Further support for the use of MRD as a surrogate comes from analysis of the Myeloma IX study, 3 in which it was demonstrated that although the proportion of patients who become MRD-negative can vary significantly between regimens, the impact of MRD negativity on survival and PFS is similar regardless of the induction therapy received; that is, patients who become MRD negative have identical subsequent survival and PFS despite being treated with different regimens, as do those who remain MRD positive. 3 This suggests that it is the MRD level itself, which is the prime determinant of the subsequent PFS and survival outcome.The example quoted where MRD status and survival were "not so plainly concordant as in the article by Munshi et al" was from the Myeloma IX trial in the nontransplant group. However only 36 of 245 patients with MRD evaluated in this study became MRD negative, so the study was underpowered to show a clear difference; the P value was in fact .10.We plan to do further analyses exploring surrogacy, using this statistical definition, from available myeloma data sets, such as Myeloma IX, Myeloma XI, and others. To establish surrogacy using these methods ideally requires a relatively wide variety and number of trials, and so this information will require considerable additional effort. The current results, at the individual patient level, can be considered an important step, and we agree that new clinical trial designs, incorporating MRD to further explore its clinical utility, are both important and necessary.