Site-Specific Variations in Circulating Tumor DNA Levels and Kinetics in Metastatic Colorectal Cancer
Varun V. Bansal*1, Ankit Dhiman2, Hunter D. Witmer3, Frederick Godley3, Shen Li3, Erika Belmont4, Cecilia T. Ong3, Blase Polite4, Oliver Eng5, Ardaman Shergill4, Kiran K. Turaga1
1Surgical Oncology, Yale University School of Medicine, New Haven, CT; 2Surgery, Medical College of Georgia, Augusta, GA; 3Surgery, University of Chicago, Chicago, IL; 4Medicine, University of Chicago, Chicago, IL; 5Surgery, University of California, Irvine, Orange county, CA
Objective: Plasma circulating tumor DNA (ctDNA) is a promising biomarker for monitoring treatment response and recurrence after surgery for metastatic colorectal cancer (mCRC); however, previous work from our group demonstrated poor sensitivity in for peritoneal-only metastases. This study aimed to investigate metastatic site-specific variations in ctDNA levels and the prognostic significance of postoperative ctDNA kinetics.
Design: Retrospective cohort study evaluating ctDNA test characteristics in the presence of confirmed mCRC (radiographic or laparoscopic).
Setting: Single academic comprehensive cancer center.
Patients: Patients with mCRC managed between 2020-22 were classified as having isolated peritoneal metastases (PM), or distant lymph node or visceral disease.
Interventions: Personalized and tumor-informed ctDNA testing (SignateraTM bespoke multiplex PCR Next Generation Sequencing) during clinical care. Disease status at the time of highest ctDNA level during treatment was noted cross-sectionally.
Main outcome measures: Primary outcome measures included comparing median ctDNA levels across mCRC sites and assessing factors associated with elevated ctDNA levels (> 1 MTM/ml) using logistic regression. The impact of ctDNA growth velocity on recurrence-free survival (RFS) from surgery, represented by the log-fold change between serial ctDNA assays within and beyond three months from optimal cytoreductive surgery/liver-directed therapy was ascertained in cox regression models. The association of preoperative ctDNA positivity with postoperative RFS and its correlation with peritoneal disease burden (denoted by the peritoneal carcinomatosis index (PCI)) was also assessed.
Results: Of 73 mCRC patients undergoing ctDNA testing, 51 (70%) had an assay drawn in the presence of clinically confirmed disease. The median age was 53 years (IQR 43-64), and most patients were male (63%), non-Hispanic white (78%), had left-colon primaries (53%), and developed synchronous metastases (63%). Patients with isolated PM had lower ctDNA levels [n = 17, median 0.12 MTM/ml (IQR 0.00-0.97)] than those with lymph node [(n = 4, median 3.1 MTM/ml (0.43-8.1)] and visceral metastases [n = 30, median 35.18 MTM/ml (9.21-211.05)] (p < 0.001). Compared with nodal and visceral disease, isolated PM were less likely to be associated with ctDNA elevation [adjusted OR 0.04 (CI 0.01-0.21)]. Increasing ctDNA growth velocity was associated with worse RFS (adjusted HR = 4.07 (CI 1.47-11.29), p = 0.007) over a median follow-up duration of 14 months (IQR 10-17). Preoperative ctDNA positivity was associated with worse RFS (median 1.5 months (CI 1.4-1.8) vs not reached, p = 0.007) and correlated poorly with PCI (r = -0.342, p = 0.303).
Conclusion: Isolated CRC PM are associated with significantly lower ctDNA levels, which correlate poorly with peritoneal disease burden. Postoperative ctDNA kinetics and preoperative positivity have prognostic value and may help inform systemic therapy use.
Back to 2023 Abstracts