53 Diagnostic utility of molecular and imaging biomarkers 2 making. We propose that a future meta-analysis should use the individual patient data from the large number of available original studies to develop an ultrasound algorithm specifically for indeterminate thyroid nodules. The existing TIRADS needs prospective validation. Even though more advanced and less operator-dependent techniques might be preferred, US features should always be assessed in current clinical practice. The presence of one or more suspicious US features in a Bethesda III or IV nodule increases the suspicion of malignancy and underpins the need for a definite diagnosis. Moreover, to centres or regions with limited access to other (molecular) diagnostics, ultrasound may definitely have clinical utility, pending local validation in the indeterminate population. Elastosonography Firm consistency of a thyroid nodule upon palpation is considered suspicious for malignancy – an established principle during physical examination [280]. Ultrasound elastosonography (USE) is a dynamic ultrasound technique that is sometimes referred to as ‘electronic palpation’. Tissue elasticity is evaluated by measuring tissue distortion while applying a standardized dosed external force by the US transducer. It was first applied to the thyroid gland by Lyshchik et al. in 2005 [281]. Classic real-time qualitative USE is performed by free-hand compression and a sine-wave or numerical scale showing how much pressure the operator applies with the probe. A color-coded elastosonography image is superimposed on the grey-scale US images: red and orange visualizes high tissue elasticity (soft tissue), green represents intermediate elasticity and blue low elasticity (firm tissue). Several score systems are available. The original score was developed by Itoh et al. in 2006 for the evaluation of breast tumours and considers scores 1-3 benign on a scale of 1 (highest elasticity) to 5 (no elasticity) [282]. Rago et al. first applied it to thyroid tumours and modified it to a 3-point score [260, 283]. Asteria et al. derived a modified 4-point score [284]. The earliest studies in thyroid nodules reported opportune results of USE as an additional modality to B-mode ultrasound, but were heterogeneous in USE technique and study population [285]. A recent meta-analysis by Nell et al. included twenty studies on qualitative USE prior to FNAC and concluded that qualitative USE is fit to diagnose benign nodules and safely dismiss FNAC, provided that the usual elasticity score cut-off is abandoned and only completely soft nodules (score 1 of all systems) are classified as benign. Pooled 99% sensitivity and 99% negative predictive value demonstrated the ability of USE to reliably rule-out malignancy in entirely soft thyroid nodules, composing 14% of their pooled study population [286]. In individual studies on USE in indeterminate nodules, sensitivity and specificity of qualitative USE ranged from 47% to 97% and from 6% to 100%, respectively [58, 251, 259, 260, 287]. Results of several qualitative USE studies stand out. Lippolis et al. showed an aberrant 6.1% specificity, because they reported only eight nodules with high elasticity – 62 of 66 benign nodules were not elastic.
RkJQdWJsaXNoZXIy MTk4NDMw