Probe turns up flaws in Alzheimer's study.

Probe turns up flaws in Alzheimer's study

When a 1986 research report described dramatic improvements in Alzheimer's patients on an experimental drug, some scientific reviewers hailed the work as a "triumph." The drug may yet fulfill that promise, but the report has since drawn a barrage of criticism from researchers troubled by its missing details. The Food and Drug Administration has now published the findings of a probe that reveals serious flaws in the 1986 study.

William K. Summers, a psychiatrist affiliated with the University of California, Los Angeles, led the team reporting the encouraging findings in the Nov. 13, 1986 NEW ENGLAND JOURNAL OF MEDICINE (NEJM). The study focused on tetrahydroaminoacridine (THA), a compound that inhibits a crucial brain enzyme required to break down the neurotransmitter acetylcholine (SN: 11/15/86, p.308). Most researchers who study Alzheimer's disease believe a lack of acetylcholine may cause memory loss, confusion and impaired thinking in Alzheimer's patients. Indeed, the 1986 report concluded that THA improved memory and cognitive abilities in all 12 men and women with Alzheimer's disease who had taken the drug for about a year.

The report initially met with positive reviews. By the spring of 1987, however, some scientists had criticized the researchers' methodology and others questioned Summers' involvement in Solo Research, Inc., a for-profit company set up shortly after the report's publication for the purpose of selling the experimental drug to Alzheimer's patients. These concerns triggered the FDA probe, designed to determine whether the researchers' laboratory records matched their published account of the study.

The results of that inquiry, published in the Jan. 31 NEJM, reveal "numerous differences" between the journal article and the laboratory records. For example, the researchers stated in their paper that they assigned volunteers to THA or placebo groups in a random and double-blind manner -- a standard experimental method that helps eliminate bias by keeping volunteers and researchers unaware of which participants receive treatment or placebo. FDA investigators, however, say they found no documentation of randomly assigned treatment or consistent blinding procedures.

"A substantial proportion of the evidence presented was not collected under blinded conditions, as implied or claimed, but was assembled after the fact from a mixture of sources," states the FDA report. "Not surprisingly, the inspection dramatically altered the agency's view of the importance of Dr. Summers' report. At best, we consider the evidence to be the equivalent of uncontrolled, anecdotal clinical information."

In a response published in the same issue of the NEJM, Summers and his colleagues deny these allegations, calling the FDA report "a personal attack on our group." In a telephone interview, Summers told SCIENCE NEWS that FDA inspectors dismissed handwritten notes that document the team's random assignment of subjects and blinding procedures.

FDA calls the results of its investigation "disturbing," noting that the flawed drug study spurred a flurry of requests for THA, an unproven drug later shown to cause liver damage when given at high doses. Indeed, soon after the 1986 study came out, phone lines at the Chicago-based Alzheimer's Association were jammed by calls from people frantic to get THA, says Creighton H. Phelps, the association's vice president for medical and scientific affairs.

Controversy notwithstanding, THA's benefits may ultimately prove real. A spokesman for THA manufacturer Warner-Lambert Co. hints that a well-designed, double-blind trial of THA has yielded "encouraging" results, which the company may release as early as March.