Relapse remains the leading cause of treatment failure in patients with refractory acute myeloid leukemia (AML) undergoing allogeneic hematopoietic stem cell transplantation (allo-HSCT). Despite advances in conditioning and supportive care, outcomes remain poor, with median overall survival (OS) frequently limited to less than one year. Donor lymphocyte infusion (DLI) has been recognized as a potent immunotherapeutic strategy to augment graft-versus-leukemia (GVL) effects. Both prophylactic and preemptive approaches have demonstrated efficacy in reducing relapse risk and improving survival, as highlighted in large European, Asian and American studies. However, there is absence of data from Latin America, where real-world outcomes are underrepresented. We report a Brazilian single-center experience evaluating the impact of prophylactic/preemptive donor lymphocyte infusion (pDLI) on survival outcomes in patients with refractory AML undergoing allo-HSCT.

We retrospectively analyzed adult patients with refractory AML, defined by the presence of active disease as pre-transplant status before initiation of conditioning, who underwent allo-HSCT at a tertiary transplant center in Brazil between January 2018 and July 2025. Patients were stratified according to receipt of pDLI versus no DLI. Exclusion criteria included prior grade IV acute graft-versus-host disease (GVHD) and death before day +28 post-transplant to avoid immortal time bias. Baseline demographic, clinical, and transplant-related characteristics were compared between groups. Survival outcomes were evaluated using Kaplan–Meier estimates with log-rank tests, and hazard ratios (HR) with 95% confidence intervals (CIs) were calculated using Cox regression models.

Twenty-seven patients met inclusion criteria, with 19 assigned to the non-DLI group and 8 to the pDLI group. Baseline characteristics were comparable between cohorts with respect to age, sex, European LeukemiaNet 2022 risk classification, conditioning intensity, donor source, and incidence of acute GVHD. Patients in the pDLI group were younger (median age 41 vs. 58 years, p = 0.054), and may represent a potential confounder. Conditioning was predominantly reduced-intensity in both arms, and approximately half of the cohort underwent haploidentical transplantation.

The median follow-up was 14 months (12 months in the non-DLI group and not reached in the pDLI group), indicating sustained survival and ongoing follow-up for more than half of these patients. The median OS was 432 days, in the non-DLI group was 385 days (95% CI, 43–727), whereas the pDLI group was not achieved. The relative risk of death was significantly reduced in the pDLI group (HR 0.25; 95% CI, 0.056–1.11; p = 0.049), representing a 75% relative reduction in mortality. Despite the wide CI due to small sample size, the observed benefit is consistent with prior multicenter reports, particularly in high-risk AML subsets.

The relapse analysis also favored pDLI. The median time to relapse was 599 days in the non-DLI group compared with 1060 days in the pDLI group. Although this did not reach statistical significance (HR 0.41; 95% CI, 0.11–1.54; p = 0.17), the trend strongly suggested reduced relapse risk, in line with the hypothesis that early immunological pressure from DLI can suppress leukemic proliferation.

Patients receiving pDLI underwent a mean of three infusions, with a mean first-dose cell count of 2.5 × 10⁶ CD3⁺ T cells/kg. This strategy did not result in an increased incidence of severe grade III–IV GVHD (HR 1.16; 95% CI, 0.11–12.79; p = 0.904). Although HR suggests a 16% relative increase in risk, our findings remain consistent with prior reports showing 11–20% incidence of acute grade III–IV GVHD, and no grade IV was found in our cohort. These results support the feasibility of incorporating pDLI into post-transplant strategies in high-risk AML without incurring unacceptable toxicity.

This study provides a Latin American data that pDLI may significantly improve survival outcomes in pos-HSCT refractory AML, without a detrimental increase in GVHD risk. Despite limitations, the clinical signal observed aligns with international data and highlights the necessity of expanding collaborative efforts to validate these findings across Latin American centers. Larger multicenter studies are needed to confirm these findings and optimize patient selection, timing, and dosing in our population.

This content is only available as a PDF.
2025
Sign in via your Institution