Pitfalls of post-treatment PET after de-intensified chemoradiotherapy for HPV-associated oropharynx cancer: Secondary analysis of a phase 2 trial

Objectives: We evaluated patterns of nodal response and positive predictive value (PPV) of 3 month post-treatment PET in patients with HPV-associated oropharyngeal cancer treated on a multi-institutional de-intensification trial. Materials and methods: Eligibility criteria included: (1) T0-3, N0-2c, M0, (2) HPV+/p16+ oropharyngeal squamous cell carcinoma, and (3) <= 10 pack-years smoking or <= 30 pack-years and abstinent >= 5 years. Patients received 60 Gy radiation alone (T0-2, N0-1) or with concurrent weekly cisplatin 30 mg/m(2) and surveillance PET three months post-radiation. Nodal responses were categorized as complete (CR), equivocal (ER), or incomplete (IR) using both local and central radiographic review. A true positive was ER/IR with clinical/radiographic progression or positive pathology. Results: 79 node-positive pts (84% N2) were analyzed. Distribution of nodal CR, ER, and IR was 44 (56%), 27 (34%), and 8 (10%), respectively. 29 (37%) had ER/IR in pre-treatment node-positive neck levels, whereas 14 (18%) had ER/IR in pre-treatment node-negative levels. Of patients with ER/IR, 5 were observed clinically, 19 received repeat imaging, and 11 received either biopsy (1) or neck dissection (10). The PPV was 9% for ER/IR and 13% for IR, with 3 patients found to have persistent disease on neck dissection. There was no difference in nodal relapse rate in patients with nodal CR vs. nodal ER/IR. Conclusion: Post-treatment PET may not accurately predict the presence of persistent disease in patients with favorable-risk oropharynx cancer. These results support close surveillance rather than surgical evaluation in most favorable-risk patients.