Abstract
Background: The management of indeterminate thyroid nodules (Bethesda III/IV) has evolved with molecular testing, aiming to reduce unnecessary surgeries. However, the comparative effectiveness of different platforms in influencing surgical decision-making remains unclear. This systematic review and meta-analysis evaluate the impact of molecular testing on surgical avoidance rates. Methods: A systematic literature search was conducted across eight electronic databases, including Embase, PubMed, and Cochrane Library, from January 2019 to December 2024, following PRISMA guidelines to encompass most recent advancements in the last 5 years. Studies evaluating Afirma Gene Expression Classifier (GEC), Afirma Genomic Sequencing Classifier (GSC), ThyroSeq V2, ThyroSeq V3, and ThyGenX/ThyraMIR were included. The primary outcome was surgical avoidance, analyzed using a random-effects model. Results: Thirty-one studies comprising 4464 indeterminate thyroid nodules met inclusion criteria. Pooled surgical avoidance rates varied across platforms: ThyroSeq V2 (50.3%, 95% CI: 20.8-79.6%), ThyroSeq V3 (62.5%, 95% CI: 54.8-70.0%), Afirma GEC (58.8%, 95% CI: 43.6-73.1%), Afirma GSC (50.6%, 95% CI: 34.3-66.8%), and ThyGenX/ThyraMIR (68.6%, 95% CI: 63.1-73.9%). ThyGenX/ThyraMIR had the highest surgical avoidance rate and lowest heterogeneity (I(2) = 51.2%), while ThyroSeq showed improvement from V2 to V3. Conclusions: Molecular testing reduces unnecessary thyroid surgeries, with avoidance rates ranging from 50.3% to 68.6%. While ThyGenX/ThyraMIR showed the highest avoidance rate, its limited representation warrants cautious interpretation. Standardized protocols are needed to optimize clinical application. Further prospective studies should compare platforms and assess long-term outcomes and cost-effectiveness.