To the Editor: Colorectal cancer (CRC) is the third most commonly diagnosed cancer worldwide and the second leading cause of cancer-related mortality, underscoring the critical need for effective non invasive screening.1 Current tools like the fecal immunochemical test (FIT) are suboptimal for detecting precancerous advanced adenomas (AAs).2 Alterations in the gut microbiome are strongly implicated in colorectal tumorigenesis, offering a promising avenue for novel biomarkers.3 The prospective multi center study aimed to evaluate the diagnostic performance of a panel of four fecal microbial biomarkers: Fusobacterium spp. (Fn), Hungatella hathewayi (Hh), Lachnoclostridium sp. (m3), and Christensenella hongkongensis (Chk) combined with FIT for the non invasive detection of CRC and AA. This prospective, multi center diagnostic study was approved by the Ethics Committee of Beijing Friendship Hospital, Capital Medical University (No. 2022-P2-084). We consecutively recruited adults scheduled for colonoscopy from five centers located across northern, southern, eastern, and western China, including Beijing Friendship Hospital affiliated with Capital Medical University, the First Affiliated Hospital of Kunming Medical University, Renji Hospital affiliated with Shanghai Jiaotong University School of Medicine, Xijing Hospital, and the Sixth Affiliated Hospital of Sun Yat-sen University, from July 2022 to December 2023. Exclusion criteria included a history of inflammatory bowel disease or antibiotic use within 30 days before stool collection. Informed consent was obtained from all participants. Control status was defined as participants with normal colonoscopy findings, or with findings limited to non-AAs or non-adenomatous polyps, as confirmed by the gold-standard method of colonoscopy and histopathological examination. Stool samples were collected within 30 days before bowel preparation and immediately stored at −80°C. Genomic DNA was extracted from each sample using the Baypure Magnetic Stool Nucleic Acid Kit (Baybio Biotek Corp, Guangzhou, China). The detection reagent “M3CRC” (MicroSigx Biotech Diagnostic, Guangzhou, China) was used for simultaneous quantification of microbial gene markers and FIT analysis according to the manufacturer’s protocol. Quantitative polymerase chain reaction (qPCR) was performed for microbial quantification, and FIT was conducted on the DFIA300 fluorescence immunoanalyzer (Guangzhou Biotron Technology Co., Ltd., Guangzhou, China), with technicians blinded to the clinical diagnoses. The primary outcome was the ability of fecal microbial biomarkers combined with FIT to detect CRC and AA (adenomas with a diameter of ≥1 cm, with a tubulovillous or villous component, or with high-grade dysplasia). The secondary outcome was the performance of this combined model in identifying early-stage CRC (AJCC American Joint Committee on Cancer Stages I-II) and high-risk AA (defined by high-grade dysplasia). A prototypical STARD (Standards for Reporting of Diagnostic Accuracy Studies) diagram Supplementary Figure 1, https://links.lww.com/CM9/C784, detailed laboratory protocols, and statistical analysis methods are provided in the Supplementary Materials, https://links.lww.com/CM9/C784. A total of 821 participants were enrolled, comprising 285 patients with CRC, 70 with AA, and 466 controls. Baseline characteristics are summarized in Supplementary Table 1, https://links.lww.com/CM9/C784. The relative abundances of the fecal microbial biomarkers Fn, m3, Hh, and Chk differed significantly among these groups. All four biomarkers were significantly elevated in the CRC group compared to controls (P <0.001). Furthermore, the levels of m3 and Hh were also higher in patients with AA than in controls (P <0.05). Critically, a significant increasing trend in the abundance of these microbial markers was observed across the groups, progressing from controls to patients with AA, with the highest levels seen in patients with CRC (P <0.001, Figure 1A–D).Figure 1: (A–D) Relative abundances of the four microbial biomarkers in controls, patients with advanced adenomas (AAs), and colorectal cancer (CRC). (E–H) Receiver operating characteristic (ROC) curves of the combined biomarkers and the fecal immunochemical test (FIT) for detecting all CRC (E), early-stage CRC (F), and Stage I (G) and II (H) disease. (I–J) ROC curves for detecting any AA (I) and high-risk AA (J). Solid lines depict the original ROC curves; and shaded areas represent 95% confidence intervals (CI) from 1000 bootstrap replicates. AUC: Area under the curve.The diagnostic performance of individual microbial markers was assessed by receiver operating characteristic analysis. For CRC detection, Fn demonstrated the highest individual area under the curve (AUC) of 0.73 (95% confidence interval CI: 0.70–0.77), followed by m3 (AUC: 0.63, 95% CI: 0.59–0.67) and Hh (AUC: 0.63, 95% CI: 0.59–0.67), and Chk (AUC: 0.57, 95% CI: 0.53–0.60). For AA detection, m3 and Hh achieved an AUC of 0.59 (95% CI: 0.52–0.66) and 0.59 (95% CI: 0.51–0.66), respectively Supplementary Figure 2, https://links.lww.com/CM9/C784. Given the moderate performance of individual markers, we developed a combined model integrating all four biomarkers with FIT. This combined model demonstrated superior diagnostic performance. For overall CRC detection, it achieved an AUC of 0.96 (95% CI: 0.94–0.97) with 88.4% sensitivity and 90.1% specificity. Bootstrap validation confirmed the model’s robustness (corrected AUC: 0.96, 95% CI: 0.94–0.97, Figure 1E). Performance remained excellent for early-stage disease, with an AUC of 0.97 (95% CI: 0.95–0.98, Figure 1F) for Stage I–II CRC. Notably, Stage I CRC was identified with 96.2% sensitivity (AUC: 0.98; 95% CI: 0.96–1.00, Figure 1G), while Stage II CRC was detected with 89.5% sensitivity (AUC: 0.97; 95% CI: 0.95–0.98, Figure 1H), maintaining consistent specificity (90.1%). All bootstrap-corrected AUC values consistently demonstrated the model’s robust discriminatory ability across different disease stages. Similarly, the combined model showed significant diagnostic value for AA, achieving an AUC of 0.83 (95% CI: 0.80–0.86) with 64.3% sensitivity and 90.1% specificity. Bootstrap correction further validated the model’s robustness (corrected AUC: 0.83, 95% CI: 0.76–0.88, Figure 1I). Moreover, its performance was further enhanced in high-risk AA, with an AUC of 0.87 (95% CI: 0.84–0.90), a sensitivity of 71.7%, and a specificity of 90.1%. Bootstrap analysis confirmed its discriminatory power (corrected AUC: 0.88, 95% CI: 0.78–0.92, Figure 1J). Previous studies have implicated Fn, m3, Hh, and Chk in colorectal carcinogenesis.4,5 Our prospective multicenter study validates a combined model of these four fecal microbial biomarkers and FIT that achieves high diagnostic accuracy for CRC and AA, particularly for early-stage CRC (AUC 0.97) and high-risk AA (AUC 0.87). This performance represents a notable improvement over existing non invasive methods, which have limited sensitivity for detecting precancerous lesions Supplementary Table 2, https://links.lww.com/CM9/C784. This study has several strengths and limitations. The prospective multicenter design and the model’s enhanced detection of early-stage CRC and AA are key strengths. However, the study is limited by the sample size in Stage I CRC and high-risk AA groups, the single-ethnicity cohort, the lack of external validation, and the inability to rule out potential confounding effects from dietary habits and health supplements on the gut microbiota. Future research should therefore focus on validating the model in broader, multiethnic populations with controlled potential confounding effects. Collectively, the integration of fecal microbial markers with FIT represents a promising, non invasive strategy for improving the detection of both CRC and its precancerous lesions, with clear potential to enhance early intervention and reduce disease-related mortality. Funding None. Conflicts of interest None.
Shi et al. (Fri,) studied this question.