Differential expression profiles of immunoregulatory genes in anaplastic thyroid carcinomas with a coexistent papillary carcinoma component

Thyroid carcinoma is the most common endocrine neoplasm, in which thyroid follicular cells transform through a multi-step process, resulting in several histological types, from well-differentiated thyroid carcinoma (WDTC) (including papillary (PTC) and follicular thyroid carcinomas (FTC)) to high-grade follicular cell-derived non-anaplastic thyroid carcinoma (including poorly differentiated carcinoma (PDTC) and high-grade differentiated thyroid carcinoma) and to anaplastic thyroid carcinoma (ATC). PDTCs and ATCs are rarer than WDTCs, but clinically more aggressive, with a short median survival time (5 years and 6 months, respectively) [1, 2]. In the past 40 years, several studies have recognized a well-differentiated component in a significant number of PDTC and ATC cases, suggesting the existence of evolutionary pathways of progression from well-differentiated to poorly differentiated or undifferentiated forms. In particular, a PTC component represents the most commonly observed histotype in association with ATC [3‐6]. At the molecular level, a progressive increase of tumor mutational burden exists from WDTC to PDTC and ATC [7‐9]. BRAF^V600E and other molecular drivers, such as RAS mutations or RET and NTRK rearrangements, are typically associated with PTCs. The same alterations can be observed in PDTCs and ATCs with similar prevalence rates [8]. Moreover, in high-grade carcinomas, additional mutations occur (including TERT-promoter and TP53 mutations) that have been proposed to play a major role in tumor progression [10, 11]. A study by Ragazzi et al. highlighted that coexistent PTC and ATC components are characterized by common early driver (BRAF, PIK3CA, RAS) and late (TERT-promoter) mutations, whereas TP53 mutations occur almost exclusively in the ATC component [12]. These findings suggest that non-genomic mechanisms may concur in tumor progression-related mechanisms in thyroid cancer. Among them, a relevant aspect to unravel the progression mechanisms of well-differentiated to poorly differentiated and undifferentiated carcinomas is the study of the tumor immune microenvironment. Giannini et al. showed histotype-specific signatures of immune-related genes, with a significantly upregulated in ATCs [13]. Moreover, in the tumor microenvironment of the majority of the analyzed ATC cases and in around half of the PTCs, an infiltrate of macrophages and CD8+ T cells was observed, with a functionally exhausted profile (the so called “hot” tumors). Interestingly, PDTCs and part of PTCs showed lower levels of immune-related gene expression and low-to-absent immune cell infiltration, enhanced by the presence of non-inflamed CD8+ T cells (referring such cases as “cold” tumors). In addition, it was observed that in ATCs and PTCs, an upregulation of immune checkpoint (i.e., PDL1, PDL2, PD1, LAG-3, TIM3, PVR, and TIGIT) inhibits the immune response. 

Based on the above, our work aims (i) to study the differential immune-related gene expression profile of ATC and PTC components in combined tumors, employing nCounter® PanCancer Immune Profiling Panel; (ii) to evaluate the status of the checkpoint inhibitor PD-L1 in both components; and, finally, (iii) to assess if miRNA may interfere with the regulation of the immune-related gene expression signature.

In this study, we investigated the transcriptional signature of cases of ATC associated with PTC component using Nanostring-based gene expression and miRNA global profiling. We observed that only 11 out of more than 700 immune-related genes were significantly de-regulated compared to the two carcinoma counterparts. This mild difference in gene expression profiles may reflect an intrinsic aggressiveness already owned by the papillary component, which is already committed to progression and loss of differentiation. However, the differentially expressed genes showed a similar profile in most cases, suggesting that their up- or downregulation represents a potential biologically meaningful feature. Moreover, the histopathological assessment of TILs did not significantly differ when comparing PTC and ATC components within the same tumor but was rather case-specific and was not directly correlated with immune-related gene signatures. Among the six upregulated genes in ATC components, RIPK2 (receptor-interacting protein kinase 2) has been reported to be highly expressed in some cancer types, such as bladder urothelial, breast invasive, and thyroid carcinomas [18‐20]. RIPK2 not only plays an important role in inflammatory and immune diseases [21] but is also involved in tumor invasion and metastasis [22‐24] and promotes immune cell infiltration, especially in thyroid carcinoma, renal clear cell carcinoma, and testicular germ cell tumor [25]. Song et al. demonstrated that RIPK2 was positively correlated with the expression of some immune checkpoint markers, such as PD-1, PD-L1, CTLA-4, and TIGIT, suggesting that RIPK2 expression correlates with immune escape strategies [20]. ITGA5 and ITGB1 (αV and β1 integrins) mediate cell adhesion and promote survival, proliferation, and migration of tumor cells, thus contributing to tumor progression and metastasis [26, 27]. Aberrant upregulation of integrin αVβ1 has been observed in several human malignancies, and it is closely correlated with a poor prognosis [28]. CCL3L1 encodes for a pro-inflammatory chemokine, and in addition to being involved in the chemotaxis of immune cells, an upregulation of this molecule has been observed in association with PD-L1 expression in high-grade muscle-invasive urothelial carcinoma of the bladder, indicating a potential role of this in tumor immune tolerance and tumor progression [29, 30]. PLAUR encodes for the urokinase plasminogen activator receptor (uPAR). Some data show that, in aggressive thyroid forms, there is a trend towards an increased expression of uPAR as compared to WDTC [31]. Li C.W. et al. demonstrated that PLAUR and three other immune-related genes (PRKCQ, PSMD2, and BMP7) play an important role in the de-differentiation process in thyroid carcinomas [32]. Accordingly, they observed an upregulation of PLAUR in ATC compared to PTC. In addition, this molecule showed a positive correlation with the expression of PD-1, CTLA-4, TIGIT, LAG-3, and TIM-3, indicating its involvement in the immune exhaustion process in the tumor microenvironment. TICAM2 encodes for an adaptor protein of TLR, in particular for TLR4. TICAM-2/TICAM-1 complex (MyD88-independent pathway) is necessary for the activation of TLR4 signaling, leading to the activation of NF-κB and interferon regulatory factor (IRFs) signaling, resulting in the regulation of innate and adaptive immune responses and inflammation [33, 34]. In particular, dysregulation of TLR signaling is associated with an exacerbated production of pro-inflammatory cytokines involved in tumor progression, as described in several neoplastic processes [35]. Overall, these data support the role of the above-mentioned genes in PTC as mediators favoring tumor progression to ATC. Regarding the downregulated genes in ATC, three of them, namely PRKCD, CYFIP2, and BLNK, were described to have tumor suppressor activity [36‐38] and their downregulation is in line with the more aggressive behavior that characterizes an anaplastic cancer. The downregulation of MAP3K1 in the ATC component was partially unexpected, since data in thyroid cancer show its role in promoting PTC formation [39]. However, a tumor suppression role has also been postulated for MAP3K1, at least in the breast cancer model [40]. Finally, EPCAM (epithelial cell-adhesion molecule) plays a role in proliferation, differentiation, and migration. Interestingly, ATC has a loss of the EpCAM extracellular domain and an increased accumulation of the intracellular domain in nuclear and cytoplasmic compartments. In contrast, PTC possesses only the EpCAM extracellular domain on the cell membrane. This may suggest a possible involvement of EpCAM in the progression from an indolent to an aggressive phenotype in thyroid carcinoma [41, 42]. These data suggest the existence of different immunoregulatory mechanisms playing in the process of loss of differentiation in PTC, resulting in an enrichment of immune mediators in ATC. In line with the study by Giannini et al. [13], the immune signature related to the macrophages score was the most significantly upregulated pathway in ATC compared to PTC samples. Cancer immunotherapy has proven to be an effective therapeutic option in some human carcinomas over the last decade [43, 44]. The use of immune checkpoint inhibitors, such as anti-CTLA-4, anti-PD-1, and anti-PD-L1 antibodies, allows the reactivation of an adaptive anti-tumor immune reaction and opens the development and clinical validation of immunotherapy in thyroid cancer [45]. Accordingly, in ATC, in addition to the current therapeutic options, which include surgery, radiotherapy, chemotherapy, and multi-kinase inhibitors, the use of immune checkpoint inhibitors has been explored in cases showing positive PD-L1 expression [46‐48]. In our series, no statistically significant difference in CD274 gene expression between the anaplastic and papillary components was observed. However, using immunohistochemistry, we observed an increase in PD-L1 expression in most ATCs compared to the respective PTC components, a finding that was not significantly correlated with the presence of TILs. PD-L1 expression in ATC samples was also more frequent compared to PDTC. Based on the observation that, in the present series, PDTC samples generally had a downregulated profile of immune-related genes, we also checked whether PD-L1 status was interfering with the immune gene expression profiles between ATC and PDTC. Interestingly, PD-L1-negative PDTC and ATC showed a rather stable gene expression signature, whereas PD-L1-positive ATC cases had a significant upregulation of a higher number of genes compared to PD-L1-positive PDTC. The final aim of the study was to assess the differential expression of miRNA in ATC compared to PTC components and the possible interaction with immune-related gene signatures. We clearly showed that miRNA expression profiles were mostly histotype-specific rather than case-specific. However, our data reinforced a major role for the de-regulation of the MAPK pathway as a hallmark of tumor progression from PTC to ATC. In fact, a subset of miRNAs upregulated in ATC is specifically targeting MAP3K1, which we found downregulated in PTC samples. Moreover, the MAPK signaling pathway is the second most relevant pathway impaired by significantly de-regulated miRNAs in ATC samples.

The study, however, presents a main Limitation that deserves a few considerations. The current series of only 12 cases may limit the overall significance of data, with special reference to statistical comparisons. However, it reflects the low incidence of ATC [49] (being cases with an associated PTC component even rarer), with special reference to cases with localized/regional disease, thus with a potential indication for surgery, that worldwide represent less than 25% of overall ATC cases. Similar to our study, the paper by Ragazzi et al. [12], conducted at an institution with a high volume of thyroid cancer diagnoses, collected only eight cases of ATC with an associated PTC component within a 42-year period (1978 to 2020).

In conclusion, we explored the immune-related gene expression profiles using NanoString Technology in a series of thyroid carcinomas with coexisting ATC and PTC components. Overall, our study revealed a subset of genes and pathways that are consistently impaired during PTC to ATC progression. However, further studies are needed to explore their functional mechanisms and to evaluate their role as potential biomarkers.