Patients with esophageal cancer showed differences in fecal metabolites and gut bacteria compared with people without cancer, according to a study published in Scientific Reports. The findings may guide future biomarker research, but they do not establish a diagnostic test.
Researchers studied 34 patients with esophageal cancer and 29 controls matched for age, sex and dietary habits. The cancer group included patients with stages I through IV disease. Stool samples were analyzed using liquid chromatography-tandem mass spectrometry and 16S ribosomal RNA gene sequencing. They also measured five serum tumor markers: CEA, CA19-9, CA72-4, CYFRA21-1 and squamous cell carcinoma antigen.
For the metabolomic analysis, compounds were extracted from 20-milligram fecal samples and analyzed using high-performance liquid chromatography coupled to quadrupole time-of-flight mass spectrometry. Data were collected in positive and negative ionization modes. Metabolites were identified by comparing their accurate mass and fragmentation spectra with an in-house database of authenticated standards.
The researchers identified 792 metabolites with different abundances between the two groups. The most prominent changes involved fatty acid, amino acid, and nucleic acid metabolism. Lipid-related findings included altered levels of oleic, linoleic, dodecanoic, and stearic acids. The study also identified changes in phosphatidylcholine and phosphatidylethanolamine, which are important components of cell membranes.
Pathway analysis highlighted fatty acid transport, linoleic acid metabolism, and lipid regulation involving peroxisome proliferator-activated receptor alpha. An exploratory comparison across cancer stages also identified a group of metabolites associated mainly with lipid metabolism.
Microbiome sequencing generated approximately 6.5 million valid reads, averaging more than 103,000 per sample. Several measures of microbial richness and diversity differed between the cancer and control groups.
At the genus level, bacteria including Akkermansia, Dialister, and Prevotella were more abundant in the cancer group in comparative analyses. Other bacteria, including Bacteroides, Romboutsia, and Fusicatenibacter, were more abundant among controls.
The researchers used PICRUSt2 to predict possible microbial functions from the 16S sequencing data. The analysis suggested differences in pathways associated with steroid and fatty acid biosynthesis, glycerolipid metabolism, and amino and nucleotide sugars. These results represent computational predictions rather than direct measurements of microbial genes or activity.
Correlation analysis also linked several bacterial groups with metabolites involved in lipid, indole, and nucleoside metabolism. However, these associations do not show that microbial changes caused the metabolic differences or contributed directly to cancer development.
The study has several limitations. It included only 63 participants from one hospital, and the small number of patients in each cancer stage limited the progression analysis. The cross-sectional design could not determine whether the observed changes developed before or after the cancer.
Patients in the cancer group had also received treatments including chemotherapy, radiotherapy, or combined regimens. These treatments may have affected their microbiomes and metabolism. Dysphagia, malnutrition, and dietary changes associated with advanced disease could also explain some findings.
In addition, 16S sequencing generally identified bacteria only to the genus level. The study did not develop or independently validate a diagnostic model or report sensitivity and specificity.
Larger prospective studies involving untreated patients and early-stage disease are needed to determine whether fecal metabolites or microbial profiles could complement endoscopy and tissue biopsy. Future work will also need to assess analytical reproducibility, potential confounders, and performance in independent populations.
