Incidence, Risks, and Outcomes of Relapse Following Reduced-Intensity Allogeneic Hematopoietic Stem Cell Transplantation for Non-Hodgkin's Lymphoma.

Abstract 3451 Relapse is the major cause of treatment failure and death following reduced-intensity (RI) allogeneic hematopoietic stem cell transplantation (HSCT) for non-Hodgkin's lymphoma (NHL), yet there are no reports that focus specifically on the natural history of relapse in this patient population. We assessed relapse risks and outcomes on 120 consecutive patients (median age = 52 years; range, 28–71 years) with NHL (indolent = 43; aggressive = 77) who all received a T-cell replete allograft from HLA-matched siblings following a reduced-intensity conditioning regimen (fludarabine/cyclophosphamide). All patients were assessed for disease status at 1, 3, 6, 12, 18, 24 months post-transplant, and annually thereafter or as clinically indicated. Median event-free survival (EFS) from transplant date for all 120 patients was 11.3 months (range, 0.5–105.5+ months) with a 5-year EFS = 35%. Median EFS (months) was assessed for the following pre-transplant characteristics: chemo-resistant vs. chemo-sensitive = 3.3 vs. 39.1, p = 0.0001; pre-transplant disease status - CR vs. PR vs. SD vs. PD = not reached (NR) vs. NR vs. 11.2 vs. 1.7, p <0.0001; aggressive vs. indolent histology = 6.6 vs. 15.8, p = 0.13; number of disease sites: 0–1 vs. 2 vs. 3 vs. 4+ = 4.8 vs. 7.4 vs. 18.4 vs. 7.1, p = 0.85. Median overall survival (OS) from transplant date for all 120 patients was 71.1 months (range, 0.5–120+ months). We identified 55 patients (46%) who either relapsed or progressed post-transplant. Among those who have progressed/relapsed to date, median time to progression/relapse was 3 months (range, 0.5–46 months) with 72% of progressions/relapses occurring prior to 6 months post-transplant. The overall cumulative incidence (CI) probabilities for relapse/progression, adjusted for competing non-relapse mortality, at 3, 6, 12, 24 and 36 months were 0.169, 0.329, 0.394, 0.447, and 0.460, respectively. The association of 3-year relapse CI probabilities with pre-transplant characteristics were as follows: aggressive vs. indolent histology = 0.530 vs. 0.315; chemo-resistant vs. chemo-sensitive = 0.571 vs. 0.316; disease status prior to transplant – CR vs. PR vs. SD vs. PD = 0.210 vs. 0.217 vs. 0.597 and 0.627; number of prior treatments: 1–2 vs. 3 vs. 4 vs. 5+ = 0.375 vs. 0.418 vs. 0.613 vs. 0.450; sites of disease: 0–1 vs. 2 vs. 3 vs. 4+ = 0.458 vs. 0.516 vs. 0.331 vs. 0.579. Median OS from date of progression was 8.3 months (0.5 - 94+ months) with a 5-year OS = 32%. No patient who progressed/relapsed within 3 months post-transplant has survived beyond 12 months, with one patient still alive at 6 months. Median OS (months) from date of progression was assessed for the following characteristics: progression < 3 months vs. ≥ 3 months post-transplant = 2.6 vs. NR, p <0.0001; progression < 6 months vs. ≥ 6 months post-transplant = 5.3 vs. NR, p = 0.0002; aggressive vs. indolent histology = 7.8 vs. 18.3, p = 0.53; number of disease sites: 0–1 vs. 2 vs. 3 vs. 4+ = 5.8 vs. 13.8 vs. 4.9 vs. 12.9, p = 0.75; number of prior treatments: 1–2 vs. 3 vs. 4 vs. 5+ = 8.4 vs. NR vs. 11 vs. 5.6, p = 0.27; response vs. no response to relapse/progression treatment = NR vs. 4.7, p = 0.0087 (determined by a landmark method, beginning 30 days after progression to allow time for determination of response to relapse treatment). These data demonstrate that a significant minority of NHL patients, who relapse after RI allogeneic HSCT, can achieve long-term survival, including patients with aggressive histology. These analyses, which utilized standard clinical characteristics, identified NHL patients at higher risk for relapse and poorer outcomes once relapse occurred. In particular, disease progression/relapse occurring less than 3 months post-transplant was associated with an extremely poor prognosis; novel strategies and trials are needed for such patients. These results need to be confirmed by other groups, and these analyses need to be performed in other transplant settings (e.g. unrelated donors and myeloablative conditioning). The use of these patient characteristics, alone or in combination, may ultimately lead to a method of estimating the risk for relapse and the subsequent prognosis, if relapse should occur, in NHL patients undergoing RI allogeneic HSCT. Disclosures: No relevant conflicts of interest to declare.