39 Diagnostic utility of molecular and imaging biomarkers 2 that demonstrated the strength of the Afirma® GEC [164]. After the overwhelming results from this key-publication, popularity of the GEC took flight. It is marketed as a highly accurate rule-out test for malignancy in thyroid nodules with indeterminate cytology. In 2014, the first multicentre study that retrospectively assessed the clinical utility of the Afirma® GEC was published. Only 6% of reported GEC-negative Bethesda III, IV and V nodules eventually underwent surgery, of which one resulted in a 6 mm mPTC. Unfortunately, data on GEC negative nodules were only reported on an aggregate level; exact test performance rates in Bethesda III and IV nodules cannot accurately be extracted from the publication. Less than half of the GECnegative nodules without surgery (71/163, 44%) had clinical or radiological follow-up, ranging from 1 to 24 months (median 8 months) – a limited duration compared to the natural, indolent course of differentiated thyroid carcinoma. The published paper does not describe whether the remaining 92 patients with GEC-negative nodules received any follow-up at all. Despite evident limitations to the applied reference standards, Alexander et al. concluded that their results confirm both the accurate test performance from their prior study as well as the large impact that the Afirma® GEC has on clinical decision-making for cytologically indeterminate thyroid nodules [165]. Yet, physicians indeed seemed reassured by a negative GEC result based on the first studies alone [164, 166]. In many institutions in the USA the Afirma® GEC was immediately implemented in clinical practice. The retrospective studies that followed were mere post-implementation utility studies, and generally reported very high but moderately consistent sensitivities. GEC-negative nodules were largely managed without surgery and considered true-negative, resulting in possible overestimation of test sensitivity. Long-term follow-up is not yet available to endorse a benign diagnosis in these cases [167, 168]. The high degree of missing histology was recognized by most of these studies as a major limitation [165, 167, 169-172]. This was confirmed by the 2015 ATA guidelines: recognizing the Afirma® GEC as a promising diagnostic tool, the guidelines stress that it is a major shortcoming that external clinical validation studies with full histological follow-up of Afirma® GEC-negative nodules are still lacking [17]. Not all studies were able to confirm the potential of the Afirma® GEC. Some struggled with a low benign call rate (i.e. useful negative test result that could lead to management change) [168, 173]. McIver et al. questioned the cost-effectiveness of the Afirma® GEC in their population, as the mere 22% (16/72) negative test rate was much lower than anticipated. Moreover, a quarter of these GECnegative patients rejected the proposed conservative treatment of ultrasound-based follow-up and underwent surgery anyway; one of them was diagnosed with a 3.2 cm FTC with focal capsular and vascular invasion. Also, 84% of GEC-positive nodules proved histopathologically benign, overall resulting in a disappointing 83% sensitivity and 10% specificity [168]. Besides concerns regarding adequate clinical validation of test performance, the post-implementation influence of the GEC on surgical decision-making for individual patients was also questioned. In line
RkJQdWJsaXNoZXIy MTk4NDMw