Abstract 583P
Background
The detection of colorectal cancer (CRC) in early stages can enable the removal of pre-cancerous lesions and allow earlier treatment. Most CRCs develop from adenomas, with “advanced” adenoma (AA) considered to be clinically relevant precursor of CRC. Screening for CRC is recommended for average-risk adults using faecal immunochemical testing (FIT), but the performance of FIT is limited. FIT reportedly has false positive rate of ∼93%, and ∼10% of CRC diagnoses have a negative FIT result. Moreover, the sensitivity for detecting AA is only 24%, highlighting the need for alternative strategies.
Methods
The Dxcover® Colorectal Cancer Liquid Biopsy is a spectroscopic test that interrogates a blood sample with infrared light and generates a distinctive signature that represents the whole biomolecular profile of the sample. Patient samples (n=296) were retrospectively collected for this discovery study, and the full cohort consisted of 100 CRC, 99 adenoma (A) and 97 colonoscopy screening controls with a non-cancer (NC) diagnosis. Machine learning algorithms have been developed with the resultant spectral profiles to predict CRC and AA. Feature importance analysis has also been carried out to identify the spectral regions found to be most significant for each classification.
Results
The CRC v NC model achieved an area under the receiver operating characteristic curve (AUROC) of 0.93. At 90% specificity, the overall CRC sensitivity was 80%, and the detection rates were consistent across stages: 83% (I), 73% (II), 76% (III), and 100% (IV). The CRC+A v NC model reported an AUROC of 0.88, and at 90% specificity, the overall CRC sensitivity was 78%. The test successfully detected 83% of stage I, 73% of stage II, 76% of stage III and 87% of stage IV CRC. Furthermore, 59% of AA patients were identified with this classifier, which is greater than the current performance levels of FIT.
Conclusions
A rapid blood test that is sensitive to AA and early-stage CRC could substantially improve patient outcomes. This liquid biopsy can be fine-tuned to give a different balance of sensitivity and specificity depending on international healthcare markets. Current screening programmes have addressable shortcomings, and the emergence of new technologies is essential to support earlier CRC detection.
Clinical trial identification
Editorial acknowledgement
Legal entity responsible for the study
Dxcover Limited.
Funding
Dxcover Limited.
Disclosure
J.M. Cameron, A. Sala, G. Antoniou, J.J.A. Conn, R.G. McHardy: Financial Interests, Personal, Full or part-time Employment: Dxcover Limited. D.S. Palmer: Financial Interests, Personal, Advisory Role: Dxcover Limited; Financial Interests, Personal, Full or part-time Employment: Dxcover Limited; Financial Interests, Personal, Leadership Role: Dxcover Limited; Financial Interests, Personal, Research Funding, Dxcover, GSK, Endophotonics: Dxcover Limited; Financial Interests, Personal, Stocks or ownership: Dxcover Limited. M.J. Baker: Financial Interests, Personal, Full or part-time Employment: Dxcover Limited; Financial Interests, Personal, Leadership Role: Dxcover Limited; Financial Interests, Personal, Stocks or ownership: Dxcover Limited. All other authors have declared no conflicts of interest.
Resources from the same session
548P - Diffuse midline gliomas H3K27-altered in adults, the experience of an oncology center
Presenter: Carolina Trabulo
Session: Poster session 10
561P - Risk of colorectal cancer and premalignant lesions after kidney transplantation
Presenter: Koen Zwart
Session: Poster session 10
562P - Quantifying the effect of colonoscopy and polypectomy rates in stage shift of colorectal cancer and reduced need for adjuvant chemotherapy
Presenter: Mina Sarofim
Session: Poster session 10
563P - The impact of participation in colorectal cancer screening: A population-based results from Lithuania
Presenter: Tomas Poskus
Session: Poster session 10
564P - A large-scale real-world study for colorectal cancer screening
Presenter: Song LIU
Session: Poster session 10
565P - Real-world data of early-onset colon cancer (early-onset CC) from the AIO colopredict plus (CPP) registry: Clinical and molecular characteristics, treatment and outcome
Presenter: Doreen Gisder
Session: Poster session 10
566P - Long-term survivorship outcomes of adolescents and young adults (AYA) with cancer of the lower gastrointestinal tract
Presenter: Winette Van Der Graaf
Session: Poster session 10
567P - Impact of concomitant medication on recurrence, survival and tolerability of chemotherapy in early colon cancer patients: Results from the PETACC 8 study
Presenter: Elisabeth Bergen
Session: Poster session 10
568P - Single cell AI-based detection of DNA mismatch repair deficiency in 1,988 colorectal cancers reveals prognostic and predictive value in the SCOT trial
Presenter: Faiz Jabbar
Session: Poster session 10
569P - Artificial intelligence for detection of mismatch repair deficiency in colon carcinomas (alliance)
Presenter: Frank Sinicrope
Session: Poster session 10