Statistical and practical considerations for clinical evaluation of predictive biomarkers

Abstract

Predictive biomarkers to guide therapy for cancer patients are a cornerstone of precision medicine. Discussed herein are considerations regarding the design and interpretation of such predictive biomarker studies. These considerations are important for both planning and interpreting prospective studies and for using specimens collected from completed randomized clinical trials. Specific issues addressed are differentiation between qualitative and quantitative predictive effects, challenges due to sample size requirements for predictive biomarker assessment, and consideration of additional factors relevant to clinical utility assessment, such as toxicity and cost of new therapies as well as costs and potential morbidities associated with routine use of biomarker-based tests.

Figure 1.

Examples of qualitative interactions. Gefitinib vs carboplatin + paclitaxel for first-line treatment of non–small cell lung cancer patients with EGFR mutation–positive tumors (A) and EGFR mutation–negative tumors (B) [adapted from Figure 2 of Mok et al (11). Reprinted with permission. Copyright 2009 Massachusetts Medical Society.]. Cetuximab + chempotherapy vs chemotherapy for first-line treatment of non–small cell lung cancer patients with high-expressing EGFR immunohistochemistry (IHC)–positive tumors (C) and low-expressing EGFR IHC–positive tumors (D) [adapted from Figure 4 of Pirker et al. (13). Reprinted with permission. Copyright 2012 Elsevier]. PFS = progression-free survival.

Figure 2.

Examples of quantitative interaction: pazopanib vs placebo for locally advanced or metastatic renal cell carcinoma patients with high interleukin 6 (IL-6) values (A) and low IL-6 values (B) [adapted from Figure 2 of Tran et al. (14). Reprinted with permission. Copyright 2012 Elsevier]. Erlotinib maintenance therapy vs placebo for non–small cell lung cancer patients with EGFR mutation–positive tumors (C) and EGFR wild-type tumors (D) [adapted from Figure 3 of Brugger et al. (15). Reprinted with permission. Copyright 2011 American Society of Clinical Oncology]. Note that data were not available from Brugger et al. (15) to provide the number of patients at risk for (C) and (D). CI = confidence interval; HR = hazard ratio; PFS = progression-free survival.

Figure 3.

Combined therapy (temozolomide + radiotherapy) vs radiotherapy for glioblastoma patients with methylated MGMT (A) or unmethylated MGMT (B). The numbers of patients at risk are below the graphs. P values are based on two-sided log-rank tests [reprinted from Figure 4 of Stupp et al. (22). Reprinted with permission. Copyright 2009 Elsevier].