Allogeneic hematopoietic stem-cell transplantation (HCT) remains a key component of consolidation therapy for many adults with acute lymphoblastic leukemia (ALL). The role of HCT is especially critical in settings where pediatric-inspired protocols may be excessively toxic or difficult to implement, and novel agents for B-cell ALL are not readily available. In recent years, the use of haploidentical donors and reduced-intensity conditioning (RIC) regimens expanded treatment options for selected patients. However, there are currently no comprehensive data describing HCT outcomes for ALL in Latin America (LA).

In this study, we aimed to assess survival and response outcomes in a large, multicenter cohort, with an additional focus on less-studied subgroups of the disease.

This is a retrospective registry-based study that included HCT data from 18 centers in LA. We included patients aged ≥ 15 years with ALL or ambiguous lineage leukemia who underwent a first HCT between January 2007 and December 2024. Primary endpoints were overall survival (OS), disease-free survival (DFS), cumulative incidence of relapse (CIR), and non-relapse mortality (NRM). Multivariate Cox regression analyses (MVA) were performed. Country was incorporated as a frailty term in the models.

A total of 748 patients were included (median age 29; 84% B-cell ALL). Transplants occurred in Brazil (49.9%), Mexico (22.7%), Argentina (15.4%), and Chile (12%). Ph+ ALL was present in 28.5%, CNS disease in 16.9%. HCT was performed in CR1 (57.2%), CR2+ (37.4%), or refractory disease (5.4%); pre-HCT MRD was positive in 22.8%. Donors were haploidentical (40.6%), MSD (39.3%), MUD (11.6%), MMUD (5.5%), or cord blood (2.7%); 78.1% used peripheral blood. Myeloablative conditioning (MAC) was predominant (84.3%), with TBI in 69.6%. At 47.7 months’ median follow-up, 4-year OS and DFS were 46.8% and 42.5% (45.4% excluding refractory). CIR and NRM were 29.9% and 28.5%. Donor type did not affect OS, but MUD lowered relapse risk (HR 0.551, p = 0.027). OS predictors included age (HR 1.014, p = 0.005), CR2+ (HR 1.398, p = 0.01), TBI (HR 0.749, p = 0.02), donor age (HR 1.01, p = 0.035), and HCT from 2019 (HR 0.759, p = 0.038). TBI improved DFS (p = 0.0003) and reduced relapse (p = 0.003) without raising NRM, persisting in MAC (p = 0.044) but not RIC (p = 0.25). Donor age predicted DFS (p = 0.036) and NRM (p = 0.03). HCT from 2019 protected against relapse (HR 0.546, p = 0.0005). MRD positivity worsened all outcomes (p < 0.05). In Ph+ ALL (n = 196), donor age impacted OS (HR 1.02, p = 0.004) via higher NRM; TBI and age were not significant. In patients ≥ 50 (n = 116), RIC reduced NRM (HR 0.49, p = 0.032) without affecting relapse; TBI lowered relapse (HR 0.413, p = 0.046) without increasing NRM. Country of transplant was not significant in any model.

These findings demonstrate that allogeneic HCT for ALL in LA yields survival outcomes comparable to international benchmarks despite marked heterogeneity in practice. TBI-based conditioning and more recent transplant periods were associated with improved disease control without increasing NRM, while donor age remains an important predictor of adverse outcomes. RIC regimens reduced toxicity in older patients without apparently compromising relapse risk. These results highlight the importance of optimizing conditioning strategies, careful donor selection, and regional collaboration to improve HCT outcomes in LA.