*2.3. Copy Number Variations*

In oncology, copy number variations (CNVs) are well characterized as prognostic factors for recurrence and death [47]. This evidence has been confirmed also in advanced thyroid cancer [18]. If they are quite rare in differentiated thyroid cancer (less than 10%) [19], in PDTC and ATC they are widespread, especially in cancers without known driver mutation (losses of 1p, 8p, 13q, 15q, 17p, 22q, and gains of 1q and 20q) [18]. Interestingly, they seem to be hystotipes-specific: 8p and 17p losses and 20q gains are more frequent in ATC while loss of 1p was substantially more recurring in PDTCs [18]. Moreover, CNVs correlate with gene context where occur: 1p, 13q, and 15q losses were enriched in PDTCs without known driver mutation while loss of 22q was associated with RAS-mutated PDTCs [18]. In ATC, beyond large chromosomal variations, Pozdeyev et al. reported more restricted CNVs such as losses of CDKN2A and CDKN2B or amplification of KIT, PDGFRA and KDR, further confirmed by other authors [17,24].

Finally, since CNVs have recently been related to resistance to target therapies in thyroid cancer [48], it would be very interesting to ascertain if some of them (e.g., PDGFRA amplification) could induce resistance to target therapy (e.g., multikinase inhibitors, MKIs) in ATC and PDTC.
