CHICAGO – Circulating tumor DNA (ctDNA), or DNA shed from tumors that is detected in the bloodstream, has shown increasing promise as a prognostic tool in gastrointestinal cancers, allowing investigators to make real-time assessments of treatment response and the likelihood of recurrence.
Depending on the type of assay and analysis used, ctDNA can provide a wealth of information about cancer genetic variants. ctDNA assays can be used for primary screening, to track tumor burden, or to detect minimal residual disease (MRD) after cancer surgery.
However, ctDNA’s role in guiding clinical decisions is still being defined.
The same group also presented exploratory findings showing that positive ctDNA is a significant predictor of recurrence in people with early-stage pancreatic cancer following surgery. However, the investigators concluded, ctDNA status should not be used to inform treatment decisions concerning duration of adjuvant chemotherapy in these patients.
DYNAMIC Trial Results
Jeanne Tie, MD, of the Peter MacCallum Cancer Centre in Melbourne, presented 5-year survival results at ASCO from the DYNAMIC randomized controlled trial, whose 2-year findings had already shown ctDNA to be helpful in stratifying stage II colon cancer patients for adjuvant chemotherapy or no treatment.
Because surgery is curative in 80% of these patients, it is important to identify the minority that will need chemotherapy, Dr. Tie said.
At 5 years’ follow-up, Dr. Tie reported, patients randomized to a ctDNA-guided approach (negative ctDNA post surgery resulted in no treatment, and positive ctDNA led to adjuvant chemotherapy) did not see differences in overall survival compared with conventionally managed patients, who received chemotherapy at the clinician’s discretion.
Among ctDNA-guided patients in the study (n = 302), 5-year overall survival was 93.8%. For conventionally managed patients (n = 153), overall survival was 93.3% at 5 years (hazard ratio [HR], 1.05; 95% CI, 0.47-2.37; P = .887).
Further, the researchers found that a high ctDNA clearance rate was achieved with adjuvant chemotherapy in postoperative patients who were ctDNA positive. And 5-year recurrence rates were markedly lower in patients who achieved ctDNA clearance, compared with those who did not: 85.2% vs 20% (HR, 15.4; 95% CI, 3.91-61.0; P < .001).
“This approach of only treating patients with a positive ctDNA achieved excellent survival outcomes, including in patients with T4 disease. A high ctDNA clearance rate can be achieved with adjuvant chemotherapy, and this in turn was associated with favorable outcomes,” Dr. Tie said during the meeting. “And finally, the precision of the ctDNA approach may be further refined by increasing [the number of genetic variants] tracked and by incorporating ctDNA molecular burden. However, these findings will require further validation.”
DYNAMIC-Pancreas Study Results
In a separate presentation during the same session, Belinda Lee, MD, also of the Peter MacCallum Cancer Centre, showed results from the DYNAMIC-Pancreas study, which looked at ctDNA testing a median 5 weeks after surgery in 102 people with early-stage (Eastern Cooperative Oncology Group 0-1) pancreatic cancer. Patients who were ctDNA positive received 6 months of adjuvant chemotherapy of the physician’s choice (FOLFIRINOX or gemcitabine/capecitabine) while those who were ctDNA negative after surgery had the option to de-escalate to 3 months of chemotherapy treatment at the physician’s discretion.
At a median 3 years’ follow-up, Dr. Lee and colleagues found that the median recurrence-free survival was 13 months for patients who were ctDNA positive after surgery and 22 months for those who were ctDNA negative (HR, 0.52; P = .003), showing that positive ctDNA is prognostic of earlier recurrence independent of other factors.
Dr. Lee said that, given the high recurrence risk also seen in ctDNA-negative patients, reducing duration of chemotherapy was not recommended based on ctDNA-negative status.
In an interview, Stacey Cohen, MD, of Fred Hutch Cancer Center in Seattle, Washington, the discussant on the two presentations at ASCO, said that, until these results are further validated in stage II colon cancer patients,t it is unlikely that they will change clinical practice guidelines.
“They did an amazing job,” Dr. Cohen said of the researchers. “They’re at the forefront of the field of actually doing prospective analysis. And yet there are still some gaps that are missing in our understanding.”
The assays used in both studies, Dr. Cohen noted, are used only in research and are not available commercially in the United States. That, plus the fact that physicians were allowed to choose between chemotherapy regimens, made it harder to parse the results.
“Provider choice increases bias,” Dr. Cohen said. “And I think that’s the problem of having two chemo regimens to choose from, or in the case of the colon cancer trial, not selecting whether patients got a single chemotherapy agent or a doublet. These are pretty big differences.”
But the field is moving quickly, “and it is an exciting time to improve patient selection for chemotherapy treatment,” she continued.
Allowing physicians to choose chemotherapy regimens reflected real-world clinical practice, “especially given that this study is designed to test a strategy rather than a specific treatment, said Dr. Tie in an interview. “More work will need to be done to specifically address the question of which chemotherapy regimen is more effective to treat ctDNA-positive disease.”
Dr. Cohen noted that, while evidence is mounting to support the value of ctDNA in colon cancer, there is far less evidence for pancreatic cancer.
Dr. Lee and colleagues’ study “adds to the literature, and I think what it teaches us is that ctDNA remains a prognostic risk factor,” she said. “But we saw that even patients who are negative have a high recurrence risk. So we’re not ready to act on it yet. As with the colon cancer study, different chemotherapy regimens were used, and for different time lengths.”
Whether in colon cancer or pancreatic cancer, ctDNA results, “are highly tied to which assay you’re using and which scenario you’re testing them in,” Dr. Cohen said.
Dr. Tie and colleagues’ study was sponsored by her institution, with additional funding received from the Australian government, the National Institutes of Health, and other foundations. She disclosed speaking and/or consulting fees from Haystack Oncology, Amgen, Novartis, Bristol-Myers Squibb, Merck, AstraZeneca, and others. Dr. Lee’s study was sponsored by the Marcus Foundation. She disclosed receiving honoraria from Roche. Dr. Cohen reported no conflicts of interest.