Identification of microbial markers across populations in early detection of colorectal cancer

Colorectal cancer (CRC) is a leading cause of cancer-related deaths globally. Early detection at the adenoma stage significantly improves survival rates. While the gut microbiome's role in CRC is increasingly recognized, identifying replicable microbial markers for early-stage adenoma diagnosis across diverse populations remains challenging. Previous studies have identified some fecal microbial markers associated with CRC, but their ability to specifically detect adenomas, the early-stage precursors to CRC, remains unclear. Moreover, significant variations in identified microbial markers exist across studies due to biological factors influencing gut microbiome composition and inconsistencies in data processing protocols. Meta-analysis provides a powerful approach to overcome such limitations by integrating data from multiple studies. However, universal microbial markers specific for colorectal adenomas are less frequently reported and demonstrate relatively low diagnostic accuracy. This study aims to address these limitations by performing a comprehensive meta-analysis of publicly available 16S rRNA sequencing data from multiple studies to identify robust and reliable microbial markers for the detection of colorectal adenomas.

Numerous studies have investigated the association between gut microbiota and colorectal cancer (CRC). However, the identification of consistent and reliable microbial markers for early-stage adenoma diagnosis across diverse populations remains a challenge. Existing studies have identified some fecal microbial markers for CRC, but their effectiveness in specifically detecting adenomas has been inconsistent. Variability in results may be attributed to biological factors influencing gut microbiome composition, differences in study methodologies, and inconsistencies in data processing and analysis techniques. Meta-analyses have been employed to identify universal microbial markers across different diseases, but the accuracy of adenoma-specific markers remains relatively low. Previous meta-analyses utilizing whole metagenome shotgun (WMS) sequencing have shown limited accuracy in differentiating adenomas from healthy controls or CRC, possibly due to limited taxonomic coverage and reliance on reference genomes. This underscores the need for a comprehensive meta-analysis of 16S rRNA sequencing data to identify accurate and widely applicable microbial markers for colorectal adenoma detection. The current non-invasive stool-based screening test, fecal immunochemical test (FIT), has limitations in detecting early and advanced adenomas, emphasizing the urgency of identifying more accurate and efficient stool-based microbial markers.

This study integrated 16S rRNA sequencing data from four published studies encompassing 1056 fecal samples from individuals with colorectal adenomas, CRC, and healthy controls. Data were preprocessed using QIIME2 and DADA2 to ensure high-quality data. Potential confounding factors, including age, BMI, sex, antibiotic use, and study origin, were carefully considered. A two-sided blocked Wilcoxon rank-sum test was used to identify differentially abundant ASVs (amplicon sequence variants) between different disease groups while adjusting for confounding effects. Random Forest (RF) classifiers were constructed to differentiate adenomas from healthy controls and CRC using differentially abundant ASVs as features. The performance of the classifiers was evaluated using the area under the receiver operating characteristic curve (AUC). Study-to-study transfer validation and leave-one-dataset-out (LODO) validation were performed to assess the robustness and generalizability of the identified markers across different populations and experimental settings. Functional analysis using PICRUSt2 was conducted to infer the functional pathways associated with altered microbiome composition. Quantitative real-time PCR (qRT-PCR) was used to validate key findings in a newly collected cohort of 43 samples. Additionally, the specificity of the identified markers was evaluated by assessing their performance in differentiating adenoma from other microbiome-linked diseases.

The study identified 11 microbial markers that effectively distinguished colorectal adenomas from healthy controls (AUC = 0.80) and 26 markers that distinguished adenomas from CRC (AUC = 0.89) using Random Forest classifiers. These results were validated in two independent cohorts, achieving AUCs of 0.78 and 0.84 respectively. Functional analysis indicated that the altered microbiome in adenoma was characterized by increased ADP-L-glycero-β-D-manno-heptose biosynthesis, a key metabolic intermediate in LPS biosynthesis, potentially contributing to inflammation and tumorigenesis. In CRC, elevated menaquinone-10 biosynthesis was observed, potentially representing a compensatory response to the tumor microenvironment. These findings were validated in a new cohort using qRT-PCR. The study also demonstrated the high specificity of these adenoma-specific markers against other microbiome-linked diseases. The combination of the identified microbial markers with FIT significantly improved the diagnostic accuracy (AUC = 0.81) compared to either test alone.

This study successfully identified a set of microbial markers that accurately distinguish colorectal adenomas from healthy controls and CRC across multiple populations. The high AUC values obtained in both the discovery and validation cohorts, along with the consistent performance across different studies, demonstrate the robustness and generalizability of these markers. The functional analyses provide mechanistic insights into the role of microbial dysbiosis in adenoma and CRC development. The identification of adenoma-specific markers, combined with the improved accuracy when used in conjunction with FIT, suggests a potential non-invasive diagnostic strategy for early CRC detection. This could significantly improve patient outcomes by enabling earlier intervention and treatment. The findings of this study contribute valuable knowledge to the field of CRC research and highlight the potential of microbiome-based approaches for improved early detection and treatment strategies.

This study successfully identified and validated a panel of microbial markers capable of accurately distinguishing colorectal adenomas from healthy controls and CRC across diverse populations. These markers, combined with existing non-invasive screening methods like FIT, hold significant potential for improving early CRC detection. Future research should focus on larger, more diverse cohorts and the development of clinical diagnostic tools based on these findings. Investigation into the precise mechanisms by which these microbial alterations contribute to adenoma and CRC development will further enhance our understanding of CRC pathogenesis and potentially lead to novel therapeutic strategies.

While this study demonstrates the potential of microbiome-based biomarkers for early CRC detection, certain limitations should be acknowledged. The reliance on publicly available datasets may introduce biases due to variations in study designs, sample collection methods, and sequencing technologies. The relatively small sample size in some validation cohorts may limit the generalizability of the findings. Further studies with larger and more diverse populations are necessary to confirm the robustness of these findings and evaluate their clinical utility. Additionally, functional analysis relies on computational prediction; experimental validation of functional pathways is needed.