Abstract
Objectives The aim of this study was to determine the diagnostic accuracy of contrast-enhanced magnetic resonance imaging (CE-MRI) for the detection of perineural spread (PNS) in head and neck cancer patients. Methods A systematic review of PubMed, Embase, Scopus, Web of Science and Cochrane Library databases was performed up to May 20, 2022. We included diagnostic accuracy studies that used CE-MRI for the diagnosis of PNS in patients with head and neck cancer, using histopathology from surgical specimens as the reference standard. Potential bias and applicability of the included studies was assessed using the Quality Assessment of Diagnostic Accuracy Studies 2 (QUADUS-2) tool. Pooled joint effect sizes of sensitivity and specificity were calculated by applying bivariate random-effects meta-analysis model. Results Nine studies with 259 patients were included. The pooled sensitivity and specificity of CE-MRI for detecting PNS were 89% (95% confidence interval [CI]: 73-96) and 83% (95% CI: 73-90), respectively. Stratifying by MRI strength, 1.5 T had a higher sensitivity of 97% (95% CI: 47-100) compared with 3 T, which had a sensitivity of 83% (95% CI: 72-90). Both 1.5- and 3-T MRI had a similar specificity in detecting PNS of 85% (95% CI: 63-95) and 84% (95% CI: 75-91), respectively. Conclusions CE-MRI provides good diagnostic test accuracy for the detection of PNS in head and neck cancer. Current evidence suggests 1.5-T MRI provides greater sensitivity compared with 3-T MRI.