Relapse after allogeneic stem cell transplantation (Allo-SCT) remains a serious concern and its prognosis is poor. Although 2nd Allo-SCT is the only curative strategy for relapsed patients after Allo-SCT, it is not easy to proceed to 2nd Allo-SCT because of the next donor availability, tumor cell control, and infectious and organ complications. Cord blood (CB) is rapidly available in Japan, but its clinical efficacy is not clearly clarified yet.


We retrospectively analyzed the outcome of 85 consecutive 2nd CB recipients with relapsed myeloid malignancies following allogeneic transplantation at Toranomon hospital between January 2005 and March 2014.


Their median age was 52 years (range, 19-71). Donors at 1st Allo-SCT were related peripheral blood /bone marrow (BM) (n=24), unrelated BM (27) or CB (34), respectively. Underlying diseases were AML in 79, CML in 4, MDS in 1 and MPD in 1. The median duration of remission after 1st Allo-SCT was 197 (29-2586) days, and the median time from relapse to 2nd CBT and from 1st Allo-SCT to 2nd CBT were 85 (13-1667) days and 370 (56-2680) days, respectively. Twenty-three (27%) received re-induction chemotherapy after relapse after 1st Allo-SCT, whereas 42 (71%) received less intensive chemotherapy (low dose Ara-C and/or hydroxyurea (n=37), gemtuzumab ozogamicin (9), 5-azacitidine (2), and DLI (8)), and 20 (24%) did not receive any treatment. All except 5 patients (94%) were not in remission and 15 (18%) had active infections at the start of conditioning regimen of 2nd CBT. Performance status (PS) at 2nd CBT were 0 in 8 patients, 1 in 30, 2 in 29 and 3 in 7. Thirty-eight patients (45%) were conditioned with myeloablative regimens, whereas 47 patients received reduced-intensity. Calcineurin inhibitor (CI) plus mycophenolate mofetil were used in 36 cases (42%) as GVHD prophylaxis, while CI alone in 49 (58%). All patients received single CB unit with 2.84 (1.85-5.87) x 107/kg median number of total nucleated cell. Sixty-three patients achieved neutrophil recovery on median of 18 (11-39) days post-transplant with a cumulative incidence of 74.1%, and, among 22 who failed to achieve neutrophil recovery, 15 died before engraftment, 4 had early disease progression and 3 were rejected. Fifty-five patient (68%) developed infectious complications due to bacteria (n=42), virus (3), fungus (2) and unknown pathogens (8) within 30 days after 2nd CBT at median of 5 days post-transplant (range, -7 - 21). Median observation period of survivors was 734 (101-3023) days post-transplant. Cumulative incidences of NRM at 100 days and 2 years were 43.5% and 53.6%, respectively. Causes of NRM were Infections (n=19), idiopathic pneumonia syndrome (7), MOF+infections (4), VOD (2), GVHD (2), graft failure (1), and others (11). Higher age, myeloablative-conditioning at 1st Allo-SCT and active infection at 2nd CBT were negatively affected NRM in multivariate analysis. Twenty-seven patients relapsed at a median of 209 (16-2597) days. Cumulative incidences of relapse at 100 days and 2 years were 11.2% and 32.4%, respectively. Overall survival (OS) and progression free survival (PFS) at 2 years were 14.4% and 10.5%, respectively (Figure). Reduced-intensity conditioning at 1st Allo-SCT, younger age (<50), CI alone for GVHD prophylaxis, better PS (0-1) and absence of active infection at 2nd CBT showed a superior survival rate in multivariate analysis. Based on the results of these risk factor analysis, a group of patients who were in younger age (<50) with better PS (<2) (n=29) showed better OS (33.4%) and PFS (23.8%) at 2 years post-transplant (Figure).


Although CBT has provided more opportunities of 2nd transplant for relapsed patients, high NRM was noted in this study, resulted in unsatisfactory survival. Multiple factors associated with poor outcome were identified, mostly related to patients' poor background conditions that are not easy to be ameliorated. There were, however, certain population of patients (age<50 and PS 0-1) who showed higher overall survival.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.