Prognostic Value of 2-[18F] Fluoro-2-deoxy-D-glucose Positron Emission Tomography-Computed Tomography Scan Carried out During and After Radiation Therapy for Head and Neck Cancer Using Visual Therapy Response Interpretation Criteria

Abstract Aims To evaluate the prognostic utility of 2-[ 18 F] fluoro-2-deoxy- d -glucose positron emission tomography-computed tomography (FDG PET-CT) carried out in the third week (iPET) and after completion (pPET) of definitive radiation therapy in patients with mucosal primary head and neck squamous cell carcinoma (MPHNSCC) and to investigate the optimal visual grading criteria for therapy response assessment. Materials and methods Sixty-nine consecutive patients with newly diagnosed MPHNSCC treated with radical radiation therapy with or without systemic therapy underwent staging. PET-CT, iPET and pPET were included. All PET-CT images were reviewed by using a visual grading system to assess metabolic response for primary tumour: 0 = similar to adjacent background blood pool activity; 1 = more than background but  Results The median follow-up was 28 months (range 6–62), the median age was 61 years (range 39–81) and AJCC 7th edition clinical stage II, III and IV were six, 18 and 45 patients, respectively. The optimal threshold for non-complete metabolic response (non-CMR) was defined as focal uptake ≥ liver (grade 3) for iPET and focal uptake ≥ mediastinum (grade 2) for pPET. The 2 year Kaplan-Meier LRFS, DFS and overall survival estimates for primary CMR and non-CMR in iPET were 89.8% versus 71.5% ( P  = 0.062), 80.1% versus 65.3% ( P  = 0.132), 79.1% versus 72.1% ( P  = 0.328) and in pPET 86.2% versus 44.6% ( P  = 0.0005), 77.6% versus 41.2% ( P  = 0.006), 81.2% versus 40.6% ( P  = 0.01), respectively. The negative predictive value (NPV) for LRFS for patients achieving both primary and nodal CMR in iPET was 100%. No locoregional failure was observed in patients with both primary and nodal iPET CMR ( P  = 0.038), whereas those with nodal iPET CMR had no regional failure ( P  = 0.033). However, the positive predictive values (PPV) for LRFS and DFS for iPET and pPET were found to be poor: 30% and 36% for iPET and 35% and 39% for pPET, respectively. Conclusion Standardised criteria using visual assessment are feasible. The metabolic response using visual assessment with standardised interpretation criteria of iPET and pPET can be useful predictors of tumour control. Dose de-escalation can be considered on the basis of a high NPV for iPET. However, the PPV of iPET is poor, indicating that additional discriminative tools are needed.