Table 2.
Studies of the esophageal microbiome and esophageal cancer.
| Main Findings | |||||||||
|---|---|---|---|---|---|---|---|---|---|
| Author (Year) | Study Design | Country | Cases (N) | Controls (N) | Sample Type | Method | Alpha Diversity | Beta Diversity | Differentially Abundant Taxa |
| Yu 2014, [150] | Cross-sectional case-control | China | 142 ESD | 191 | Balloon | HOMIM | Microbial richness lower in ESD | First principal component associated with ESD status | None. |
| Shao 2019, [166] | Cross-sectional | China | 67 ESCC, 36 gastric cardia adenocarinoma | – | Biopsy | 16S rRNA, NGS | No difference between ESCC tumor and paired nontumor samples | Separation between ESCC tumor and nontumor samples | ESCC tumor samples had higher Fusobacterium and lower Streptococcus compared with paired nontumor samples. Inverse association between Streptococcus and ESCC tumor stage. |
| Li 2020, [167] | Cross-sectional case-control | China | 17 ESCC, 11 esophagogastric junction cancer, 15 post-esophagectomy | 16 | Biopsy | 16S rRNA, NGS | Overall lower alpha diversity in all case groups compared with controls | Separation between ESCC and controls | Streptococcus, Lactobacillus, Prevotella, and Fusobacterium enriched in ESCC compared with controls. Fusobacteria was higher and Actinobacteria was lower in ESCC compared with controls. Key taxa distinguishing ESCC patients identified to be Clostridiales, Pseudomonas, and Selenomonadales. |
| Nasrollahzadeh 2015, [168] | Cross-sectional case-control | Iran | 37 early ESCC and ESD | 17 diseased controls (mid-esophageal esophagitis), 37 healthy controls | Biopsy | 16S rRNA, NGS | No difference | Some differences in principal coordinates between cases and controls | Clostridiales and Erysipelotrichales higher in cases compared with healthy controls. |
| Macfarlane 2007, [176] | Cross-sectional case-control | UK | 7 BE | 7 | Biopsy, aspirate | Culture | – | – | Campylobacter (C. concisus and C. rectus) in most BE patients but in none of the controls. |
| Yang 2009, [145] | Cross-sectional case-control | US | 12 esophagitis, 10 BE | 12 | Biopsy | 16S rRNA gene clones | – | Separation between cases (combined) and controls | Streptococcus predominant in controls. Increased Gram-negative bacteria in BE and esophagitis. |
| Blackett 2013, [159] | Cross-sectional case-control | UK | 37 GERD, 45 BE, 30 EAC | 39 | Biopsy | Culture/qPCR | – | – | Higher Campylobacter (C. concisus was the dominant species) in GERD and BE compared with controls but not with EAC. |
| Liu 2013, [157] | Cross-sectional case-control | Japan | 6 esophagitis, 6 BE | 6 | Biopsy | 16S rRNA gene clones | – | – | Streptococcus predominant in controls and reflux esophagitis. Veillonella, Neisseria, and Fusobacterium present in reflux esophagitis and BE but not controls. |
| Amir 2014, [158] | Cross-sectional case-control | Israel | 13 GERD with esophagitis, 6 BE | 15 GERD without esophagitis | Biopsy, gastric fluid | 16S rRNA, NGS | – | No separation between esophageal biopsies from esophagitis, BE, and controls; separation between gastric fluid samples from controls and abnormal esophagus (esophagitis and BE combined) | No differential taxa for esophageal biopsies from abnormal esophagus (esophagitis and BE) and controls. Higher Enterobacteriaceae (specifically the genus Escherichia) and Methylobacteriaceae in gastric fluid of esophagitis and BE compared with controls. Pasteurellaceae and Porphymonodaceae higher in controls compared with esophagitis and BE. |
| Gall 2015, [151] | Cross-sectional | US | 12 BE | – | Biopsy, brushing | 16S rRNA, NGS | – | – | Inverse correlation between Streptococcus:Prevotella ratio and hiatal hernia length (risk factor for BE and EAC) in the stomach corpus, stomach antrum, and BE. |
| Elliott 2017, [149] | Cross-sectional case-control | UK | 24 non-dysplastic BE, 23 dysplastic BE, 19 EAC | 20 | Biopsy, brushing, Cytosponge | 16S rRNA, NGS/qPCR | Alpha diversity lower in EAC compared to controls | Separation between EAC and controls | BE had higher Proteobacteria compared with controls and EAC. Campylobacter, Veillonella, Megasohaera, Granulicatella, Atopobium, Actinomyces, and Solobacterium lower in EAC compared with BE and controls. Lactobacillus fermentum enriched in EAC compared with BE and controls, and Lactobacillales (Lactobacillus spp. and Streptococcus spp.) predominant in EAC. |
| Deshpande 2018, [152] | Cross-sectional case-control | Australia | 29 GERD, 7 glandular mucosa, 5 BE, 1 EAC, 1 eosinophilic esophagitis | 59 | Biopsy, brushing | 16S/18S rRNA, NGS/shotgun | No difference by disease status | No difference by disease status | Gram-negative bacteria (e.g., Leptotrichia, Fusobacterium, Rothia, Campylobacter, Capnocytophaga) enriched in GERD, glandular mucosa, and BE. Microbial lactic acid production was increased in GERD and BE. Streptococcus:Prevotella ratio defined functionally distinct esophageal communities. |
| Snider 2019, [163] | Cross-sectional case-control | US | 14 BE without dysplasia, 6 low grade dysplasia, 5 high grade dysplasia, 4 EAC | 16 | Brushing | 16S rRNA, NGS | No difference between BE cases (combined) and controls; among BE cases, EAC had decreased Simpson index | No difference | Combined group of high-grade dysplasia and EAC had decreased Firmicutes and increased Proteobacteria compared with the group of BE without dysplasia and low-grade dysplasia. Group of high-grade dysplasia and EAC had increased Enterobacteriaceae and Akkermansia muciniphila and reduced Veillonella. |
| Okereke 2019, [147] | Cross-sectional | – | 12 BE | – | Biopsy | qPCR | – | – | Haemophilus abundant in BE tissue but relatively absent elsewhere in the esophagus. BE tissue dominated by a larger percentage of Gram-negative organisms compared with other sites in the esophagus. |
| Lopetuso 2020, [177] | Cross-sectional case-control | Italy | 10 BE, 6 EAC | 10 | Biopsy | 16S rRNA, NGS | Higher alpha diversity in BE and EAC compared with controls, but not statistically significant | Separation between EAC and controls and between EAC and BE | Progressive reduction in Streptococcus and corresponding increase in Prevotella in BE and EAC. Leptotrichia was a distinguishing taxon for EAC. |
| Zhou 2020, [160] | Cross-sectional case-control | Australia | 11 NERD, 20 RE, 17 BE, 6 EAC | 16 | Biopsy, brushing | 16S rRNA, NGS | Lower Chao1 in NERD compared with controls and RE; no difference in Shannon index | Some differences between EAC and controls | Controls had higher Firmicutes and Actinobacteria compared with other groups. NERD had higher Proteobacteria and Bacteroidetes and lower Fusobacteria and Actinobacteria, along with decreases in several Firmicutes genera including Dorea. RE and BE had lower Firmicutes and increased Gram-negative Fusobacteria and Proteobacteria compared with controls. EAC shifted towards Firmicutes (mainly Staphylococcus aureus, Streptococcus infantis, Moryella sp. and Lactobacillus salivarius) and Proteobacteria, while shifting away from Actiobacteria (Rothia mucilaginosa) compared with controls. |
– absent or not applicable. BE, Barrett’s esophagus; EAC, esophageal adenocarcinoma; ESCC, esophageal squamous cell carcinoma; ESD, esophageal squamous dysplasia; GERD, gastroesophageal reflux disease; HOMIM, Human Oral Microbe Identification Microarray; NERD, non-erosive reflux disease; NGS, next-generation sequencing; RE, reflux esophagitis.