F‐18 FDG PET‐CT for response evaluation in head and neck malignancy: Experience from a tertiary level hospital in south India

BACKGROUND: Head and neck squamous cell carcinoma (HNSCC) accounts for 90% of head and neck cancers. There has been no established qualitative system of interpretation for therapy response assessment using PET‐CT for HNSCC. AIM: To assess response evaluation of nodal status in post‐treatment PET‐CT scans in HNSCC using a 5‐point Likert scale (Deauville score [DS]). METHODS AND RESULTS: Retro‐prospective analysis was performed of the nodal status of pre and post‐RT PET‐CT in patients diagnosed with HNSCC (n = 43) from May 2013 to March 2018. All eligible patients underwent a pre‐RT PET‐CT scan before the start of RT. Another post‐RT PET‐CT scan was performed 12 weeks after the completion of RT. The median time from completion of radiotherapy (RT) to post‐RT PET‐CT was 92 days; 80% of the patients had their post‐RT PET‐CT scan between 77 and 147 days after therapy. Of 43 patients (M/33, F/10, age range 18 to 80 years (median 54 years) selected for the study, good concordance was noted between DS and clinical response in these patients. The change in SUV from pre‐RT PET to post‐RT PET was analyzed using a paired t‐test. The P‐value was found to be statistically significant while comparing pre and post‐RT SUVmax levels showing that RT had significantly reduced the SUVmax levels of the nodes in DS 2‐3 groups whereas the number of patients was too small to allow a reliable calculation in DS 4‐5 groups. It was found that 36/39 patients with DS 1‐3 had no nodal recurrence showing a high NPV of 92.3%. Of the four patients with DS 4‐5, all had active disease showing PPV of 100%. Applying Fisher's exact test, the P‐value was found to be .004. CONCLUSION: DS seems to satisfy the requirements for a simple qualitative method of interpreting PET scans and for identifying patients requiring neck dissection. Consensus regarding qualitative assessment would facilitate standardization of PET reporting in clinical practice and enable comparative multicentric studies