ABSTRACT Esophageal cancer (EC) results in high mortality due to difficulty in early diagnosis, particularly in low- and middle-income countries, including the African EC belt typified by high prevalence, early onset, and poor prognosis. While the precise etiological factors remain unknown, emerging data suggest links to the oral microbiota. In this study, we conducted a secondary analysis using V4 16S rRNA sequencing from a cross-sectional study of treatment-naive, newly diagnosed EC patients ( N = 103) and healthy controls ( N = 108) residing in agricultural regions of Ethiopia. We report that the salivary microbiota in the healthy Ethiopian controls is highly diverse, forming two functionally distinct community clusters differing in diversity, composition, and absolute abundance. Microbiota composition was associated with sex and alcohol consumption, but not age. Comparisons against groups from geographically distinct populations representing Tanzania, Uganda, Venezuela, and the United States ( N = 641) showed that cluster 2 resembled other East African populations, while cluster 1 was unique to the Ethiopian cohort. Both EC subtypes, esophageal squamous cell carcinoma (ESCC) and esophageal adenocarcinoma, were associated with a loss of microbial diversity and an increased probability of having a cluster 2 microbiota (adjusted OR = 2.9 95% CI 1.5–5.9). Classifiers trained to discriminate healthy and EC samples were further validated on two external EC cohorts from China ( N = 161). Models trained on the Ethiopian cohort could predict disease status in an external cohort of mid- and late-stage ESCC from China (AUROC = 0.70 ± 0.03 mean ± SD), demonstrating the generalization of microbial features of ESCC across populations. IMPORTANCE Recent reports in North America and China have correlated oral microbiota composition with esophageal cancer, although the translation of this knowledge into the African esophageal cancer belt is hampered by a lack of data on the oral microbiota of East Africans and limited cross-cohort comparative analyses validating the utility of these biomarkers. We report that the human salivary microbiota is a meaningful biomarker of later-stage esophageal cancer that transcends geography and ethnicity and may provide utility for large-population screening. A lower-diversity and lower-abundance salivary microbiota correlated with esophageal cancer warrants further investigation to understand the role of oral microbes in mediating carcinogenesis.
Mulisa et al. (Tue,) studied this question.