Wednesday, May 12, 2010

Why Malcolm Gladwell Is Wrong About Drug Development

Cross-posted from The Science Business on Forbes.com.


UPDATE: Gladwell’s response to my article can be found here.

Ever since Watson and Crick first elucidated the structure of DNA in 1953, medical science has made enormous strides in understanding the basic biological mechanisms of disease. This has allowed scientists all over the world to develop innovative and effective approaches to previously incurable conditions.

Malcolm Gladwell isn't buying it. In the May 17 issue of The New Yorker (subscription required), he implies that we should go back to the old way of doing things: of throwing mud on a wall and seeing what sticks. But his case study for that approach, Synta Pharmaceuticals' elesclomol, is in fact an object lesson in its deficiencies.

Gladwell has become a household name by, as the New York Times put it, seeking to "undermine the ideals of talent, intelligence and analytical prowess in favor of luck, opportunity, experience, and intuition." There is no doubt that luck plays a big role in successful drug development. After all, very few compounds get out of the lab and into human clinical trials. And a drug that manages to makes it into human trials has a mere 8% chance of being approved by the FDA. Contra Gladwell, however, recent history shows that talent, intelligence, and analytical prowess are just as important, if not more so, in the development of successful new medicines.

Synta is commonly described as a biotechnology company because it's a small company focused on the traditional biotech domains of cancer and immunology. But Synta's approach to drug development is as old-school pharma as the 21st century gets.

Most biotech companies try to identify the genes and enzymes that cause a disease, and "rationally" design chemicals or antibodies that specifically target those causes. Synta, on the other hand, uses the same basic approach that drug companies have used for centuries: taking a library of thousands of chemical compounds, and testing those compounds one by one in the lab until something seems to work.

Gladwell is a lucid writer, and he elegantly details Synta's thus-far futile quest to turn its chemical library into everyday medicine for patients. He describes the moment in 2006 when, just as Synta is about to enact a round of layoffs, a company scientist tells CEO Safi Bahcall and colleagues that elesclomol has shown promise in a mid-stage melanoma trial:

"So the lights go down," Bahcall continued. "Clinical guys, when they present data, tend to do it in a very bottoms up way: this is the disease population, this is the treatment, and this is the drug, and this is what was randomized... They go on and on and on, and all anyone wants is, Show us the [bleeping] Kaplan-Meier! Finally [Chief Medical Officer Eric Jacobson] said, 'All right, now we can get to the efficacy.' It gets really silent in the room. He clicks the slide. The two lines separate out beautifully--and a gasp goes out, across a hundred and thirty people. Eric starts to continue, and one person goes like this"—Bahcall started clapping slowly—"and then a couple of people joined in, and then soon the whole room is just going like this—clap, clap, clap. There were tears. We all realized that our lives had changed, the lives of patients had changed, the way of treating the disease had changed. In that moment, everyone realized that this little company of a hundred and thirty people had a chance to win. We had a drug that worked, in a disease where nothing worked. That was the single most moving five minutes of all my years at Synta."

The following Feburary, Synta sucessfully launched an IPO, opening its first day of trading at $10 per share. But a number of hedge funds were skeptical, and began shorting Synta's stock. Their criticism was that, upon further analysis, elesclomol's performance in melanoma was not as strong as it seemed. For one thing, the patients in the placebo arm of the study appeared to be sicker, older, and more refractory to prior chemotherapy than those in the elesclomol arm. Therefore, the skeptics reasoned, the elesclomol-treated patients were destined to look like they were faring better, even if the drug itself was doing nothing.

For the next two years, Wall Street continued to energetically debate the merits of elesclomol in melanoma. Synta's share price gyrated accordingly.

On February 26, 2009, the company made a surprising announcement: that they were halting the phase III trial of elesclomol in melanoma, because a board of independent reviewers found that more people were dying on elesclomol than in the control arm. Synta shares plunged 79%, from $6.39 to $1.36. The shorts had won. Melanoma patients and Synta investors had lost.

Though Synta continues to work on elesclomol, and is encouraged by the activity of a new drug called STA-9090, thus far, the company's old-school pharma approach has yet to prove itself. Despite the best efforts of the able and idealistic team at Synta, its chemical library, thrown on the wall, hasn't stuck.

One gets the impression that Gladwell followed elesclomol over the years in the hopes of using its story for one of his books: the story of how luck and intuition can lead to pharmaceutical success. But the opposite happened, and the account ended up in The New Yorker instead. In Gladwell's final assessment, the story remained one about luck: but about bad luck instead of good.

One can't fault Synta nor its partner, GlaxoSmithKline, for investing in elesclomol's phase III melanoma trial. After all, the phase II data seemed compelling. But the companies were not merely unlucky. Their optimism stemmed from trial results that were less conclusive than enthusiasts believed, with a drug that bore no definitive biological relationship to the disease of melanoma. A better-designed phase II study may have identified these problems earlier, and saved both companies tens of millions of dollars.

Rational drug design, the hallmark of new-school biotech, is no panacea, as Gladwell rightly points out. It took decades for Judah Folkman's insight into how tumors gain a blood supply, called angiogenesis, to turn into an FDA-approved therapy. And it's worth noting that, in melanoma, the new-school method hasn't done any better than Synta's. But most of the new drugs that have successfully beaten back cancer—drugs like Genentech's Avastin and Herceptin, and Novartis' Gleevec—have used this analytical, scientific, rational approach.

As the FDA has gotten more stringent about approving new drugs, the cost of developing new medicines has exploded. And while it remains better to be lucky than good, every investor in biotech research must ask the question: if you are going to bet hundreds of millions of dollars on developing a new drug, why wouldn't you try to optimize your chances by attacking a reasonably well-validated biological target? You may still fail—but you will have given yourself the best opportunity to succeed. And that is why the old-school, Gladwell approach is on the ropes.

No comments:

Post a Comment