38 chapter 2 The high diagnostic accuracy is also a downside to NGS. Highly sensitive, NGS is able to identify mutant alleles at very low levels (<10%). A low percentage of mutant alleles might reflect a subclone within the nodule, which is not histopathologically identified as carcinoma. This detection of germline or clinically insignificant low-level somatic mutations in benign nodules could decrease NGS specificity [61, 118]. Nikiforov et al. suggested that the next improvement of the NGS-related tests should therefore be to determine accurate threshold levels for the various gene variations [118]. NGS encompasses crucial technology that is rapidly advancing. The ThyroSeq® v3 was recently announced, promoting to encompass no less than ~95% of genetic alterations occurring in PTC. Extraordinary diagnostic accuracy above 90% is anticipated, including high accuracy in Hürthle cell lesions. Results of the prospective studies validating this new version will likely be published shortly. Nonetheless, NGS techniques currently have limited global availability, with the exception of some European countries and the USA. The ThyroSeq® is available for $3,200 per test [161]. In contrast, the thyroid non-specific AmpliSeq™ panel can be ordered online for only €230 [162]. Independent prospective studies are needed to validate its performance and predicted cost-utility in different patient populations, and confirm the superior position of the ThyroSeq® and other NGS techniques. Afirma® gene expression classifier In molecular diagnostics, the chief competitor of the 7-gene mutation panel is the commercial Afirma® gene expression classifier (GEC) (Veracyte Inc., South San Francisco, CA, USA). The GEC uses quantification of the mRNA-expression of 167 genes and a proprietary classification algorithm to determine the probabilities of malignancy in the samples’ expression patterns. The classification algorithm to discern a ‘benign’ (negative test) from a ‘suspicious’ (positive test) thyroid nodule results from a successful designer study that trained the GEC in both a tissue set and diverse FNAC sample sets with known histopathology [163]. Alexander et al. performed the first prospective, blinded, industry-sponsored clinical study to validate this Afirma® GEC in patients with indeterminate thyroid nodules [164]. From 49 hospitals 577 Bethesda III, IV and V FNAC samples were collected, obtained by two additional needle aspirates from thyroid nodules with a diameter of at least 1 cm. After exclusion of over half (312/577, 54%) of the samples for reasons such as nodules that were not surgically resected, duplicate specimens from the same nodule, and issues with specimen shipments to Veracyte, finally 265 FNAC samples were included in the analysis. Sensitivity of the Afirma® GEC was 90% in the 129 Bethesda III as well as the 81 Bethesda IV nodules with a useful GEC-negative test result in 38% (100/265), but specificity was merely 53% and 49%, respectively (52% on average). Despite the relatively high malignancy rate in Bethesda III nodules and the high number of exclusions, this study is well conducted and recognized worldwide as the landmark study
RkJQdWJsaXNoZXIy MTk4NDMw