Pharma’s Cutting Edge

I was on the Evista Product Team when we spiritedly debated, designed and launched the ambitious RUTH trial, so I maintained a strong interest in its outcome. For the record, nothing said here will violate my oath of confidentiality with my former employer. RUTH was a large Phase 3b trial designed initially to test the hypothesis that Evista (raloxifene, a SERM) would reduce the risks of cardiovascular complications, including cardiovascular death, in high-risk postmenopausal women. Initially, breast cancer risk reduction was a key secondary endpoint and became a co-primary endpoint after the trial began enrolling. Apparently, Evista did not meet the CV endpoint but perhaps met the breast cancer endpoint (it’s not clear from the press release).

I think there are a couple of good lessons for sponsors that relate to this news. First, medical opinion leaders and increasingly ordinary practitioners (and managed care, er, managers) want hard outcomes to support biomarker/surrogate marker data in all diseases, but especially in chronic diseases when medicines must be used continuously to prevent disease and when disease manifestations accrue over years as opposed to weeks or months. Knowing this, strategists can factor the need for large outcomes trials into long-range development planning. Hard outcomes are no longer a nice-to-have to gain indications for usage in major markets. On the other hand, when the surrogate marker is perceived as being strongly predictive of outcomes (as is too often sanguinely inferred from observational studies…a topic for another posting), an argument can be made for “slow-playing” the indication. The decision requires balancing many factors, key of which are the regulatory environment, the reimbursement/coverage/pricing environment, and the perceived ethics of the situation. There’s no hard and fast rule that can be applied in every case.

The other lesson is broadly applicable but doesn’t make life any easier for sponsors. It’s the need to consider very carefully the balance between the desire to meet a primary endpoint and to do so in a reasonable period of time without breaking the bank. Now, you might think, “well…duh.” Fair enough, but I’ll assert that in nearly every case, sponsors err on the side of saving time and money. Now…I am not going to make a strong argument here that RUTH would have had a different outcome as regards CV risk reduction if it had erred on the side of focusing on the optimal design for meeting its CV primary endpoint. Do I believe it would have been different? In my heart I do, but the evidence to support my intuition isn’t strong today. It was even weaker back when the study was being designed. The point being that very expensive decisions must be made on very little data, and, as a result, senior pharma management is forced to rely on asymmetric information that ALWAYS FAVORS the decision to save time and money. Stated differently, when senior managers accept the need to do something, like an outcomes trial, and they perceive the risk of failure as high, they will weigh the costs and benefits of the relatively unknown (the study outcome) and the known (the costs of running the trial) quite differetly. The known risks (i.e. costs) will be weighed more heavily, resulting in the tendency to emphasize expediency in the study design. As a result, the study design might not be optimal to support the scientific hypothesis. It’s exactly the opposite of what industry is usually accused of: unfair tests that favor the study drug. How to better balance known costs with unknown (but perceived great) risks? That will be fodder for a future discussion.