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.

Increased Body Mass Index Augments Endothelial Injury and Clinical Outcomes after Hematopoietic Stem Cell Transplantation.

Higher body mass index (BMI) is characterized as a chronic inflammatory state with endothelial dysfunction. Endothelial injury after allogeneic hematopoietic stem cell transplantation (allo-HSCT) puts patients at risk for such complications as transplantation-associated thrombotic microangiopathy (TA-TMA) and acute graft-versus-host-disease (aGVHD). To evaluate the impact of increased BMI on endothelial injury after allo-HSCT in pediatric and young adult patients, we conducted a retrospective cohort study evaluating 476 consecutive allo-HSCT children and young adult recipients age 0 to 20 years. Our analysis was subdivided based on distinct age categories (<2 years and 2 to 20 years). BMI was considered as a variable but was also expressed in standard deviations from the mean adjusted for age and sex (z-score), based on established criteria from the World Health Organization (age <2 years) and the Centers for Disease Control and Prevention (age 2 to 20 years) to account for differences associated with age. Primary endpoints included the incidences of TA-TMA and aGVHD. Increased BMI z-score was associated with TA-TMA after allo-HSCT in patients age <2 years (median, 18.1; IQR, 17 to 20; P = .006) and in patients age 2 to 20 years (median, 18.7; IQR, 16 to 21.9; P = .02). Higher BMI z-score correlated with TA-TMA risk in both age groups, with a BMI z-score of .9 in the younger cohort and .7 (IQR, -.4 to 1.6; P = .04) in the older cohort. Increased BMI z-score was associated with an increased risk of TA-TMA in a multivariate analysis of the entire cohort (odds ratio [OR], 1.2; 95% confidence interval [CI], 1.05 to 1.37; P = .008). Multivariate analysis also demonstrated that patients with BMI in the 85th percentile or greater had an increased risk of developing TA-TMA compared to those with a lower BMI percentile (OR, 2.66; 95% CI, 1.62 to 4.32; P < .001). Baseline and day +7 ST2 levels were elevated in subjects with TA-TMA compared to those without TA-TMA in both age groups. Baseline sC5b-9 concentration was not correlated with BMI z-score, but sC5b-9 concentration was increased markedly by 7 days post-allo-HSCT in patients age <2 years who later developed TA-TMA compared to those who never developed TA-TMA (P = .001). The median BMI z-score was higher for patients with aGVHD compared to patients without aGVHD (.7 [range, -3.9 to 3.9] versus .2 [range, -7.8 to 5.4]; P = .03). We show that high BMI is associated with augmented risk of endothelial injury after HSCT, specifically TA-TMA. These data identify a high-risk population likely to benefit from early interventions to prevent endothelial injury and prompt treatment of established endothelial injury.

Intestinal permeability in patients undergoing stem cell transplantation correlates with systemic acute phase responses and dysbiosis.

Intestinal permeability may correlate with adverse outcomes during hematopoietic stem cell transplantation (HSCT), but longitudinal quantification with traditional oral mannitol and lactulose is not feasible in HSCT recipients because of mucositis and diarrhea. A modified lactulose:rhamnose (LR) assay is validated in children with environmental enteritis. Our study objective was to quantify peri-HSCT intestinal permeability changes using the modified LR assay. The LR assay was administered before transplant, at day +7 and +30 to 80 pediatric and young adult patients who received allogeneic HSCT. Lactulose and rhamnose were detected using urine mass spectrometry and expressed as an L:R ratio. Metagenomic shotgun sequencing of stool for microbiome analyses and enzyme-linked immunosorbent assay analyses of plasma lipopolysaccharide binding protein (LBP), ST2, REG3α, claudin1, occludin, and intestinal alkaline phosphatase were performed at the same timepoints. L:R ratios were increased at day +7 but returned to baseline at day +30 in most patients (P = .014). Conditioning regimen intensity did not affect the trajectory of L:R (P = .39). Baseline L:R ratios did not vary with diagnosis. L:R correlated with LBP levels (r2 = 0.208; P = .0014). High L:R ratios were associated with lower microbiome diversity (P = .035), loss of anaerobic organisms (P = .020), and higher plasma LBP (P = .0014). No adverse gastrointestinal effects occurred because of LR. Intestinal permeability as measured through L:R ratios after allogeneic HSCT correlates with intestinal dysbiosis and elevated plasma LBP. The LR assay is well-tolerated and may identify transplant recipients who are more likely to experience adverse outcomes.