170 Background: The effectiveness of screening using either fecal immunochemical testing (FIT) or colonoscopy in reducing colorectal cancer (CRC) mortality has been demonstrated in previous studies. Nevertheless, the single test sensitivity of FIT for relevant precursor lesions, especially high-risk neoplasms like invasive cancer or advanced adenomas (AA), is a concern. Improving the sensitivity for high-risk neoplasm is critical to improving the effectiveness of screening in reducing CRC incidence and mortality. Dysregulation of histone post-translational modifications (PTMs) has been linked with several cancers including CRC. We investigated the clinical performance of circulating nucleosome levels containing specific histone PTMs in blood, in combination with FIT from symptomatic subjects referred to colonoscopy to evaluate their discriminant power towards CRC and AA. Methods: 10mls whole blood was obtained prior to colonoscopy from 476 patients referred to colonoscopy in National Taiwan University Hospital for surveillance or secondary to bowel symptoms. Patients were classified into 5 groups based on their colonoscopy reports: (i) patient with CRC (n = 67), (ii) patients with AA (n = 60, including 22 with AA≥2cm); (iii) patients with non-AA (n = 123); (iv) patients with non-neoplastic polyps (n = 29); (vi) healthy controls with no endoscopic lesion (n = 197). Plasma samples were analyzed for nucleosome levels using 7 different quantitative immunoassays (Nu.Q assays; Belgian Volition SRL, Belgium) targeting H3.1-nucleosomes or different histone modifications (H3K27Me3-, H3K36Me3-, H3K9Me3-, H3K14Ac-, H3K27Ac- or H3R8Cit-nucleosomes). All study subjects were asked to collect a stool sample at home within 48-hours before the colonoscopy. The fecal samples were assayed with the OC-SENSOR FIT kit (Eiken Chemical Co., Ltd., Tokyo, Japan) for all subjects. Results: At a cut-off of 20µg/g feces, the FIT test showed a sensitivity of 83.5% at 82.1% specificity for CRC +AA vs. Controls. All the CRC patients were detected but 35% of AA were missed, and 7 of which were high-risk adenomas (AA≥2cm). A combination of 2 Nu.Q biomarkers: H3K36Me3 and H3K9Me3 with FIT in a decision tree model showed an improved sensitivity of 98.4% allowing the detection of all CRC patients and 97% of the patients with AA including all high-risk adenomas. The 2 missed AA were one AA below 1cm and one between 1 and 2cm. Unnecessary colonoscopies could be reduced by 28.9% of the control and 20.7% of the non-neoplastic polyps subgroups as both are found negative with this panel of assays. Conclusions: At present, all the symptomatic patients are sent to colonoscopy in clinical practice. Our results indicate that H3K36Me3- and H3K9Me3-nucleosome levels, in combination with FIT in a decision tree model, could detect all CRC patients and all high-risk adenomas and help reduce unnecessary colonoscopies.
Read full abstract