Background Paper 1 Page 3 The MMPR in 1979 endorsed the comments of the 1977 meeting regarding the quality of toxicological data submitted for consideration by the Joint Meeting. The Meeting was informed that several studies utilized in assessing the toxicology of a number of pesticides considered by previous meetings were being repeated because of questions regarding the quality of the original data generated by some contract laboratories. The meeting recommended that all daca generated from replacement studies should be drawn to the attention of WHO and that compounds supported by such studies should be considered for early re-evaluation by future Joint Meetings. (JMPR, 1979, p.4). Finally, the MMPR in 1980 reaffirmed the opinion expressed by the 1977 Joint Meeting that studies on the safety evaluation of pesticides should comply with currently accepted scientific standards and with acceptable laboratory practice. Although all data submitted concerning safety evaluation will in normal circumstances be considered valid, occasions may arise when submitted data are unacceptable because of low scientific standards or undesirable laboratory practices. Studies from laboratories where such unacceptable standards and practices are known to exist, or to have existed for a definice period, would still be considered but might require further validation before ́acceptance. The meeting considered two ways in which data had been validated and accepted the results of these reviews. The two validation procedures were 1) the sponsoring industry employed an independent scientific consultant to assess all data and other relevant factors, who concluded that the background data coincided with the data in the final report, and 2) the sponsoring industry, at the conclusion of the study collected and reassessed all the data including evaluating the histopathology programme with an independent pathologist, Again the original data coincided with those in the final report. In both these programmes, final reports were approved by responsible scientists and the reports were accepted as presented. In one other instance a sponsoring industrial company repeated an entire series of studies with no reference to the validity of the original studies. The meeting recommended that when data are of questionable validity and cannot be verified, the studies should be repeated if they are significant to the toxicological review and evaluation. The meeting emphasized that there are no fixed procedures for validating data and a variety of validation procedures may be acceptable. These must be considered individually as the need arises. (JMPR, 1980, pp. 3-4). ACTION TAKEN BY THE JECFA The JECFA in 1981 noted that the question of data validation was discussed by the Joint FAO/WHO Meeting on Pesticide Residues and itself commented on the problem. It observed that normally the committee considers that studies submitted for toxicological evaluations are a true representation of the actual work carried out. However, occasions may arise when the data submitted are unacceptable because of low standards or unsatisfactory laboratory practices. The Committee has occasionally encountered data from laboratories where unsatisfactory standards or practices are known to exist. The Committee considered that in such instances data should be validated prior to their use for evaluation. The Committee agreed not to establish fixed procedures for validation of data and considered that a variety of procedures may be acceptable. Each study must be considered individually as the need arises. The Committee considered that when data are of questionable validity and cannot be verified, the results of adequate relevant studies must be made available if they are crucial to the safety evaluation (JECFA, 1981, draft report). CODEX COMMITTEE ON PESTICIDE RESIDUES (CCPR) At the thirteenth Session of the CCPR, in June 1981, the delegation of Canada drew attention to the fact that toxicological data for a number of pesticides had been evaluated on the basis of toxicological studies carried out by Industrial Bio-Test Laboratories. Many of these studies had now found to be invalid. Until replacement studies had been evaluated they reserved their position on the pesticides concerned. The attention of the Committee was drawn. to paragraph 2.4 of the 1980 JMPR Report which had dealt with validation of toxicological data. (CCPR, 1981, Draft Report). -News and Comment The Murky World of Toxicity Testing Four scientists are on trial for fraud in a case that has cast doubt on the safety of 200 pesticides and on EPA's monitoring procedures Chicago. Four scientists are being tried here on criminal charges that they faked studies on drugs and chemicals during the 1970's when they ran a major private laboratory, Industrial Bio-Test Inc. (IBT) of Northbrook, Illinois. The government has taken a hard line in this case for two reasons. Before its collapse, IBT was one of the most prestigious contract labs in the country, with around 22,000 studies to its credit, the basis for safe product ratings for hundreds of drugs and pesticides. In addition to being prominent, IBT seems to have been uncooperative. The Food and Drug Administration (FDA) and the U.S. Attorney's office charge that IBT's top scientists were so deeply mired in the scandal that they resisted a federal investigation in 1976 and plotted to hide evidence of their cheating. The trial began on 13 April in the U.S. District Court in Chicago and has been crawling along ever since. It may not be over until September. The pace is set by the defendants and their battery of nine expert criminal lawyers. The leader is a prominent Chicagoan, George Cotsirilos, who recently represented a Teamsters Union official in a notorious pension fund case. His law firm says its top rate is $150 an hour, per lawyer. In this trial Cotsirilos represents Joseph C. Calandra, the founder and expresident of IBT, also a professor of pathology at Northwestern University since 1942. The other defendants are Moreno Keplinger, a Fulbright scholar and former general manager of the Northbrook lab; Paul Wright, a former toxicology department head at IBT now employed by Monsanto; and James Plank, Keplinger's former assistant. They are charged with mail fraud: passing phony data to the government and calling it scientific information. Each defendant has his own lawyers, but the fees are being paid by IBT. The company is just a shell of its former self. Its chief reason for existing now is to minimize damage to its parent, the Nalco Chemical Company. Calandra sold IBT to Nalco 17 years ago, long before this trouble arose, and remained president until shortly after the federal investiga 1130 tion began in 1976. He left in March 1977. The current president of IBT, Yvonne Bonahoom, says the firm is paying legal bills on the understanding that employees who carry out company business in good faith must be indemnified. However, she says that IBT's continued support depends on what comes out in the trial. She would not elaborate. Nalco has set aside a multimilliondollar reserve to take care of IBT's problems. One reason for doing so is that IBT's former clients are thinking of filing civil damage suits. These would be strengthened by a criminal conviction in the current trial. IBT has already settled seven civil suits and is helping former Even IBT's wealthiest clients, such as Monsanto, Inc., are still digging out from the rubble. clients dig through the IBT archives to salvage what they can of old test data. The defendants see the government's case as an exaggeration of some minor slip-ups that occurred a very long time ago. But one reason there has been delay is that the defendants have fought hard for it. Two years passed between the indictment and the trial. At one point the court had to evaluate Cotsirilos's claim that his client's aneurism of the aorta made him too weak to stand trial. In the end, Calandra was judged strong enough. Even now, hardly a day passes when Cotsirilos does not move for a mistrial. As for the significance of the data fudging, suffice it to note that it has called into question the safety reviews of more than 200 pesticides, many of which are being retested at great expense by the manufacturers. Even IBT's wealthiest clients-such as Monsanto, Inc., which claims to have spent $12 million replacing bad IBT studies-are still digging out from the rubble. The John Dean of this case is a pudgy technical report writer named Philip Smith. As the government's lead witness he testified steadily from 5 to 19 May, enduring a sustained probe by the defense aimed at finding flaws in his character as well as his story. Smith showed remarkable self-control. One attorney complained acidly that Smith, who shattered more than one theatrical drum-roll of defense queries with a plea to repeat the question because of poor hearing, seemed less deaf in cross examination than in direct testimony. But the witness was not shaken. Smith's testimony is critical because he had an overview of all that went wrong. After the FDA began investigating in April 1976, Calandra put Smith in charge of an in-house audit. Smith told the jury that he was asked to find out where all the problems were and summarize them in a private report to Calandra, left untyped to avoid showing it to the secretaries. Smith provoked an angry protest from the defense when he said. that after this report was finished, he felt "threatened" by Calandra. Indeed, the "threats" seem mild, but reveal something about the atmosphere at IBT. Late in 1976, according to Smith, after Nalco learned that there was trouble at IBT and decided to send an investigator, Calandra urged Smith not to confess anything that he was not "100 percent sure of," and to consider 95 percent certainty as inadequate. Calandra also allegedly told Smith he would deny this conversation had taken place. In early 1977 after Smith talked to the Nalco agent, Smith claims that Calandra "suggested I might be subject to charges of libel for things I believed were true and for things I was telling him." Calandra left in March 1977. Three months later Smith was fired and given 20 minutes to clean out his office. When confronted by Justice Department investigators the next year, Smith invoked the Fifth Amendment against self-incrimination, was given immunity, and became a paid government witness. Much of the defense's interrogation has been aimed not at the substance of the charges but at suggesting that Smith is the kind of person who might be bullied into making up stories about his boss. But the government plans to substantiate Smith's ac SCIENCE, VOL. 220 count with testimony from other IBT staffers, including one who will testify about the shredding of data, another who will state that raw data cited in two studies never existed, and one who will say that his findings of a health effect were arbitrarily deleted from a report. Although there were dozens of cases to choose from, the government decided to prosecute on studies of just four compounds: an antibacterial agent used in "deodorant soap" (TCC), an arthritis drug (Naprosyn), a pesticide (Nemacur), and a herbicide (Sencor). All have been retested and found reasonably safe. Adrian Gross, the former FDA investigator who launched this case with a visit to IBT in April 1976, says that his suspicions were aroused by the fact that IBT's data were "unbelievably clean," proving the safety of products a little too convincingly. Going over some of IBT's raw data, Gross saw a term he had not come across before. "TBD, TBD, I kept seeing it and I wondered, what the hell is this?" Gross recalls. It stood for "too badly decomposed," meaning that test animals had died and rotted in their cages before yielding any data. The total breakdown of animal care at IBT, well described in a recent article by Keith Schneider, is an important part of the prosecution's argument. TCC is a Monsanto antibacterial agent which the company hoped to prove safe enough to add to bath soap in larger quantities than the FDA had allowed before. Rats in a long-term (24-month) study in 1971 were being fed various quantities of TCC to learn at what levels of exposure the compound would cause atrophy of the testicles. The room where the rats lived came to be known as "the swamp" because it housed a $120,000 automatic watering and cleaning machine, a new gadget that never worked properly. Bits of feed and feces clogged the water nozzles and drain hoses, drenching some animals in a cold spray while others died of thirst. In these foul conditions, the mortality rate was high, 80 percent by one estimate. But according to the prosecution, IBT's report on TCC did not reflect the premature deaths or the fact that dead rats were replaced with many healthy ones which had not been fed the same test chemical. The government also claims that Calandra ordered an important change in a pathologist's report on TCC, making it seem less hazardous. Donovan Gordon, IBT's former pathologist, has become a "Faking It: The Case Against Industrial Bio-Test Laboratories," in the Spring 1983 issue of The Amicus Journal, published by the Natural Resources Defense Council, New York, N.Y. 10 JUNE 1983 nthal-Newsweek government witness and is expected to say that he found evidence that TCC was affecting the testis at the lowest dose being given, but that Calandra ordered him to interpret the tissue slides differently, finding no effect. A particularly sensitive question is the degree to which Monsanto was aware of what was happening. It arises because Paul Wright, who originally came from Monsanto, helped run the TCC study while at IBT for 18 months, then returned to Monsanto at a higher level and oversaw the report's drafting and publi Joseph Calandra Ran IBT from 1952 to 1977. cation. Monsanto denies any complicity and declines to comment on Wright. Wright's attorney said during the trial that it is "amazing" to charge that Wright would defraud his own employer by faking a report on TCC. Just the opposite, the attorney said, Wright kept his Monsanto superiors "well informed" of all aspects of the study. While the government is not exploring Monsanto's involvement, the trial inevitably touches the company. Already, an exchange of letters between Calandra and Monsanto's manager of toxicology in 1975 has become public, revealing that the company at least once did tell IBT how to express its findings. The case involves Aroclor 1254, a compound containing PCB's, not a part of the trial. Monsanto had IBT change the wording on a report from "slightly tumorigenic" to "does not appear to be carcinogenic." It was done, a spokesman says, to ensure that all reports on Aroclor were phrased consistently. Three other products cited by the prosecution had problems similar to TCC's. The government charges that the Nemacur and Sencor studies were cut short because they were running past the deadline, and that the last 4 months' data were fabricated. In addition, the government says that the positive control mice which were supposed to get tumors did not, so Keplinger lifted tumor data from a study at another lab and used them instead. Smith testified that in doing this, Keplinger transformed skin-painting data into feeding data, figuring that the former would convert roughly to a feeding rate of 1 part per 1000. In the case of Naprosyn, the govern Adrian Gross Found IBT's data too clean to be true. ment charges that all the final blood and urine data were fabricated, because no blood and urine samples were collected. The government also claims that an appendix on gross pathological observations was entirely invented. The animals had died and been thrown out before it was written. Evidence has been introduced indicating that Smith would not include these false data in the Naprosyn report, but was tricked into signing his name on another piece of paper, which was attached later. IBT grew from obscurity in 1952 to a huge multistate laboratory in the 1970's. It seemed to have a knack for making chemicals come out clean in tests. The prosecution is working on the assumption that IBT was greedy, taking on too much with too little expertise. In the late 1960's, as a new environmental consciousness awakened in the nation, federal regulators demanded more laboratory proof that chemicals were safe. IBT leapt at the opportunity, not knowing how to fulfill its commitments. Smith testified that in the early 1970's he recalls that there were only seven 1131 technicians to take care of 10,000 to 15,000 rodents, each of which was supposed to be monitored individually. When Calandra quit, IBT had a business backlog of $25 million. By then, management controls and staff morale had broken down. The company collapsed from within. After the fiasco became known, the U.S. and Canadian governments jointly began reviewing all compounds that relied on IBT studies for marketing approval. The few affected drugs have been cleared, but the record on pesticides is not as good. A summary prepared in February by Kevin Keaney of the Environmental Protection Agency (EPA) shows that of 1205 key studies, only 214 have been found valid. Many are being replaced, but at present. 737 are listed as invalid with no immediate prospect of replacement. Other officials say this summary does not reflect the great number of replacement studies that have been sent to EPA in the last 2 years. On the other hand, Keaney's summary says nothing about the quality of long-term toxicological studies supporting the IBT problem pesticides. Ninety-five percent of these studies are very poor, according to an official with first-hand knowledge. The first full report on all of this is expected in mid-June," according to EPA. Federal laboratory inspectors agree that the cheating they found at IBT was in a class by itself. After Senator Edward Kennedy (D-Mass.) held hearings on these problems, Congress passed a Good Laboratory Practices Act in 1978. The two responsible agencies (EPA and FDA) began an inspection program to prevent future IBT's from happening. The new regulations have driven marginal labs out of the market, in particular, out of the complex business of running long-term studies. A couple of firms withdrew only after being hit with criminal indictments, similar to the one in Chicago. An experienced lab watcher at EPA says that no other company since then has tried to grow as rapidly in as many areas as IBT. Today, any outfit that seems to build momentum quickly is watched. Despite the attention this issue has received, one EPA official concedes, there are still three or four testing centers that have a record of submitting sloppy work or of losing data. They are seen as living in a kind of limbo. EPA has not yet decided how it should regard the work they did in the past in support of pesticide registrations. It is a knotty problem, and one EPA is unlikely to solve by mid-June.-ELIOT MARSHALL 1132 On most days during this long hot summer a jury in Chicago's federal district court has heard evidence of what some have called the "most massive scientific fraud" in this nation's history. According to federal prosecutors, executives of the Industrial Bio-Test Laboratories (IBT, Northwood, Ill.), once this country's largest independent toxicity testing facility, systematically falsified test data collected on scores of drugs and chemicals. The charges read like passages in a horror novel. The stench of the IBT animal room, known as "The Swamp," was so noxious that government inspectors armed themselves with gas masks before entering it. And, government lawyers charge, when test chemicals caused massive animal deaths, IBT officials either substituted live animals in mid-stream, or fabricated favorable test data out of whole cloth. Most independent laboratories insist that IBT was an abberation. "The prob lems aren't with established labs, but The government charges that But neither chemical industry executives nor Washington regulators are ready to take that on faith. They have been taking a closer look at testing facilities and they have not always liked what they have found there. "When you look at most of the major laboratories you see problems," says Thomas Hamm, director of the toxicol. ogy department of the Chemical Industry Institute of Toxicology (CIT), which A new generation of computer is helping to eliminate a host of lab error in records. recently opened its own toxicity labs. "Contract labs have a difficulty with the quality and timing of their reports that is almost universal." Other industry specialists complain of the physical facilities at many of the labs. "Sometimes you'll find labs where the animal rooms are not washed down and sealed off from one another," says Steven J. Barbee, toxicology manager for Olin's environmental hygiene and toxicology department. The push is on from all sides to rectify those problems. Last year the testing industry itself closed ranks around the National Assn. of Life Sciences Institutes (NALSI), which currently is working on a professional code of ethics and a laboratory accreditation program. Several chemical companies have started their own toxicity labs, while others have stepped up the level of scrutiny they give to the labs they use. Charles H. Frommer, Velsicol's director of regulatory affairs, sums up the prevailing view: "There [can be] no room for IBT to happen again." Lower quality. When pressed, most chemical industry specialists concede that the independent laboratories do not bear all of the blame for shoddy practices. For one thing, Hamm notes, there is a tendency toward bargain-hunting among some companies that pressure labs to stress cheap rather than good. "There is very strong competition among labs for work right now and some companies will try to get labs to submit the lowest possible bids so that they can get their work done as cheaply as possible," Hamm says. The labs, as often as not, will put forth less than their best effort, he says. Even more disturbing to most observers, companies often tacitly encourage labs to proceed with an experiment that has gone awry, simply because so much has already been invested. It is often difficult to place blame: Labs may have been provided with diseased animals from their own contractors; inadvertent accidents may have scuttled an experi |