Skip to main content
Premium Trial:

Request an Annual Quote

Circulating Tumor DNA Utility in Post-Surgery Stage II Colon Cancer Cemented by New Data


NEW YORK – Years after the first hints that liquid biopsy could guide adjuvant treatment decisions in early-stage colorectal cancer, some of the field's leading researchers have shared new, persuasive clinical utility data supporting this molecular residual disease detection as a tool to direct the use of post-surgery chemotherapy to those who will actually benefit, and away from those likely to be cured by surgery alone.

The group's primary investigator, Jeanne Tie, an oncologist at the University of Melbourne's Peter MacCallum Cancer Centre, presented the findings this week at the American Society of Clinical Oncology's annual meeting, with a publication of the full study appearing simultaneously in the New England Journal of Medicine.

Among other things, the study, called DYNAMIC, demonstrated the non-inferiority of circulating tumor DNA (ctDNA)-guided treatment, with a clear de-escalation benefit. In the ctDNA-guided arm, use of adjuvant chemotherapy was nearly halved compared to the standard treatment arm, without compromising average recurrence-free survival.

According to Tie, the three-year recurrence-free survival seen in ctDNA-positive patients treated with adjuvant chemotherapy also suggests that this well-defined high-risk subgroup does indeed benefit. And although the recurrence risk is clearly not zero for ctDNA-negative patients, the results did confirm a very low risk in untreated ctDNA-negative patients, most notably in those also classified as clinically low risk.

"The stage II colon cancer space has really been fraught with challenges and difficulties of who should get and who will benefit from adjuvant chemotherapy, and the DYNAMIC trial really begins to assess this [and] pushes this to the next step of clinical utility," Vanderbilt-Ingram Cancer Center researcher Ben Ho Park, who was not involved in the study said discussing the results.

While the role of adjuvant chemotherapy is well established for stage III CRC patients, the management of stage II colon cancers after curative intent surgery has continued to be a "clinical predicament," Tie said.

Because surgery alone can cure most patients and the survival benefit isn't as clear as in stage III disease, current guidelines suggest oncologists consider adjuvant chemotherapy for patients with high-risk features. But, the definition of these features varies between guidelines, and not all elements carry the same prognostic weight.

Past observational studies by Tie and others in recent years have confirmed that patients with lingering tumor DNA in their blood after surgery (minimal residual disease, or MRD) have a very high recurrence risk. This would suggest that they stand to benefit from adjuvant chemotherapy, while those with no signs of tumor DNA can avoid unnecessary overtreatment.

But so far this hypothesis hadn't been comprehensively tested in a prospective, randomized setting.

Tie and colleagues' DYNAMIC trial recruited and randomized a total of 455 stage II CRC patients across 23 Australian centers. Two-thirds went to the ctDNA-guided therapy arm, where investigators collected plasma samples and performed MRD testing at week four and seven after surgery.

Any patient with a positive result at either week four or week seven was treated with chemo, the specific agent being at clinicians' discretion. Those with a negative ctDNA result at both time points received no adjuvant treatment.

For the remaining third of patients in the standard management arm, adjuvant treatment decisions were based on conventional clinical pathological criteria, and researchers compared outcomes, with the primary endpoint being two-year recurrence-free survival.

Tie reported that ctDNA analysis was successful in 99 percent of cases, with only three individuals unable to be tested due to no mutation being found in their tumor tissue. Like many commercial companies now vying for a position in MRD testing, the DYNAMIC study employed a tumor-informed, patient-specific methodology, in which tumor tissue sequencing was used to inform personalized ctDNA assays for each subject. In this case, testing was performed at Johns Hopkins University using Safe-seq, a barcode-based error reduction method designed to allow detection of very low-frequency mutations.

The study authors reported that the ctDNA-guided approach resulted in significantly fewer patients receiving adjuvant chemotherapy — 15 percent, compared to 28 percent in the standard management arm.

"Perhaps not unexpectedly, among those who received adjuvant chemotherapy, oxaliplatin-based doublet chemotherapy was given to a much higher proportion of patients in the MRD arm, at 62 percent compared to 10 percent in the standard management arm," Tis said.

Meanwhile, recurrence-free survival rates were similar in both arms. At two-years post-surgery, 93.5 percent of the MRD arm remained disease-free, compared to 92.4 percent of the standard management group.

"If I go back to the clinic tomorrow and this test is actually available, I will be using it to add to my tool of prognostication to guide adjuvant therapy," Tie said during a discussion of the results.

"Based on the data that we have, I certainly think that it now has utility, particularly in avoiding chemotherapy in those with clinically low-risk disease and also to pick patients who actually need adjuvant chemotherapy," she added. "Sometimes we have patients with low-risk disease that we would not be offering adjuvant chemotherapy, [and] in our study 10 percent of them … have detectable ctDNA. Those are the patients where this would make a difference in outcome."

Discussing the study at the ASCO meeting, Park noted that while the overall non-inferiority of MRD-guided care was confirmed in the study, "ctDNA may still miss patients who are, in fact, going to relapse."

However, he said, "what I was really heartened to hear and really got a lot of people thinking … is [whether] we can combine this with other risk markers like clinical risk, in which case the negative predictive value becomes acceptable."