What can you say in a press release about a clinical trial? “Darn near anything, apparently” will be the response from many people who’ve been seeing them over the years. But really, what can you say, legally? Is there some point where you’ve clearly crossed the line into fraud, or are all these things just varying interpretations of scientific data?
That uncomfortable question has been working its way through the court system in the person of W. Scott Harkonen, former CEO of Intermune. This case is back in the news thanks to a long article in the Washington Post (pointed out to me by a reader of this blog in the comments section here). Here’s the background: Intermune was selling Actimmune (interferon gamma-1b) for two rare-disease indications, but wanted to break into the much larger market for idiopathic pulmonary fibrosis (IPF), for which there were basically no therapies at all.
Unfortunately, the trial didn’t go the way that everyone had hoped. Here’s the Washington Post‘s take on it, complete with quotation marks around the phrase “statistical significance”:
In all, 330 patients were randomly assigned to get either interferon gamma-1b or placebo injections. Disease progression or death occurred in 46 percent of those on the drug and 52 percent of those on placebo. That was not a significant difference, statistically speaking. When only survival was considered, however, the drug looked better: 10 percent of people getting the drug died, compared with 17 percent of those on placebo. However, that difference wasn’t “statistically significant,” either.
Specifically, the so-called P value — a mathematical measure of the strength of the evidence that there’s a true difference between a treatment and placebo — was 0.08. It needs to be 0.05 or smaller to be considered “statistically significant” under the conventions of medical research.
Technically, the study was a bust, although the results leaned toward a benefit from interferon gamma-1b. Was there a group of patients in which the results tipped? Harkonen asked the statisticians to look.
It turns out that people with mild to moderate cases of the disease (as measured by lung function) had a dramatic difference in survival. Only 5 percent of those taking the drug died, compared with 16 percent of those on placebo. The P value was 0.004 — highly significant.
But there was a problem. This mild-to-moderate subgroup wasn’t one the researchers said they would analyze when they set up the study. Subdividing patients after the fact and looking for statistically significant results is a controversial practice. In its most extreme form, it’s scorned as “data dredging.” The term suggests that if you drag a net through a bunch of numbers enough times, you’ll come up with something significant sooner or later.
Yes indeed. In fact, the term suggests that because that’s absolutely true, and it can be proven mathematically. If you take a large enough pile of clinical data, with enough variables, and break it down into enough subgroups, the odds get better and better than something will look significant simply by chance. You can say when you’ve done enough to have a 50% chance of that happening, or a 90% chance, or whatever cutoff you like. It isn’t voodoo, although statistics are poorly understood enough to make it sound so if you’ve never had to think about the issues.
The situation Intermune found itself in is a tough one, but it’s not uncommon. This is why clinical trial design is so critical. You want to run one that has the best chance of showing a clinically relevant effect, but you’re not going to be able to cover every sort of patient and subgroup, either, because time and money get out of control very quickly at this point and can absolutely sink your whole effort. The “might have worked” result is a hard place to be.
Unfortunately, Harkonen does not seem to have reacted well to it. He personally wrote a press release with these headlines: “InterMune Announces Phase III Data Demonstrating Survival Benefit of Actimmune in IPF” and “Reduces Mortality by 70% in Patients with Mild to Moderate Disease.” If you read the rest of the text, it did mention that the study missed its primary endpoints, but there was no mention that the mild-to-moderate group was a post hoc analysis. He ended up charged with wire fraud, and was convicted. That conviction has been appealed (here’s the appeals court ruling against him), and now there’s talk of going to the Supreme Court, although I don’t know if his legal team has asked for certiorari yet or not.
So, what’s the line between protected free speech and fraud? The newspaper article mentions the recent United States v. Alvarez decision, which ended up overturning the “Stolen Valor Act” making it illegal to claim (falsely) to have military decorations. The Alvarez in question was an idiot who claimed, in his job on a water district board, to have won the Medal of Honor. Sure thing. But although I Am Not a Lawyer, this seems like a different case to me. One of the big points of contention in the Alvarez case was whether or not it was a defamation, and if so, who was being defamed, and how they were damaged. The Supreme Court was split up all over the place in their own opinions, with even justices that agreed with the majority disagreeing over what they were agreeing about.
This case seems more straightforward, at least to me. I think it’s more a case of commercial speech. There’s probably case law by the shelfload on this, and I may be totally wrong, but I would regard all press releases by publicly traded companies as commercial speech, simply due to their effects on the stock prices involved. If the case makes it to the Supreme Court, I’d expect it to be so that the justices can try to untangle that issue (as in this previous case, which the Court dodged due to a procedural matter).
Oh, one other thing, which doesn’t necessarily bear on this case but is worth mentioning as an illustration of its issues. Intermune actually did go forward with another trial of Actimmune against mild-to-moderate IPF patients. And sure enough, it didn’t work.