152 - Integration of Circulating Tumor HPV-DNA with PET/CT for Improved Surveillance of HPV-Associated Oropharyngeal Cancer: A Secondary Analysis of a Prospective Trial
Presenter(s)
A. R. R. Kawakibi1, M. J. Gutman1, E. Izumchenko2, A. Juloori1, D. J. Haraf1, R. Philips3, H. Arshad3, E. A. Blair3, A. Starus4, F. Jones4, N. Choudhury2, A. Pearson2, E. E. Vokes2, N. Agrawal3, A. J. Rosenberg2, and R. R. Katipally1; 1Department of Radiation and Cellular Oncology, University of Chicago, Chicago, IL, 2Department of Medicine, Section of Hematology/Oncology, University of Chicago, Chicago, IL, 3Section of Otolaryngology-Head and Neck Surgery, Department of Surgery, University of Chicago, Chicago, IL, 4Sysmex Inostics, Inc., Baltimore, MD
Purpose/Objective(s): National guidelines recommend PET/CT 12 weeks after definitive chemoradiation for head and neck cancers. However, post-treatment PET uptake often reflects benign changes rather than true recurrence and leads to unnecessary invasive interventions. Circulating tumor HPV-DNA (ctDNA) is a promising biomarker in HPV+ oropharyngeal cancers (OPC) for diagnosis, therapy, and surveillance. We hypothesized that the integration of HPV ctDNA with PET improves the specificity of identifying recurrence and may reduce unnecessary invasive interventions.
Materials/Methods: This was a secondary analysis of a prospective trial (NCT04572100) in locally advanced HPV+ OPC that utilized induction chemotherapy (carboplatin/paclitaxel) followed by response-adapted de-escalation per RECIST 1.1 with either: radiation alone (50 Gy) or transoral robotic surgery (Arm A), intermediate dose chemoradiation (CRT; 50 Gy with weekly cisplatin; Arm B), or regular dose CRT (70 Gy with weekly cisplatin, Arm C). ctDNA was prospectively collected serially during and after treatment utilizing a next-generation sequencing assay (HPV-SEQ, Sysmex Inostics, Inc) but did not guide de-escalation. Detectable PET uptake was determined per Nuclear Medicine. Test characteristics, including sensitivity (SENS), specificity (SPEC), positive predictive value (PPV), negative predictive value (NPV), were computed and compared using the exact McNemar’s test.
Results: Of 45 patients, 21 (47%) had detectable avidity on PET at 12 weeks post-treatment (PET+) and 3 (6.7%) patients had detectable ctDNA. Of 21 PET+ patients, 13 (62%) received additional workup (i.e. interval PET scans, biopsy, and/or surgery), including 7 (33%) who underwent invasive procedures. Over the follow-up period (median 33 months), 5 (11%) patients recurred: 2 (4.4%) locoregional recurrence (LRR) only, 2 (4.4%) distant metastasis (DM) only, and 1 (2.2%) with locoregional failure followed by distant recurrence. For any relapse, ctDNA after 12 weeks demonstrated superior specificity (100% vs 55%, McNemar P < 0.0001) and superior PPV (100% vs 14%) to PET. For LRR, ctDNA similarly demonstrated superior specificity (98% vs 57%) and PPV (67% vs 14%). Among PET+ patients, SUVmax was not correlated with relapse (OR 1.05, P = 0.8 by logistic regression; SUVs in relapsing patients were 2.0, 3.7, and 11.7), and 12 week PET only localized the eventual recurrent site in a single patient.
Conclusion: Integrating ctDNA with surveillance PET at 12 weeks improves the specificity of detecting recurrence in HPV+ OPC and may reduce unneeded invasive intervention. Future larger studies will clarify the optimal schedule to integrate ctDNA surveillance and explore radiomic models to further refine PET-based recurrence detection.
Abstract 152 - Table 1| 12 week PET | 12 week ctDNA | ||
| Any Relapse | SENS | 60% | 60% |
| SPEC | 55% | 100% | |
| PPV | 14% | 100% | |
| NPV | 92% | 95% | |
| LRR | SENS | 100% | 67% |
| SPEC | 57% | 98% | |
| PPV | 14% | 67% | |
| NPV | 100% | 98% |