he latest findings for circulating tumor DNA (ctDNA) in colorectal cancer were discussed at the 2024 ASCO Gastrointestinal (GI) Cancers Symposium by Aparna R. Parikh, MD, a gastrointestinal oncologist at Massachusetts General Hospital Cancer Center and Assistant Professor of Medicine at Harvard Medical School, Boston. She concluded from the latest data, “We certainly have many answers, but I will make the argument that we have many more questions that are still left to answer.”
She added: “Circulating tumor DNA as a biomarker in oncology is rapidly evolving…. In study after study, ctDNA has emerged as the most powerful prognostic biomarker to date, compared with any other clinical pathologic features.”
Aparna R. Parikh, MD
Clinical Implications
For ctDNA assays to be useful in the clinic, they need analytic validity (acceptable performance and reproducibility), clinical validity (correlation with a clinical endpoint), and clinical utility (improvement in clinical outcomes). Studies presented at the ASCO GI Cancers Symposium aimed to answer these questions, among them an update of the GALAXY trial, the first large observational study of this technology.
“What GALAXY demonstrated about clinical validity is that ctDNA positivity is undoubtedly prognostic,” Dr. Aparna said, but she questioned whether it has adequate sensitivity. At the 8- to 12-week window during which treatment decisions are made, the test had a sensitivity < 50% (at a specificity of 94%). Of concern, she said, “Postoperative ctDNA is only capturing 40% to 50% of recurrences in non–stage IV patients in multiple data sets.”
Although surveillance testing can improve sensitivity (to 73% in GALAXY, with specificity of 97%), “it is not yet clinically actionable outside of clinical trials,” she said. Duration of follow-up, stage breakdown, sampling frequency, response to therapy, distribution of metastases, and other factors can also influence sensitivity. More work needs to be done to improve the positive predictive value of these tests in early disease, she emphasized.
The study confirmed that sustained clearance of ctDNA over time heralded very good survival, but interestingly, for patients with only transient clearance, the 2-year outcomes were no better than for those with no clearance at all. Chemotherapy may be suppressing but not eliminating measurable residual disease in this subset of patients, she said.
Results were more disappointing for the COBRA trial, which failed to demonstrate ctDNA as predictive for a chemotherapy benefit. However, the study enrolled low-risk patients and involved an older LUNAR platform, which was a predecessor to the newer REVEAL assay with broader epigenomic coverage. These factors may have influenced the outcome.
Altogether, she concluded: “The first-generation tests are certainly promising, but I would argue they are just not sensitive enough” to be used to de-escalate treatment. Based on current data, and outside of clinical trials, Dr. Parikh said she considers dose escalation in patients who are ctDNA-positive but does not use ctDNA negativity to de-escalate treatment in stage III disease. “When monitoring patients for recurrence, shared decision-making—and explaining [to patients] what these tests can and cannot do—is critical.”
DISCLOSURE: Dr. Parikh has served as a consultant or advisor to AbbVie, Bayer, Biofidelity, Checkmate Pharmaceuticals, CVS, Delcath Systems, Foundation Medicine, Guardant Health, Illumina, Eli Lilly, SAGA Diagnostics, Scarce, Seagen, Taiho Oncology, Takeda, UpToDate, and Value Analytics Labs; has received institutional research funding from Array BioPharma, Bristol Myers Squibb, Daiichi Sankyo, Erasca, Genentech, Guardant Health, Eli Lilly, Mirati Therapeutics, Novartis, Plexxikon, PMV Pharma, and PureTech; has received reimbursement for travel, accommodations, or other expenses from Karkinos Healthcare; and has had other financial relationships with C2i Genomics, Parithera, and Xact Robotics.
