Last week, the Wall Street Journal ran an article (if you have a subscription, it's here)on the Novartis research center in Cambridge (the MA one) and its director, Mark Fishman. I wrote a couple of blog posts back when the place was first starting up, which I'll have to unearth at some point and compare to reality. So far, that reality looks a bit different from the way that the rest of Novartis (and much of the rest of the drug industry) are doing things.
The article makes much of the emphasis on genetic causes of disease that Fishman has spoken of, which is only natural, since he's known as a force behind the decipherment of the zebrafish genome. (If you're not in the biomedical sciences, you may wonder what the big deal is about zebrafish. Actually, they're one of the most heavily studied creatures in developmental biology, sort of a fruit fly with fins.)
That genomic strategy is probably going to be either a big winner or a big loser, but it's too soon to say which. He has several others which fall into that category, too. From the article:
"Dr. Fishman also is changing how Novartis selects diseases to study. Traditionally, it and other big companies have decided by calculating the size of the potential market. Dr. Fishman thinks Novartis has a better chance of increasing sales and curing people if it goes after illnesses whose genetic makeup it has the best chance of deciphering, and diseases for which few treatments are available. In some cases, that has meant studying illnesses that affect relatively few people, such as multiple sclerosis.
Most drugs today are tested on large groups of patients suffering from a common illness. But Dr. Fishman believes that many diseases, such as hypertension and diabetes, can be divided into subgroups. So Novartis has begun conducting more specific drug trials in smaller groups of patients, which Dr. Fishman says has the added benefit of saving money and time."
That first one is sort of the Gleevec story writ large, but I've said several times that Gleevec's status as a billion-dollar orphan drug says more about the state of the oncology market than it does about the compound. In general, drugs are going to sell in proportion to the size of their market. (Now, you can underestimate that size or overestimate it, of course - this is far from being a science. Pfizer underestimated Viagra's potential at first, and Pfizer's competitors seem to have turned around and then overestimated the exact same market.) I worry that some of these decipherable genomic targets will be for diseases that affect very few people. The resulting drug will end up with quite a price tag. But, still, Novartis is a Swiss company, and I assume that the Swiss have thought this through.
And as for the smaller trials, that's indeed where the industry would like to go. But there's a shortage of biomarkers (so far) for being sure that you're dividing up your patients into the right groups of responders and non-responders. Validating a new genetic marker could, in some cases, be just as expensive as developing a drug. And if you pick the wrong one, as one of my commentators once pointed out, you could find yourself testing your drug against a group that's actually worse than random chance would have given you.
I don't want to give the impression that Fishman's ideas won't work. They're good ones, and worth trying. I think, though, that some other drug companies may be just as glad that someone else's money is trying them out first. But perhaps Novartis will eventually have their revenge.