New insights into adaptive enrichment designs

Abstract

The transition towards personalized medicine is happening and the new experimental framework is raising several challenges, from a clinical, ethical, logistical, regulatory, and statistical perspective. To face these challenges, innovative study designs with increasing complexity have been proposed. In particular, adaptive enrichment designs are becoming more attractive for their flexibility. However, these procedures rely on an increasing number of parameters that are unknown at the planning stage of the clinical trial, so the study design requires particular care. This review is dedicated to adaptive enrichment studies with a focus on design aspects. While many papers deal with methods for the analysis, the sample size determination and the optimal allocation problem have been overlooked. We discuss the multiple aspects involved in adaptive enrichment designs that contribute to their advantages and disadvantages. The decision-making process of whether or not it is worth enriching should be driven by clinical and ethical considerations as well as scientific and statistical concerns.

Optimal design for inference on the threshold of a biomarker

“…Within our setting, the impact of the enrichment strategy can be assessed by the sample proportion of patients with biomarker values above the threshold; values close to 0 or 1 lead to the question of whether or not is worth enriching, depending on the gravity/rarity of the disease and/or cost of evaluating the biomarker, and/or side effects of the treatment. 6,16,17 Note that the optimal design for the heteroscedastic model requires prior knowledge of the ratio of the outcome variances in both treatments. However, variances are rarely known in designing any clinical trial and, in practice, we make a best guess based on the literature or the investigator's knowledge, and this is what we use in sample size computations.…”

“…Within our setting, the impact of the enrichment strategy can be assessed by the sample proportion of patients with biomarker values above the threshold; values close to 0 or 1 lead to the question of whether or not is worth enriching, depending on the gravity/rarity of the disease and/or cost of evaluating the biomarker, and/or side effects of the treatment. 6,16,17…”

“…This is an important aspect of enrichment with potentially disastrous consequences if this is not done accurately 17 and this work provides the optimal design for an inference on the threshold of a quantitative biomarker. Almost all the recent papers on enrichment have been dedicated to analysis, while the optimal design problem has received very little attention 6 ; indeed, most of the existing proposals deal with dichotomous markers 18 and balance the allocation among the treatment groups.…”

“…Its utility arises from the subgroup analysis commonly carried out at the end of each clinical trial and extends to adaptive enrichment designs. 7,6 Enrichment can be traced to Simon and Maitournam's 8 and Wang et al, 9 but the idea likely originated in oncology and several earlier papers. 10,11 An initial phase of a clinical trial is conducted on a full population and, at the conclusion of the first phase, the biomarker threshold is determined and consequently the benefitting patient subgroup.…”

“…Despite the well-known potential benefits of an enrichment design, the complex experimental framework poses new challenges and several complications need to be taken into account, even leading to questioning whether and when it is worth enriching. 6 One of the major concerns of these procedures is how "to make optimal decisions for updating the enrollment criteria," 16 which is even more complicated when the estimation of the threshold of a continuous biomarker is involved. This is an important aspect of enrichment with potentially disastrous consequences if this is not done accurately 17 and this work provides the optimal design for an inference on the threshold of a quantitative biomarker.…”