Italian consensus for the classification and reporting of thyroid cytology: the risk of malignancy between indeterminate lesions at low or high risk. A systematic review and meta-analysis

Background: Italian consensus for the classification and reporting of thyroid cytology has proposed to discriminate the cancer prevalence of high (Tir 3B) vs. low (Tir 3A) risk indeterminate nodules. To obtain more robust evidence on this topic, we performed a meta-analysis of the Odds Ratio (OR) of malignancy of Tir 3B vs. Tir 3A nodules. Methods: A comprehensive literature exploration of online databases was conducted until May 2018. Original articles reporting histology of nodules cytologically classified as Tir 3A and Tir 3B were eligible. Pooled cancer prevalence in Tir 3A and Tir 3B, and OR of Tir 3B vs. Tir 3A were calculated. Results: The search revealed 95 articles, and 10 were included for the meta-analysis. Overall, 1168 indeterminate lesions were reported (441 Tir 3A and 727 Tir 3B), of which 391 were cancers. The pooled cancer prevalence was 17% in Tir 3A and 47% in Tir 3B. The OR of Tir 3B vs. Tir 3A was 4.24 (95% CI 2.75 to 6.53) with mild heterogeneity and without publication bias. When we considered non-invasive follicular thyroid neoplasms with papillary-like nuclear features (NIFTP) as non-malignant, cancer rate was lower, and OR of Tir 3B against Tir 3A was 2.93 (95% CI 1.60 to 5.37), with no heterogeneity but with publication bias. Conclusions: The Italian system for thyroid cytology is reliable to assess indeterminate lesions at low and high risk, being Tir 3B associated with a cancer risk significantly higher than Tir 3A, also when considering NIFTP as non-malignant entity.