Circulating Tumor DNA Assay Detects Merkel Cell Carcinoma Recurrence, Disease Progression, and Minimal Residual Disease: Surveillance and Prognostic Implications

Abstract

Purpose: Merkel cell carcinoma (MCC) is an aggressive skin cancer with a 40% recurrence rate, lacking effective prognostic biomarkers and surveillance methods. This prospective, multicenter, observational study aimed to evaluate circulating tumor DNA (ctDNA) as a biomarker for detecting MCC recurrence.

Methods: Plasma samples, clinical data, and imaging results were collected from 319 patients. A tumor-informed ctDNA assay was used for analysis. Patients were divided into discovery (167 patients) and validation (152 patients) cohorts. Diagnostic performance, including sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV), was assessed.

Results: ctDNA showed high sensitivity, 95% (discovery; 95% CI, 87 to 99) and 94% (validation; 95% CI, 85 to 98), for detecting disease at enrollment, with corresponding specificities of 90% (95% CI, 82 to 95) and 86% (95% CI, 77 to 93). A positive ctDNA during surveillance indicated increased recurrence risk, with hazard ratios (HRs) of 6.8 (discovery; 95% CI, 2.9 to 16) and 20 (validation; 95% CI, 8.3 to 50). The PPV for clinical recurrence at 1 year after a positive ctDNA test was 69% (discovery; 95% CI, 32 to 91) and 94% (validation; 95% CI, 71 to 100), respectively. The NPV at 135 days after a negative ctDNA test was 94% (discovery; 95% CI, 90 to 97) and 93% (validation; 95% CI, 89 to 97), respectively. Patients positive for ctDNA within 4 months after treatment had higher rates of recurrence, with 1-year rates of 74% versus 21% (adjusted HR, 7.4 [95% CI, 2.7 to 20]).

Conclusion: ctDNA testing exhibited high prognostic accuracy in detecting MCC recurrence, suggesting its potential to reduce frequent surveillance imaging. ctDNA also identifies high-risk patients who need more frequent imaging and may be best suited for adjuvant therapy trials.

FIG 1.

Flow diagram for patients with MCC enrolled and analyzed in the discovery and validation cohorts. Patients enrolled at University of Washington and Stanford University were designated the discovery cohort. Patients enrolled at Dana-Farber Cancer Institute, Northwestern Memorial Hospital, University of California San Francisco, and Moffitt Cancer Center were designated the validation cohort. ctDNA, circulating tumor DNA; MCC, Merkel cell carcinoma.

FIG 2.

Diagnostic performance of ctDNA at enrollment for MCC disease status in the discovery and validation cohorts. (A) Flowchart for sensitivity and specificity calculations on the basis of disease status and ctDNA status at enrollment. (B) Diagnostic accuracy of ctDNA at enrollment. The sensitivity of ctDNA for CED, defined as detection of MCC on imaging or physical examination, at enrollment was 63/66 (95%; 95% CI, 87 to 99) in the discovery cohort and 59/63 (94%; 95% CI, 84 to 99) in the validation cohort. The specificity of ctDNA at enrollment was 86/96 (90%; 95% CI, 82 to 95) in the discovery cohort and 75/87 (86%; 95% CI, 77 to 93) in the validation cohort. (C) Relationship of primary tumor size (median, 1.8 cm; IQR, 1.0-2.2 cm; range, 0.4-12 cm) and the corresponding ctDNA level at enrollment (median, 4.2 [MTM/mL; IQR, 0.7-16 MTM/mL; range, 0.03-4490 MTM/mL) in stage I-II patients with detectable ctDNA, local disease only, and enrolled before initial treatment (n = 20; Spearman's r = 0.83; P < .001). CED, clinically evident disease; ctDNA, circulating tumor DNA; MCC, Merkel cell carcinoma; MTM, mean tumor molecule. ^aSpecificity was based on the absence of clinically evident disease at the time of enrollment, without consideration of subsequent recurrences.

FIG 3.

Risk of recurrence by ctDNA status and diagnostic accuracy of ctDNA during the surveillance period. (A) Recurrence-free probability stratified by ctDNA status. The recurrence-free probability after a positive ctDNA test (orange) at any point during disease course was significantly lower than when ctDNA tests were persistently negative (blue) in both the discovery cohort (dashed curves; HR, 6.8 [95% CI, 2.9 to 16]; P < .001) and the validation cohort (solid curves; HR, 20 [95% CI, 8.3 to 50]; P < .001). (B) PPV and NPV for subsequently detected recurrence over different time frames after each ctDNA test. Error bars indicate 95% CIs. NPV remained high at 135 days (4.5 months) after each negative ctDNA test in the discovery cohort (gray points; NPV, 94%; 95% CI, 90 to 97) and the validation cohort (black points; NPV, 93%; 95% CI, 89 to 97). ctDNA, circulating tumor DNA; HR, hazard ratio; NPV, negative predictive value; PPV, positive predictive value.

FIG 4.

Likelihood of clinical detection of recurrence at different quantitative levels of ctDNA. (A) ctDNA levels from positive tests, stratified by whether the positive test was within 90 days of a clinical recurrence. Units are MTM per mL. Positive ctDNA levels drawn within 90 days of a recurrence (n = 79 tests) were significantly higher than levels drawn >90 days before a recurrence (n = 67 tests; P < .001). (B) Estimated likelihood of a recurrence being clinically detectable within 90 days before or after a positive ctDNA test, stratified at different ctDNA levels. Gray bars show risk of clinical recurrence when the ctDNA level was at or higher than the given threshold on the x-axis and the white bars show the risk of clinical recurrence when the ctDNA level was positive but below the given threshold. ctDNA, circulating tumor DNA; MTM, mean tumor molecule.

FIG 5.

Recurrence risk stratified by initial post-treatment ctDNA status in the combined cohort. There were 84 patients with local or regional disease who underwent surgery or RT for initial treatment, became clinically negative for disease after treatment, had ctDNA measured within a 4-month post-treatment window, and had follow-up after the ctDNA test was drawn. (A) Recurrence rates were significantly higher if the first post-treatment ctDNA test was positive than if it was negative (1-year recurrence-free probability, 26% v 79%; P < .001). (B) Post-treatment ctDNA status remained significantly associated with recurrence after adjusting for stage, immunosuppression status, sex, and age (HR, 7.4 [95% CI, 2.7 to 20]; P < .001), and had the strongest association with outcome among these risk factors. AJCC, American Joint Committee on Cancer; ctDNA, circulating tumor DNA; HR, hazard ratio; RT, radiation therapy.

FIG A1.

Protocol flowchart for ctDNA testing and imaging. Ordering of ctDNA testing and imaging was guided by the flowchart at all sites in the discovery and validation cohorts throughout the study period. CT, computed tomography; ctDNA, circulating tumor DNA; MCC, Merkel cell carcinoma; PET, positron emission tomography. ^aFor every patient, including after a recurrence is noted. ^bCoupled is defined by ctDNA collection and imaging ordered within 4 weeks of each other.

FIG A2.

Swimmer plot of patients in the discovery cohort, categorized by local MCC, regional disease, or distant MCC at the time of enrollment. All 167 patients included in the discovery cohort are depicted, starting from the time of enrollment. Categorization into local, regional, and distant MCC was based on AJCC eighth edition staging at the initial MCC diagnosis or most recent recurrence. All ctDNA tests performed during each patient's study follow-up are shown using orange circles (positive test) and blue circles (negative test). Surveillance periods, between when a patient was determined clinically to be negative for disease and a recurrence or end of follow-up, are indicated by the green lines. Recurrences or disease progression (X), death (upside-down triangle), and other relevant treatments and procedures are also depicted. AJCC, American Joint Committee on Cancer; ctDNA, circulating tumor DNA; MCC, Merkel cell carcinoma.

FIG A3.

Swimmer plot of patients in the validation cohort, categorized by local MCC, regional disease, or distant MCC at the time of enrollment. All 152 patients included in the validation cohort are depicted, starting from the time of enrollment. Categorization into local, regional, and distant MCC was based on AJCC eighth edition staging at the initial MCC diagnosis or most recent recurrence. All ctDNA tests performed during each patient's study follow-up are shown using orange circles (positive test) and blue circles (negative test). Surveillance periods, between when a patient was determined clinically to be negative for disease and a recurrence or end of follow-up, are indicated by the green lines. Recurrences or disease progression (X), death (upside-down triangle), and other relevant treatments and procedures are also depicted. AJCC, American Joint Committee on Cancer; ctDNA, circulating tumor DNA; MCC, Merkel cell carcinoma.

FIG A4.

ROC curve for performance of ctDNA at enrollment for Merkel cell carcinoma disease status in the discovery cohort. The AUC of ctDNA in the discovery cohort was 0.95 (95% CI, 0.92 to 0.99). Sensitivity decreased rapidly as the ctDNA threshold was increased from 0.00 MTM/mL with minimal improvement specificity. To avoid this disproportionate loss of sensitivity, the ctDNA threshold for positivity of ctDNA >0.00 MTM/mL was used for validation in the validation cohort. ctDNA, circulating tumor DNA; MTM, mean tumor molecule; ROC, receiver operating characteristic.

FIG A5.

HRs relating ctDNA status and other risk factors with recurrence during surveillance. HRs are adjusted for all other factors shown. Patients who were ctDNA-positive at any point during surveillance compared with those who remained ctDNA-negative in both the discovery cohort (HR, 7.0 [95% CI, 2.6 to 18.7]; P < .001) and the validation cohort (HR, 19 [95% CI, 7.1 to 51]; P < .001) after accounting for other risk factors. Detailed numeric results are shown in Appendix Table A4. CI, confidence interval; ctDNA, circulating tumor DNA; HR, hazard ratio.

FIG A6.

ROC curve for recurrence within 90 days of a positive test on the basis of quantitative ctDNA levels. All positive ctDNA tests from the surveillance period with at least 90 days of follow-up after the test or within 90 days of a recurrence, before or after, were included (n = 146 tests). Quantitative ctDNA levels were compared between tests drawn within 90 days of a recurrence (n = 79 tests) and tests drawn >90 days before a recurrence (n = 67 tests). The area under the curve was 0.86 (95% CI, 0.80 to 0.92). Three ctDNA thresholds (10 MTM/mL, 1 MTM/mL, and 0.1 MTM/mL) are marked on the curve (magenta, gray, and green points, respectively) to illustrate the sensitivity and specificity of different cutpoints. ctDNA, circulating tumor DNA; MTM, mean tumor molecule; ROC, receiver operating characteristic.

FIG A7.

Flow diagram for the patients included in the analysis of outcomes after stratifying by post-treatment ctDNA. Patients from the discovery cohort and the validation cohort were combined in the analysis. To be eligible for the analysis, patients needed to have AJCC eighth edition stage I-III, be treated using surgery or radiation therapy, be enrolled within the 4-month post-treatment window, and be documented as clinically negative for disease during the 4-month post-treatment window. Of the 126 patients who met the preceding criteria, 28 patients were excluded because of no ctDNA testing within the 4-month post-treatment window and 14 were excluded because of lack of follow-up after their qualifying ctDNA test, leaving 84 available for analysis. AJCC, American Joint Committee on Cancer; ctDNA, circulating tumor DNA; MCC, Merkel cell carcinoma.