Case‐Only Analysis of Treatment–Covariate Interactions in Clinical Trials

Abstract

Differential effectiveness of treatments across subgroups defined by pretreatment variables are of increasing interest, particularly in the expanding research field of pharmacogenomics. When the pretreatment variable is difficult to obtain or expensive to measure, but can be assessed at the end of the study using stored samples, nested case-control and case-cohort methods can be used to reduce costs in large efficacy trials with rare outcomes. Case-only methods are even more efficient, and reliable under a range of circumstances.

Genome-wide approaches to identify pharmacogenetic contributions to adverse drug reactions

“…One useful strategy could be the case-only analysis, developed to identify gene-environment interactions. 35 Instead of comparing adverse event cases versus controls in a cohort of treated patients-a comparison in which drug-and nondrug-related genetic factors would be confounded-adverse event cases among drugtreated patients would be compared with adverse event cases among untreated patients. The approach would identify genes that exert a greater-than-multiplicative effect on risk of the adverse event in the presence of drug.…”

A Targeted Maximum Likelihood Estimator for Two-Stage Designs

“…They have the potential to reduce the costs associated with collecting data on the full cohort with minimal losses in efficiency (Ernster, 1994;Rothman and Greenland, 1998;Hak et al, 2004;Vittinghoff and Bauer, 2006). We introduced the IPCW-TMLE for estimation of causal effects in two-stage sampling designs, with a focus on nested case-control sampling designs.…”

“…This is of particular importance when the candidate patient characteristic effect modifier of the treatment effect is difficult or expensive to measure (Vittinghoff and Bauer, 2006).…”

“…Advantages can include reduction in costs associated with collecting data on the entire cohort and minimal losses in efficiency (Ernster, 1994;Rothman and Greenland, 1998;Essebag et al, 2003;Hak et al, 2004;Vittinghoff and Bauer, 2006). Much of the literature focuses on logistic regression for effect estimation (Breslow and Cain, 1988;Flanders and Greenland, 1991;Ernster, 1994;Barlow et al, 1999;Szklo and Nieto, 1999).…”

“…Nested case-control studies within clinical trials and observational studies are increasingly popular when researchers are interested in effect modification (Rothman and Greenland, 1998;Essebag et al, 2003Essebag et al, , 2005Prentice and Qi, 2006;Vittinghoff and Bauer, 2006;Polley and van der Laan, 2009). This is of particular importance when the candidate patient characteristic effect modifier of the treatment effect is difficult or expensive to measure (Vittinghoff and Bauer, 2006).…”