FDA has two competing priorities. On the one hand, they want to collect as much data as possible to insure drugs are safe and effective; on the other hand, patients want access to innovative medicines as quickly as possible. To try to balance these priorities, FDA often will approve treatments based on evidence from surrogate outcomes. In oncology, for instance, many treatments are approved based on surrogate endpoints (e.g., progression free survival, response rates) rather than the key endpoint of interest (i.e., overall survival). FDA often recommends or mandates that drug manufacturers to conduct confirmatory trials using the endpoint of interest even if the drug is conditionally approved based on the surrogate endpoints.
A key question then, is whether drug manufacturers are able to extract higher prices from payers once they have confirmed efficacy using the preferred endpoint rather than the surrogate endpoint. Based on the results from a paper by Frank, Shahzad, and Emanuel (2022), the answer is ‘no’.
Theory suggests that because confirmatory trials reduce uncertainty, they should provide an economic reward in the form of higher prices for a positive finding. We used a sample of physician-administered cancer drugs and data on average sales price to test this hypothesis. We found no significant relationship between confirmatory trial completion with a positive outcome and elevated prices. This represents a failure of the market to reward reduced uncertainty about a cancer drug’s true benefits. This inefficiency would be mitigated if major payers such as Medicare built price schedules that directly rewarded completion of confirmatory trials. More completed trials would ensure that patients are receiving truly effective chemotherapies and not suffering the adverse effects of drugs that are ultimately not effective.
Do read the full article here. Also, my own analysis of the relationship between surrogate endpoints on overall survival in cancer can be found at Lakdawalla et al. (2017) and Shafrin et al. (2016).