Implementation of Blood-Based Colorectal Cancer Screening: Real-World Clinical Experience

Introduction: Despite multiple colorectal cancer (CRC) screening options, screening adherence has plateaued well below the 80% goal set by leading health organizations, with more than one-third of eligible individuals not up to date. Implementation of a blood-based CRC screening test can enhance effectiveness of population-based screening programs by achieving increased adherence, especially in those unscreened or not up to date. We report on real-world adherence rates of a blood-based CRC screening test. A version of the test was recently validated in a large prospective CRC screening study (ECLIPSE, NCT04136002) and demonstrated 83% sensitivity and 90% specificity for CRC. Methods: The blood-based test (Shield) is a qualitative Laboratory Developed Test (LDT) validated in a CLIA/CAP-accredited laboratory (Guardant Health, CA) for CRC screening in average risk, guideline recommended individuals, age 45+. Results are returned as “normal signal detected” or “abnormal signal detected” and not intended to be the sole basis for a CRC diagnosis. Patients (pts) with an abnormal result should be referred for colonoscopy evaluation. Laboratory orders of the first 10,000 screening age-eligible pts were retrospectively reviewed. Test completion is defined as both clinical test order and blood sample received. A cross-sectional survey sent to ordering providers and staff (N=1,524) collected data on ordering behaviors (e.g., pts unscreened, not up to date) and acceptance rate (defined by practice behavior to continue clinical use of the test). Results: 9,584/10,000 pts completed the blood-based test, yielding a 96% adherence. 59% were female. Median age: 60 years (range: 45-99). 93% (N = 8,867) of tests were issued results. Reasons for results not reported include provider/pt factors (e.g., lack of insurance coverage, inaccurate sample labeling, subsequently deemed not at average risk) or laboratory related reasons (e.g., sample received was quantity not sufficient (QNS)). The cross-sectional survey completed by ordering providers and staff identified that 89% of respondents ordered the blood-based test for individuals never previously screened (65%) or not up to date with screening (24%). A binary-response model investigating practice behavior yielded a 95% provider acceptance rate with the blood-based test. Conclusion: Implementation of this blood-based CRC screening test in 10,000 pts yields an adherence rate (96%) that exceeds rates with existing options (< 67%).