Relationship of Race/Ethnicity and Survival After Single Umbilical Cord Blood Transplantation

Abstract 224

Umbilical cord blood transplantation (UCBT) is an alternative stem cell source for patients with leukemia in need of an allogeneic transplant without a suitable marrow or peripheral blood stem cell donor. Extensive resources have been mobilized to increase the number of UCB donations from patients of diverse ethnic and racial backgrounds, in an effort to provide an adequate stem cell source for patients of ethnic minorities who may have a difficult time finding a compatible unrelated marrow or peripheral blood stem cell donor. We analyzed the association of race/ethnicity and outcomes of 907 patients receiving unrelated single unit UCBT from 1995 to 2006 for AML, ALL, CML, and MDS and who were reported to the CIBMTR. Race/ethnicity was reported by the transplant centers and included 612 Whites, 145 Blacks and 128 Hispanics. Patients received a variety of conditioning and graft versus host disease (GVHD) prophylaxis regimens. Most patients (72%) received ablative conditioning. Median ages at transplant were 8 years (<1-78), 8 years (<1-57), and 6 years (<1-56) for Whites, Blacks and Hispanics, respectively. Overall, 63% White, 30% Black and 33% Hispanic patients received UCB units from donors of the same race. Black patients were more likely to receive UCB units with a lower cell dose and inferior HLA match. A total nucleated cell (NC) count infused of ≥ 2.5 × 10⁷/kg was given to 69% of White patients, 62% Blacks, and 77% Hispanics. A 6/6 matched unit was transplanted into 16% of White patients and 9% of Hispanic patients but only 4% of Black patients. Blacks received a ≤ 4/6 matched cord in 68% of cases, compared to 46% for Whites and 47% for Hispanics. Probability of overall survival (OS) at 2 years was 44% (95%CI 40–48%) for Whites, 34% (26-42%) for Blacks and 46% (37-55%) for Hispanics (P=0.117). Two-year leukemia-free survival (LFS) was 42% (38-46%), 34% (26-42%) and 43% (34-52%) (P=0.337) and cumulative incidence of 1-year treatment-related mortality (TRM) was 29% (26-33%), 38% (31-46%) and 27% (19-35%) for the three racial groups, respectively (P=0.107). Cumulative incidence of acute GVHD Grades II-IV and chronic GVHD, rates of engraftment, and relapse were similar across the racial/ethnic groups. Degree of HLA had no significant effect on OS. Causes of death were also similar among the three groups. In multivariate analysis, Blacks, as compared to Whites, had inferior OS (RR=1.3, P=0.02) while OS of Hispanics was similar to that of Whites (RR=1.03, P=0.81). The elevated risk of death among Blacks, compared with Whites, did not differ by cell dose (≥ 2.5 × 10⁷ NC/kg, RR for OS=1.26, P=0.13; < 2.5 × 10⁷ NC/kg, RR for OS=1.32, P=0.17). Young age at UCBT, early-stage disease, receipt of UCB with higher cell dose, and performance status ≥ 80% at UCBT were independent predictors of improved survival. Overall, the risks of LFS, TRM and relapse among Blacks and Hispanics were comparable to those among Whites. In conclusion, 1) Single unit unrelated UCBT for leukemia is associated with approximately 40% survival at 2 years; 2) As anticipated, higher cell dose, younger age, good performance status and early-stage disease were associated with improved survival across all groups; 3) Hispanic patients had similar survival to White patients; and 4) White patients have higher survival than Black patients, a difference that persists even after adjustment for cell dose and HLA, suggesting other factors may be important. Further investigation of other reasons for racial differences in the success of UCBT will benefit from increased numbers of patients for analysis.