WHEN MEMBERS OF THE UPJOHN CO.'S HALCION BUSINESS team convened on Dec. 4, 1991, they knew they had trouble on their hands. After dominating the global market for nearly a decade, the company's blockbuster sleep medication was under attack. Scientists and journalists were questioning Halcion's safety. British regulators had just suspended its license. Litigious consumers were accusing Upjohn of concealing the drug's hazards. And the U.S. Food and Drug Administration (FDA) was investigating the firm for possible misconduct. In this air of crisis, the business team concluded that the company needed to make a dramatic public gesture: it needed to sue someone. "We are not in a position to assess the legal ramifications of such action," a committee spokesman wrote in a memo to superiors, "but we can assess the business ramifications. . . . Initiation of legal action would publicize our intent to defend Halcion against unjust action. This message would encourage ... physicians to continue writing Halcion prescriptions."
Prof. Ian Oswald, a retired Scottish psychiatrist, was a natural target. Ever since governments started licensing Halcion in the late '70s, some users had reported psychiatric side effects ranging from anxiety and memory loss to paranoia and hallucinations. Upjohn had dismissed the complaints, arguing that controlled studies showed no cause for concern. But Oswald had published controlled studies documenting just such reactions. To make matters worse, he had recently appeared in a documentary by the British Broadcasting Corp. (BBC), accusing Upjohn of recording similar problems in one of its own studies and failing to report them.
When Oswald showed up on the front page of The New York Times on Jan. 20 describing the Halcion affair as "one long fraud"-Upjohn's lawyers were ready to pounce. Within days they filed a libel suit in Britain, seeking unspecified damages and an injunction against further charges. "We can only guess at Dr. Oswald's motives for these attacks," Dr. Theodore Cooper, then Upjohn's chairman, declared in a press statement, "but the fact is that he is wrong."
Was Oswald wrong, or has Upjohn deceived consumers and governments to keep a profitable drug on the market? The question is now before Britain's High Court of justice, in London, and the outcome will shape the fate of the world's most widely prescribed sleeping pill. (Though Halcion's sales have fallen by half since 1991, doctors continue to write more than 3 million prescriptions a year in the United States alone.) Legally, the burden is on Oswald to substantiate his charges. But by suing the scientist-and the BBC, for its 1991 documentary-Upjohn has placed itself on trial. Lawyers are now publicly dissecting records that the company has fought for years to keep secret. Those documents may or may not exonerate Oswald and the BBC. But together with records now becoming public in U.S. lawsuits, they suggest that Upjohn has been sloppy at best in testing its product, and evasive at best in defending it.
Halcion's problems didn't start with Oswald or the BBC. In order to get a new drug approved, the maker has to establish its safety and efficacy in clinical trials. Rather than conduct those trials themselves, pharmaceutical companies typically pay physicians to recruit patients, administer a treatment and record the effects. The doctors send their findings back to the company, which analyzes them and submits them to the FDA. Long before critics started questioning the integrity of Upjohn's Halcion studies, the FDA discovered problems with some of the private doctors who conducted them. But instead of distancing itself from questionable clinicians, Upjohn has consistently downplayed concerns about their work.
The first Upjohn contractor to fall was Dr. William C. Franklin, a Houston internist whose work was central to Halcion's 1982 approval in the United States. Just days after licensing the drug, the FDA notified Upjohn that Franklin had confessed to falsifying data in drug tests for other companies. The agency asked Upjohn to discard any data Franklin had contributed and determine whether it could still "justify all claims [of] safety and efficacy." Franklin had led the largest of three studies the FDA deemed "pivotal" to Halcion's approval. But Upjohn didn't mention that in its reply. It reported that his work was negligible. The company stood by that strange assertion two years ago, when NEWSWEEK revealed the true extent of Franklin's work, but it has now changed its story. In testimony for the Oswald lawsuit, Dr. Robert Straw of Upjohn says he was simply unaware of some of Franklin's work when he drafted the misleading letter.
Upjohn has been equally slow to distance itself from Dr. Samuel I. Feurst, a Mississippi clinician who died several years ago. Upjohn contracted with Feurst to conduct a large Halcion study in 1978 and included his findings in its successful 1982 application. Feurst reported virtually no complaints among Halcion users. But when FDA investigators audited his clinic in 1983, they found no evidence that the people he had listed as patients had ever taken the drug. Dr. Alan Lisook of the agency's Division of Scientific Investigations now calls the study one of the most blatant cases of fraud his department has ever seen. Upjohn agreed in 1984 to exclude Feurst's findings from its database-yet the company has continued to tout them. Straw cited Feurst's work as definitive during a 1991 interview with the BBC (an oversight, according to the company). And in court testimony this month, Upjohn president Ley Smith acknowledged that the company submitted Feurst's data to British regulators in 1991 without mentioning the FDA's 1984 action. (Upjohn refuses to comment further while the lawsuit is in progress.)
Unlike Franklin and Feurst, Dr. Louis Fabre has not been disqualified by the FDA, and he strongly denies any wrongdoing. But records from lawsuits now underway in the United States suggest that Upjohn has confirmed, and downplayed, problems with his data. Fabre, a Houston psychiatrist, conducted more than a dozen Halcion studies for Upjohn during the 1970s. In 1992, when one of his former patients accused him of scientific misconduct, Upjohn sent a compliance manager to investigate. She found that "certain patients were enrolled in studies more than once" and that some patients had identical birth dates and strange names (allegedly including "Clark Kent" and "Ellis Dee"). Yet in a September 1993 letter to the FDA, Upjohn concludes that "there is not sufficient evidence to question the validity of the data collected from this site."
Even when its contractors have followed the rules, Upjohn hasn't always reported their findings accurately. The company has acknowledged disparities between clinical reports it received from the field and summaries it submitted to licensing agencies. Upjohn maintains that any errors have been innocent. But court papers are bringing new disparities to light and casting doubt on the firm's explanations.
In the summer of 1991, as lawyers for Upjohn and a former Halcion user prepared for a court battle in Utah, both sides discovered major flaws in a 1973 study known as Protocol 321. In its reports to regulatory agencies, Upjohn had mentioned just four events of nervousness or anxiety among 28 healthy prisoners who received large nightly doses of the drug. Yet original records from the study revealed more than 70 such events. The records also showed that seven of the 28 prisoners had complained of paranoia, though Upjohn had mentioned only two. In letters alerting agencies to the newly exposed flaws, the company blamed innocent "transcription errors."
But Upjohn's internal correspondence contradicts that claim. In an Aug. 22,1991, memo, an Upjohn official named Mayo Lawrenz reported that an in-house inquiry was making unwelcome progress. "There are some real concerns," he wrote, "because of some manipulation of data by one of our people and ... the involvement of this individual in other product submission." A plausible inference is that Upjohn found evidence of wrongdoing but reported just the opposite. Testifying in the Oswald trial, company officials have dismissed Lawrenz's statement as "inappropriate use of language," but they haven't explained what he meant to say.
However they came about, t e discrepancies in Protocol 321 aren't unique. For example, "restlessness-nervousness" typically showed up as a single side effect on the forms that doctors used to assess patients' responses to Halcion. But in at least two studies, Upjohn's scientists split restlessness and nervousness into two separate categories as they tabulated findings. Patients who suffered "restlessness-nervousness" were arbitrarily assigned to one category or the other. By including the data on "nervousness" in Halcion's package insert-but making no mention of "restlessness"-Upjohn effectively obliterates affected patients and (intentionally or not) improves the drug's side-effect profile.
Some patients may have vanished by less subtle means. Federal law bars drug companies from omitting any clinical information from license applications, but court papers raise the possibility that more than two dozen participants in a 1973 study called Protocol 6048 were never reported to the FDA at all. Upjohn maintains that Dr. James Donovan of Bakersfield, Calif, treated only 15 patients in the study. But witnesses who have reviewed the company's accounting records say Donovan was paid for treating at least 40. And two experts who have reviewed diaries kept by the 15 reported patients say that most complained of classic side effects that neither Donovan nor Upjohn disclosed.
Upjohn did report side effects in 16 patients treated by a second doctor in Protocol 6048, but it may have misreported the causes. According to Upjohn's report on the study, half of the patients with severe side effects were on Halcion and half had received a comparison drug. But that isn't what the doctor, Roy Wiggins, reported to the company. According to the doctor's forms, three fourths of the side-effect sufferers (12 of 16) had received Halcion. In court testimony, Upjohn has acknowledged that one of its employees altered codes on the forms, effectively moving several Halcion takers on to the comparison drug. The company claims that the altered forms are the accurate ones-that Wiggins misclassified the patients and that the changes merely set the record straight. But Wiggins and his secretary insist they made no such errors.
Upjohn has paid dearly for its lapses since the first ones came to light two years ago. At least five countries have banned the drug, yet the company has managed to save some face in the United States. In May 1992, an FDA advisory committee reviewed Halcion's safety record and concluded that the drug should stay on the market With stronger warnings, despite its side effects. And in November, FDA officials Wrote Upjohn a three-sentence letter to say it bad concluded a separate fraud investigation and found that "no further action by this agency is warranted at this time." in a press statement, Upjohn's chairman boasted that after reviewing thousands of pages of documents, the FDA had exonerated the company and its employees.
That's not quite what happened. In fact, FDA staffers were still busy investigating Upjohn's handling of Protocol 321 when a committee of agency officials decided to cut the inquiry short-and the committee's decision was no endorsement of Upjohn. In the FDA's official "close-out memo," which NEWSWEEK obtained under the Freedom of Information Act, George Prager of the Division of Scientific Investigations writes:
FDA's investigation strongly suggests that the firm was aware that the report [on Protocol 321] was inaccurate and incomplete but continued to withhold notification to the FDA that the report was flawed.
Prager goes on to explain that the FDA committee considered giving the case to the Justice Department, for investigation by a grand jury, but decided to let it drop after determining "the facts withheld by the firm" weren't critical to the drug's approval.
Upjohn has vet to voice any public regret over the Halcion affair. Instead, it continues to paint itself as the victim of a few misguided critics and a scandal-hungry press. The company's internal documents "can be understood when taken in context," spokeswoman Kaye Bennett wrote when asked to comment for this article. "We've not been given the opportunity to describe that context in the news media, where charges and accusations are vastly more interesting than the painstaking minutiae of scientific discussion. For precisely that reason, we chose the courtroom, rather than the media, to tell our story." The company will have several months to tell its story as the Oswald trial continues in London.
