Background: The present study was undertaken to systematically review the literature on the reliability of using contrast-enhanced ultrasound (CEUS) to assess thyroid nodules. To avoid the potential bias in studies using a cytological standard of reference, here we aimed to meta-analyze data from studies adopting histological diagnosis as the gold standard.
Methods: A comprehensive literature exploration of PubMed and Scopus was conducted. The search was updated until June 2018 and references of the retrieved articles screened. Only original articles reporting the histological follow-up of nodules previously undergone CEUS evaluation were eligible for inclusion. Pooled sensitivity, specificity, PPV, and NPV of CEUS were calculated by DerSimonian and Laird method (random-effects model).
Results: The literature search retrieved 1885 articles, and 14 were included for the study. There were Chinese, Italian, German, and Austrian authors. All studies used SonoVue. The overall number of reported nodules was 1515, of which 775 were classified as positive at CEUS and 740 as negative. Pooled sensitivity, specificity, PPV, and NPV of CEUS were 85% (95% CI 83-88), 82% (95% CI 77-87), 83% (95% CI 77-88), and 85% (95% CI 81-88), respectively. Moderate inconsistency was present for specificity and PPV. There was publication bias for sensitivity and NPV.
Conclusions: CEUS reaches good performance in discriminating between malignant and benign thyroid lesions.
Keywords: Carcinoma; Contrast-enhanced ultrasound (CEUS); Meta-analysis; Nodule; Thyroid.