Prediction of failure-free survival in the context of HDT and ASCT. (A) Forest plot depicting HRs and 95% CIs of interaction terms of monogenic Cox regressions for failure-free survival. Rows represent results of individual Cox regressions considering treatment (as-treated), mutations status of the respective gene, and the interaction term. HRs and 95% CIs of the interaction terms are visualized. Only genes mutated in ≥10% of cases (n = 11) were included in the analysis. (B) Development of a multivariable model to predict failure-free survival benefit from HDT/ASCT (PReDiCt-FL). (C) Comparison of mutation frequencies between PReDiCt-FL low (yellow) and PReDiCt-FL high (blue) cohorts. Asterisks indicate significance levels assessed by Fisher’s exact test with Holm correction. (D) FFS of cases predicted to have a high FFS benefit (PReDiCt-FL high, left) and predicted to have no benefit from HDT/ASCT (PReDiCt-FL low, right), respectively. *P < .05; **P < .01; ***P < .001.