Pharmaceutical Statistics has just published a special issue entitled “Addressing and Communicating Heterogeneous Treatment Effects for Patient Subpopulations: Challenges and Opportunities” (20:5). The special issue, guest edited by former Editor-in-Chief John Scott, combines papers originating at a symposium held on November 28, 2018 at the Food and Drug Administration’s White Oak campus to address some of challenges related to the heterogeneity of treatment effects. A portion of the introduction written by Ravi Varadhan and a list of papers included in the special issue is included below.
There is often a considerable variation in the responses of individuals to a given treatment. Heterogeneity of treatment effects (HTE) is defined as the explainable variation in how individuals respond to a treatment. Even if the treatment is efficacious on average, some subgroups may derive large benefit, some groups little benefit, and some may even be harmed by the treatment. Thus, it may be argued that the “evidential chasm” between the goal of randomized trials and the needs of patients is primarily due to HTE. HTE is a vital and yet underexplored concept of importance not only to patients and care providers, but also to other players in the healthcare arena including regulators, manufacturers, and payers.
While there is an obvious need for evidence that takes HTE into account, it is a major challenge to determine the sources of HTE and to reliably estimate heterogeneous treatment effects. This problem has vexed trial methodologists for several decades. Individual treatment effect, which is the sought after goal, is not directly identifiable from the available data, since we cannot observe the response of the same individual to different treatments. Hence, the most common way to address HTE is through subgroup analysis. These analyses have to contend with a number of challenges including small sample sizes, and high rates of spurious findings. These and various other challenges have been discussed extensively in the literature.
Another important challenge, one that is much less appreciated, but is especially critical for a regulatory agency such as the FDA, is the communication of the evidence on HTE to patients and care providers. For example, the FDASIA Act of 2012 mandated the FDA to report on the inclusion and analysis of demographic subgroups in applications for drugs, biologics, and devices. The FDA does this reporting through the drug trial snapshots (DTS). A major challenge is to make such communications both easily understandable and accurate.
A symposium was held on November 28, 2018 at the FDA’s White Oak campus to address some of these challenges. The symposium discussed three main topics: (1) Major sources of HTE including subpopulations (e.g., sex, race, pediatric and geriatric populations); (2) HTE considerations in study design and statistical analysis; and (3) How HTE may be communicated to stakeholders via various avenues such as drug trial snapshots and labeling.
Encouragement of subgroup assessment by the FDA by Robert Temple
Heterogeneity in treatment effects across diverse populations by Bridget M. Nugent, Rajanikanth Madabushi, Barbara Buch, Vasum Peiris, Victor Crentsil, Virginia M. Miller, Jonca Bull and Marjorie R. Jenkins
Assessing and communicating heterogeneity of treatment effects for patient subpopulations: Panel discussion on considerations in design and analysis by Mark Rothmann, William Crown, Thomas A. Louis, Thomas Permutt, Stephen Ruberg, Jodi Segal and John Scott
Assessing and communicating heterogeneity of treatment effects for patient subpopulations: Keynote and panel discussion on communicating heterogeneous treatment effects across populations by James Heyward, Milena Lolic, Catherine Spong, David Atkins, Gene Pennello, Ravi Varadhan (Open Access article)