Diagnostic performance of MRI to detect metastatic cervical lymph nodes in patients with thyroid cancer: a systematic review and meta-analysis

Abstract

To evaluate the diagnostic performance of magnetic resonance imaging (MRI) in the diagnosis of metastatic cervical lymph nodes. Ovid-MEDLINE and EMBASE databases were searched up until 12 June 2018. Eleven articles were included in the qualitative systematic review and nine of the 11 in the quantitative analysis. Two radiologists independently performed data extraction and methodological quality assessment using the Quality Assessment of Diagnostic Accuracy Studies-2 tool. A qualitative systematic review and quantitative analysis were performed, followed by a meta-regression analysis to determine factors causing heterogeneity. The pooled sensitivity and specificity in the diagnosis of metastatic cervical lymph nodes were 80% (95% confidence interval [CI]: 68-88%) and 85% (95% CI: 63-95%), respectively. The sensitivity and false-positive rate (correlation coefficient, 0.655) showed a positive correlation due to a threshold effect, which was responsible for heterogeneity across the studies, as indicated by a Q-test (p<0.01) and Higgins I2 statistic (sensitivity, I2=90.11%; specificity, I2=92.49%). In the meta-regression analysis, fat-suppressed imaging, and the analysis method were significant factors influencing the heterogeneity in diagnostic performance. MRI shows moderate diagnostic performance in the diagnosis of metastatic lymph nodes in patients with thyroid cancer in the neck. MRI may be an optional or complementary imaging method to ultrasound or computed tomography (CT) in thyroid cancer patients.