Diagnostic performance of NIRADS on F-18 FDG PET/CT in the post-treatment assessment of head and neck malignancies - a systematic review and meta-analysis

Purpose This review aimed to evaluate the diagnostic performance of Neck Imaging Reporting and Data System (NIRADS) categories using 2-deoxy 2-(Fluorine-18) fluoro D-glucose positron emission tomography integrated with computed tomography (F-18 FDG PET/CT) to diagnose recurrence or treatment failure in the post-treatment assessment of head and neck squamous cell carcinoma (HNSCC). Methods A systematic search of the indexed medical literature was conducted till 7 November 2024 using PubMed, Scopus, Cochrane Library and Google Scholar to include studies reporting post-treatment recurrence rates on F-18-FDG PET/CT as per the NIRADS criteria. Studies were qualitatively assessed for risk of bias using the QUADAS-2 tool. We categorized patients with 'NI-RADS <= 2' as low risk of recurrence and 'NI-RADS >= 3' as high risk of recurrence. Diagnostic performance of NI-RADS was evaluated through weighted-pooling of standard metrics of diagnostic accuracy, diagnostic Odd's ratio (DOR) and summary receiver operator characteristic (SROC) curve analysis. Results Out of 1632 records identified, 8 studies with 7 datasets were included, with over 1200 patients and over 1300 F-18 FDG PET/CT scans. All studies were retrospective, with presence of a risk for bias, publication bias and data heterogeneity. The time between treatment and F-18 FDG PET/CT assessment was 8-17 weeks. For combined sites (primary, nodal or distant sites), the ability of 'NIRADS >= 3' over 'NIRADS <= 2' to detect a recurrence was acceptable with a pooled sensitivity, specificity, positive and negative likelihood ratios of 0.68 (95% Confidence Interval, CI 0.63-0.73), 0.90 (95% CI, 0.88-0.92), 6.01 (95% CI, 2.9- 12.6) and 0.47 (95% CI, 0.31- 0.71) respectively, with a DOR of 14.13 (95% CI, 9.78-20.42) and an area under the curve, AUC of 0.859 (Standard Error, SE -0.018) on SROC analysis Conclusion In F-18 FDG PET/CT done on post-treatment HNSCC patients, the diagnostic performance of NIRADS >= 3 categories over NIRADS <= 2 categories in detecting treatment failure or recurrence at combined primary, nodal or distant sites was acceptable, with a low certainty of evidence. The NIRADS categories, especially NIRADS >= 3 and NIRADS <= 2 categories should be routinely reported in post-treatment F-18 FDG PET/CT scans in HNSCC patients.