*2.1. Somatic Driver Mutations*

Vogelstein et al. considered a driver mutation as a genomic variant that directly or indirectly induces a selective growth advantage [20]. As shown in Table 1, ATC and PDTC harbor many driver mutations, occurring mainly in both MAPK and PI3K-AKT pathways.

BRAF and RAS genes (HRAS, KRAS, and NRAS) are main members of MAPK pathway. Both of them occur in more than 25% of ATCs, according to catalogue of somatic mutations in cancer (COSMIC) database [21], while 15.38–33.33% and 6.8–41.2% of PDTCs harbor BRAF and RAS mutations, respectively (Table 1) [18,22–24]. Interestingly, although BRAF and RAS mutations are present in a relevant percentage of both ATC and PDTC cases, they seem to play different roles. In ATC, neither BRAF or RAS mutations seem to be sufficient to induce neoplastic cell anaplasia. McFadden et al. produced a thyroidspecific CreER transgenic mouse in order to specifically induce BRAFV600E mutation in thyroid cells; although this mutation induces PTC foci, it was capable to promote ATC tumorigenesis only in the presence of p53 mutation [25]. Likewise, KRASG12D mutation developed anaplastic foci with complete deregulation of normal thyroid follicular morphology in mice model only in association with a homozygous mutation of TSH receptor [26]. However, BRAF-RAS signaling retains a crucial role in ATC cells and its inhibition by siRNA anti-BRAF produces growth arrest in ATC cell lines [27], even stronger in combination with MEK inhibition [25]. Otherwise, the mechanisms seem to be different in PDTC: Vitagliano et al. were able to promote progression of FTC foci into PDTC in mouse model by NRASG61K mutation [28].


**Table 1.** ATC and PDTC genetic landscape: somatic mutations.

In addition to mutations of MAPK pathway, next generation analysis showed that ATC harbors higher prevalence of mutations in PI3K-AKT pathway compared to other histotypes [30]: according to COSMIC database, PI3KCA and PTEN were found mutated in 11.24% and 9.27%, respectively (Table 1) [21]. Likewise, also PDTC harbors frequently PIK3CA or AKT1 mutations (2.38–19.51% and 0–8.70%, respectively) (Table 1) [18,22–24].

Interestingly, in ATC series provided by Liu et al., the 81.3% of samples presented genetic alterations affecting both MAPK and PI3K-AKT pathways [31]. Accordingly, in mouse model, the presence of mutations occurring in both pathways induced ATC foci, confirming the synergistic interactions between these pathways [32]. On one hand, MAPK pathway has a crucial role in cell proliferation and survival, and, on the other hand, upregulated PI3K-AKT pathway has been related to tumor aggressiveness [33].

Beyond mutations occurring at members of MAPK and PI3K-AKT pathways, many variants have been reported in cell cycle regulators. Many reports showed that mutations occurring in p53 and TERT promoter (pTERT) are highly prevalent in ATC, occurring even simultaneously [17,18,21] (Table 1). Likewise, PDTC presents both mutations, even if less frequently than ATC [18,34]. Intriguingly, in the presence of an impaired cell-cycle checkpoint pathway (e.g., p53), the occurrence of a concomitant mutation in telomerase activity (e.g., pTERT) could induce an indefinite cell proliferation [35]. In addition, the interplays between the duet BRAF-pTERT have recently been described by Tan and colleagues [36]. In particular, in case of mutation of both of them, cancer cells suppress apoptosis mainly thank to pTERT activity, while in case of mutation occurring only on BRAF gene, apoptosis activity seems to be not significantly affected [36]. Accordingly, the inhibition of TERT activity could represent an Achilles heel, as recently shown in-vitro and in-vivo model by Bu et al. In these models, BIBR1532 (a TERT inhibitor) significantly inhibited tumor growth as well as cell invasion, migration and angiogenesis [37].

If regulation of cell cycle has a crucial role in oncogenesis, also protein metabolism control has been deeply involved in tumorigenesis [14,38]. Not surprisingly, both PDTC and ATC harbor EIF1AX mutations in about 10% of cases (Table 1) [18,22–24]. EIF1AX is a member of 43S preinitiation complexes, responsible of translation initiation, and its mutation has recently been involved in preinitiation complex stabilization and, further, in deregulating protein synthesis [39,40]. Interestingly, EIF1AX mutations are mutually exclusive with other drivers in PTC [19], while they co-occur with RAS mutations in ATC and PDTC [18]. Recently, Krishnamoorthy et al. showed a positive feedback relationship between RAS and EIF1AX proteins, which reinforces c-MYC gene expression [40].
