Graft Vs Host Disease (GVHD) in Pediatric Hematopoietic Stem Cell Transplant (HCT) Recipients and Impact on Overall Survival: A CIBMTR Analysis

Recent multi-institutional studies of GVHD incidence and impact in pediatric patients are lacking. In this CIBMTR analysis, we compare the incidence of acute and chronic GVHD in pediatric patients by recipient age and compare overall survival (OS) and transplant related mortality (TRM) among patients with and without GVHD.We included 14,099 patients [7,582 with malignancy] ≤ 21 years receiving their first allogeneic HCT from 2002-2020. Recipient age was categorized as: 0-2 years (infants and toddlers, I/T), 3-12 years (school-aged, S/A), and 13-21 years (adolescent and young adult, AYA). In the malignant cohort, 92% received myeloablative conditioning (MAC), and 42% received bone marrow (BM) grafts. In the nonmalignant cohort, 55% received MAC, 79% received in vivo T-cell depletion (TCD) and 56% received BM grafts.Unadjusted incidence of grade 2-4 or 3-4 aGVHD by Day 100 did not differ by recipient age within the malignant cohort; in the nonmalignant cohort, incidence of grade 3-4 aGVHD was statistically but perhaps not clinically different among age groups: 9% in I/T, 10% in S/A, and 12% in AYAs [P<0.001]. Unadjusted incidence of any cGVHD by 1y was higher with age: 20%, 26%, and 33% in the malignant cohort [P<0.001] and 17%, 17%, and 23% in the nonmalignant cohort [P<0.001] in I/T, S/A, and AYAs, respectively. Unadjusted 5-year OS significantly differed by age, with the S/A children having the highest OS [malignant: I/T 56% (53-59), S/A 58% (56-60), AYA 51% (49-53), P<0.001; nonmalignant: I/T 75% (73-77), S/A 82% (81-84), AYA 75% (73-77), P<0.001].In multivariate analysis, risk of grade 3-4 aGVHD was not significantly associated with age in the malignant cohort; risk of grade 3-4 aGVHD was higher in older children in the nonmalignant cohort (Figure 1). Adjusted incidence of cGVHD was significantly higher in older children among both cohorts (Figure 2). Donor type, sex match, graft type, in vivo TCD, and race were potential risk factors for development of cGVHD in the malignant cohort (Figure 3). Considering GVHD as a time varying covariate, in the malignant cohort, both grade 3-4 aGVHD and any cGVHD were associated with OS [HR, 1.87, P<0.0001, and HR 1.32, P<0.0001, respectively] and with TRM [HR 2.76, P<0.0001, and HR 2.93, P<0.0001, respectively]. Similarly in the nonmalignant cohort, grade 3-4 aGVHD and any cGVHD were significantly associated with OS.In conclusion, the risk of cGVHD increases with age. The risk of severe acute GVHD is increased among older patients with nonmalignant conditions but this association was not observed in patients with malignancies possibly due to factors unaccounted for such as relapse risk. Despite the low overall incidence of acute and cGVHD in children, both are associated with lower survival. This is essential to consider at the time of HCT consultation, particularly in the nonmalignant cohort.