Predicting response to radical (chemo)radiotherapy (R-CRT) with circulating HPV DNA and tumor DNA (ctDNA) analysis in locally-advanced head and neck squamous cell carcinoma (LAHNC).
6043 Background: Following R-CRT for human papilloma virus positive (HPV+) and negative (HPV-) LAHNC, patients frequently undergo unnecessary neck dissection (ND) and/or repeated biopsies for abnormal PET-CT findings even in the presence of a complete pathological response (pCR), which causes significant morbidity. We assessed the role of circulating tumor DNA analysis in identifying patients with true residual disease. Methods: We prospectively recruited development (DC, n=55) and test (TC, n=33) cohorts of LAHNC patients having R-CRT. For HPV+ tumors we developed a novel amplicon based next generation sequencing assay (HPV-detect) to detect circulating HPV DNA and for HPV- tumors we used personalised droplet digital PCR assays of somatic mutations. Circulating tumor DNA levels at 12 weeks post-R-CRT were correlated to residual disease assessed by PET-CT and surgery. Results: In the DC (27 HPV+), baseline HPV-detect demonstrated 100% sensitivity and 93% specificity, confirmed in the TC (20 HPV+). 37 HPV+ patients (DC&TC) had complete samples-set. 36 had a negative HPV-detect at end of treatment, including 6 patients who underwent ND (3) and repeat primary site biopsies (3) for positive PET-CT but had pCR on surgical/biopsy specimen. 1 patient had positive HPV-detect and positive biopsy, indicating 100% agreement for HPV-detect and residual cancer. In a 10 HPV- patients with complete sample-set, there was 90% agreement between ctDNA and residual disease in HPV- tumors (3 ctDNA positive and tumor present, 1 ctDNA negative but tumor present, and 6 negative ctDNA negative tumor) with 80% sensitivity for residual disease and 100% specificity. Combined agreement between ctDNA testing (HPV+ and -) & residual disease was 98% (Table). Conclusions: Circulating HPV DNA quantified using HPV-detect and ctDNA identifies patients with residual disease post-R-CRT in LAHNC. Further studies are required to validate these findings. [Table: see text]