Back

Methylation-Based ctDNA for Post-Treatment Surveillance of Non-Viral Head and Neck Cancer

Tomer, D.; Brodin, P.; Malik, N.; Mayat, S.; Velten, C.; Rhieu, B.-H.; Castellucci, E.; Schiff, B.; Mehta, V.; Smith, R.; Kalnicki, S.; Garg, M.; Kabarriti, R.

2026-09-20 oncology
10.64898/2026.09.17.26363361 medRxiv
Show abstract

Importance: Recurrence rates in head and neck squamous cell carcinoma (HNSCC) remain high, yet current post-treatment surveillance relies primarily on clinical examination and imaging with known limitations. Circulating tumor DNA (ctDNA) has shown promise for recurrence detection in HPV-positive HNSCC, but less is known about its role in non-virally associated disease. Objective: To evaluate the diagnostic performance of a methylation-based ctDNA assay for detecting recurrence during post-treatment surveillance of non-virally associated HNSCC and to assess its prognostic association with survival outcomes. Design, Setting, and Participants: This retrospective single-institution cohort study included patients with non-metastatic non-virally associated HNSCC who underwent definitive therapy and had at least one Guardant360 ctDNA test during post-treatment surveillance between January 2024 and October 2025. Guardant360 identifies tumor-specific DNA methylation patterns via next-generation sequencing of plasma cell-free DNA. A positive result was defined by a methylation signal of 0.05% or greater. Main Outcomes and Measures: Sensitivity, specificity, positive predictive value (PPV), negative predictive value (NPV), and area under the receiver operating characteristic curve (AUC). Progression-free survival (PFS) and overall survival (OS) were estimated using the Kaplan-Meier method. Results: Fifty patients were included; most had larynx (38%) or oral cavity (36%) primaries, AJCC stage IVA-IVB disease (72%), T3-T4 tumors (69%), and N2-N3 nodal classification (48%). Treatment was chemoradiation (54%) or surgery (46%). At a median follow-up of 18.0 months, 13 patients recurred, and 12 had positive ctDNA results. Sensitivity, specificity, PPV, and NPV were 76.9%, 94.6%, 83.3%, and 92.1%, respectively (AUC, .858; 95% CI, .733-.982). Positive ctDNA was associated with worse PFS (1-year: 25.0% vs 96.4%; P < .001) and OS (1-year: 75.0% vs 100%; P < .001). Among 10 patients with recurrence and positive ctDNA, ctDNA preceded clinical detection in 6 (60%) by a median of 131 days. An exploratory analysis identified TP53, SETD2, and ROBO2 mutations as associated with recurrence. Conclusions and Relevance: A methylation-based ctDNA assay demonstrated high sensitivity and specificity for detecting disease recurrence in non-virally associated HNSCC. Positive post-treatment ctDNA was associated with inferior survival and preceded clinical detection in most cases, supporting ctDNA integration into surveillance strategies, though prospective validation is warranted.

Matching journals

The top 8 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.