A popular weight-loss pill was buoyed by studies that understated its harms

A weight-loss pill taken by millions of patients in the last two decades has been propped up by problematic clinical studies that “systematically understated” the drug’s potential harms, according to a new analysis. Danish researchers who reviewed data summaries and published journal articles found that seven drug trials funded by the drug maker Roche in the 1990s downplayed the frequency of apparent side effects like diarrhea or incontinence.

The drug, known to scientists as orlistat, and marketed in the US as Alli, has generated hundreds of millions in sales but has slumped in recent years in part because of a reputation for unpleasant gastrointestinal side effects.

The analysis shines light on “something that clinicians often are suspicious of: that the adverse event reporting in clinical trials doesn’t give the whole picture,” said Dr. Raj Padwal, a clinical pharmacologist and internist at the University of Alberta, who was not involved in the analysis.

The new research, published Tuesday in PLOS Medicine, didn’t examine the data reports that were submitted to the Food and Drug Administration when orlistat was approved, first in 1999 as a prescription drug sold by Roche as Xenical or later as a low-dose, over-the-counter pill sold by GlaxoSmithKline as Alli. (The FDA did, however, review the data from the trials in question.)

Those trials, conducted in the 1990s by researchers at academic or medical centers, resulted in journal articles. To evaluate the thoroughness of these articles, the Danish researchers compared them to data summaries submitted to European regulators, which were obtained by filing Freedom of Information Act requests. They found that the publications disclosed just a fraction — between 14 and 33 percent — of patients’ so-called “adverse events” as compared to the data summaries.

For example, two of the publications left out apparent side effects that occurred in fewer than 5 percent of trial participants; another counted multiple reports from the same patient as just one adverse event.

Another two articles, which are listed currently on Alli’s website as evidence of the drug’s safety and effectiveness, were also problematic. One of them only reported apparent side effects deemed “common”; the other omitted reports of apparent side effects that weren’t at least twice as common among trial participants taking the drug as those taking a placebo.

All of the studies in question still stand in the scientific literature.

Anja von Treskow, a spokeswoman for Roche, said the reporting on the trials “was done according to the standards of the time” and that the company “continuously update[s] our methods” to comply with changing standards.

Joanmarie Goddard, a spokeswoman for Alli’s marketer, GlaxoSmithKline, said the company “take[s] adverse event reporting for our products very seriously and routinely monitor[s] safety information in conjunction with health care professionals and regulatory agencies worldwide.”

The rise and fall of a weight-loss pill

In the first year after Roche brought Xenical to market in the US, the drug brought in $600 million worldwide. Over the course of the 2000s, the drug was prescribed 11 million times in the US, according to an analysis of data from the consumer group Public Citizen. When GSK launched the drug over-the-counter as Alli in 2007, it blew through the gates, selling $155 million in its first weeks on the market.

The drug, which is meant to be paired with diet and exercise, works by preventing the body from absorbing fat. But excreting lots of fat makes the stool oily, leading to gastrointestinal side effects for many patients.

Even at the drug’s peak, warning signs were apparent. US prescriptions of Xenical declined every year. A 2007 study of nearly 17,000 patients found that just 6 percent of participants were still taking Xenical after one year; after two years, it was down to 2 percent.

That’s an indication that “the balance between efficacy and harms is not very beneficial,” said Dr. Jeppe Schroll, lead author of the new analysis and a researcher at the Nordic Cochrane Centre in Copenhagen.

Then in 2010, the FDA added a warning to both Xenical and Alli cautioning about “rare reports” of liver damage; a 2012 study found such alarming toxicity effects that the researcher immediately reported his findings to the FDA.

Roche no longer markets Xenical in the US, though generic versions of the prescription drug are still available. GSK no longer breaks out Alli sales in its financial reports, and wouldn’t comment on current sales figures for the drug. A pack of 120 Alli pills sells online for $56.

In spite of orlistat’s struggles, drug makers aren’t giving up on the pill. A Canadian drug company, M Pharmaceutical, is working on a reformulated alternative that will steer clear of the existing drug’s “socially unacceptable ‘underwear issues.’”

The problem of hidden harms

It’s hard to say how common the harms reporting discrepancies uncovered in the new orlistat analysis are.

The Danish researchers noted that space restrictions in journals may have played a role in what was disclosed, and that data reporting standards that have since been developed might make their findings less applicable to newer drugs.

Dr. Rohan Khera, a cardiology fellow at UT Southwestern Medical Center who has studied the effectiveness of different weight-loss drugs, called the new analysis “a good start in figuring out that there is a need to study this in further detail.”

Dr. Sidney Wolfe, a consumer advocate with Public Citizen who has unsuccessfully petitioned US regulators to take orlistat off the market, believes the stakes are high. “If you’re a doctor and rely on medical journals, you may be misled. Doctors don’t have the time to go reading the actual raw data,” he said.