Abstract
INTRODUCITON: We reported the outcomes of double-unit cord blood transplantation (dCBT) after myeloablative conditioning performed in a prospective multicenter phase II study (C-SHOT0507) (Biol Blood Marrow Transplant 19, 2013: 812-819). However, the benefit of using double unit compared to single unit was not clearly shown. We performed matched control analysis to compare dCBT and single-unit CBT (sCBT), using C-SHOT0507 data of dCBT and registry data of sCBT.
METHODS: Between Apr.2006 and Jan.2010, 61 cases of dCBT in C-SHOT0507 phase II clinical study and 932 cases of sCBT were performed in Japan. Because all cases of dCBT perfomed by TBI12Gy containing myeloablative conditioning (CA/Cy/TBI +/- G-CSF for myeloid and Cy/TBI for lymphoid) and uniform GVHD prophylaxis (CSA/MTX), we excluded reduced intensity conditioning and myeloablative conditioning without TBI 12 Gy from the cases of sCBT. Finally 307 sCBT patients who had hematological malignancies (AML170, ALL80, MDS24, CML14 and ML20) were extracted. All the sCBT recipients received uniform preconditioning described above and same GVHD prophylaxis (CSA/MTX), selected as a matched control cohort of dCBT C-SHOT0507 study. All statistical analyses were performed using EZR (Kanda Y, Saitama Medical Center, Jichi Medical University), a graphical user interface for R (The R Foundation for Statistical Computing, version 2.13.0).
RESULTS: Median observation period of survivors was 1431 (106-2315) days post-transplant. Backgrounds of dCBT group and matched-control sCBT group were comparable except gender: (dCBT:M/F = 53 /8 versus sCBT:M/F = 154/153). Median age were 37 (10-54) in dCBT and 40 (14-54) in sCBT. Disease risk included standard 27, advanced 34 cases in dCBT, and standard 142, advanced 165 cases in sCBT. sCBT were grouped according to infused TNC number; 148 cases of low TNC (<2.5x107/kg) group (lowTNC) including 35 cases of <2.0x107/kg, and 159 cases of high TNC (>=2.5x107/kg) group (highTNC). Cumulative incidence (C.I.) of neutrophil engraftment were 87.2% (69.5-95.0) in dCBT, 84.7% (76.2-90.3) in sCBT of low TNC and 86.8% (79.3-91.7) in those of high TNC (n.s.) [Figure] and median days of engraftment were 25, 23, 22 days, respectively. 3yEFS were 44.3% (31.8-56.7), 41.6% (33.5-49.7), 44.0% (36.1-51.9) [Figure], and 3yOS were 52.3% (39.7-64.8), 48.7% (40.5-56.9), 51.0%(43.1-59.0), respectively. There were no statistically significant differences in both of them. C.I. of NRM were 24.1% (13.9-35.7), 31.0% (22.0-40.3), 40.0% (18.0-61.3), acute GVHDII-IV were 30.8% (18.2-44.4), 41.9% (33.1-50.4), 54.8% (45.5-63.1), acute GVHDIII-IV were 10.8% (3.2-23.6), 12.7% (6.8-20.6), 19.0% (11.4-28.1), chronic GVHD were 31.6% (19.8-44.1), 41.4% (32.6-50.0), 33.8% (26.2-41.6), and relapse incidences were 24.6% (14.6-36.1), 24.5% (17.7-31.8), 26.3% (19.6-33.5), respectively. All parameters assessed showed no significant differences between dCBT, sCBT, sCBT of low TNC and high TNC, except C.I of acute GVHDII-IV and NRM. In multivariate analysis, disease risk was an only significant factor affected on EFS (HR0.58; 95%CI: 0.400-0.833, P=0.003).
DISCUSSION: Although this is a retrospective analysis, double-unit use for CBT showed no statistically significant impact on engraftment rate and transplant outcomes compared to single-unit use of CBT even in patients of TNC number lower than 2.5x107/kg after myeloablative conditioning in Japan. Since more than 90% of the transplant candidates can find single CB unit >2x107/kg TNC, advantage of adding another CB unit may be quite limited in Japan both scientifically and financially, until meaningful benefit is demonstrated by randomized phase III study such as BMT-CTN 0501.
No relevant conflicts of interest to declare.
