Telling a half-truth doesn’t work for drugmaker

Under securities law, a publicly disclosed half-truth is worse than no truth at all, according to an appeals court opinion filed this week involving Arena Pharmaceuticals Inc. The decision has an intuitive moral appeal. But it’s not at all clear that it makes sense from the standpoint of investors, who might be misled just as thoroughly by failure to disclose material information as they would be by partial disclosure.

The case arose from a scenario painfully familiar to anyone who follows or trades in pharma stocks. Arena was developing a weight-loss drug called lorcaserin, which had entered the stage of human clinical trials. At the same time, the Food and Drug Administration required lorcaserin to be put through a separate, nonclinical trial in rats to see whether it created an unacceptably elevated risk of cancer.
The human clinical trials went fine—the drug worked, insofar as such studies can reliably demonstrate. The rat study was another story. Lorcaserin caused strongly elevated levels of cancer in rats.
From a scientific standpoint, however, that wasn’t the end of the story. It’s possible that a drug that causes cancer when given to rats in high doses might still be safe in humans. Much smaller doses might be less dangerous, or a rat’s biology might cause a distinctive interaction with the drug.
Arena thought that both of those theories were relevant to lorcaserin. It argued both that the drug was safe at lower doses and that the cancer in rats was because of the drug’s interaction with a hormone known to cause cancer in rats, but not in humans. Arena engaged in extensive private discussions with the FDA and in further rat testing. Eventually, the FDA came around, and lorcaserin was approved. It’s now available on the market and in use.
In March 2009, after Arena had gathered the data to make its scientific argument to the FDA, but before the FDA had accepted it, Arena’s then-chief executive officer, Jack Lief, told investors on a call that he was confident in FDA approval “based on the Phase II data, the Phase I data, the preclinical studies that was done, [and] all the animal studies that have been completed.” The company repeated similar claims in an SEC filing that said “the long-term safety and efficacy” of the drug had been “demonstrated” by “preclinical, animal studies.”
On Sept. 14, 2010, the FDA posted Arena’s submitted briefing on its website. The stock crashed at the disclosure of the possible cancer risk, falling 40 percent in a single day. A shareholder derivative suit followed, alleging that Arena misled the public with its earlier statements about the animal studies.
A federal district court dismissed the suit, but the US Court of Appeals for the 9th Circuit reinstated it. Under SEC Rule 10b-5, the most basic and arguably the most important in securities regulation, it’s unlawful to knowingly make a material misrepresentation of fact in connection with the marketing or sale of securities.
The technical legal issue in this case is whether the suit sufficiently alleged a knowing violation of the securities laws.
Arena’s position was simple: It told the truth as it understood it. Yes, it had an ongoing conversation with the FDA about whether lorcaserin would cause cancer in humans. But based on its own testing, including rat testing it had done to alleviate worries about the early cancer finding, the company was confident that it had shown the drug wasn’t likely to cause cancer in humans. Arena was therefore telling the truth when it said it was confident of approval. To top off the argument, Arena was right: The FDA eventually did approve the drug.
The 9th Circuit disagreed with Arena’s position. It said that Arena may not have had any obligation to disclose anything about animal trials. But once it mentioned the animal trials, the court said, it was potentially misleading for Arena not to disclose that some of the rat studies showed elevated cancer risks. Therefore, the court concluded, the shareholders allege facts that were sufficient to show a material misrepresentation. In practice, that almost certainly means Arena will have to settle the suit.
The decision follows a moral intuition that it’s worse to tell a half-truth than to remain silent. Under ordinary circumstances, that might be true. When the FDA posted the briefing, it seems highly likely that the stock would have fallen in precisely the same degree.
Put another way, it seems unlikely that the market was giving special weight to Arena’s reference to animal studies, weight that meaningfully increased the value of the stock.
The upshot is that the true problem with the securities laws isn’t the issue of half-truths, but the background assumption that nondisclosure is legally permitted. Arena kept a secret, and when the FDA disclosed it, the share value fell. There’s something wrong with this picture—but it isn’t half-truth, it’s the suppression of material information about the drug testing process.
— Bloomberg
Noah Feldman is an American author and the Felix
Frankfurter Professor of Law at Harvard Law School