9: Risk Factors Associated with Graft Failure After Umbilical Cord Blood Transplantation (UCBT): A Single Center Analysis in 539 Patients

Abstract

UCBT has proved to be a viable alternative for pediatric and adult patients who do not have an HLA matched related or unrelated donor. However, the incidence of graft failure (GF), an often fatal complication, has been reported to be as high as 30% in some series. Therefore, we attempted to identify potential risk factors associated with GF in 539 patients (adult, n = 297; pediatric, n = 242) transplanted with UCB at University of Minnesota between 1994–2006. Patients received one (n = 261) or two (n = 278) UCB units after a myeloablative (MA, n = 338) or reduced intensity (RIC, n = 201) conditioning regimen for malignant (n = 418) or non-malignant (n = 121) disease. GF was defined as evidence of <5% chimerism or failure to achieve neutrophil recovery by day 42. Median patient age and weight was 22 years (r: 0–69) and 61 kg (r: 4–149 kg), respectively, with 66 having had a prior autologous transplant. Time from diagnosis to transplant was 13.3 months (1.26–250 months), 57 patients were CMV+, 186 received a sex-matched graft, 132 ABO-matched, 168 minor ABO-MM, and 229 major ABO-MM. Median graft cell doses were 3.8 × 107 nucleated cells (NC)/kg (r: 0.7–48.9) and 4.6 × 105 CD34/kg (r:0.4–275.3). Fifty-seven received 6/6 HLA-matched units, 196 received at least one 5/6 matched unit and 286 received at least one 4/6 matched unit. GF was observed in 43/338 (12.7%) after MA and 36/201 (17.9%) after RI UCBT. Risk factors associated with GF in logistic regression were transplantation of HLA mismatched units (5/6 match, OR 9.13, 95%CI, 1.19–70.38; 4/6 match, OR 11.66, 95%CI 1.53–89.00; p < .01), treatment of non-malignant disease (OR 2.55, 95%CI, 1.35–4.81, p < .01), low CD34 cell dose (top low quartile OR 2.33, 95%CI, 1.07–5.05; bottom low quartile, OR 3.11, 95%CI, 1.44–6.70; p < .01), and use of a RIC (OR 2.40, 95%CI 1.35–4.26; p < .01). Patients who received a prior autologous transplant had lower odds of GF (OR 0.12, 95%CI, 0.03–0.52, p < .01). Factors not impacting risk of GF were recipient age, sex-match, ABO-match, CMV serostatus, performance status, diagnosis, disease risk, interval from diagnosis to transplant, NC dose, and CD3 cell dose. In addition to CD34 cell dose and HLA match, patients with non malignant disease and recipients of a RIC are at higher risk of GF. These results support the need for greater investment in banking UCB with larger cell doses and HLA diversity and development of improved strategies for overcoming GF in high risk patient populations.