In earlier-stage melanoma, it can be more difficult to collect an adequate sample, Dr. Salama says. She and her colleagues at Duke conducted a feasibility study of serial ctDNA monitoring in 30 patients with stage II/III melanoma (Rhodin KE, et al. Ann Surg Oncol. 2025;32[7]:5292–5299). The assay was generated successfully for all participants; tumor from the initial biopsy or surgical pathology was sent to Natera for creation of a tumor-informed, multiplex PCR next-generation sequencing assay that was then used to create multiplex PCR primers used to track ctDNA in plasma samples. “We did demonstrate it was feasible,” she says. “But that hasn’t been the case in every study.”
Of 13 patients who had disease recurrence, eight had detectable ctDNA levels predating their clinical or radiographic occurrence. These elevations were highly specific for recurrence, with a specificity approaching 95 percent, a finding the authors noted was consistent with other studies. Sensitivity, however, was only 61 percent. Though one study of stage III and IV melanoma reported a higher sensitivity (Rhodin, et al., noted their own study’s lower sensitivity is likely owing to their inclusion of stage IIB/C patients), “Overall, ctDNA elevations or lack thereof do not appear to be very sensitive for ruling out recurrence.” Says Dr. Salama, “A positive test is probably clinically meaningful. A negative test can’t take the place of routine monitoring.”
In the Duke study, elevated ctDNA at the first postoperative timepoint was associated with worse recurrence-free survival. Most patients, however, are negative immediately after resection. “Certainly less than 20 percent of patients are positive,” Dr. Salama says, and they tend to be the higher-risk patients at baseline by disease stage. “Even though a meaningful proportion will ultimately demonstrate unfavorable kinetics and convert to positive, we can’t use that to make a decision about therapy,” she says.
What action might a positive ctDNA call for?
“Perhaps more intensive scan surveillance,” she says. “We know that person is now at increased risk.” And the data suggest ctDNA positivity can precede a positive scan by a few months, in some cases. “If they were due for a six-month scan, maybe we can get them a two- or a three-month scan,” she says.
Not all recurrences are preceded by a positive ctDNA, however. “One study noted a patient developed brain metastases and was ctDNA-negative,” Dr. Salama says. “That’s important because it says we can’t necessarily replace traditional imaging, especially in patients who are considered high risk by traditional clinicopathologic criteria.”
What investigations are happening now, and which should happen?
“One big question is, do we treat at the time of ctDNA positivity?” Dr. Salama asks. “Traditionally, the answer has been no, as we worry that we may just be exposing patients to the potential downsides of treatment—side effects—for longer, and not necessarily extending survival.” One possibility, though, would be to forgo treatment after neoadjuvant therapy and monitor with ctDNA, beginning therapy if the patient converts to a positive ctDNA. In the metastatic population, where most patients are ctDNA-positive, there’s the possibility of treating based on ctDNA levels rather than traditional radiographic requirements. Early discontinuation of therapy, too, would be important to investigate in the metastatic population. “If they become negative, can you stop therapy?” Some studies have investigated whether therapy can be discontinued early based on PET imaging, but ctDNA might provide an even earlier indication that treatment can be stopped, she says.