A universal recommendation does not exist for thyroid FNA suspicious for malignancy (SFM). In this context, the guidelines have estimated a risk of malignancy (ROM) from 50 to 80% and both total thyroidectomy and lobectomy may be indicated. This study aimed to (1) retrospectively evaluate the SFM (i.e., TIR4) in a single institution to estimate their cancer prevalence at histology, and (2) systematic review the literature to obtain more robust information. The study period was 2015-2018. As a major inclusion criterion, both cytology and histology had to be performed in our institution. Histological diagnosis was the gold standard. For the systematic review, the online databases of Google Scholar, PubMed/MEDLINE, and Scopus were searched for papers using the same classification for thyroid FNA. A proportion meta-analysis was performed to obtain the pooled histological cancer rate among TIR4 and TIR5 (random-effects model). In the institutional database, there were 271 nodules with both histology and FNA and the cancer rate of TIR4 was 88.9%. By systematic review, five studies were selected for the meta-analysis. The pooled cancer rate was 85% in TIR4 and 99% in TIR5 (I2 = 0%; no publication bias). In conclusion, these new findings should prompt the guidelines board to fully revise the estimated ROM of SFM category. Clinical thyroidologists and thyroid surgeons should be aware of these data and the patients with SFM should be informed of their ROM.
Keywords: Carcinoma; Fine-needle aspiration (FNA); Suspicious for malignancy; Thyroid.