In this study, we demonstrated that the genetic profiles of T4 staged WDTC has genetic alterations, somatic mutations, and a relevant pathway compared to the less aggressive WDTC by analyzing a TCGA cohort. T4 staged WDTC presented 16 genetic alterations compared to less aggressive thyroid cancers, and we identified somatic mutations associated with locally invasion transformations. We also identified seven pathways related to 16 genetic markers. Moreover, we enhanced the current knowledge about the genetic characteristics of locally advanced and resectable WDTC.
WDTC, such as follicular and papillary carcinomas, account for 95% of all thyroid cancer cases and are generally have a survival rate of almost 100% when diagnosed early.(20) Approximately 25% of patients with WDTC develop locally advanced or metastatic disease and locally advanced WDTC is managed differently with adjuvant therapy and surgical treatment.(21) Post-surgical management includes RAI therapy to reduce these locoregional and distant metastasis. When RAI therapy becomes ineffective against differentiated thyroid cancer (DTC), the five-year survival is < 50% and 10-year survival is < 10%.(21, 22) Of 80 patients with radioactive iodine-refractory DTC, 38 had somatic mutations (47.5%), including BRAFV600E, RAS, TP53, MET, PIK3CA, GNAS, and TPO.(23) The current study identified the incidence of somatic mutation in BRAFV600E at 59.8%, which is similar to the TCGA cohort (58.0%) in 2014, whereas only 8.4%, 3.5%, and 0.7% of tumors had RAS mutations (NRAS, HRAS, and KRAS, respectively). The first TCGA cohort discovered two molecular subtypes, BRAFV600E-like and RAS-like, in the papillary thyroid carcinoma.(13) Dedifferentiation is likely to play a role in mitigating responses to RAI treatment and is consistent with the BRAF mutation.(24) However, our results showed no difference in the BRAF mRNA expression level and the frequency of somatic mutations between the two clusters.
In our study, 23 patients with pathological T4 classification had significantly elevated expression of CDKN2A and significantly lower JAK2 compared with Cluster2. Expression of BRAF, RAS, RET, and ALK was not different between the two subgroups. CDKN2A and CDKN2B are negative cell-cycle regulators, and their loss due to copy number alterations, inactivating mutations, or epigenetic silencing is one of the most frequently encountered genetic events in human cancer.(25) In thyroid cancer, genetic alterations of CDKN2A and CDKN2B were seen more frequently in anaplastic thyroid cancer compared with PTC, suggesting a potential role in anaplastic transformation.(17) The prognostic significance of JAK2 has been well described for a variety of cancers, but not advanced thyroid cancer. JAK2 plays a critical role in the signaling of prolactin (PRL) hormone, which may be involved in the development of medullary thyroid carcinoma.(26) Peng et al.(27) examined the TCGA database containing 9,315 tumor samples from 31 cancer types to study the relationship between the mRNA expression of JAK2 and PD-L1 expression. They found that high JAK2 expression was associated with high mRNA expression of PD-L1, which is likely a good indicator for immunotherapy response, including anti-PD-1/PD-L1 therapy. Our results of significantly lower JAK2 expression in Cluster1 compared to Cluster2 may contribute to a treatment strategy that can screen patient populations that may benefit from immunotherapy.
A recent study identified 676 genes associated with an increased risk of PTC recurrence in TCGA data.(28) Chien et al. demonstrated that downregulation of the sodium-iodide symporter SLC5A5 (NIS) was the strongest predictor of a decreased recurrence-free survival.(28) Our study and previous studies have shown the associations of KIT, TFF3, and TG low with increased recurrence rates. Most of the recently discovered targeted therapies inhibit the known oncogenic mechanisms in thyroid cancer initiation and progression, such as the MAPK pathway, PI3K/Akt-mTOR pathways, or VEGF. (29) In clinical trials using trametinib with VEGF inhibitors in advanced DTC patients, 33% of subjects showed a partial response (PR) and 50% had stable disease.(30) Bible at el. reported significant activity of single-agent VEGF inhibitor in progressive RAI refractory DTC, with an overall confirmed PR rate of 49%.(31)
The present study revealed that mutations of TERT, EIF1AX, and ATM were found more frequently in Cluster1 than Cluster2. The TERT gene locus was amplified in various cancers, including lung, breast, and cervical cancer,(32) and the prevalence of TERT promoter mutations in thyroid cancer was more frequent in poorly-differentiated and anaplastic thyroid cancers.(33) A recent study identified an EIF1AX mutation in a nodule ultimately diagnosed as oncocytic thyroid carcinoma.(34) EIF1AX mutations are predictive of worse survival in WDTC.(35) This suggests that the treatment strategy can be predicted by markers found in advanced WDTC, unlike the mutations that are the major drivers of WDTC.
Table 3 shows the KEGG pathways of our 16 genes. The growing body of evidence demonstrating the involvement of the PI3K-AKT signaling pathway in thyroid carcinogenesis and drug resistance led to the discovery of several agents targeting key members of this cascade.(29) Multi-kinase inhibitors targeting highly expressed tyrosine kinases in thyroid cancer cells demonstrate a promising anti-tumor activity in vitro and in vivo.(36, 37) The few targeted therapies that have been proven effective for advanced WDTC, sorafenib, selumetinib, pazopanib, and sunitinib, have promising results. Due to its high affinity to the VEGF receptor, lenvatinib showed high anti-angiogenic effects in the thyroid cancer mouse model.(38)
This study has some limitations. High post-ablation stimulated thyroglobulin level (≥ 1ng/mL) and macroscopic ETE were reported as negative prognostic factors in T4a staged WDTC.(39) It is difficult to consider these clinical and pathological factors using TCGA data. Specific clinical factors may help delineate the genomic landscape related to aggressive tumor behavior. Another limitation of this study included not controlling for histologic variants or tumor nodals positive of the well that are differentiated for the thyroid cancer studied. Our study aimed at identifying driver genes and somatic mutations instead of developing a validated prognostic factor in locally advanced WDTC. We hypothesized that using pathway analysis would provide a better understanding of the thyroid carcinogenesis involved in T4 stage tumors.