Diagnostic Performance of Ultrasound for Differentiating Malignant From Benign Cervical Lymphadenopathy in Children: A Systematic Review and Meta-Analysis.

Abstract

Objectives: To evaluate the diagnostic performance of ultrasound (US) features used to differentiate malignant from benign cervical lymphadenopathy in children.

Methods: PubMed and EMBASE were searched for articles reporting the diagnostic performance of US features to differentiate benign and malignant cervical lymphadenopathy in children. Article quality was evaluated using the QUADAS-2 tool. Sensitivity and specificity were pooled using a bivariate random-effects model. Further sensitivity analyses and meta-regression were performed to determine the potential source of heterogeneity. The area under the receiver operating characteristic curve was calculated using the hierarchical summary receiver operating characteristic model.

Results: Ten articles (1077 children) were included. Among the retrieved US features, abnormal vascularity, heterogeneous echogenicity, abnormal hilum echogenicity, and long-axis/short-axis (L/S) ratio were significantly associated with malignant lymphadenopathy, with pooled diagnostic odds ratios of 36 (95% confidence interval [CI]: 14-92), 17 (3-91), 16 (5-54), and 7 (5-9), respectively. The most sensitive US features were abnormal hilum echogenicity (0.86, 95% CI: 0.66-0.95) and heterogeneous echogenicity (0.84, 95% CI: 0.25-0.99). Abnormal vascularity (0.91, 95% CI: 0.82-0.97) was the most specific. Substantial heterogeneity was found in both sensitivity and specificity (I² > 50%; P < .01), although the source was not revealed.

Conclusion: Among US features, abnormal vascularity, heterogeneous echogenicity, abnormal hilum echogenicity, and L/S ratio are useful for differentiating malignant from benign cervical lymphadenopathy in children, showing good diagnostic performance. These findings should be carefully interpreted due to unexplained heterogeneity, which may lower the validity of the pooled estimates.