68 chapter 2 setting. With sufficient validation studies with complete histopathological follow-up, it demonstrated consistent high sensitivity and a benign test result in 40% of the patients, although the number of currently published patients is moderate. Cost-effectiveness of [18F]FDG-PET over other diagnostics is presumed [53]. Its popularity in the USA is more limited, although the efficacy of this molecular imaging technique could likely compete with molecular biomarkers panels, even if the costs per scan are somewhat higher than in Europe. The main drawback of [18F]FDG-PET/CT is its – admitted minor – risk to the patient by using a limited dose of ionizing radiation. The recently announced version 3 of the ThyroSeq® may become a prime contender. Dependent on the case mix, the ThyroSeq® v2.1 anticipated high negative predictive value [60]. However, the number of studies to confirm test performance and clinical utility in different patient populations is limited. Clinical results for the ThyroSeq® v3 are eagerly awaited. Semi-quantitative elastosonography could be a suitable alternative, in particular in case a more economic test is required. However, overfitting and lack of external cut-off validation likely overestimated the performance of this technique in the limited number of available studies. If future prospective studies can confirm its performance and thresholds of this operator-dependent but globally accessible method, USE could become a more important diagnostic in this field. None of the diagnostic techniques under investigation in this review has a perfect NPV or fulfils the threshold proposed by the ATA. A number of malignant nodules will be misdiagnosed as benign on first assessment. Considering the typical indolent clinical course of differentiated thyroid cancer, follow-up of these initially false-negative nodules will most likely still result in timely diagnosis without relevant treatment delay and dismal prognostic consequences. Recommendation for clinical use of rule-in tests The best rule-in performance was unmistakably demonstrated by BRAF mutation analysis, which showed perfect 100% specificity in an abundance of studies. Yet, strong regional differences in prevalence of BRAF mutations have a major impact on its clinical utility, especially when comparing South Korea to other countries. Moreover, the analysis most likely has very low yield in Bethesda IV nodules, in which the mostly follicular type malignancies are more frequently RAS-mutated [69, 88, 108, 114]. Testing for individual genetic alterations other than the BRAFV600E point mutation is not useful. In American and European settings, a gene mutation panel is likely preferred over any individual mutation analysis. Promising rule-in capacity was also demonstrated for Galectin-3 immunocytochemistry. An infrequently applied technique with limited validation studies, further prospective studies are warranted to validate its performance in indeterminate thyroid nodules and endorse its possible clinical use. Besides BRAF mutation analysis, none of diagnostics meet the 2015 ATA requirements of an ideal rule-in test. Compared to ruling-out tests, ruling-in tests face an additional challenge. With a
RkJQdWJsaXNoZXIy MTk4NDMw