A randomized phase 2 trial of oral vitamin A for graft-versus-host disease in children and young adults

Abstract Vitamin A plays a key role in the maintenance of gastrointestinal homeostasis and promotes a tolerogenic phenotype in tissue resident macrophages. We conducted a prospective randomized double-blinded placebo-controlled clinical trial in which 80 recipients of hematopoietic stem cell transplantation (HSCT) were randomized 1:1 to receive pretransplant high-dose vitamin A or placebo. A single oral dose of vitamin A of 4000 IU/kg, maximum 250 000 IU was given before conditioning. The primary end point was incidence of acute graft-versus-host disease (GVHD) at day +100. In an intent-to-treat analysis, incidence of acute GVHD was 12.5% in the vitamin A arm and 20% in the placebo arm (P = .5). Incidence of acute gastrointestinal (GI) GVHD was 2.5% in the vitamin A arm (P = .09) and 12.5% in the placebo arm at day +180. Incidence of chronic GVHD was 5% in the vitamin A arm and 15% in the placebo arm (P = .02) at 1 year. In an “as treated” analysis, cumulative incidence of acute GI GVHD at day +180 was 0% and 12.5% in recipients of vitamin A and placebo, respectively (P = .02), and cumulative incidence of chronic GVHD was 2.7% and 15% in recipients of vitamin A and placebo, respectively (P = .01). The only possibly attributable toxicity was asymptomatic grade 3 hyperbilirubinemia in 1 recipient of vitamin A at day +30, which self-resolved. Absolute CCR9+ CD8+ effector memory T cells, reflecting gut T-cell trafficking, were lower in the vitamin A arm at day +30 after HSCT (P = .01). Levels of serum amyloid A-1, a vitamin A transport protein with proinflammatory effects, were lower in the vitamin A arm. The vitamin A arm had lower interleukin-6 (IL-6), IL-8, and suppressor of tumorigenicity 2 levels and likely a more favorable gut microbiome and short chain fatty acids. Pre-HSCT oral vitamin A is inexpensive, has low toxicity, and reduces GVHD. This trial was registered at www.ClinicalTrials.gov as NCT03202849.