Factors Associated with Long-Term Risk of Relapse after Unrelated Cord Blood Transplantation in Children with Acute Lymphoblastic Leukemia in Remission.

Journal Article (Clinical Trial;Journal Article;Multicenter Study)

For pediatric patients with acute lymphoblastic leukemia (ALL), relapse is an important cause of treatment failure after unrelated cord blood transplant (UCBT). Compared with other donor sources, relapse is similar or even reduced after UCBT despite less graft-versus-host disease (GVHD). We performed a retrospective analysis to identify risk factors associated with the 5-year cumulative incidence of relapse after UCBT. In this retrospective, registry-based study, we examined the outcomes of 640 children (<18 years) with ALL in first complete remission (CR1; n = 257, 40%) or second complete remission (CR2; n = 383, 60%) who received myeloablative conditioning followed by a single-unit UCBT from 2000 to 2012. Most received antithymocyte globulin (88%) or total body irradiation (TBI; 69%), and cord blood grafts were primarily mismatched at 1 (50%) or 2+ (34%) HLA loci. Considering patients in CR1, the rates of 5-year overall survival (OS), leukemia-free survival (LFS), and relapse were 59%, 52%, and 23%, respectively. In multivariate analysis (MVA), acute GVHD (grades II to IV) and TBI protected against relapse. In patients in CR2, rates of 5-year OS, LFS, and the cumulative incidence of relapse were 46%, 44%, and 28%, respectively. In MVA, longer duration from diagnosis to UCBT (≥30 months) and TBI were associated with decreased relapse risk. Importantly, receiving a fully HLA matched graft was a strong risk factor for increased relapse in MVA. An exploratory analysis of all 640 patients supported the important association between the presence of acute GVHD and less relapse but also demonstrated an increased risk of nonrelapse mortality. In conclusion, the impact of GVHD as a graft-versus-leukemia marker is evident in pediatric ALL after UCBT. Strategies that promote graft-versus-leukemia while harnessing GVHD should be further investigated.

Full Text

Duke Authors

Cited Authors

  • Page, KM; Labopin, M; Ruggeri, A; Michel, G; Diaz de Heredia, C; O'Brien, T; Picardi, A; Ayas, M; Bittencourt, H; Vora, AJ; Troy, J; Bonfim, C; Volt, F; Gluckman, E; Bader, P; Kurtzberg, J; Rocha, V

Published Date

  • August 2017

Published In

Volume / Issue

  • 23 / 8

Start / End Page

  • 1350 - 1358

PubMed ID

  • 28438676

Pubmed Central ID

  • PMC5569913

Electronic International Standard Serial Number (EISSN)

  • 1523-6536

Digital Object Identifier (DOI)

  • 10.1016/j.bbmt.2017.04.015


  • eng

Conference Location

  • United States