Navigating Measurable Residual Disease as a Surrogate Endpoint in Chronic Lymphocytic Leukemia

ABSTRACT

Measurable residual disease (MRD) has established itself as an essential post‐treatment biomarker in chronic lymphocytic leukemia (CLL). While achieving undetectable MRD (uMRD) reliably predicts prolonged progression‐free survival (PFS) for individual patients across clinical trials, establishing MRD as a universal, trial‐level regulatory surrogate endpoint across contemporary targeted therapies remains a formidable biostatistical hurdle. This perspective addresses the “atomistic fallacy” underlying the apparent erosion of trial‐level surrogacy observed in aggregate meta‐analyses. We demonstrate that the attenuation of trial‐level surrogacy stems from fundamental biological effect modification—specifically the distinction between cytostatic therapies (e.g., continuous Bruton tyrosine kinase inhibitors) and cytotoxic, deep‐remission regimens (e.g., fixed‐duration venetoclax combinations)—as well as operational confounders introduced by modern adaptive trial designs. Reconciling these individual‐ and trial‐level observations requires transitioning from a binary endorsement model to a mechanism‐dependent, context‐specific framework for surrogate endpoint validation in CLL.