= 0.86) were more accurately predicted PFS than blood. In sensitivity analyses stratified by treatment approach we found a correlation of treatment-effect in fixed-duration therapy approaching the pre-specified validity threshold (R = -0.80 -0.93, -0.53; R² = 0.43, P < 0.01). Our data indicate the accuracy of MRD-testing results to surrogate PFS is context-dependent, influenced by treatment approach, therapy type, and the sampling source, and should be interpreted cautiously when informing clinical or regulatory decisions.
Wang et al. (Thu,) studied this question.