To evaluate the patterns of vitamin and herbal supplement use among patients with advanced gastrointestinal (GI) cancers and the association of such behavior with the efficacy and toxicity of systemic anticancer treatment.
Project data sphere (PDS) was used to access de-identified datasets of eight clinical trials of advanced GI cancers. Multivariable logistic regression analysis was used to identify factors predicting the use of supplements. Kaplan–Meier survival estimates were used to evaluate the association of supplement use with overall and progression-free survival. Results were stratified according to the site of the primary tumor [pancreatic, gastric, colorectal or hepatocellular carcinoma (HCC)] The association between supplement use and selected chemotherapy side effects was evaluated through Chi-squared testing and subsequent logistic regression.
A total of 3441 patients were included in the analysis. Of these, 775 patients reported use of supplements and 2666 patients reported no use of supplements. Higher ECOG performance score (Odds ratio: OR for ECOG 1 versus 0: 1.629; 95% CI 1.363–1.947; P < 0.001) and pancreatic primary site (OR for gastric cancer versus pancreatic cancer: 0.538; 95% CI 0.408–0.709; P < 0.001) was associated with greater use of these supplements. Supplement use was associated with a better overall survival among patients with pancreatic cancer (P = 0.002) but not other GI malignancies. Supplement use was associated with a higher probability of anemia and diarrhea among patients with pancreatic cancer (P < 0.001 for both), gastric cancer (P = 0.016; P = 0.036, respectively) and colorectal cancer (P < 0.001 for both).
There is an association between the use of vitamin and herbal supplements and a higher probability of hematologic and gastrointestinal toxicities. There is a need for more studies to confirm the association between such behavior and better overall survival among patients with pancreatic cancer.