Abstract

69 Background: Blood-based CRC screening has yielded performance comparable to stool-based screening tests and may be a modality that addresses the adherence gap in CRC screening. To investigate the diagnostic performance and its clinical relevance, we evaluated the test performance of a blood-based CRC screening test in a large cohort of individuals enriched for advanced colorectal neoplasia and early CRCs. Methods: 501 individuals presenting to ten institutions in Japan from 2020 to 2021 for treatment of advanced adenomas (AAs) or CRCs were consented and provided a pre-treatment blood sample (40mL whole blood in Streck cfDNA blood collection tubes). All eligible individuals underwent surgery or endoscopic resection within 30 days after sample collection and the pathological specimens were reviewed to determine cancer histology and stage, as well as AA histology. Pre-treatment blood samples were analyzed with a multi-modal (genomic and epigenomic) cfDNA CRC detection assay (Shield, Guardant Health, USA). Final cfDNA results were correlated with clinical data to determine the association of clinicopathological characteristics with test sensitivity. This study is registered as a clinical trial under UMIN000037765. Results: 451/501 (90%) individuals met inclusion/exclusion criteria and provided adequate blood samples that passed testing quality control criteria. Median age at consent was 67.4 years (range 31 – 92). 44% were female. 214/451 (47%) had a prior positive fecal occult blood test (FOBT). Of individuals with a negative or unknown FOBT testing, 161/237 (68%) had GI symptoms (e.g. abdominal pain, bloody/narrowing stool, constipation/diarrhea). The test identified 88% (268/305) of CRCs, including 71% (79/111) Stage I, 98% (87/89) Stage II, 97% (90/93) Stage III, and 100% (12/12) Stage IV. Overall sensitivity in detecting Stage I-III was 87%. The test identified 37% (54/146) of AAs, including 49% (33/68) of carcinoma in situs and 24% (9/37) of AAs with high-grade dysplasia. Sensitivity for right-sided CRC was 84% (61/73), 87% (59/68) for left-sided CRC, and 90% (148/164) for rectal tumors. Stage I test sensitivity was significantly associated with tumor size (>20mm; p<0.01), left-sided or rectal tumor location (p<0.05), and tumor depth (p<0.01) in univariate analysis, whereas the tumor depth (T2) was the sole factor associated with test positivity (p<0.05) in multivariate analysis. Specificity on this assay is 90% as determined in the ECLIPSE clinical study (NCT04136002). Conclusions: This study evaluates the performance of a blood-based test for CRC and its clinicopathological relevance in a large-scale cohort. In this population enriched for early colorectal neoplasia, the sensitivity was 88% for CRC and 37% for AAs. This performance is consistent with previous reports. Clinical trial information: UMIN000037765 .

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call