Second Allogeneic Hematopoietic Stem Cell Transplantation (allo-HSCT) for Relapse of Hematological Malignancies after First Allo-HSCT

[Background] Patients with hematological malignancies who relapse after their first allogeneic hematopoietic stem cell transplantation (allo-HSCT1) have a dismal prognosis. Second allo-HSCT (allo-HSCT2) with bone marrow or peripheral blood stem cell from matched related or unrelated donor has been already used as a potentially curative treatment. Furthermore, umbilical cord blood (UCB) or haplo-identical transplantation as allo-HSCT2 has been increasingly used, because rapid availability of UCB or haplo-identical transplantation may provide a great advantage for patients who need urgent allo-HSCT2 to treat aggressive relapsed disease after allo-HSCT1. However, there is limited information of alternative donor sources on the efficacy of allo-HSCT2.

[Methods] Data from the Japan Society for Hematopoietic Cell Transplantation database were retrospectively analyzed.

[Results] A total of 1150 patients with hematological malignancies had received allo-HSCT1 after January 2000 and subsequently received allo-HSCT2 for their relapsed or refractory disease until December 2011. The median interval between allo-HSCT1 and allo-HSCT2 was 368 days (range: 17-4028). Patients had acute myeloid leukemia (AML, n = 683), myelodysplastic syndrome (MDS, n = 62), acute lymphoblastic leukemia (ALL, n = 275), malignant lymphoma (ML, n = 61), chronic myeloid leukemia and myeloproliferative disease (CML/MPD, n = 32), and other diseases (multiple myeloma/adult T-cell leukemia/chronic lymphocytic leukemia, n = 37). The intensity of the conditioning regimen was classified as full-intensity conditioning in 171 patients (15%) and reduced-intensity conditioning in 979 patients (85%). The number of mismatches was counted among HLA-A, HLA-B (low-resolution typing), and DRB1 (high-resolution typing). Donors at allo-HSCT2 were HLA-matched or 1 antigen HLA-mismatched related donor (MRD, n = 116), unrelated bone marrow donor (URD, n = 339), single-unit UCB (n = 515), and 2 antigens or more HLA-mismatched donor (mmRD, n = 180). Anit-thymoglobulin (ATG) was administered for GVHD prophylaxis in 125 patients (70%) of patients with mmRD. Among 1150 patients, 242 (21%) were alive with a median follow-up period of 628 days (range, 0-3627). The 2-year overall survival (OS) and progression-free survival (PFS) rate were 19.8% [95% confidence interval (CI), 17.5-22.5%] and 19.1% (95% CI, 16.7-21.7%), respectively. The cumulative incidence of relapse and transplant-related mortality (TRM) at 2-year were 40.4% (95% CI, 37.5-43.4%) and 40.4% (95% CI, 37.5-43.4%), respectively. Confounding factors that might affect OS were as follows: age, gender, disease, disease status at allo-HSCT2, intervals from allo-HSCT1 to allo-HSCT2, date at transplantation, conditioning regimen, donor source, and GVHD prophylaxis. Multivariate analysis revealed that higher OS correlated with complete remission/partial remission/chronic phase in disease status at allo-HSCT2 [hazard ratio (HR) 0.55, 95% CI 0.47-0.66; P < 0.001], longer interval (>12 months) between allo-HSCT1 and allo-HSCT2 (HR 0.68, 95% CI 0.47-0.78; P < 0.001), and CML/MPD (HR 0.56, 95% CI 0.36-0.87; P = 0.010). In contrast, mmRD without ATG was associated with inferior OS (HR 1.55 to MRD, 95% CI 1.17-2.06; P = 0.003). The OS at 2 years according to donor source was as follows: 18.0% in MRD, 28.2% in URD, 18.1% in UCB, 12.5% in mmRD with ATG, and 5.5% in mmRD without ATG.

[Conclusions] Allo-HSCT2 is a potential curative therapy especially in patients who have a sensitive disease and an interval time of >12 months to allo-HSCT2 even after allo-HSCT1, whereas the outcomes of allo-HSCT2 from alternative donor sources such as URD, UCB, and mmRD with ATG are comparable to that of MRD. Overall, the outcome of allo-HSCT2 has been still unsatisfactory because of the high rate of TRM and relapse. Novel strategy is required to further improve the outcome of allo-HSCT2, particularly for patients with refractory or relapsed disease despite allo-HSCT1.