332 chapter 4 The earlier [18F]FDG-PET/CT studies repeatedly demonstrated sensitivities up to 100%, while more recent studies did report some missed cancer diagnoses [38-40, 485]. A recent meta-analysis hypothesised that the progress from stand-alone PET to hybrid PET/CT techniques likely increased the false-negative rate because PET/CT provides a better anatomical correlation [486]. Although [18F] FDG uptake in ipsilateral multinodular disease could complicate exact anatomical correlation, we consider this a highly unlikely explanation. The improved spatial resolution and decreased detection limit (now ~10 mm diameter to reliably exclude [18F]FDG-uptake) of newer PET/CT scanners likely results in fewer false-negative as well as more false-positive readings. Rather than the impact of improved technology, between-study heterogeneity may result from varying thresholds for the definition of an [18F]FDG-negative nodule, in combination with global variations in case-mix, ranging from variable malignancy rates to differences in histopathological subtypes, genomic patterns, and altered protein expression levels related to the glycolysis pathway [25, 487]. Other diagnostics can be considered for indeterminate nodules [25]. Even though they were initially developed for pre-FNAC risk assessment of thyroid nodules, various ultrasound classification systems, such as the American Thyroid Association (ATA) and Thyroid Imaging Reporting and Data System (TIRADS) classifications, have increasingly demonstrated their added diagnostic value in nodules with indeterminate cytology [488]. TIRADS assessment may also improve the diagnostic accuracy of [18F]FDG-PET/CT [39, 40]. The current study focussed purely on [18F]FDG-PET/CT: patients were only included after their indication for diagnostic surgery was established based on cytology, clinical and ultrasound parameters (in accordance with international guidelines) to prevent undesirable interference of considerations regarding ultrasound characteristics when aiming to assess the impact of [18F]FDG-PET/CT-driven management, our primary objective. At the time when the current study was initiated in 2015, TIRADS was less established and it was only implemented very limitedly in the Netherlands. Its prospective assessment was not part of the study procedures. We considered it inappropriate to retrospectively reassess baseline stored ultrasound captures as ultrasound is a dynamic technique. Molecular diagnostics are undeniably gaining traction in clinical practice and are increasingly applied in the pre-operative workup of thyroid nodules. Besides aiding the differentiation between benign and malignant, these have an added advantage of risk stratification based on the type of genetic alteration found [489]. However, few tests meet the rule-out and/or rule-in requirements for safe implementation of an ancillary test [17, 25]. [18F]FDG-PET/CT meets this rule-out criterium (i.e., false-negative rate lower than or equal to a benign (Bethesda II) cytological diagnosis), as do some commercial gene mutation classifiers with similar sensitivity. These panels appear to outperform [18F]FDG-PET/CT on specificity and benign call rate, but have major downsides with regard to their limited global availability and very high costs per patient (a Medicare reimbursement rate of $3,600=€3,109; €1=$1.18 on 01-10-2021), in addition to practical challenges concerning the required quality, quantity, and storage of the cytological material [490, 491]. In a European
RkJQdWJsaXNoZXIy MTk4NDMw