[Correspondence] Quantifying Immunoscore performance – Authors’ reply
We thank Ryan Sun and colleagues for their interest in our Article.1 In their comments, the authors state that the use of hazard ratios (HRs) to estimate the Immunoscore performance in the prediction of the time to colon cancer recurrence could be invalid since the proportional hazards assumption (PHA) might not be met. The authors subsequently propose the restricted mean survival time (RMST) as an alternative statistical parameter. We appreciate the statisticians’ comments and hypothesis. In our study1, we tested the PHA and found no violation of it (p value of the Cox proportional hazard test) in any Immunoscore subgroup for time-to-recurrence, disease-free survival, and overall survival in all cohorts (training set [TS], internal validation set [IVS], external validation set [EVS], and all [appendix]).