Despite continued improvement in outcomes after allo-HCT, acute (aGVHD) and chronic GVHD (cGVHD) remain major causes of morbidity and transplant related mortality (TRM). UCB has emerged as an important alternative donor source with clinical outcomes similar to those after HCT with 8/8 allele matched unrelated and related donor. While prior studies have indicated differences in risks and natural history of GVHD between hematopoietic stem cell (HSC) sources, we sought to provide a more detailed analysis taking advantage of the homogeneity in grading criteria, treatment plans, and graft selection criteria at a single center. To identify incidence and risk factors for acute and chronic GVHD, we retrospectively studied 1180 adult (≥18 years, n=801) and pediatric (n=379) recipients of single (sUCB, n=295), double (dUCB, n=416) UCB and MRD (n=469) allografts transplanted between 2000-2012. We estimated cumulative incidence of aGVHD and cGVHD with non-GVHD deaths modeled as competing risks. Analysis of risk factors for GVHD was performed separately in the three cohorts. sUCB, dUCB, and MRD differed in age (median 8, 44, 47 years, respectively), gender (54%, 63%, 62% males), conditioning (myeloablative in 80%, 40%, 55%; ATG was used in 60%, 23%, 16%), CsA/MMF GVHD prophylaxis (66%, 99%, 48%), total nucleated cells (median 0.5, 0.4, 7.7 x 108 cells/kg), CD34 dose (median 0.5, 0.5, 5.4 x 106 cells/kg), and CD3 dose (median 0.2, 0.1, 3.6 x 108 cells/kg). sUCB transplants had lower incidence of grade II-IV aGVHD (26%), grade III-IV aGVHD (7%) and cGVHD (7%) as compared to dUCB (56% aGVHD II-IV; 21% aGVHD III-IV; 26% cGVHD) and MRD (37% aGVHD II-IV; 16% aGVHD III-IV; 40% cGVHD) (all p<0.01, Figure). In competing risk multivariable analysis of sUCB allografts, age≥18 years was associated with higher incidence of grade II-IV aGVHD (HR=1.8, 95%CI, 1.1-2.9; p=0.01), grade III-IV aGVHD (HR=3.4, 95%CI 1.6-7.2; p<0.01) and cGVHD (HR=5.7, 95%CI 1.9-16.5; p<0.01). Better HLA match influenced grade II-IV aGVHD with lower incidence for 6/6 match (HR=0.5, 95%CI 0.2-1.0; p=0.04 vs. 4/6), but did not influence risks of grade III-IV aGVHD or cGVHD. GVHD prophylaxis with CsA/MMF was associated with reduced incidence of cGVHD (HR=0.3, 95%CI 0.1-0.9; p=0.03 vs. CsA/Steroid). In dUCB allografts, the use of ATG (HR=0.5, 95%CI 0.4-0.7; p<0.01) and year of transplant ≥2006 (HR=0.6, 95%CI 0.5-0.8; p<0.01) were associated with less grade II-IV aGVHD. Better HLA match limited both grade III-IV aGVHD (HR=0.3, 95%CI 0.1-0.9; p=0.03 for 6/6 vs. 4/6 match) and cGVHD (HR=0.6, 95%CI 0.4-0.9; p=0.02 for 5/6 vs. 4/6). Reduced intensity conditioning was associated with less grade III-IV aGVHD (HR=0.6, 95%CI 0.4-0.9; p=0.02) and cGVHD (HR=0.6, 95%CI 0.4-1.0; p=0.03). Prior grade II-IV aGVHD led to more frequent cGVHD (HR=2.0, 95%CI 1.3-3.1, p<0.01). In MRD analysis (79% received peripheral-blood stem cell [PBSC] grafts), risk of grade II-IV aGVHD (HR=2.5, 95%CI 1.1-5.5; p=0.02), grade III-IV aGVHD (HR=7.1, 95%CI 2.3-22.2; p=0.001) and cGVHD (HR=2.6, 95%CI 1.1-6.3; p=0.04) were higher in PBSC vs. marrow allografts. Older MRD recipients had higher risk of cGVHD (HR=5.8, 95%CI 1.7-19.8; p<0.01 for age≥18 vs. <18). This large analysis establishes contemporary incidence and prognostic determinants of GVHD for UCB in parallel to benchmark MRD allo-HCT. Our data argue that better HLA match may further mitigate the risks of grade III-IV aGVHD and cGVHD to maximize the benefits of dUCB transplantation. Increasing the UCB inventory or developing strategies (e.g., ex vivo expansion culture) that reduce the cell-dose threshold of 2.5 x 107 nucleated cells/kg and thereby increase the chance of identifying an adequately dosed, better HLA matched single UCB unit would further limit risks of acute and chronic GVHD after UCB transplantation.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.