalton (kDa) lytic antigen. Neither of these assays is practical for screening purposes, but both seem to indicate that the KSHV virus is not ubiquitous. A third assay based on recombinant ORF65, a lytic antigen, has been developed by Thomas Schulz and colleagues at the University of Liverpool, and shows similar results. The latent antigen immunofluorescence assay (IFA) is encoded by open reading frame (ORF) 73 of the virus, and has an actual molecular weight of about 135 kDa but migrates on polyacrylamide gels as a doublet at 220 kDa. Using these assays, various groups have found that the seroprevalence among KS patients for the latent antigen assays is about 80-90 percent.

• The western blot assay, because it uses a high molecular weight antigen, is extremely difficult to transfer. Discussions with other investigators suggest that the westerm blot is not being widely used directly from KSHV-infected cells. However, the IFA, based on an EBV-negative cell line--either body cavity-based lymphoma-1 (BCBL-1) cells with a nuclear preparation, or whole cells--shows 80-90 percent positivity among KS patients and is in widespread use as a screening assay.

• Several other lytic antigens have been studied, such as a recombinant ORF 65 western blot that also appears to be about 80 percent sensitive. Its developer (T. Schultz) observed that some cross-reactivity with EBV ORF 65 occurs for the full-length protein. Other investigators identified the same antigen by immunoscreening, without any cross-reactivity for the full-length protein.

• EBV-negative cell lines can also be TPA-induced and used in IFAs; this assay provides the highest sensitivity for detecting patients who are positive for KSHV, but there are important questions about whether the cytoplasmic reactivity seen with these assays are due to nonspecific cross-reactivity.

• IFAs for KSHV exhibit a nuclear staining pattern characterized by a specular pattern with apparent colocalization of the antigen to spliceosomes in infected cells. This pattern and immunofluorescence reactivity together help in identifying specifically reactive sera. IFAs have the disadvantage of being somewhat subjective, particularly if one is unfamiliar with the specific staining pattern; they also are difficult to use for high throughput and have only about an 80 percent sensitivity rate.

• The latency-associated nuclear antigen (LANA) is encoded by ORF 73 on a polycistronic transcript that also includes ORF K 13 (VFLIP), and ORF 72.

• Dr. Moore's data on 40 KS patients followed through the Multicenter AIDS Cohort Study (MACS) suggest that seroconversion in patients who eventually develop KS represents de novo infection rather than reactivation of a long-term infection.

• When the positivity rate for various population groups between the LANA IFA and the ORF 65 ELISA are compared, the values are comparable for most groups except U.S. blood donors, in whom a higher positivity rate is found using ORF 65 compared with the IFA (5% vs 0%).

• An interesting comparison between the ORF 65 ELISA and LANA assay suggest that there are about as many discordant patients for each assay, i.e., those that are positive by ORF 65 and negative by the LANA, as there are negative for the ORF 65 and positive for LANA. This pattern has also been seen in other assays used by Dr. Moore's group, suggesting that these two assays might be complementary.

• The lytic phase IFAs have the highest positivity rates for KS patients. Unlike the other assays, they yield a high positivity rate for low-risk populations such as blood donors, which has given rise to come controversy as to whether this represents the true value of seropositivity among blood donors in the U.S., or if this rate reflects cross-reactivity. It is Dr. Moore's view that the observed rate probably reflects cross-reactivity. A modification of this assay, not tested by Dr. Moore's group, is used with Evans blue counterstaining that may reduce some of the apparent cross-reactivity.

• A key question is whether the existing KSHV assays can be improved and used in a systematic fashion for screening. At this time, there is little commercial interest in developing these assays further for serologic screening. It should also be acknowledged that the value of identifying an individual as seropositive for KSHV is limited unless some intervention in the disease process is possible or a useful public health message can be provided.

Key Points

• It was suggested that commercial interest in KSHV screening was unlikely unless KSHV seropositivity becomes an issue in blood banking.

• Experience with the other seven herpesviruses indicates that people who are immunosuppressed and have persistent replication or frequent recurrences tend to have the highest antibody titers; this is likely the case with KSHV.

• From a worldwide epidemiologic perspective, an assay will be needed that can detect someone who has had a mild or asymptomatic infection that may have been latent for decades and in whom the antibody titer will be relatively low compared with KS patients. The technical issue is to find antigens that are specific for KSHV and design the assays such that they are robust enough to discriminate the true seropositives from the true seronegatives. This has been achieved over a long period of time with at least some of the other herpes viruses. In KSHV, distinguishing clearly those who are infected from those who are not remains an issue; in addition, geometric mean titers have yet to be developed to distinguish those who are positive and develop KS from those who are positive but do not develop KS.

• A recent paper (R. Weiss) suggested that HIV-infected people who have detectable HHV-8 sequences in their peripheral blood are at very high risk for developing KS within approximately two years. If this is borne out in other studies, this subgroup may be appropriate for interventional efforts. Dr. Moore presented data on a group of 40 men who eventually developed KS, in which some entered the study seropositive for KSHV, approximately 50 percent seroconverted over the three and one-half year follow-up period, and some left the study still seronegative. These and other data suggest a long latency period; potentially, patients could be screened at some point during this period, but at this time, little can be done to change the clinical course of disease for those at high risk. Evidence is accumulating that, at least for gay men in the United States, KSHV is a sexually transmitted agent; counseling for discordant couples may be a useful public health intervention. Data from Africa suggest that although KSHV seropositivity prevalence is relatively high, KS rates may vary substantially among populations.

• The use of PCR-based assays for KSHV in epidemiologic studies has been troubled by PCR contamination. A consistent finding in PCR-based assays has been a positivity rate of about 50 percent (using approximately a 100 nanogram sample of PBMCs from patients). These different positivity rates using the various assays complicate assay interpretation, since it is unclear what constitutes a true negative.

• Experience with EBV assays in a Chinese population at high risk for nasopharyngeal cancer indicated that virtually all of the population was infected by age four, but then most subsequently became IgA or total early antigen (EA) seronegative or had very low titers for a substantial period of time. However, it appears that viral replication may increase later in life, leading to nasopharyngeal cancer in adulthood. Assays conducted on the adult population found that those with a titer greater than 1 to 160 were at risk for the malignancy over the next three to four years. By contrast, many of the populations at risk for KS that were studied by Dr. Moore's group remained consistently negative on similar assays for KSHV, yet some developed KS if KSHV is an ubiquitous infection, it is unclear why this should be so.

• In the time course data presented by Dr. Moore, all of the patients who developed KS seroconverted at least five years earlier. It is possible that these patients had incubating replication of the virus during that period and the development of KS represents viral reactivation; however, this explanation seems biologically implausible.

• Using the greater than 1 to 160 titer positivity level employed in the latent IFA, about 80 percent of HIV-positive KS patients (AIDS KS) were consistently positive, while among those who are HIV-negative and have KS (classic KS), sensitivity of the test approached 95-100 percent. Further, the 20 percent of AIDS KS patients (in the MACS cohort)who are negative for this assay developed KS in a significantly shorter period of time than the patients who were HIV seropositive. It is unclear if these findings reflect the presence of protective antibodies, a protective CTL epitope, or other factor. Dr. Moore cautioned that these observations are based on a small sample and may be proven incorrect over time.

• Replicative synergism between KSHV and EBV may exist at the level of the peripheral blood lymphocytes (PBL); however, it is unclear if such synergism between the two viruses exists in vivo in most patients. Thus, a concomitant increase in antibody response on latent antigen or other assays cannot be presumed to enhance the specificity of the assay for KSHV.

Key Points

• While HHV-8 DNA is detectable in all KS tissue, antibody tests are important for detecting infection in the absence of KS. Recently developed assays have been based on immunofluorescence, immunoblot, and enzyme immunoassay. Lack of standardization between these assays, however, hampers HHV-8 seroepidemiology.

• Five independent laboratories tested the same set of sera for HHV-8 antibodies. The blinded set included sera from 143 patients (10 with classic KS, 33 with AIDS-related KS, 15 HIV-1 seropositive but without KS, and 85 normal blood donors). Correlations between labs were analyzed by kappa statistics (with kappa scores considered significant for p .05).

• The assays evaluated were:

  Immunofluorescence Assays:

  • Lab 1--Latent IFA with the BCP-1 cell line (HHV-8+/EBV-). Sera tested at 1:100 dilution, and scored for specific nuclear staining (Simpson et al., Lancet 1996; 348:1133-8)

  • Lab 2--Lytic and latent IFA using BCBL-1 cells (HHV-8+/EBV-) pre-treated with phorbol-ester; sera tested at 1:10 dilution. Cytoplasmic (lytic) and nuclear (latent) staining scored separately (Lennette et al., Lancet 1996; 348:858-61)

  • Lab 3--Latent IFA; similar to Lab 1, except sera tested at 1:150 dilution

  Enzyme Immunoassays:

  • Lab 3--ORF65 capsid protein EIA. Recombinant lytic-cycle protein (HHV-8 open reading frame 65) with low sequence similarity to EBV. Sera tested at 1:100 dilution. Sera discrepant on EIA and IFA were resolved with third test (ORF65 Western blot for EIA+/IFA-: "positive" if WB+, or, repeat IFA for EIA-/IFA+: "positive" if repeat IFA+) (Simpson et al., Lancet 1996; 348:1133-8)

  • Lab 4--Minor capsid peptide EIA (withdrawn)

  • Lab 5-Minor capsid protein EIA; recombinant HHV-8 minor capsid protein. Sera tested at 1:1000 and 1:10,000 dilutions.

  • Lab 6--Whole virus EIA

• Strikingly, the various assays were all similar in their performance individually, although there were exceptions, such as the minor capsid protein ELISA. All of the others followed a general track of increasing positivity (antibody prevalence) by patient group in the order: blood donors, HIV+/no KS, AIDS KS, classic KS. The most sensitive assays were the whole virus ELISA, which found no positives in the blood donors, and the Lennette lytic assay.

• Kappa measures of agreement were also used to determine if some set of assays together could better describe true seropositivity than any measure alone. Except for the minor capsid protein ELISA, most of the assays had kappa scores of agreement in the good to very good range when all of the samples were considered together. The best agreement (kappa = .86) was between the Lennette assay and the whole virus ELISA, and was highly statistically significant.

• Agreement was less good, however, within subgroups with KS or at moderate risk for KS. For example, for the sera from classic KS patients, agreement between the Lennette assay and the whole virus ELISA dropped to .57. No agreement was found with low KS risk (i.e., blood donors).

• Overall, the assays were found to have good sensitivity for KS risk at the group level, with moderate disagreement on the individual level. Antibody prevalence in low risk populations, however, remains uncertain. No two tests were found to be confirmatory.

Key Points

• In considering all of the samples together, the observed level of agreement between assays is driven somewhat by the blood donor samples, not because they represent the majority of the samples, but because there is virtually nothing to agree or disagree about. A mixture of samples from different patient groups will have fairly good agreement overall between the two most sensitive assays, but agreement differs more as to whether they find the same patients or blood donors to be positive.

• It can be reasonably assumed that AIDS KS patients have the KSHV virus; if an assay is good at detecting those with the virus, AIDS KS patients who test negative are very likely false negatives.

• The ability to discriminate KSHV seropositives from seronegatives is particularly important in the population that is HIV seropositive but does not have KS. Dr. Rabkin expressed his belief that even the level of agreement in this group observed between the BCBL cytoplasmic and whole virus ELISA assays is unlikely to be replicated in a larger set.

• From an epidemiologic perspective, identifying individuals who are KSHV-positive in the general population is also important. Assay agreement on which individual blood donors are positive has been essentially nonexistent to date. It was pointed out that the cut-off point selected (i.e., below which a test is considered negative) is somewhat arbitrary; thus a set of samples from a blood bank may all test negative, yet contain some true but undetected positives. Such a result demonstrates only that the assay performed according to its design.

• KSHV titer levels measured for the same patient group and individuals within those groups (classic KS and AIDS KS) also do not agree between assays.

• Dr. Moore urged caution in conclusions as to how good or bad the existing KSHV antibody assays may be. Laboratories that originated some of the assays and have the most experience with them did not participate in the interest comparison, and the assays were modified by laboratories that subsequently performed them. In addition, it is unclear how to distinguish the relative importance of observed interest and interlaboratory variation.

• Experience with the other herpesviruses has shown that the key to developing a sufficiently sensitive ELISA is to find the immunodominant capsid antigens or glycoproteins, identify those that do not cross-react with other herpesviruses, and express them.

• Achieving this goal has proven to be a vexing practical problem--by western blot one is measuring the naturally occurring antigen made in an infected cell; whereas using recombinant technology, this dimension can be almost infinitely expanded depending on the amount of protein applied to a plate or other inert surface to detect antibody response. More research needs to accumulate on the use of different protein and different recombinant assays to identify which are the best serologic creations.

• It was suggested that phage display of the entire virus using known KSHV positive high-titer sera, or some analogous approach, might enable a more rapid identification of appropriate epitopes for development into assays. For example, ORF 65 was identified through cDNA screening of whole cell ELISA--a somewhat analogous approach in that distinguishing antigens are identified by the immune system and can then be defined. Dr. Kieff indicated that the phage display will point up useful epitopes for some of the herpesviruses, but their cumulative value in the context of the larger protein can differ in important ways. Thus far, because of the volume of possible epitopes, this approach has not proven to be practical for the herpesviruses.

• Extrapolating from previous work with the other herpes viruses, an alternative approach suggested would be to target the dozen or so major capsid proteins and glycoproteins, express them, and then set up the ELISAs. It is known that western blot loses much of the antigenic reactivity of some proteins.

• Another possibility is to use an Epstein-Barr nuclear antigen (EBNA)-like antigen and a recombinant EBNA antigen. However, the problem with using this antigen on an EIA is that it is large (preventing good expression) and has an acidic region that makes it hard to work with. Using truncated constructs of ORF 73, it is possible to sharpen sensitivity compared to the whole antigen; this indicates that the epitopes are primarily in an amino terminus region.

Key Points

Dr. Feigal asked the Working Group subcommittees to focus their discussions on two major areas:

Key Points

• It is important to recognize that even before the availability of HAART, some change was already occurring in the incidence of AIDS-defining malignancies in adults. For example, it appears that KS incidence and death began dropping in the early 1990s, well before the availability of HAART. Thus, any trends should be interpreted in the context of such underlying changes. Data becoming available through certain cohort studies suggest that with the availability of better antiretroviral therapy, KS incidence has dropped dramatically several-fold. An effort is underway to ensure that all investigators with data on this subject are present at the April 1998 National AIDS Malignancy Conference.

• It is currently unclear why better treatment of HIV should decrease KS incidence, (i.e., does it reflect some restitution of the immune system? Does it indicate something about the direct role of HIV in causing KS?). A related question is the effect of antiherpes compounds, in addition to HAART, on trends in AIDS-defining malignancies. Data from several studies indicate that some of these drugs (especially foscarnet and ganciclovir) seem to be decreasing KS risk. Similarly, existing data sets should be studied to determine if acyclovir is having any impact on trends in NHL incidence.

• Anecdotal evidence suggests that up to 50 percent of adults with KS who are put on combination therapy for HIV have their KS go into remission. This should be documented though clinical or other studies. If it is true, it may have implications for KS treatment recommendations. It should also be determined why some people go into remission under these circumstances, while others do not.

• The number of HIV-related malignancies in children is actually quite small. The number of children with AIDS can be expected to decrease because of the use of antiretroviral therapy during pregnancy. Cases of AIDS resulting from perinatally acquired HIV infection in children are now down about 40 percent. There is a great deal of interest in studying children born to mothers who received antiretrovirals because of the possibility that these children will develop malignancies at some point in their lives related to the use of RT or protease inhibitors. The need for very long-term follow-up of this population has been recognized; a collaborative effort is underway to conduct a workshop in 1998 on this issue. The National Institute on Child Health and Development (NICHD) and the NIAID have already begun intensive follow-up of children whose mothers participated in the ACTG 076 or other ACTG trials.

• NCI-sponsored studies of mice born to mothers exposed to antiretrovirals during pregnancy suggest that the mice are at higher risk for tumors in middle age. The question remains whether children born to mothers who received antiretroviral therapy are at similar risk. Clearly, the risk of HIV far outweighs the hypothetical risk of prenatal exposure to the antiretrovirals. It has also been determined that it is beneficial for HIV-infected pregnant women to continue taking AZT. It is unclear how best to follow this population--using the DES registry model (e.g., enrolling children born to HIV-infected mothers in ACTG trials and following them through doctors' offices), or through more passive surveillance mechanisms (e.g., CDC and birth defect registries). In addition, NIH is considering following these children for liver toxicities as well as cancer outcomes.

• The subcommittee also discussed whether at least some forms of Hodgkin's disease should be considered an AIDS-defining malignancy, and if additional studies are needed to make this determination.

• It was the subcommittee's conclusion that further epidemiologic studies on HHV-8 and KS could not be recommended in the absence of better diagnostic tests. Although many assays exist, we do not know if they are sensitive and specific enough to be useful in support of efforts to determine prevalence and modes of transmission.

Key Points

• Currently, data on the effect of antiretroviral therapy on the course of KS or NHL (i.e., patients going into remission) are largely anecdotal; the question was raised as to whether these data are accurate, if they can be quantified, if additional data can be accumulated, and if so, how or from whom. A participant indicated that data sets exist and could be accessed. In the Abbott 247 study of patients with very advanced KS, an approximately 50 percent reduction in KS incidence was observed between the placebo and the protease inhibitor arms. Based on these results and evidence of two patients going into remission of their KS in Australia, a study of 32 patients with KS on protease inhibitors was conducted; there were no positive responses. Discussants expressed skepticism concerning the study showing a 50 percent response rate and the likelihood of replicating such results consistently. Data from antiviral therapy was also discussed. Data from three prospective randomized prophylaxis trials of ganciclovir demonstrated a 52-63 percent reduction of relative risk in KS patients who received ganciclovir for prevention of cytomegalovirus (CMV) disease.

• It was noted that some remission of KS was seen in a Phase I trial of zidovudine (AZT); it was unclear whether this effect was the result of AZT anti-tumor activity, or a drop in viral load resulting from high doses of AZT. Secondarily, other immune reactions may be involved with the pathogenesis or exacerbation of existing KS. KS may be highly sensitive to T-cell function, since it can occur early in AIDS patients and in many studies, has been observed to be one of the first diseases that improves with antiviral therapy (either AZT or a protease inhibitor).

• It was suggested that the ACTG might further explore the remission of KS with administration of antivirals. Dr. Richman noted that patients with disseminated disease usually are excluded from study, but those with limited cutaneous involvement might be included. Dr. Feigal noted as another possibility the ongoing AIDS Malignancy Consortium (AMC) trial of a protease inhibitor and two RT inhibitors in patients with KS. Although this study would not provide a baseline of patients not on protease inhibitors, it would show if there were some KS responses with antiviral therapy alone. The study does, however, also include interferon alpha, which would complicate interpretation of responses observed. The AMC is also conducting a trial of triple therapy (a protease inhibitor and two RT inhibitors) plus chemotherapy in lymphoma patients which may or may not be able to provide information on tumor effect attributable to the antivirals.

• Other anecdotal data indicate that patients whose KS has stabilized on taxol or another agent have been able to stop chemotherapy for a considerable period of time when they were on relatively effective antiretroviral therapy. There was general agreement among the AMWG members that anecdotes such as these are sufficiently pervasive that a relationship between antiretroviral therapy and KS probably exists. The question remains how to quantify the relationship and translate it into treatment guidelines. A small ongoing study is measuring shrinking or disappearance of dermal lesions and performing quantitative assays of HHV-8 in cells and plasma; because of the small number of patients, however, these data will also be anecdotal.

• Dr. Richman will determine if the ACTG database (now tracking several thousand patients) will permit a cross-sectional analysis to assess whether there is a subpopulation in whom a rise in CD4 counts (indicative of some immune reconstitution) correlates with an improvement in KS response. Recognizing the need to follow HIV/AIDS patients even when they go off protocols, ACTG is initiating longitudinal monitoring of patients so that it will be possible to answer questions such as these over time. It is already clear that current therapies are eliminating cryptosporidiosis, CMV retinitis, and other opportunistic conditions; it will also be necessary to track impact on KS and lymphomas.

Key Points

• There were differing views within the subcommittee concerning HHV-8 seroprevalence in the normal non-HIV infected population; these ranged from 1-3 percent to 25 percent. The group did agree that better serologic tests are needed, and that RO1 funding (perhaps an RFA) should most likely be the mechanism for evaluation and funding of additional research in this area. The specific goal of such research would be the development of more robust sensitive and specific serologic tests using recombinant DNA or other technologies to obtain viral proteins or antigens. PCR-based diagnosis is also important as a method for establishing the prevalence of HHV-8 infection and should be pursued in conjunction with or the development of serologic assays. A proposal could be focused on optimizing both approaches or use standard approaches for one technique (serology or PCR) while focusing on optimizing the other. Proposals could also be considered for one of the two approaches. Another related area of inquiry best addressed through the usual RO1 funding mechanism is epidemiologic studies.

• The second area addressed by the subcommittee concerned the need to maintain a program announcement regarding active and passive immunization strategies for preventing and treating viral-associated malignancies in HIV-infected people. An ongoing program announcement in this area would send the message that there is a lot of interest in this area. This area of clinical research is viewed by the subcommittee as a very important target area for research in general because so few people currently are working in it and because there have been promising initial developments. A suggestion was made, though not thoroughly debated within the subcommittee, that a program announcement may be needed to increase interest in looking for new infectious agents in those tumors for which there is an epidemiologic suggestion of an infectious etiology.

• BLC-6 translocated malignancies are not associated with EBV, but may be associated with other viral infections. The subcommittee agreed this may be a topic for future discussion and possibly the focus of a program announcement.

Key Points

• Existing serologic tests for HHV-8 are not robust; there is a need for higher level of protein expression that in normal people contain epitopes that are immunoreactive. However, the work of identifying and expressing these proteins, though essential to the goal of developing sensitive serologic tests for HHV-8, is not considered particularly innovative or creative, either technologically or analytically. As a result, some study sections tend not to look favorably on grant applications for this kind of work. The Working Group suggested that some degree of targeting or set-aside within the traditional grant mechanisms, or an alternative funding mechanism such as the Small Business Innovation Research (SBIR) program, may be needed to support this work. The NCI Cancer Diagnostics Program might also be a vehicle for supporting the development of better HHV-8 serologic assays. The NIH SBIR announcement has, for the last two or three years, included sections on serodiagnostic assays for HHV-8 and other suspected cancer viruses; a few applications have been received, but none have been funded. Similarly, investigators can partner with biotechnology companies to pursue funding under the Small Business Technology Research (SBTR) program.

• It was emphasized that without the basic reagents needed to detect HHV-8 antibody, it will remain exceedingly difficult to define HHV-8's epidemiologic niche, the mode of transmission, and risk factors. There were divergent views as to the best way to fund this work; the RO1 and SBIR/SBTR approaches could be complementary.

• If a primary infection syndrome for HHV-8 could be identified, as can be done with herpes simplex or chicken pox (herpes zoster), it would be possible to collect serial sera on people during the course of infection. This would provide a baseline against which true seronegatives could be measured. Short of this ideal approach to identifying true seronegatives, one must guess what population is least likely to be infected. Without knowing the mode of transmission, this is still highly problematic (e.g., young children may be infected perinatally, or subsequently by their mothers or in day care settings, and may or may not remain seropositive).

• There is a need to define a "gold standard" panel of sera on which there is agreement as to which are positive and which are negative. Another approach is to have a battery of high-level expression of proteins that by scanning analysis appear to be immunogenic and non-cross-reactive with other herpesvirus protein; achieving this would then enable epidemiologic test development. However, even if patients who have KS are consistently positive on a test for the presence of KS, that test may or may not necessarily be a test for KSHV/HHV-8 prevalence. It will be up to the investigators applying for grants in this area to sort out how to attack this problem, and defining how positives and negatives may prove to be an iterative process. Further study of latent and lytic antigens may help in better defining the KS versus latent populations.

• At this point, not enough is known about the reservoir(s) of HHV-8 in the body (unlike, for example, herpes simplex, which resides in ganglia) to make effective use of PCR-based technologies. Presentations are likely to be made at the spring 1998 National AIDS Malignancy Conference.

Key Points

• A major topic of discussion in the subcommittee was whether there were parameters that could be of predictive value in determining a patient's risk for the development of tumors. Specifically, it was thought that there is a need for intensive laboratory study of a subpopulation of patients to assess parameters of viral load in both plasma and cells for KSHV, EBV, and CMV, to correlate these with HIV, RNA, and CD4, and to determine if any of these parameters correlate with clinical outcome. Candidate populations could include patients diagnosed with KS who subsequently went on antiretroviral therapy (as a sizable fraction of these patients will have tumor response), partners of KS patients, or serologically positive patients who could be followed long-term. There was a proposal forwarded to evaluate this in the context of a prophylaxis study; data presented by Dr. Miles suggested that a prospective study of serologically positive patients might be possible, treating patients at a two to one randomization. Concerns about the logistics of accomplishing this type of study and the ultimate clinical utility of such information were voiced. It was suggested that perhaps this could be addressed by evaluating whether the incidence of KS is affected in patients who receive proganciclovir or another KSHV-active antiherpesvirus drug, compared with acyclovir or another herpesvirus drug known not to be active against KSHV. Patients might be more willing to participate in such a study if they knew they would be continuing therapy that would prevent emergence of KSHV, and it would still be possible to assess whether KS incidence was affected.

• Assays are needed that can identify indirect measures of immune function that may relate directly to tumor incidence. Such assays might detect low frequency antigen-specific events (e.g., CTL or proliferative responses) related to EBV, HPV, or KSHV. Given the low incidence of EBV-related tumors, it may be necessary to devise a study assessing patients with EBV-related malignancy versus comparable HIV, RNA and CD4 populations that either had, or were free of, malignancy.

• The NCI should establish a mechanism by which to acquire the samples collected in previous antiretroviral studies. Many of these studies were industry-sponsored; the sample banks associated with these studies are in danger of being destroyed because of a lack of interest in their long-term maintenance. These sample banks might be a valuable asset for studies of the correlation of HIV, CD4, and herpesvirus serologies in patients who have a well-documented set of clinical outcomes.

• It is also important to try to capture better data about the full spectrum of HIV-related tumors, including both the incidence of these tumors and key aspects of their clinical behavior. Accomplishing this objective may require establishing a new registry, or finding ways to capture these data from existing databases.

• Support mechanisms are needed for novel therapeutics. The immune-based tumors are the best platform for determining whether immune-based therapeutics will ultimately have a clinical role in oncology. Support is also needed for the development of immune manipulation strategies, especially cell-based therapies, and for novel mechanisms of altering viral kinetics or virus-host interactions to try to affect virus-related tumors. It was felt that support for such approaches would require mechanisms other than investigator initiated RO1s. Interactive efforts between clinical investigators, support laboratories and basic laboratories will require alternative granting mechanisms.

Key Points

• Despite the limitations of existing assays, the Prevention and Treatment subcommittee felt that the LANA and whole virus assays could be used to identify high-risk patients or to study prospectively patients with coincident KS at the time they present with HIV, as they go on anti-HIV therapy. For example, homosexual or bisexual men who were positive by the two tests could be reasonably assumed to be a high risk group in which a prophylaxis study could be performed.

• Data from the three retrospective studies and one prospective study conducted to date show clearly that ganciclovir is active against KSHV. However, pilot studies of derivatives such as penciclovir, famvir, lobucavir , or cidofovir, may be useful, but with so little data on the activity of these fairly toxic drugs against KSHV, patients may hesitate to participate, especially if they are going to receive triple-drug antiretroviral therapy that appears to lower the risk of developing KS. Some of the newer agents are less toxic than ganciclovir and have shown anti-KSHV activity in vitro, but it is not yet known how well the in vitro results correlate with in vivo activity. It will be important to encourage drug companies to assess the efficacy of these drugs with both CMV and HHV-8.

• It was also noted that the current cost of oral ganciclovir ($20,000/year) prohibits its prophylactic use for KS, even in a population at high risk. It will be necessary to wait for a more potent oral form that is priced in a more reasonable way to conduct a separate trial or perhaps, one nested in a CMV prophylaxis trial.

• The subcommittee considered the possibility of substituting acyclovir for ganciclovir in one arm of a study, which could then be conducted as a herpes prevention trial simultaneously with a KS prevention study. The trial would assess the rate of all of the herpes opportunistic infections (OIs), with KS being the primary endpoint. In this design, all participants would receive therapy. It was Dr. Richman's view that interest in such a trial would be limited to those who have not responded to HAART (currently, 40-60 percent of patients treated outside of the drug company-sponsored trials). Alternatively, drug-naive patients with KS could be entered onto a antiretroviral study; clinical outcome of their KS could be assessed in conjunction with their laboratory parameters. This information might help determine if the laboratory parameters could be applied to another population in a prospective fashion. Another possibility would involve a two-step process whereby small pilot-like clinical studies would assess changes in viral DNA in response to the three or four best antivirals. Those that prove most efficacious could then be tested in a larger trial. Group members noted that support for even this approach was extremely limited.

CLOSING AND NEXT STEPS

• Dr. Feigal asked participants to indicate other areas of research related to HHV-8/KSHV that need further attention and discussion. She summarized the research areas and problems highlighted throughout the day by the Working Group that need to be addressed: diagnostic issues, technology expression issues and related research, conducting pilot-type screening and prevention studies and determining laboratory correlates prior to conducting a large clinical trial. Other areas requiring further exploration may include in vitro and animal preclinical models, and study of immunologic responses to viruses other than KSHV.

• The clonality of KS, a fundamental issue in understanding its biology, seems to be understudied and somewhat controversial. If it is clonal, it must be based on loss of heterozygosity (LOH) or chromosomal abnormality; there have been some applications submitted to further investigate this question.

• Continued research is also needed to identify strain differences in HHV-8. Three or four groups of investigators are working in this area; it is clear that there are differences in biologically important genes, but most of the genome is highly conserved. At issue is whether latency and disease are associated with strain differences.

• Dr. Feigal noted that the second National AIDS Malignancy Conference will be held in April 1998. She encouraged the Working Group to identify any specific areas that they believe should be highlighted so that these can be added to the program.

LIST OF ATTENDEES:

Dr. Ellen G. Feigal, Deputy Director, Division of Cancer Treatment and Diagnosis, National Cancer Institute

Dr. Richard Ambinder, Associate Professor of Oncology, Johns Hopkins Oncology Center

Dr. Robert Biggar, International AIDS Coordinator, Viral Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute

Dr. Jaswant S. Bhorjee, Program Director, Cancer Centers Branch, National Cancer Institute

Dr. Kenneth J. Cremer, Program Director, AIDS Virus Study, Biological Carcinogenesis Branch, National Cancer Institute

Dr. Elaine S. Jaffe, Chief, Hematopathology Section, Lab of Pathology, Division of Sciences, National Cancer Institute

Dr. Harold Jaffe, Associate Director for HIV/AIDS, National Center for Infectious Diseases Centers for Disease Control and Prevention

Dr. Elliott Kieff, Professor of Microbiology and Molecular Genetics, Infectious Disease Division, Channing Laboratory, Brigham and Woman's Hospital; Professor of Medicine, Harvard University

Dr. Nancy Kiviat, Professor of Pathology and Medicine, University of Washington, Human Papillomavirus Research Group

Dr. Mary Lou Lindegren, Division of HIV/AIDS Prevention, Centers for Disease Control

Dr. Douglas Lowy, Deputy Director, Division of Basic Sciences, National Cancer Institute

Dr. H. Kim Lyerly, Professor of Surgery, Assistant Professor, Pathology and Immunology, Duke University Medical Center

Dr. Douglas L. Mayer, Head, Viral and Rickettsial Diseases, National Naval Medical Center

Dr. Sandra L. Melnick, Program Director, Infectious Diseases, Epidemiology and Genetics Program, Division of Cancer Control and Population Sciences, National Cancer Institute

Dr. Steven A. Miles, Associate Professor of Medicine, Director, UCLA CARE Center Clinic

Dr. Gregory Milman, Director, Pathogenesis, Division of AIDS, National Institute of Allergy and Infectious Diseases

Dr. Patrick Moore, Associate Professor, Department of Pathology, Columbia University

Dr. Gary J. Nabel, Professor of Internal Medicine and Biological Chemistry, Howard Hughes Medical Institute, University of Michigan

Dr. Charles Rabkin, HIV Cancer Coordinator, Viral Epidemiology Branch, Division of Cancer Epidemiology and Genetics, National Cancer Institute

Dr. Scot C. Remick, Director, Developmental Therapeutics Program, Case Western Reserve University/University Hospitals of Cleveland

Dr. Douglas D. Richman, Professor of Pathology and Medicine, Departments of Pathology and Medicine, 0679, University of California at San Diego

Dr. David T. Scadden, Associate Professor of Medicine, Harvard Medical School, Massachusetts General Hospital

Dr. Richard M. Selik, Surveillance Branch, Division of HIV and AIDS Prevention, National Center for HIV and AIDS Prevention, Centers for Disease Control

Dr. Roy S. Wu, Health Science Administrator, Cancer Therapy Evaluation Program, Division of Cancer Treatment and Diagnosis, National Cancer Institute

Dr. Robert Yarchoan, Chief, HIV and AIDS Malignancy Branch, National Cancer Institute