Personalized circulating tumor DNA dynamics predict survival and response to immune checkpoint blockade in recurrent/metastatic head and neck cancer

BackgroundRecurrent/metastatic head and neck squamous cell carcinoma (R/M HNSCC) is an aggressive cancer with a median overall survival of only 12 months. Existing biomarkers have limited ability to predict treatment response or survival, exposing many patients to the potential toxicity of treatment without certain clinical benefit. Circulating tumor DNA (ctDNA) has emerged as a non-invasive, real-time biomarker that could address these challenges. MethodsWe analyzed 137 plasma samples from 16 patients with R/M HNSCC undergoing immune checkpoint blockade (ICB)-based therapy. A tumor-informed, highly sensitive next-generation sequencing liquid biopsy assay (RaDaR, NeoGenomics Laboratories, Inc.) was applied to track ctDNA changes at baseline and throughout treatment. Univariable and multivariable analyses were used to assess the association between ctDNA negativity and key clinical outcomes: disease control (best objective response of stable disease, partial response, or complete response), three-year overall survival (OS), and three-year progression-free survival (PFS). We also assessed a machine learning model to predict disease progression based on ctDNA dynamics. ResultsMultivariable analysis revealed that ctDNA negativity during treatment was significantly associated with improved disease control (OR 21.7, 95% CI 1.86-754.88, p=0.0317), three-year OS (HR 0.04, 95% CI 0.00-0.47, p=0.0103), and three-year PFS (HR 0.03, 95% CI 0.00-0.37, p=0.0057). The machine learning model predicted disease progression with 88% accuracy (AUC 0.89). ConclusionSerial ctDNA monitoring predicted disease control, survival, and progression in patients with R/M HNSCC receiving treatment with ICB, suggesting that incorporation of ctDNA into clinical practice could enhance treatment decision-making for clinicians and improve patient outcomes.