Validating the allogeneic stem cell transplantation disease risk index: sample size, follow-up, and local data are important.

ABSTRACT

BACKGROUND: Reporting allogeneic hematopoietic stem cell transplantation (alloHSCT) outcomes by disease and disease stage can limit statistical power. Recently, the disease risk index (DRI) was developed and validated to stratify clusters of patients with various combinations of disease and stage for overall survival (OS), progression-free survival (PFS), and cumulative incidence of relapse (CIR). However, the DRI has not been tested for smaller cohorts or cohorts with shorter follow-up. METHODS: Data from recipients of a first alloHSCT between 2000 and 2011 (n=466; median follow-up, 55.2 months) were extracted from our database. Each patient was assigned to one of four risk categories according to the DRI. RESULTS: The DRI was validated as being significantly predictive of OS, PFS, and CIR (P<0.001 for all). The OS, PFS, and CIR from a contemporaneous cohort (n=324) from our institution were superior to the original training cohort. Using randomized patient subsets, the DRI stratified a cohort of 100 patients (OS, P=0.010; PFS, P=0.016; CIR, P=0.027), but failed to stratify a cohort of 50 patients (OS, P=0.385; PFS, P=0.167; CIR, P=0.026). When simulating shorter follow-up, the DRI stratified a cohort (n=322) with a median follow-up of 40.6 months for OS and PFS, but failed to stratify a cohort (n=242) with a median follow-up of 33.1 months. CONCLUSION: The DRI is a simple, robust pre-alloHSCT risk stratification tool. However, users should calibrate with local data before using the DRI to estimate absolute OS, PFS, and CIR and understand its limitations when applied to smaller cohorts or cohorts with shorter follow-up.