Risk Stratification of Outcomes Among Patients with Adult T-Cell Leukemia/Lymphoma Receiving Allogeneic Hematopoietic Cell Transplantation: A Retrospective Analysis of the JSHCT ATL Working Group

Introduction: Disease status at allogeneic hematopoietic cell transplantation (HCT) is the most important prognostic factor in HCT for adult T cell leukemia/lymphoma (ATL), but other factors also need to be considered, including relatively older recipient ages, performance status (PS), and comorbidities. Several scoring systems, such as the HCT-comorbidity index (CI) and the modified European Group for Blood & Marrow Transplantation (EBMT) risk score (mEBMT), are available for predicting outcomes in HCT recipients; however, their prognostic relevance in ATL is not well defined.

Materials and Methods: We retrospectively evaluated the HCT-CI and mEBMT in 635 ATL patients registered to the Japan Society for Hematopoietic Cell Transplantation (JSHCT) TRUMP database from 2008 until 2012. We further developed a new prognostic index (PI) for ATL patients receiving HCT (ATL-HCT-PI) using each factor from the HCT-CI and the mEBMT, as well as other known risk factors.

Results: In multivariate analyses, higher HCT-CI was associated with inferior overall survival (OS) in the following comparisons: HCT-CI 0 (n = 407) vs. 1–3 (n = 190) (HR 1.38; 95% CI 1.09–1.76; P < 0.01 and HCT-CI 0 vs. ≥4 (n = 38) (HR 2.82; 95% CI 1.95–2.50; P < 0.001. Further, multivariate analyses identified arrhythmia, diabetes mellitus, cardiovascular comorbidities, infection, hepatic comorbidities, and renal comorbidities as significant risk factors among the HCT-CI variables. In univariate analyses, higher mEBMT was associated with inferior OS in the comparison of mEBMT 0–3 with 4–6 (HR 1.59; 95% CI 1.27–2.00; P < 0.001). HCT from a female donor to a male recipient (FDMR) and disease status were significant risk factors among the mEBMT variables. We developed a new risk score (ATL-HCT-PI) based on known risk factors (PS, human leukocyte antigen mismatch, ABO mismatch), age (≤50, 51–62, ≥63 years), arrhythmia, diabetes mellitus, cardiovascular comorbidities, infection, hepatic comorbidities, renal comorbidities, FDMR, and disease status (CR vs. non-CR). ATL-HCT-PI scores were collapsed into three risk groups: ATL-HCT-PI 0–7 (n = 285), 8–14 (n = 237), and ≥15 (n = 92), which statistically predicted 3-year OS (48.3%, 29.2%, and 5.5%, respectively) (Figure). The better predictive performance of ATL-HCT-PI for OS was confirmed using c-statistics (c-statistics: ATL-HCT-PI = 0.705, HCT-CI = 0.592, and mEBMT = 0.601). Conclusions: The newly developed ATL-HCT-PI can be a useful tool for predicting mortality in HCT for patients with ATL.