Diagnostic efficacy of PET-CT, CT, and MRI in preoperative assessment of mandibular invasion caused by head and neck cancer: A systematic review and meta-analysis

Objective: This study aims to conduct a systematic review and meta-analysis of the performance of PET-CT, CT, and MRI in diagnosing mandible invasion induced by head and neck cancer (HNC). Materials and methods: The MEDLINE, Embase, Science Direct, CNKI and CQVIP databases were searched from inception until August 1, 2020. Then, a meta-analysis was conducted to calculate the combined diagnostic values with the corresponding 95% CIs. Two independent researchers completed the full text screening, data abstraction, and risk assessment. Results: This meta-analysis included 53 studies (N = 2 946 participants). For the pooled sensitivity (SEN), MRI (SEN: 0.88, 95% CI: 0.81-0.93) was found to have a significantly higher SEN (P = 0.0045), when compared to CT (SEN: 0.77, 95% CI: 0.71-0.82), while compared with PET-CT (SEN: 0.88, 95% CI: 0.64-0.97), the SEN was approximately equal (P > 0.05). The analysis revealed that the combined specificity (SPE) of MRI (SPE: 0.83, 95% CI: 0.74-0.89) and PET-CT (SPE: 0.81, 95% CI: 0.57-0.93) was lower than that of CT (SPE: 0.87, 95% CI: 0.83-0.90), but there was no statistical significance among these (P > 0.05). The comparison of the area under curve (AUC) reflected that PET-CT, CT and MRI have approximately equal summary diagnostic power in detecting mandibular invasion (P > 0.05). Conclusion: The findings suggest that compared with CT, MRI is significantly superior for higher SEN in diagnosing mandibular invasion. The SEN of MRI and PET-CT were approximately equal. For the summary of diagnostic power, more prospective clinical trials that directly compare these three methods are needed in the future.