72 chapter 2 from indeterminate thyroid nodules. Besides insuperable epidemiological variations, the selection of indeterminate cytology, and the retrospective nature of many studies may have contributed to these discrepancies. The type of indeterminate cytology included by individual studies varied, likely leading to betweenstudy heterogeneity. Besides global variations and known intra- and interobserver discordance, diverse definitions of indeterminate cytology were adhered [26]. Nowadays, the Bethesda system differentiates indeterminate from benign and suspicious cytology in a more standardized manner in both literature and clinic. Bethesda III and/or IV and similar categories from other classification systems were frequently applied. Unfortunately, some studies also included small numbers of Bethesda V nodules without presenting results for individual categories separately [165]. Many other studies adhered to their own definition of indeterminate cytology. This especially, but not exclusively, concerns studies published before the introduction of the Bethesda system in 2009. Retrospective study designs and subsequent selection bias – only including indeterminate thyroid nodules that had undergone both thyroid surgery and (routine) pre-operative testing – likely also caused overestimation of the true efficacy of certain techniques (e.g. BRAF mutation analysis or ultrasound). Conclusion and Recommendations In current-day practice, there are numerous additional diagnostics available to further assess thyroid nodules with indeterminate cytology, all with advantages and disadvantages. This review provided a comprehensive overview of the available literature on these techniques, addressing both molecular and imaging biomarkers, aiming to provide an objective and nuanced comparison of their performance and cost-effectiveness with regard to rightful surgical decision-making. Many of these diagnostics have either an adequate rule-in or rule-out capacity, but no single currently available test seems to serve both purposes well. Diagnostics from the different research fields likely complement each other in a multimodality stepwise diagnostic approach towards. Notwithstanding, test performance is always population-dependent. To correctly interpret the results, the prevalence of malignancy and the performance, costs and feasibility of the desired diagnostic in the local patient population should be known beforehand. Local implementation studies are strongly recommended to confirm clinical utility. Most importantly, the local decision favouring or opposing a certain diagnostic should be a deliberate and multidisciplinary one. Cooperation between clinical endocrinologists, endocrine surgeons, pathologists, radiologists and nuclear medicine physicians is crucial.
RkJQdWJsaXNoZXIy MTk4NDMw