Adjusting Disease Risk Index for Regional Variations and Donor Selection: TRUMP cohort study

The Disease Risk Index (DRI) is a widely used prognostic tool following hematopoietic stem cell transplantation (HSCT), based on disease classification and status. However, the refined DRI (rDRI) does not fully account for regional variations in disease distribution and treatment practices. We analyzed 7,720 patients with hematologic malignancies who underwent their first allogeneic HSCT between 2014 and 2016 and developed a region-adjusted DRI (raDRI) using the Japanese registry database. We then validated predictive performance in 5,108 patients transplanted between 2017 and 2018. The raDRI demonstrated superior accuracy compared with the rDRI (time-dependent AUC: 0.705 vs. 0.670; integrated calibration index: 0.016 vs. 0.017). In patients with low to intermediate raDRI, matched related donors were associated with improved overall survival compared with matched unrelated bone marrow donors (bone marrow: hazard ratio (HR) 0.50, 95% CI 0.26 - 0.96; peripheral blood HR 0.54, 95% CI 0.36-0.83), whereas no differences were seen in patients with high to very high raDRI. Our findings suggest that disease risk should be carefully interpreted considering disease distribution and donor source.