Utility of response assessment PET-CT to predict residual disease in neck nodes: A comparison with the Histopathology.

Abstract Objective To assess the ability of Positron Emission Tomography-Computed Tomography (PET-CT) scans to detect residual disease in neck nodes with the Histopathology (HPR) as the gold standard. To obtain a Standardized Uptake Value max cutoff in these patients to predict residual disease in neck. Methods Head and neck squamous cell carcinoma patients who underwent Salvage neck dissection with or without primary site surgery post Concurrent Chemo-Radiotherapy (CCRT) during the period January 2008–December 2017 were included. All patients had response assessment PET-CT scan at 10–14 weeks. Agreement analysis was performed between PET-CT and HPR, fine needle aspiration cytology and HPR. Positive predictive value, Negative predictive value of PET-CT to detect residual neck nodal disease in comparison to HPR was analyzed. A Receiver Operating Characteristic (ROC) curve was plotted between the SUV max values and the HPR. A SUV max cutoff value was obtained from the ROC curve. Results A total of 75 patients were included. Thirty-one underwent salvage neck dissection along with surgery for primary disease and 45 underwent salvage neck dissection alone. PET-CT showed good agreement with the HPR to detect residual disease in neck nodes (Kappa = 0.604). PET-CT had a PPV and NPV of 87.5% and 79.15% respectively as compared against the HPR. A SUV max cutoff of 4.62 had a specificity of 92.3% and sensitivity of 73.5% to detect residual disease in neck nodes on the HPR. Conclusion PET-CT surveillance is an accepted treatment strategy. A neck node with SUV max of 4.62 and above is most likely to harbor residual nodal disease. Level of evidence: Level 2b