FDA's Forensic Center: Speedy, Sophisticated Sleuthing

By the time the mouse and its Pepsi-can coffin reached FDA, Fred L. Fricke and his team of chemists and microbiologists didn't have much to work with.

The mouse, found dead in a can of Diet Pepsi in New York, had already been examined by Pepsi officials. From there it went to a veterinarian on the East Coast. Next stop was a pathologist in Utah. Finally, the dissected mouse was sent to Fricke at FDA's Forensic Chemistry Center in Cincinnati.

Fortunately, the mouse's teeth were still intact. "We measured the spacing between the teeth and the pattern of bite marks on the can," explains Karen A. Wolnik, director of the center's inorganic chemistry branch. "From those measurements it was determined that his lower teeth had left marks on the inside of the can and his upper teeth had gnawed the outside, right at the pull-tab opening."

Wolnik says that pattern demonstrated the mouse had been inside the can when it bit the lid. But because the can lid with the pull-tab opening is in one intact piece throughout manufacturing, the mouse couldn't have bitten the lid or gotten into the can until after it was opened. The evidence was used to convict a tamperer who had falsely claimed to have found the mouse inside the can when she opened it. (Under the Federal Anti-Tampering Act, it is a felony to tamper with foods, drugs, devices, cosmetics, and other consumer products.)

"Every week we get something that's suspected tampering," says Fricke, director of the forensic center. "It never slows down."

FDA established the center in 1989 to provide the agency with a team of forensic science experts who can respond immediately to all tampering incidents and provide expert advice and scientific evidence to FDA officials. The 30 chemists and three biologists unravel the scientific mysteries of tamperings and other criminal activities involving FDA-regulated products through careful observation and high-tech instruments.

The center has inorganic chemistry and organic chemistry branches. The organic branch uses organic analytical detection methods--such as infrared spectroscopy, gas and liquid chromatography, and mass spectrometry--to separate and identify the components of mixtures. The inorganic branch uses tools such as digital image analysis and scanning electron microscopy to detect physical evidence of tampering or counterfeiting, and ion chromatography, atomic absorption, and inductively coupled plasma spectrometry to measure inorganic components of mixtures. (See accompanying article.)

Most cases require the expertise of both branches. "There's no division [of responsibilities] that's really sacred," says R. Duane Satzger, director of the organic branch. "When we get a case, both branches sit down and talk about it and decide how to address the situation."

For example, he explains, a syringe might first be examined by the inorganic group. Wolnik's staff would use light microscopy to examine the syringe. If, during this examination, they observed some kind of liquid in the needle of the syringe, Satzger's staff would use chromatographs or mass spectrographs to identify the liquid. Electron microscopy might be used to detect decomposition and other physical changes to the syringe that might have occurred if the syringe had been submerged in soda or come in contact with poison.

Tylenol, Grapes and Cyanide

In 1980, Fricke, Satzger and Wolnik worked at the forensic center's predecessor, FDA's Elemental Analysis Research Center in Cincinnati, where they conducted research and developed procedures for detecting toxic and nutritional trace elements in foods and drugs.

In 1982, when the first Tylenol tampering occurred, FDA chemists developed elemental "fingerprinting" techniques that allowed the authorities to trace the cyanide back to the manufacturer and the distributor. "The identity and relative amounts of various elemental constituents in a suspect sample form a distinct pattern that can be used for comparison with other samples, much like actual fingerprints," explains Wolnik.

The next few years saw more cases of cyanide in Tylenol and other pain relievers, as well as other types of tampering. "We applied the "fingerprinting" techniques to various poisons," says Fricke. They also developed "fingerprints" for inorganic substances such as metal and glass.

By the time cyanide was discovered in Chilean grapes in 1989, the center had developed expertise in detecting cyanide and other poisons in drugs and processed foods. But they had very little knowledge about what effect the cyanide would have on the fruit and vice versa. Would the poison become more or less toxic? Would it do something to the fruit that would be obvious to consumers?

"We didn't have a lot of answers at that time about what would happen," says Fricke. To keep FDA from being caught off-guard in the future, the agency redirected the focus of Fricke's lab from elemental to forensic research.

The lab's primary function shifted to research on what happens when poisons are added to foods and drugs. The "fingerprinting" technique used for comparing items of evidence was expanded to include many chemicals. In addition, the lab began developing screening methods for poisons so it could respond rapidly to any suspected tamperings. Since then, the lab has developed techniques to screen for more than 250 of the most toxic poisons commonly available to the public.

The forensic lab is also the only laboratory facility in FDA especially equipped for, and experienced in, ultra-trace elemental analysis. Using a specialized type of mass spectrometry called inductively coupled plasma/mass spectrometry, the lab's chemists can find contaminants in amounts as small as parts per trillion.

Tracking Down the Source

There are three points at which a foreign object or other contaminant can get into a product.

* During manufacturing. "There are legitimate things that can go wrong during manufacturing," says Wolnik, "and there are rare occasions of employee sabotage."

* While the product is in distribution. These are cases in which someone tampers with a product and returns it to the store shelf looking as untouched as possible so the purchaser is unlikely to detect the tampering. Frequently the perpetrator has a single victim in mind but tries to make the crime look like random tampering. "It's a crime similar to blowing up an airplane to kill one person," says Wolnik.

* After purchase. Those are the false report cases, such as the false claims of Pepsi tamperings during the summer of 1993.

Many clues help the forensic lab's people zero in on the time and place of contamination, including the amount of physical deterioration of foreign objects, the breakdown of poisons into chemical components, and physical measurements of containers. But they won't tell the public what those clues are. "We don't want to hand out a blueprint to would-be tamperers," says Fricke.

The process for identifying poisons and other chemical contaminants is something the scientists will share. Separation of the different chemical components of a mixture is one of the most frequent techniques used and requires some type of chromatography.

Chromatography separates complex mixtures by measuring migration rates of component molecules through columns and through coatings on chromatography plates. "We'll compare those components with a control and we'll be looking for differences," explains Satzger. "We won't try to identify everything in the sample. So if there are 20 components that we can separate in the suspect sample and only 18 in a control, we'll zero in on those two extra components."

Identification usually requires infrared spectroscopy or some form of mass spectrometry. Gas chromatography/mass spectrometry is used for volatile components. Liquid chromatography/mass spectrometry is the usual choice for nonvolatile ones. Ion chromatography and atomic spectrometry are used for inorganic components.

Even after suspect components are identified, the sleuthing may not be over. Sometimes components that show up in lab tests are part of a bigger picture.

"We've done studies to show that when you put sodium hypochlorite (bleach) in soft drinks it breaks down into several different components," says Wolnik. "So we look for elevated breakdown products of the sodium hypochlorite. That's where the sophistication [of our lab] comes in. Any lab can run a test for bleach in soft drinks. But if they don't find any bleach it doesn't necessarily mean bleach wasn't in there. It just means that [the bleach] may have been changed by the material."

Even when the final lab results are in, FDA's work may not be done. "A lot of what we do isn't the be all and end all," says Wolnik. "It just really helps focus the investigation. We work closely with [FDA] agents and investigators so we know what kinds of questions they're interested in answering. As we learn things, we provide information to the investigators which may help direct their investigation."

Leftovers

One of the toughest obstacles the forensic lab faces is the condition of samples when they reach the lab. Like the mouse in the Pepsi can, samples have frequently been studied by other authorities first, leaving very little for the forensic lab to work with.

"It's infrequent that we get the sample first," says Wolnik. "That alone makes our analysis difficult. We have to spend some time thinking about what analyses we want to do, and what order we want to do them in. We can't afford to waste what little sample is left."

Another problem is damage or contamination of the samples. For example, Wolnik says during the 1982 Tylenol tampering incident, the medical examiner unintentionally contaminated the cyanide from some of the poisoned capsules with sodium during his analysis. He then sent the contaminated capsules to FDA.

"In forensics, you want the evidence as close to the condition in which it was originally found as possible, and you want to preserve that," says Wolnik.

That was not the case on March 19, 1993, when Bobby Joe Johnston of Oklahoma City, suffered burns to his lips and tongue after drinking from a can of Pepsi. The hospital where he was treated took a sample of Pepsi from the can and determined it was highly caustic. The hospital called the fire department, which retrieved the can and the rest of the six-pack Johnston had purchased. The fire department treated the cans as hazardous materials instead of forensic evidence, however.

"They put the five unopened cans in a glass container, set the open can on top of the others, and went home for the weekend," says Wolnik. "When they checked it on Monday, the corrosive material had eaten through the open can and dribbled onto and through another can, causing the second one to explode. When we finally got the sample, we had to reconstruct what came from the contaminated can and what came from the other previously unopened can. It ended up being fairly tricky."

However, the lab was eventually able to confirm that sodium hydroxide (lye) had been added to the can after it was opened and could not have been in any of the unopened cans. "We did studies to see how long it would take [for sodium hydroxide] to eat through a can," says Wolnik. The FDA chemists found that highly caustic solutions such as lye ate through the can in a matter of hours. "That proved that it couldn't have happened during manufacturing." Johnston was convicted of tampering on June 3, 1993.

Beyond Tampering

While tampering cases are the main focus of the forensic lab, there are other activities--such as illegal sale and use of unapproved drugs, counterfeit drugs, and economic fraud--that require the lab's expertise.

For example, the drug clenbuterol, which isn't approved in the United States for any use in either people or animals, is sometimes used illegally in show animals, including cattle, pigs and sheep to increase muscle. If those animals are subsequently slaughtered, anyone who eats the tainted meat might experience symptoms such as increased heart rate, muscle tremors, dizziness, nausea, fever, and chills. "We developed methods to analyze animal retinas for clenbuterol residues," says Fricke. This is significant, he explains, because while clenbuterol residues may show up in various tissues, almost any use will leave residues in the retina.

Counterfeit products require a combination of high-tech analysis to compare ingredients with the real product and careful study of the labels and packages.

The lab has uncovered evidence to support charges of economic fraud as well. Two such cases involved substandard stainless steel on imported surgical instruments and purported nutritional products that didn't contain any of the listed ingredients.

The procedures and techniques the forensic lab has developed better prepares the agency to meet these types of emergencies. Still, "every sample that comes in is like a separate research project," says Fricke. "We have to decide what procedures and methods to apply, and there still are many cases that require new procedures. And, always, we have to do this as quickly as we can."

Isadora B. Stehlin is a staff writer for FDA Consumer.

High-Tech Tools

Chromatography (thin-layer, gas, liquid, and ion)
Separation of complex mixtures based on physical and chemical interaction with a solid adsorbent material (for example, activated carbon, alumina gel, or silica gel) on a plate or in a column.

Mass Spectrometry
Identification of chemical structures and quantitative elemental analysis based on the mass of ionized (charged) molecules, ionized fragments of molecules, and ionized elements.

Infrared Spectroscopy
Measurement and identification of molecules by means of the molecules' interaction with infrared radiation.

Atomic Spectroscopy
Determination of elements based on the emission or absorption of electromagnetic radiation (for example, ultraviolet and visible light) by atoms.

Capillary Electrophoresis
Separation of chemicals by movement through a glass capillary tube in a solution under the influence of an electric field.

Microscopy (light and electron)
Production of magnified images of objects using light or electrons.