Background: There is a growing appreciation for the GM in response to melanoma therapy. Recent work identified a favorable (T1) GM signature associated with checkpoint benefit. However, little is known about how lifestyle factors may influence features of the GM or about the influence of the GM on therapeutic outcomes to other common melanoma therapies.
Methods: We prospectively collected fecal samples melanoma patients at MD Anderson Cancer Center and characterized the GM via 16S rRNA sequencing (n=146). In a subset initiating systemic therapy (n= 113), we also collected baseline diet information via the NCI dietary screener questionnaire, and probiotic and antibiotic use and followed for response (RECISTv1.1). We compared GM beta diversity (BD)(Bray-Curtis) with ANOSIM and descriptive statistics and Wilcoxon tests were applied to alpha diversity (AD)(Chao1) among treatment response groups. Spearman correlation was used to test relationships of dietary groups and GM composition. Univariate and multivariate logistic regression models were used to determine the individual and joint effects of lifestyle factors on response. Receiver operating characteristic (ROC) curves were constructed for area under the curve (AUC) quantification.
Results: Considering all treatment groups, baseline AD was highest in patients with complete/partial response (mean=355), compared to patients with stable (mean=330) and progressive disease (mean=318). The GM did not significantly differ by age, sex or BMI. Report of probiotic (42%) and antibiotic (29%) use at baseline was associated with lower AD (p=0.02). Whole grains and overall diet quality correlated positively with pro-response bacteria, added sugars and processed meat negatively correlated. GM community structure (BD) differed by high and low fiber intake (p<0.05). Patients with high vs. low fiber diet had higher odds of response to aPD1 (OR=5.3, 95% CI: 1.02-26.3) and AUC for fiber in predicting response was 0.65. Joint effects analysis showed greatest odds of response among patients with a T1 signature and high fiber diet (OR=3.6, 95% CI: 0.7-17.8) vs. low fiber diet (OR=1.9, 95%CI 0.6-6.4), compared to patients without a T1 signature. Similar results were seen for overall diet score; and for both fiber and diet score, results reached statistical significance excluding patients on antibiotics.
Conclusion: AD of the GM varies by response regardless of treatment type. While preliminary, this data points to the idea that the GM in melanoma patients may be negatively influenced by probiotics and could be targeted by dietary manipulation. Larger prospective and interventional studies are needed to assess the relationships between host lifestyle factors, the GM and response to melanoma therapies.
Citation Format: Christine N. Spencer, Vancheswaran Gopalakrishnan, Jennifer McQuade, Miles C. Andrews, Beth Helmink, M.A. Wadud Khan, Elizabeth Sirmans, Lauren Haydu, Alexandria Cogdill, Elizabeth Burton, Rodabe Amaria, Sapna Patel, Isabella Glitza, MIchael Davies, Eliza Posada, Wen-Jen Hwu, Adi Diab, Kelly Nelson, Hussein Tawbi, Michael Wong, Robert R. Jenq, Lorenzo Cohen, Carrie Daniel-MacDougall, Jennifer A. Wargo. The gut microbiome (GM) and immunotherapy response are influenced by host lifestyle factors [abstract]. In: Proceedings of the American Association for Cancer Research Annual Meeting 2019; 2019 Mar 29-Apr 3; Atlanta, GA. Philadelphia (PA): AACR; Cancer Res 2019;79(13 Suppl):Abstract nr 2838.