Thomas Y. Lam

OBJECTIVE: To evaluate the potential for diagnosing colorectal cancer (CRC) from faecal metagenomes. DESIGN: We performed metagenome-wide association studies on faecal samples from 74 patients with CRC and 54 controls from China, and validated the results in 16 patients and 24 controls from Denmark. We further validated the biomarkers in two published cohorts from France and Austria. Finally, we employed targeted quantitative PCR (qPCR) assays to evaluate diagnostic potential of selected biomarkers in an independent Chinese cohort of 47 patients and 109 controls. RESULTS: Besides confirming known associations of Fusobacterium nucleatum and Peptostreptococcus stomatis with CRC, we found significant associations with several species, including Parvimonas micra and Solobacterium moorei. We identified 20 microbial gene markers that differentiated CRC and control microbiomes, and validated 4 markers in the Danish cohort. In the French and Austrian cohorts, these four genes distinguished CRC metagenomes from controls with areas under the receiver-operating curve (AUC) of 0.72 and 0.77, respectively. qPCR measurements of two of these genes accurately classified patients with CRC in the independent Chinese cohort with AUC=0.84 and OR of 23. These genes were enriched in early-stage (I-II) patient microbiomes, highlighting the potential for using faecal metagenomic biomarkers for early diagnosis of CRC. CONCLUSIONS: We present the first metagenomic profiling study of CRC faecal microbiomes to discover and validate microbial biomarkers in ethnically different cohorts, and to independently validate selected biomarkers using an affordable clinically relevant technology. Our study thus takes a step further towards affordable non-invasive early diagnostic biomarkers for CRC from faecal samples.

Objective There is a need for an improved biomarker for colorectal cancer (CRC) and advanced adenoma. We evaluated faecal microbial markers for clinical use in detecting CRC and advanced adenoma. Design We measured relative abundance of Fusobacterium nucleatum ( Fn ), Peptostreptococcus anaerobius ( Pa ) and Parvimonas micra ( Pm ) by quantitative PCR in 309 subjects, including 104 patients with CRC, 103 patients with advanced adenoma and 102 controls. We evaluated the diagnostic performance of these biomarkers with respect to faecal immunochemical test (FIT), and validated the results in an independent cohort of 181 subjects. Results The abundance was higher for all three individual markers in patients with CRC than controls (p<0.001), and for marker Fn in patients with advanced adenoma than controls (p=0.022). The marker Fn , when combined with FIT, showed superior sensitivity (92.3% vs 73.1%, p<0.001) and area under the receiver-operating characteristic curve (AUC) (0.95 vs 0.86, p<0.001) than stand-alone FIT in detecting CRC in the same patient cohort. This combined test also increased the sensitivity (38.6% vs 15.5%, p<0.001) and AUC (0.65 vs 0.57, p=0.007) for detecting advanced adenoma. The performance gain for both CRC and advanced adenoma was confirmed in the validation cohort (p=0.0014 and p=0.031, respectively). Conclusions This study identified marker Fn as a valuable marker to improve diagnostic performance of FIT, providing a complementary role to detect lesions missed by FIT alone. This simple approach may improve the clinical utility of the current FIT, and takes one step further towards a non-invasive, potentially more accurate and affordable diagnosis of advanced colorectal neoplasia.

OBJECTIVE: We aim to develop and validate a clinical scoring system to predict the risks of colorectal neoplasia to better inform screening participants and facilitate their screening test choice. DESIGN: We recruited 5220 Chinese asymptomatic screening participants who underwent colonoscopy in Hong Kong during 2008-2012. From random sampling of 2000 participants, independent risk factors were evaluated for colorectal neoplasia, defined as adenoma, advanced neoplasia, colorectal cancer or any combination thereof using binary regression analysis. The ORs for significant risk factors were used to develop a scoring system ranging from 0 to 6: 0-2 'average risk' (AR) and 3-6 'high risk' (HR). The other 3220 screening participants prospectively enrolled between 2008 and 2012 for screening colonoscopy formed an independent validation cohort. The performance of the scoring system for predicting colorectal neoplasia was evaluated. RESULTS: The prevalence of colorectal neoplasia in the derivation and validation cohorts was 31.4% and 30.8%, respectively. Using the scoring system developed, 78.9% and 21.1% in the validation cohort were classified as AR and HR, respectively. The prevalence of colorectal neoplasia in the AR and HR groups was 27.1% and 44.6%, respectively. The subjects in the HR group had 1.65-fold (95% CI 1.49 to 1.83) increased prevalence of colorectal neoplasia than the AR group. CONCLUSIONS: The scoring system based on age, gender, smoking, family history, Body Mass Index and self-reported diabetes is useful in predicting the risk of colorectal neoplasia.