464 chapter 11 Finally, the absence of CNA alone does not exclude malignancy and does not justify withholding diagnostic hemithyroidectomy for a Bethesda III/IV oncocytic cell nodule. Instead, the results of the CNA-LOH analysis are best interpreted in combination with somatic mutation and fusion analysis results (Figure 2). Various point mutations have additionally been described in OCA, infrequently in malignancies lacking the typical GH type CNA patterns. These include DAXX, NF1, ARHGAP35, MADCAM1, ATXN1, UBXN11, TSC1/2 and CDKN1A mutations, mutations characteristic of FTC including RAS, PIK3CA, and PTEN mutations, and those that are characteristic of poorly differentiated and anaplastic thyroid carcinoma including TERT promoter, PIK3CA, PTEN, EIF1AX, ATM, and TP53 [342, 484, 649, 674, 676, 680, 681]. As in other types of thyroid carcinoma, TERT promoter mutations in OCA are associated with more aggressive tumour behaviour, distant metastasis, and tumour dedifferentiation including radioiodine refractory disease [674]. The somatic mutation and gene fusion NGS panels that were used in the current study included the most important but not all of the OCA-appurtenant molecular alterations that were previously described in literature. Yet, when no CNA and no somatic mutations or gene fusions are observed, a wait-and-see strategy with active surveillance of the nodule appears oncologically safe and should be considered. During CNA-LOH analysis using the GWLOH panel, a number of additional considerations are crucial (Table 3). First, to establish the diagnostic value of the CNA-LOH analysis, ascertaining the presence of true oncocytic cells as opposed to oncocytic cell metaplasia is key [22]. A careful morphological assessment (including immunohistochemistry) by a dedicated thyroid pathologist may accurately distinguish most true oncocytic cell lesions from other neoplastic or non-neoplastic disorders that present with oncocytic changes, such as oncocytic papillary thyroid carcinoma, oncocytic medullary thyroid carcinoma, or parathyroid proliferations [672]. Such neoplasms also show different genetic alterations [484]. The observation of an atypical CNA pattern or a somatic mutation or gene fusion that is uncommon in OCN, warrants the critical re-evaluation of the cell type and (non-)oncocytic cell origin of the tumour, including consideration of alternative diagnoses [20]. Next, the tumour cell percentage of the tested tissue sample should always be taken into consideration. Whereas the tumour cell percentage of OCN is mostly 70-80% or higher for histopathological samples, often ensuring clear amplitudes in the SNP array plots, the tumour cell percentage of cytology samples can be limited to 30-50%, resulting in smaller VAF amplitudes and/or more scattered SNP plots. In the latter cases, endoreduplication may present with less extreme amplitudes and may even go unnoticed in the assessment. As such, the GWLOH panel can indicate but not always exclude endoreduplication, and – dependent on the quality of the tissue sample – sometimes no decisive answer regarding the presence of endoreduplication may be obtained. Other techniques such as flow cytometry and LAIR analysis are more reliable for this purpose, but these are not fit for daily clinical application [343, 682]. Finally, although recognizing endoreduplication may seem critical due to its association with an unfavourable prognosis and metastatic disease, it is important to realize that metastasis have occasionally also been described for tumours without genome doubling [343, 483, 484, 649, 674-676].
RkJQdWJsaXNoZXIy MTk4NDMw