54 chapter 2 The authors themselves suggest that a rather homogenous study population with predominantly small nodules with a solid US pattern, absence of cystic areas, and follicular histology with minimal colloid could be explanatory for the poor specificity rather than operator-dependent causes [251]. Such possible relations remain undescribed in other studies. A meta-analysis on the value of USE in indeterminate thyroid nodules demonstrated meagre pooled 69% sensitivity and 75% specificity [288]. As manually applied pressure is difficult to standardize, qualitative USE is strongly operator dependent [289]. Different USE techniques have been developed to improve objectivity, such as semi-quantitative tissue-to-nodule strain ratio indices (also based on manual compression). Studies investigating semiquantitative USE in indeterminate thyroid nodules reported sensitivity and specificity ranging from 82% to 100% and from 88% to 100%, respectively [253, 255, 257, 287]. Furthermore, quantitative shear wave USE measures the propagation velocity of focused acoustic pulses – shear waves – from the probe, which correlate to tissue stiffness (Young’s modulus) [57, 290]. It had 82% sensitivity and 88% specificity in a recent prospective pilot study by Samir et al.[57]. Performance of (semi-)quantitative USE seems better than qualitative USE, but results are subject to overfitting from the ROC analysis performed to determine the strain ratio cut-off value with the highest sensitivity and specificity. None of the studies applied a predefined cut-off or validated their own cut-off externally. Consequently, the resulting thresholds were hardly comparable [253, 255, 257, 287]. Altogether, the results from currently available studies cannot support surgical decision-making in thyroid nodules with indeterminate cytology using elastosonography in any of its forms. Whereas color-coded qualitative USE has insufficient sensitivity and specificity, the semi-quantitative method lacks validation. The power of the available evidence is additionally limited by both methodological heterogeneity and the use of different USE techniques, image processing and elasticity scoring methods across studies. Nevertheless, the suggested promising rule-out capacity of qualitative USE when applying an alternative cut-off score of 1 in unselected nodules, deserves clinical validation in indeterminate thyroid nodules. Major advantages of the technique are the minor extra costs of USE, as it can be performed during regular thyroid US with the same equipment, and only adds approximately 5 minutes to the procedure time per patient. Cost-effectiveness will largely depend on performance of USE, but no cost-effectiveness studies in indeterminate thyroid nodules are available to date. Computed tomography There are no studies that investigated computed tomography (CT) scanning in thyroid nodules with indeterminate cytology. Prior studies indicated that CT cannot accurately differentiate thyroid carcinoma [291, 292].
RkJQdWJsaXNoZXIy MTk4NDMw