The table shows the significant parameters of the univariate and multivariate Cox proportional hazard models assessing the impact of various HSCT features on IFS and OS. IFS included the use of any cellular product for falling chimerism or death. Additional parameters that were analyzed but nonsignificant in either analysis included year of HSCT, underlying cause (primary vs other), cumulative alemtuzumab dose (mg/kg preconditioning, periconditioning, and total exposure), use of tocilizumab during the conditioning regimen, addition of thiotepa, T-cell–depleted grafts, HLA match, TNC/kg, CD34 count per kg, and high-risk or moderate-risk TA-TMA. For the thiotepa analysis, only patients receiving alemtuzumab/fludarabine/melphalan ± thiotepa were included. The Cox proportional hazard model was used for these analyses. Significant results (P < .05) are bolded, and numeric P values are shown in the respective columns. NA indicates that the treatment failed the Cox proportional hazards assumptions with P value <.05 and should be excluded from the analysis, likely related to the small sample size.

NA, not a number (failed the proportional hazards assumption); TNC, total nucleated cells; Y/N, yes/no.