Venture Capital Fantasyland: Venrock’s Proposal to Reform Drug Development

Here come the dancers one by one,
Your mamas calling but you’re having fun
You’ll find you’re dancing on a number nine cloud
Here come the dancers one by one,
Your mamas calling but you’re having fun
You’ll find you’re dancing on a number nine cloud
Put your hands together and sing
it out loud

It’s all part of my rock ‘n’ roll fantasy.
It’s all part of my rock ‘n’ roll
dream.

-Bad Company, Rock ‘n’ Roll Fantasy (1979)

The Calculus of Cures, a Perspective in the April 17 New England Journal of Medicine, presents a provocative proposal to transform drug development and distribution. Although it makes some good points, overall it’s an unrealistic and self-serving piece by two leading venture capitalists.

Here’s the essence of their pitch:

• Cut development costs by 90% and time required by 50% by lowering the threshold for approval to focus on efficacy and “fundamental safety”
• Provide conditional approval for drugs, and then require more post-marketing surveillance to pick up rare events that aren’t captured in clinical trials anyway. Pull drugs off the market that later prove to be unsafe
• Reduce selling costs to primary care physicians by introducing drugs that are more differentiated (a supposed effect of their development reforms), use IT to reduce the need for in-person sales forces, and improve segmentation of physicians to avoid wasting resources in non-productive pursuits

Here’s what I like about the proposal:

• It represents a serious attempt to find a way to spur drug development for common conditions like high blood pressures, where there are already good, cheap treatments on the market
• It points out the inefficiency of using so many in-person salespeople

Here’s why I’m skeptical:

• The proposal is nakedly self-serving. A venture capitalist is likely to exit its investment around the time of drug approval. If approval comes faster and cheaper it aligns exactly with the venture investor’s goal and also makes it more likely that the developer can gain approval without having to give away a lot of upside to a bigger development partner
• The  authors assert that their approach will lead to the development of drugs with more differentiation. But differentiation from existing products is notoriously difficult to predict at the outset of clinical development and they don’t explain why we’d see more differentiation if the development path were shorter. The authors are confident that as their predicted flood of more differentiated drugs reaches primary care markets that selling will be cheaper since primary care docs will essentially pull these products in. I doubt there will be so much differentiation, and if a lot of new drugs arrive at once the sheer number of new products will make it harder and more expensive to break through the clutter
• The societal trend is to place more importance on drug safety, not less. The authors make a reasonable argument that “most clinical development programs go far past the point of diminishing returns for frequent safety events, but they do not go far enough to permit detection of rare events.” They might convince statisticians with this line of reasoning, but the public will be much more wary. If we’re really talking about cutting development time in half and knocking out 90% of costs there will be a difference in the safety assurances of at least some new products as they hit the market. How will you feel as a patient about being one of the first to receive a conditionally approved blood pressure medicine when you already have something that works reasonably well? I doubt I’d want my doctor to prescribe it to me
• Conditional approval with more post-marketing surveillance sounds reasonable, but in practice FDA has had trouble getting companies to comply with post-marketing promises. Why will it be different under this new scheme?

So thanks, Venrock for challenging the current drug development and marketing approaches. But we might be better served by a conversation on how to more appropriately balance benefits and risks. That might lead us to a paradigm of requiring even more safety data before approval of drugs for common conditions that already have good treatments, and less for diseases with more serious unmet needs. One could even go so far as to advocate that any drug with a very strong safety profile be allowed on the market even if its efficacy is weak or unproven. That would make it possible that the few people who could benefit from the drug would have the chance to try it.

photo credit: Express Monorail via photopin cc