Medullary thyroid carcinoma (MTC) is a form of neuroendocrine tumors originating from calcitonin‐producing parafollicular cells (C cells). These tumors account for 5%–10% of all thyroid malignancies (1). Approximately 75% of cases are sporadic, but MTC accounts for 14% of all thyroid cancer-related deaths. MTC is prone to cervical lymph node metastasis in the early stages of the disease, and distant metastasis is detected in 10% of patients with MTC at initial diagnosis. If not treated in a timely manner, the condition will rapidly deteriorate. Surgery in combination with adjunctive therapy (e.g., radiotherapy and chemotherapy) is an effective treatment option for local MTC (1, 2). In recent years, research has identified various markers which display prognostic value in MTC. Among them, calcitonin and carcinoembryonic antigen (CEA) are used to monitor the progression of MTC. Further, it is currently thought that, following the diagnosis of MTC, a steady rise in calcitonin and CEA levels following the diagnosis of MTC is indicative of poor prognosis. However, a small proportion of healthy individuals exhibit some increase in calcitonin levels, while the CEA values in patients with MTC are mostly within the normal range. Therefore, there are certain limitations in the use of these marked for disease monitoring (3). Although numerous molecular markers have been proposed, at present, there are no absolutely effective biomarkers that can predict the prognosis of MTC.

Aurora kinase A (AURKA) is an important member of the aurora kinase family, which are involved in the regulation of cell mitosis and progression through the cell cycle (4). AURKA is primarily responsible for the replication and separation of the centrosome, the aggregation of the bipolar spindle apparatus, and the entry and exit of mitosis. In addition, AURKA plays an important role in the maturation of the centrosome and the assembly of the spindle apparatus (5). Owing to the important biological functions of AURKA, phase 2 clinical trials have been launched in recent years for the to develop targeted inhibitors (6, 7). Studies have demonstrated that the expression of AURKA is upregulated in breast cancer, prostate cancer, ovarian cancer, pancreatic cancer, and other types of tumors (8). This upregulation leads to malignant proliferation, epithelial–mesenchymal transformation, migration, and invasion of tumors cells, and has been associated with poor patient prognosis (9–11). Previous studies have further suggested that high expression of AURKA (AURKA^high) in breast cancer is a significant adverse prognostic factor for overall survival (OS). Moreover, AURKA is closely related to disease recurrence in obese patients with early-stage breast cancer (12), while AURKA^high in patients with gastric cancer has been linked to a higher risk of death (13). Given this context, the present this study aimed to investigate the potential predictive value of AURKA in MTC.

Tumor-infiltrating lymphocytes (TILs) include T and B cells that can recognize infiltrating high-risk cancer cells and promote their apoptosis. Many studies have found that TILs play a biphasic role in the occurrence and development of tumors, alternatively participating in antitumor immune responses to inhibit extension of tumor, while also functioning to mediate the production of cytokines and other molecules to form an immunosuppressive microenvironment, resulting in immune escape, and driving tumor recurrence (14–16). Furthermore, research has shown that TIL number has potential as a prognostic factor for tumor cell survival. In fact, numerous studies have reported that TILs are important prognostic factors in several types of cancer, including lung, liver, esophageal, gastric, colorectal, low-grade glioma, female breast, and cervical cancers (17–24). Furthermore, studies on colorectal cancer have shown that the individual immune cell subtype is not a good predictor for cancer prognosis (25). TIL infiltration is a coordinated and interactive process, and research has shown that the overall TIL number is predictive of OS and disease free survival (DFS) of colorectal cancer (25). Recent research has further demonstrated that patients with early-stage triple-negative breast cancer and high level of TILs have significantly lower rates of distant recurrence and longer recurrence-free survival (RFS) (26). In addition, a study of the relationship between tumor deposits and TILs in gastric cancer concluded that their interaction influenced both DFS and OS (27). MTC is an immunologically active tumor with a higher organized immune infiltration than expected (28), rendering TILs a strong possibility as favorable tumor predictors.

It is well known that immunogenic cancer cells cause immune escape by modulating the inactivation of initially recruited immune components (29). In a mouse model of breast cancer, treatment with the Aurora-A inhibitor alisertib (MLN8237) caused intensive cancer cell apoptosis, resulting in a reduction in myeloid-derived suppressor cells and tumor-associated macrophages (30). This Aurora-A inhibitor disrupted the immunosuppressive environment, without harming T cells. Further, another study showed that treatment with Aurora-A inhibitors or loss of the AURKA gene could increase the number of CD8+T cells in tumors, thus improving the prognosis of patients with colorectal cancer (31). Therefore, we speculate that AURKA is related to immune cells, and can play a role in promoting cancer by affecting the number of immune cells to a certain extent.

In this study, we evaluated AURKA expression and TIL infiltration in MTC by immunohistochemistry (IHC) and hematoxylin-eosin staining(HE), and explored their associations with clinicopathological characteristics. We hypothesized that AURKA and TILs would jointly affect the prognosis of MTC, and therefore further evaluated the predictive value of combined AURKA and TILs in patients with disease recurrence.

In this study, we established an independent cohort by randomly selecting 137 patients with MTC to investigate the predictive and prognostic value of AURKA and TILs. The expression of AURKA was associated with a high degree of MTC malignancy, as well as cervical lymph node metastasis and recurrence. We further found that the combination of AURKA with TILs exhibited significantly improved prognostic value compared with AURKA only or TILs alone in MTC. In addition, AURKA^high/TILs^low was identified as an independent prognostic factor of both BCR and STR, allowing the accurate prediction of incurable disease recurrence in patients with MTC. These results are clinically significant in terms of the adoption of individualized treatment for patients with different prognoses.

Although AURKA has been extensively investigated as a carcinogen in numerous human malignancies as a carcinogen, its prognostic value in MTC remains controversial. A pooled analysis of the AURKA mRNA levels of patients with MTC did not reveal any statistically significant alterations in AURKA expression across tumors different TNM stages. However, one study notably showed that the use of AURKA inhibitors could prevent the proliferation and tumorigenicity of MTC in vitro (35). However, another study found that AURKA expression was associated with advanced TNM stage and regional lymph node metastasis in a small cohort of 71 patients with MTC (36). As such, there is an urgent need to investigate the predictive value of AURKA for the prognosis of MTC. Given this context, in the present study, we analyzed the expression levels of AURKA in 137 patients. AURKA was not found to be associated with the T stage, but was highly expressed in patients with multifocal tumor and advanced tumor stage. Particularly, we found that AURKA^high was linked to regional lymph node metastasis and disease recurrence, suggesting that AURKA is associated with MTC aggressiveness. However, AURKA was not a significant predictive factor of RFS. Higher expression of AURKA also has limited predictive ability for RFS than lower AURKA expression (P = 0.053, HR: 2.06, 95%CI: 0.97–4.35). Nevertheless, these previous studies were characterized by limitations, namely the small sample size (only 26 patients) and measurement of only mRNA levels. Thus, we think that the present findings and conclusions are reliable because of the larger sample size included in this investigation and the measurement of quantification protein levels. There are several steps at which cells can undergo changes, including mRNA translation initiation, elongation, and post-translational modification to form a protein with a specific structure and function; all these changes can result in alterations at the protein levels. Therefore, the measurement of protein levels is more reliable than that of mRNA levels.

TILs acts as a “double-edged sword” in tumor progression, as different subsets of TILs play distinct and opposing roles in various cancers. Previous studies have proposed that TILs affect lymph node metastasis and UICC stage of urinary bladder cancer, while low TILs is predictive of poor prognosis in patients with urinary bladder cancer (37). Additionally, increased numbers of TILs in the stroma before or after therapy are typically linked to a better response to therapy, and have superior predictive value for long-term prognosis in breast cancer (26, 38–40). Considering the complex pathogenesis of thyroid cancer, its recurrence is closely related to the immune microenvironment. Further discovery of the role of useful tumor markers combined with the immune microenvironment in cancer development is expected to improve on current treatment strategies for thyroid cancer.

Thus far, it has been well-established that the infiltration of immune cells affects the recurrence of thyroid cancer (41). One study which investigated 38 thyroid cancer surgical sections, including 35 cases of papillary carcinoma and 3 of myeloid carcinoma, found that the sections of thyroid cancer patients without recurrence showed varying degrees of TILs. TILs were not observed in surgical sections of patients with recurrence (42). These results indicate that TILs may be closely related to the recurrence of thyroid cancer, which needs to be further explored in more experiments in MTC. However, thus far, only a few research studies have investigated the predictive role of TILs in MTC. In one retrospective study of 95 patients with MTC, statistical investigation of RFS and OS showed that patients lacking lymphocytic infiltration were ultimately more likely to have tumor recurrence and death (43). In our study, we also explored the relationship between TILs and relapse by examining the presence of TILs in tumor tissues obtained from 137 patients with MTC Our results showed that the ability of lower numbers of TILs alone to predict RFS may be insufficient (P = 0.14, HR: 0.59, 95%CI: 0.29–1.16). as was the predictive value of AURKA expression alone. Remarkably, patients with both AURKA^high and TILs^low were associated with an obviously shorter RFS than others. Furthermore, the AURKA^high/TILs^low phenotype exhibited a strong ability to accurately predict incurable disease recurrence with high sensitivity and specificity (i.e., 85.7% and 87.0%, respectively). We thus believe that AURKA^high/TILs^low may be a prognostic biomarker that can accurately identify patients at high risk of incurable disease recurrence and poor survival outcomes. Therefore, we hypothesized that investigation of AURKA expression and TILs quantification may serve as a predictor to guide future treatment.

In one prior study investigating a tumor marker and TILs in MTC, the authors suggested that this tumor marker (HHLA2) and TILs are closely related, which together affect the prognosis and survival of MTC (44). Similar to this study, we also explored tumor markers and TILs, but the difference is that we analyzed the overall TILs instead of a specific subtype. Because tumor lymphocyte infiltration involves a coordinated infiltration of multiple inflammatory cells, they cannot be restricted solely to observing individual subtypes. This view has been confirmed in colorectal and breast cancer, and even regulated in breast cancer (25, 45). Moreover, in this study, we found that MTC tumors with AURKA^high were characterized by extensive TIL infiltration. However, AURKA^high (HR: 2.06, 95%CI: 0.97–4.35) and high TIL infiltration (HR: 0.59, 95%CI: 0.29–1.19) indicated opposite prognostic outcomes. This contradiction may be attributed to the complex interactions which occur in the tumor immune microenvironment. AURKA^high was linked to a high degree of TIL infiltration; however, the expression of various immunosuppressive molecules, such as programmed cell death-1 (PD-1) may be upregulated, which could inhibit the antitumor effects of immune cells to exert antitumor effects (46). AURKA^high was also found to be associated with the intratumoral localization of TILs, potentially indicating that AURKA is involved in immune regulation in MTC. In a prior study, combination treatment with the AURKA inhibitor alisertib and programmed cell death-ligand 1 (PD-L1) monoclonal antibody showed a synergistic effect in terms of tumor inhibition (47). During the combination therapy with alisertib and PD-L1 monoclonal antibody, CD8+ T cells were significantly enhanced, while the levels of interferon-gamma (IFN-γ) and tumor necrosis factor-alpha (TNF-α) in CD8+ T cells were also significantly increased (48). Therefore, we hypothesized that, compared with monotherapy with an AURKA inhibitor, combination therapy with immunosuppressants could achieve better treatment effects. However, further research is warranted to investigate this hypothesis.

This study also had some limitations which should be mentioned. Firstly, owing to the retrospective nature of the study, selection bias would have inevitably impacted the results. Secondly, patients who did not undergo surgery were not included because only patients who undergo surgery had access to complete pathological information. Thus, an inherent selection bias existed. In addition, we observed a relatively good overall prognosis of MTC in this study, with only ten disease‐related deaths). Therefore, it was not possible to evaluate the association between the degree of AURKA and mortality due to the insufficient number of events. Finally, AURKA may play an important role in the regulation of the tumor immune microenvironment; however, additional research studies are warranted to identify the potential mechanism involved in this process.

AURKA expression was associated with the malignant phenotypes of MTC. The prognostic value of the combination of AURKA with TILs was found to be greater than those of either individual marker. The AURKA^high/TILs^low phenotype was identified as an independent factor of BCR and STR in MTC, and could be used to accurately identify patients at high risk of incurable disease recurrence. However, the present conclusions and the actual efficacy of AURKA inhibitors should be further investigated.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/supplementary material.

The studies involving human participants were reviewed and approved by the Ethical Committee of the Tianjin Medical University Cancer Institute and Hospital. The patients provided written informed consent to participate in this study. Human samples were collected following approval by the Ethics Committee of Tian Jin Medical University Cancer Institute and Hospital, and informed consent was obtained from all patients.

ZW: Data curation, Formal analysis, Writing – original draft, Writing – review & editing. FG: Methodology, Writing – review & editing. GF: Investigation, Writing – review & editing. ZZ: Software, Writing – review & editing. NK: Validation, Writing – review & editing. XH: Supervision, Writing – original draft. XZ: Resources, Writing – original draft, Writing – review & editing.