skip to main content

Significant Interaction Test — SCE Medical Oncology MCQ

Instant feedback + full explanation. One question, done properly.

HardClinical Trials & StatisticsSignificant Interaction TestSCE Medical Oncology

A researcher is analysing a trial where the overall population result is negative (HR 0.92, p=0.25) but a pre-specified biomarker-positive subgroup shows a positive result (HR 0.55, p=0.001). The p-value for interaction between biomarker-positive and -negative subgroups is 0.003. Can the biomarker-positive result be considered reliable?

Educational content. Not a substitute for clinical judgement or local policy.

Reveal the answer and explanation

Correct answer: B — The significant interaction p-value (0.003) supports a genuine treatment-biomarker interaction — the biomarker-positive subgroup result is credible if: (1) pre-specified, (2) biologically plausible, (3) interaction test is significant, (4) effect size is clinically meaningful — this can support regulatory approval in the biomarker-selected population

Unlike the common scenario of non-significant p-interaction (where subgroup differences are likely chance), a significant p-interaction (e.g. 0.003) provides statistical evidence of a genuine treatment-effect modifier. When combined with: pre-specification in the SAP, biological plausibility (known biomarker-drug mechanism), clinically meaningful effect size (HR 0.55), and adequate subgroup sample size, the biomarker-positive result can support regulatory decisions. Examples: KEYNOTE-042 (PD-L1 ≥50% vs <50%), DESTINY-Breast04 (HER2-low vs HER2-zero).

Reference: JRCPTB Curriculum; Sun et al JAMA 2012; FDA Enrichment Guidance