18 November 2006. Perhaps some good can come from the Alzheimer’s Disease Anti-inflammatory Prevention Trial (ADAPT), which was prematurely halted almost two years ago because of safety concerns. In yesterday’s PLoS Clinical Trials, Barbara Martin, Johns Hopkins Bloomberg School of Public Health, Baltimore, Maryland, and colleagues of the ADAPT Research Group presented safety data from the trial because, as they write, “We believe we are obligated to make these results available to the scientific community and the public at large….” But the controversy surrounding the decision to halt the trial persists, with Steven Nissen, Cleveland Clinic Foundation, Ohio, claiming in an accompanying perspective that the trial was improperly terminated.
ADAPT was designed to test the theory that non-steroidal anti-inflammatory agents, in this case naproxen and the cycloxygenase-2 (COX-2)-specific inhibitor celecoxib, can prevent or slow the progression of Alzheimer disease in an at-risk population—people with a family history of the disease. The trial was terminated in December 2004 following this chain of events: The withdrawal of and public controversy surrounding another COX-2 inhibitor rofecoxib (see ARF related news story); troubling safety data on celecoxib from the Adenoma Prevention with Celecoxib (APC) trial; and preliminary data from ADAPT that raised concerns about naproxen safety (see ARF related news story). In all cases, increased risk for cardiovascular and cerebrovascular events such as myocardial infarction, congestive heart failure, transient ischemic attack, and stroke was the death knell.
Now, the full publication of the data that took its toll on ADAPT shows that in the 2,500 plus patients that were actively taking part in the trial, there were no significant increases in cardio/cerebrovascular events in those taking celecoxib. Nor was there any significant increase in individual events for those taking naproxen. But looking at the composite scores for cardiovascular death, infarction, stroke, and ischemic attack, one sees that the hazard ratio for naproxen-takers was 1.63 (95 percent confidence interval was 1.04-2.55, p = 0.03). However, it should be emphasized that the trial was neither designed nor powered to study cardio/cerebrovascular effects and the number of such events was small (37 total events in 1,070 control patients and 40 events in 713 patients taking naproxen). Nevertheless, the authors write that “…we believe it is essential to make data from long-term trials such as ADAPT available to contribute to the larger body of evidence on drug safety.”
Nissen’s major objection is that stopping the trial was inappropriate. “These actions unnecessarily created public consternation and denied the participants in the trial the opportunity for their commitment to result in any benefit to themselves or society,” he writes, citing a recent meta-analysis to conclude that the relative cardiovascular risk for naproxen is less than 1.0 (McGettigan et al., 2006). He also questions the appropriateness of having NIH leadership “unblinded” to the trial data and suggests that the trial steering committee and the drug safety and monitoring board should have “resisted efforts to improperly interrupt an ongoing trial.”
In reply, John Breitner, University of Washington School of Medicine, and Martin and Curtis Meinert, also at Johns Hopkins, write that ultimately it was the responsibility of the ADAPT steering committee and not the NIH to decide to terminate the trial. They emphasize that there was more to the decision than purely safety data, including the need to obtain revised consent, which would have had to be reviewed by seven different institutional review boards. Breitner and colleagues also emphasize that ADAPT was a primary prevention trial with seven to 10 years of planned intervention and no anticipated short-term benefit. In fact, the paper shows that the probability of a cardio/cerebrovascular event over the two years of the trial increases more rapidly in the naproxen group compared to placebo, while in the celecoxib group, that probability appeared to increase, too, in the last few months of the trial. Extrapolated over the projected length of the trial, that could have serious consequences.
Perhaps the most important question that arises from the trial and its fallout is, How much risk is too much to ask ordinary people to assume in the search for a drug that might prevent a devastating disease such as AD?—Tom Fagan.
ADAPT Research Group. Cardiovascular and cerebrovascular events in the randomized, controlled Alzheimer’s Disease Anti-inflammatory Prevention Trial (ADAPT). PLoS Clinical Trials. 17 November, 2006;1(7):e33. Abstract
Nissen SE. ADAPT: The wrong way to stop a clinical trial. PLoS Clinical Trials. 17 November, 2006;1(7):e35. Abstract
Breitner JCS, Martin BK, Meinert CL, for the ADAPT Research Group. On the suspension of treatments in ADAPT. PLoS Clinical Trials. 17 November, 2006;1(7):e35.