Diagnostic performances of the Afirma Gene Sequencing Classifier in comparison with the Gene Expression Classifier: A meta-analysis

The Afirma microarray-based Gene Expression Classifier (GEC) with its high negative predictive value (NPV) and sensitivity has been used to rule out cancer from thyroid nodules with an indeterminate cytology but not to rule in cancer because of its low positive predictive value (PPV) and specificity. The Gene Sequencing Classifier (GSC) has been reported to improve on the weakness of GEC. In this study, a meta-analysis was performed to compare the clinical impact and diagnostic performance of GEC and GSC. Relevant data were searched in PubMed and Web of Science. Meta-analyses for proportion and dichotomous outcomes were performed to compare the benign call rates (BCRs), resection rates (RRs), risks of malignancy (ROMs), sensitivities, specificities, PPVs, and NPVs of GSC and GEC. Seven studies were included for the meta-analyses. Compared with GEC, GSC had a higher BCR (65.3% vs 43.8%; P <.001), a lower RR (26.8% vs 50.1%; P <.001), and a higher ROM (60.1% vs 37.6%; P <.001). The BCR of Hürthle cell–predominant nodules was significantly elevated (73.7% vs 21.4%; P <.001). In addition, the specificity (43.0% vs 25.1%; P =.003) and PPV (63.1% vs 41.6%; P =.004) of Afirma GSC were significantly improved while it still maintained a high sensitivity (94.3%) and a high NPV (90.0%). In conclusion, this study confirms and highlighted the clinical and diagnostic significance of GSC. With an increased BCR and improved diagnostic performance, GSC could reduce the rate of unnecessary surgical interventions and better tailor the clinical decisions of patients with indeterminate thyroid fine-needle aspiration results.