Parvimonas micra as a putative non-invasive faecal biomarker for colorectal cancer

Abstract

The use of faecal microbial markers as non-invasive biomarkers for colorectal cancer (CRC) has been suggested, but not fully elucidated. Here, we have evaluated the importance of Parvimonas micra as a potential non-invasive faecal biomarker in CRC and its relation to other microbial biomarkers. The levels of P. micra, F. nucleatum and clbA + bacteria were quantified using qPCR in faecal samples from a population-based cohort of patients undergoing colonoscopy due to symptoms from the large bowel. The study included 38 CRC patients, 128 patients with dysplasia and 63 controls. The results were validated in a second consecutive CRC cohort including faecal samples from 238 CRC patients and 94 controls. We found significantly higher levels of P. micra in faecal samples from CRC patients compared to controls. A test for P. micra could detect CRC with a specificity of 87.3% and a sensitivity of 60.5%. In addition, we found that combining P. micra with other microbial markers, could further enhance test sensitivity. Our findings support the potential use of P. micra as a non-invasive biomarker for CRC. Together with other microbial faecal markers, P. micra may identify patients with "high risk" microbial patterns, indicating increased risk and incidence of cancer.

Figure 1

Increased levels of specific microbial markers are detected in faeces of CRC patients. Scatter plots are used to illustrate the relative levels of (A) P. micra (Pm), (B) F. nucleatum (Fn), and (C) clbA + bacteria (clbA) in faeces of control patients, and patients diagnosed with dysplasia or CRC from the FECSU and U-CAN cohorts. Horizontal lines indicate mean relative expression calculated by the 2^-ΔCq method with the total microbial 16S rRNA gene DNA as reference.

Figure 2

ROC curves displaying the specificity and the sensitivity for P. micra (Pm), F. nucleatum (Fn), and clbA + bacteria (clbA) to detect CRC. ROC-curves were calculated using the levels for the specific marker as indicated and cancer/no cancer. The levels of a specific marker in each sample was given as a relative quantification calculated by the 2^-ΔCt method with the total microbial 16S rRNA gene DNA as reference.

Figure 3

Performance of single faecal microbial markers or combinations of markers in CRC detection. Sensitivity and specificity for CRC detection is displayed for a test of (A) P. micra (Pm), F. nucleatum (Fn), or clbA + bacteria (clbA), as well as combined tests using several microbial markers for the FECSU and U-CAN cohort, and (B) for combined tests using microbial markers and immunochemical F-Hb (Hb) for the FECSU cohort. For test 1, a positive test result was given to samples with at least one positive marker. For test 2, a positive test result was given to samples with at least two positive markers.

Figure 4

The distribution of specific microbial markers in faeces of CRC patients. Circle diagrams are used to illustrate the abundance of P. micra (Pm), F. nucleatum (Fn) and clbA + bacteria (clbA) in faecal samples with all three markers evaluated of control patients, and patients diagnosed with dysplasia or CRC from the FECSU and U-CAN cohorts.