Monday, January 29, 2007

DNA making big immunology bet

I hadn't realized that DNA had raised their immunology efforts to such an explicit priority as mentioned in this article (and why would they?)

One interesting question for DNA: will their solutions also be biologicals, as most of their current oncology offerings? They certainly have the expertise to do so, but you could argue that other companies (particularly Amgen) have a big head start in this area.

The other interesting stat in this article: DNA went 15 for 15 in it's critical trials (Avastin, etc.), which the company pegs the odds of success at 1 in 300,000,000.

Not to diminish anything that DNA has accomplished, but I think this stat would only be correct if these 15 compounds were developed in one insanely profitable month of serial chemical synthesis. The 15/15 1:300M odds are built on a probability of one success at 1 in 9155 (.0109%), which I think is Genentech's parallel for the old saw of "1 in 10,000 compounds ever receive FDA approvals."

The best analysis would be to identify which starting point these 15 success all have in common. For fun, let's assume the accomplishment is 15 straight compounds to go from IND to FDA approval. If you believe as some do that an IND compound has a 1 in 10 chance of FDA approval, then Genentech's accomplishment has odds more like 1:32768. (I'm a little rusty, but I think the chance of doing anything 15 times in a row is 2 (to the) 15th (32768) times the probability of one success. Anyone want to check my math?)

UPDATE, based upon actual math that works: 1:300M represents 15 consecutive successful trials each with a ~27% chance of success, which is probably DNA's assumed success rate for a compound reaching Phase I trials. (The Milken template suggests 20% for NCEs, and acknowledges that biologicals have slightly higher success rates.)

It's more likely, though, that the 15 trials cited by DNA include multiple trials for single compounds, and various stages of trials (i.e. does this figure include Rituxan trials for RA? If so, Phase 1 trials really weren't risky (or perhaps even necessary).

No matter the math involved, DNA's accomplishment is impressive!

1 comment:

crimsoncanary said...

Immunology is a much more lucrative market than cancer - if you can get a patient on biologicals they may stay on them for many years - cancer patients, by the nature of the disease are never going to take drugs for so long.

I think you'll find that the 15/15 are not molecules but trials. Genentech has only had one new product launch in the last year that was Lucentis for Macular degeneration...all other trials have been supplementary for its cancer drugs, mainly Avastin and Rituxan.