Abstract
OBJECTIVE: Both allogeneic hematopoietic stem cell transplantation (allo-HSCT) and autologous HSCT (ASCT) are important therapies for extranodal natural killer/T-cell lymphoma (ENKTCL); however, no large-scale, multicenter study has compared the efficacy and safety between allo-HSCT and ASCT in these patients. Our multicenter, real-world study aimed to evaluate the outcomes of allo-HSCT vs. ASCT as consolidation in ENKTCL patients who had achieved a complete response (CR) or partial response (PR). METHODS: This was a multicenter, retrospective study with nine hospitals in China, and 114 patients with ENKTCL were enrolled. Sixty patients received ASCT and 54 received allo-HSCT. The primary outcome was progression-free survival (PFS). In the sensitivity analysis, propensity score matching (PSM) analyses were conducted to adjust for baseline prognostic factors. Landmark analysis were conducted to minimize immortal-time bias. RESULTS: Patients in the allo-HSCT group presented with more adverse prognostic factors. Allo-HSCT group showed a significantly better PFS and a lower disease progression rate compared with ASCT group in patients with Ann Arbor stage III/IV disease (PFS: 100% vs. 82.0%, P=0.023; disease progression rate: 0 vs. 25.4%, P=0.024), those with intermediate/high prognostic index of natural killer lymphoma (PINK) scores (PFS: 100% vs. 84.4%, P=0.034; disease progression rate: 0 vs. 22.1%, P=0.034), those with intermediate/high international prognostic index (IPI) scores (PFS: 100% vs. 82.0%, P=0.038; disease progression rate: 0 vs. 25.4%, P=0.038), or those receiving HSCT at PR (PFS: 100% vs. 50%, P=0.046; disease progression rate: 0 vs. 50%, P=0.046) at the 1.5-4.0 follow-up. In multivariate analysis, receiving ASCT was significantly associated with a poorer PFS [hazard ratio (HR)=2.23, P=0.038] and overall survival (OS) (HR=2.45, P=0.045). In the sensitivity analysis, patients receiving allo-HSCT showed a significantly better PFS (70.3% vs. 39.1%, P=0.039), OS (73.9% vs. 42.0%, P=0.044), and a lower disease progression rate (22.6% vs. 57.0%, P=0.017) compared with those receiving ASCT after propensity score matching. CONCLUSIONS: ENKTCL patients with high-risk characteristics could benefit more from allo-HSCT as consolidation.