Tumor-Informed ctDNA Assay May Aid Surveillance Strategies in Merkel Cell Carcinoma

Use of circulating tumor DNA (ctDNA) testing demonstrated high baseline sensitivity and specificity prior to treatment, and the presence of ctDNA showed an association with risk of recurrence post-treatment in patients with Merkel cell carcinoma (MCC), according to findings from a study published in the Journal of Clinical Oncology.

In the discovery cohort, baseline sensitivity for detecting ctDNA in patients with clinically evident disease was 95% (95% CI, 87%-99%); its specificity for detecting ctDNA negativity in patients without clinically evident disease was 90% (95% CI, 82%-95%). When investigators raised the threshold for ctDNA positivity, minimal improvement in specificity was seen at the expense of a disproportionate loss in sensitivity, so the original threshold for ctDNA positivity was kept for validation.

In the validation cohort, baseline sensitivity and specificity for detecting ctDNA in patients with clinically evident disease was 94% (95% CI, 85%-98%) and 86% (95% CI, 77%-93%), respectively. Notably, patients on immunotherapy at the time of enrollment (n = 48) vs not (n = 264), respectively, demonstrated similar sensitivity (95% vs 94%; P >.99) and specificity (81% vs 89%; P =.33).

“ctDNA testing exhibited high prognostic accuracy in detecting MCC recurrence, suggesting its potential to reduce frequent surveillance imaging,” Tomoko Akaike, MD, of the University of Washington in Seattle, and colleagues, wrote in the publication. “ctDNA also identifies high-risk patients who need more frequent imaging and may be best suited for adjuvant therapy trials.”

MCC is an aggressive neuroendocrine skin cancer that has a high mortality rate and 5-year recurrence rate of 40%. Serial full body imaging using CT or PET-CT is performed to monitor for recurrence during the first 5 years following diagnosis given the absence of an effective biomarker that can be used to stratify patient risk.

Due to the accumulating data to support the predictive and prognostic capabilities of ctDNA in the blood, including small case series in MCC, investigators sought to evaluate the utility of tumor-informed ctDNA to detect disease in patients with MCC.

The prospective, multicenter, observational study enrolled patients with histologically confirmed stage I to IV MCC between April 2020 and August 2022. The study included discovery and validation cohorts, which had data cutoff dates of July 8, 2022, and August 31, 2022, respectively.

Patients were eligible for enrollment at any point during their disease course. Blood samples were collected for ctDNA testing via a tumor-informed ctDNA assay from Signatera at the time of enrollment and every 3 months during the surveillance period.

Patients from Stanford University and University of Washington were included in the discovery cohort (n = 167), and those from Dana-Farber Cancer Institute, Northwestern University, University of California San Francisco, and Moffitt Cancer Center were included in the validation cohort (n = 152). A total of 562 ctDNA tests were performed over a median follow-up of 295 days in the discovery cohort (median, 3 tests/patient; interquartile range [IQR], 2-5), and 640 ctDNA tests were performed over a median follow-up of 284 days in the validation cohort (median, 3 tests/patient; IQR, 2-4).

The primary objectives of the study were to determine ctDNA test sensitivity and specificity for disease status at enrollment; risk of recurrence stratified by ctDNA status according to serial ctDNA testing during surveillance; and positive predictive value and negative predictive value of the ctDNA test for predicting clinical recurrence at each time point during surveillance. Secondary objectives included correlating quantitative ctDNA level with primary tumor size; quantifying risk of recurrence at different levels of ctDNA positivity; and determining whether the detection of ctDNA after completing initial therapy can predict recurrence and risk stratify patients.

During surveillance, 119 and 96 patients with no clinical evidence of disease at the onset of this period in the discovery and validation cohorts, respectively, underwent serial ctDNA testing. The median interval between tests in the discovery and validation cohorts was 91 days (IQR, 77-107) and 83 days (IQR, 56-98), respectively. The risk of recurrence was significantly higher in patients who were ctDNA positive at any point during surveillance vs those who were ctDNA negative throughout in both the discovery (HR, 6.8; 95% CI, 2.9-16; P <.001) and validation (HR, 20; 95% CI, 8.3-50; P <.001) cohorts. The authors noted that the differences in recurrence in the ctDNA-positive and -negative populations remained significant after adjusting for stage, immunosuppression status, sex, and age in both the discovery (P <.001) and validation (P <.001) cohorts.

The positive predictive value at 1 year in the discovery and validation cohorts was 69% (95% CI, 32%-91%) and 94% (95% CI, 71%-100%), respectively. The negative predictive value was high 135 days after any negative ctDNA test, reaching 94% (95% CI, 90%-97%) and 93% (95% CI, 89%-97%) in the discovery and validation cohorts, respectively. High negative predictive values were also seen in the discovery and validation cohorts, respectively, at 180 days, at 90% (95% CI, 85%-94%) and 91% (95% CI, 85%-96%).

During surveillance, 146 positive ctDNA tests from 61 patients had at least 90 days of follow-up after the positive test or were within 90 days of a clinical recurrence. Quantitative ctDNA levels were significantly higher among the 79 positive tests that were associated with recurrence within 90 days of the positive test vs the 67 positive tests where no recurrence was recorded (median, 23; [IQR, 5-134] vs median 0.9; [IQR, 0.4-2.8] MTM/mL; P <.001). The corresponding area under the curve was 0.86 (95% CI, 0.80-0.92).

The estimated risk of clinical recurrence detection within 90 days among the 146 total positive tests was 69% (n = 72/105) when ctDNA was above 1 MTM/mL and 17% (n = 7/41) when ctDNA was below 1 MTM/mL. The estimated risk was 100% (n = 23/23) when ctDNA was above 100 MTM/mL and 46% (n = 56/123) when ctDNA was positive and below 100 MTM/mL.

Prognosis according to post-treatment ctDNA status was also evaluated in patients with stage I to III disease. Only those who underwent a ctDNA test within 4 months after curative-intent therapy were included. Among 84 total patients, there were 23 recurrences over a median follow-up of 314 days. Significantly higher 1-year recurrence rates were seen among patients with ctDNA positivity (n = 14) post-treatment vs those with ctDNA negativity (n = 70; 74% vs 21%, respectively; HR, 7.6; 95% CI, 3.0-19; P < .001). This difference in recurrence was upheld after adjusting for stage, immunosuppression status, sex, and age (HR, 7.4; 95% CI, 2.7-20; P < .001).

“In this multicenter, prospective, observational study of patients with stage I to IV MCC, we formally validated the utility of a tumor-informed ctDNA assay,” the authors concluded. “This assay may be particularly impactful in the surveillance of patients with this highly lethal malignancy characterized by a high recurrence rate of 40% within 5 years.”

Reference

1. Akaike T, Thakuria M, Silk AW, et al. Circulating tumor DNA assays detects Merkel cell carcinoma recurrence, disease progression, and minimal residual disease: surveillance and prognostic implications. J Clin Oncol. Published online July 25, 2024. doi:10.1200/JCO.23.02054