Scott Gottlieb, M.D. has stepped into his new role as Commissioner of the FDA, and he’s brought with him a forward-thinking attitude that will promote more flexible approval pathways, a cornerstone of innovation and bringing novel devices to market. Key to this effort is adaptive trial design, a broad term that encompasses various strategies to optimize clinical study design while a study is ongoing – this can include modifying enrollment criteria or enriching cohorts for certain attributes, incorporating interim analyses to evaluate hypotheses mid-stream and early stopping rules for efficacy or futility. However, the various approaches to adaptive design are linked by one goal: generate stronger evidence for novel devices and drugs faster and less expensively than was previously possible.
The historical gold standard for a clinical trial specifies that trials be randomized and blinded, proceeding along a set path: trial opens, subjects are enrolled and followed as defined in the protocol and the data are reviewed only at the end, when the study has completed. While this produces strong data, it creates costly trials. Time and money can be exhausted on trials that are successful as well as the ones that fail, and an adaptive design introduces, in its simplest form, more decision-making opportunities: incorporating an interim analysis when the study is partly enrolled may demonstrate how and in whom a device is working. For example, an interim analysis demonstrating exceptional outcomes in a particular patient population may create an opportunity to elevate product claims from non-inferiority to superiority – compelling evidence for regulators, payers and the physicians and patients who will be end users. That early insight can lead to opening another study arm for patients meeting that profile to confirm or enhance the statistical rigor of the claim. Similarly, if an interim analysis finds that a device is not performing as expected, the study can immediately be stopped or redesigned.
Adaptive trial design isn’t new in medical devices. The FDA has been permitting adaptive elements in device trials for at least a decade, and formal guidance for devicemakers on adaptive trial design was released in draft in 2015 and finalized in 2016. However, of the 225 PMA submissions received by the FDA between 2007 to 2013, only about 10% incorporated an adaptive design. The Science article speculates that adaptive design is perceived too complex or intimidating, particularly from a statistical perspective.
Both philosophically and operationally, adaptive trials are a departure from the ironclad randomized controlled trial. But it’s a concept that is being advanced with support – and even enthusiasm – from the FDA, and it has the potential to empower devicemakers to develop more compelling data on their devices and shorten time to market. The question now is how the medical technology industry will rise to the challenge – and opportunity.