Similar Survival in Patients with Acute Myeloid Leukemia Relapsing after Matched Related Donor and Umbilical Cord Blood Transplantation

Patients (pts) with acute myeloid leukemia (AML) who relapse after allogeneic hematopoietic cell transplantation (AlloHCT) have limited treatment options and their expected survival is poor. Prognostic factors for survival in pts relapsing after AlloHCT are unclear. We therefore analyzed response to salvage therapy and survival.  Three hundred and forty eight pts with AML underwent AlloHCT between 2000-2012 using umbilical cord blood (UCB, 222) or an HLA matched related donor (RD, 126). Relapse after AlloHCT occurred in 104 pts: 72 of 222 after UCB and 32 of 126 after RD transplantation at a median time of 152 days (range, 77-322). UCB and RD pts were comparable in their demographic and disease characteristics including: cytogenetic risk, presence of GVHD at relapse, proportion receiving immunosuppressive therapy (IST) at relapse, time to relapse, relapse site (systemic vs. isolated extramedullary), bone marrow (BM) or percent of peripheral blood blasts at relapse, BM donor chimerism, and concurrent infectious or GVHD-related complications. Compared to RD, more UCB recipients underwent reduced intensity HCT (68% vs. 41%, p=0.01), had active infection (25% vs. 6%, p=0.03) or acute treatment related complications (e.g. veno-oclusive disease, kidney injury, cardiac arrhythmia) (33% vs. 13%, p=0.04) at the time of relapse. After relapse, all but 1 patient in each group underwent IST taper with 3 recipients (2 UCB and 1 RD) achieving a complete remission (CR). In UCB and RD recipients, post-relapse active therapy consisted of systemic chemotherapy in 21 and 6 pts, second AlloHCT in 9 and 2 pts, supportive care in 30 and 6 pts and donor lymphocyte infusions (DLI) ± systemic chemotherapy in 0 and 12, respectively (with no additional information on 18 pts [UCB=12, RD=6] who were managed elsewhere at the time of relapse). Survival at 1 year after relapse was 22% for all pts (19% UCB vs. 28% RD, p=0.32).  Only 25% of pts with post-relapse therapy achieved subsequent CR (21% UCB vs. 35% RD, p=0.16). In multivariable analysis, donor type was not an independent predictor of post-relapse mortality (HR=0.9; 95% CI 0.5-1.6, p=0.72). Post-relapse mortality, however, was higher in pts with PB blasts above the median (median=3%; HR=3.9; 95% CI 2.2-6.9, p<0.01), active infection (HR=2.0; 95% CI 1.1-3.9, p=0.03), and history of other treatment related complications at relapse (HR=2.0; 95% CI 1.1-3.4, p=0.02). Pts receiving any type of active therapy (chemotherapy, DLI or second AlloHCT) at relapse (HR=0.3; 95% CI 0.2-0.5, p<0.01) had lower mortality. Pts with AML who relapsed after AlloHCT and were in a clinical condition sufficient to tolerate any therapeutic option can achieve CR and enjoy prolonged survival compared to those that receive supportive therapy. Despite that fact DLI is not generally available to UCB recipients, salvage therapy with systemic chemotherapy or second AlloHCT provides similar results.