Effect Modification by Serum Vitamin A Levels on the Association between Vitamin D and Relapse or Death in Patients with Digestive Tract Cancer

BACKGROUND: Vitamin D interacts with vitamin A through receptor heterodimerization. This post hoc analysis of the AMATERASU randomized clinical trial (RCT) of vitamin D3 supplementation examined whether serum vitamin A levels modify the association of vitamin D with relapse or death in patients with digestive tract cancer. METHODS: The primary outcome was relapse or death. Relapse-free survival (RFS) was evaluated using Nelson-Aalen cumulative hazard functions and Cox proportional hazards models. Serum vitamin A levels were categorized stepwise using clinical reference ranges and quantile cutoffs (halves, tertiles, quartiles, quintiles, and deciles), with post hoc grouping of the most coherent quantiles for analysis. RESULTS: Among 363 patients (mean age, 66 years; 67.8% male), serum vitamin A ranged from 0.15 to 4.30 μmol/L [median (IQR), 1.38 (1.05-1.74)] and correlated positively with 25-hydroxyvitamin D (ρ = 0.31; P < 0.0001). Patients in the lowest decile had higher relapse or death risk [HR, 2.05; 95% confidence interval (CI), 1.10-3.84; P = 0.03] than those in deciles 2 to 10. In the middle range (deciles 6-8), 5-year RFS was greater with vitamin D versus placebo (81.4% vs. 54.9%; HR, 0.31; 95% CI, 0.14-0.69; P = 0.004), with no effect in lower (deciles 1-5) or higher (deciles 9-10) ranges (Pinteraction = 0.008). CONCLUSIONS: Vitamin D supplementation may reduce relapse and death among patients with middle-to-upper range of serum vitamin A levels. IMPACT: These exploratory findings provide the first RCT evidence that serum vitamin A levels modify the effect of vitamin D supplementation on relapse or death among patients with digestive tract cancer.