Immunotherapy in Thyroid Cancer: Current Strategies and Challenges.

1
Introduction
Thyroid cancer (TC) is the most prevalent endocrine malignancy [1], and its incidence has been steadily rising worldwide in recent decades. Over the past three decades, the global incidence of TC has increased by nearly 87.5% [2], driven by factors such as excessive iodine intake, overdiagnosis, exposure to ionizing radiation, environmental endocrine disruptors, and a personal or family history of thyroid disease [3]. The increasing incidence and associated morbidity of TC place a substantial burden on healthcare systems and society.
TC can be broadly classified into differentiated thyroid carcinoma (DTC), anaplastic thyroid carcinoma (ATC), poorly differentiated thyroid carcinoma (PDTC), and medullary thyroid carcinoma (MTC). DTC is the most common subtype, comprising both papillary thyroid carcinoma (PTC)—which is typically associated with BRAF/RAS‐like mutations—and follicular thyroid carcinoma (FTC), both of which arise from thyroid follicular cells. ATC is an aggressive and poorly prognostic form of TC with a rare incidence, while PDTC represents an intermediate subtype, exhibiting more aggressive behavior than DTC but a better prognosis than ATC. MTC remains a distinct subtype, originating from parafollicular C cells, as defined in the 2022 WHO Classification of Endocrine and Neuroendocrine Tumors [4].
Current treatment strategies for TC primarily include total or near‐total thyroidectomy, radioactive iodine‐131 (I131) therapy and thyroid stimulating hormone (TSH) suppression. However, these therapies have their limitations: surgical treatment carries inherent risks and potential postoperative complications; I131 therapy can impact patients' long‐term health, including an increased risk of second primary cancers [5]; and TSH suppression therapy requires long‐term medication, which may increase fracture risk and cardiovascular issues, particularly in older patients [5]. More critically, these treatments have limited efficacy in the management of recurrent or metastatic DTC and in more aggressive forms such as PDTC and ATC [6].
In response, new treatment modalities, including targeted therapies and immunotherapy, have emerged as promising options for TC management and are already being applied in the treatment of various cancers. Unlike conventional therapies, immunotherapy harnesses the body's immune system to identify and destroy cancer cells. Recent advancements in understanding the immunological landscape of TC have paved the way for the development of novel immunotherapeutic approaches. These strategies hold significant potential to improve outcomes in TC, especially in cases resistant to traditional treatments. The growing emphasis on personalized immunotherapy offers hope for more effective and targeted treatments for TC patients [7]. This review aims to provide a comprehensive update on the latest immunological mechanisms and clinical trial developments in TC, highlighting the transformative potential of immunotherapy in managing this complex malignancy.

2
Tumor‐Associated Immune Cell Characteristics
TC is infiltrated by a variety of adaptive and innate immune cells that exert both pro‐tumorigenic and anti‐tumorigenic effects. The concept of cancer immunoediting, proposed by Dunn and Schreiber in 2002, suggests that the immune system not only identifies and eliminates tumors but also aids in enabling tumor immune evasions [8]. Recent advances in immunotherapy have affirmed the potential for treating relapsed and refractory cancers. However, therapeutic efficacy remains limited, partly due to the cancer's ability to evade immune surveillance and adapt to immune pressure. Several mechanisms contribute to immune evasion in cancer, including T lymphocyte (T cell) exhaustion, upregulation of inhibitory receptors such as PD‐1, cell‐mediated repression (e.g., by regulatory T cells [Tregs]), secretion of suppressive cytokines, nutrient depletion, metabolic dysfunction, immune escape, and immunosuppression within the tumor microenvironment (TME) [9]. As integral components of the TME, studies have shown that Tregs, neutrophils, dendritic cells, M2 macrophages, and resting mast cells play a tumorigenic role in the PTC microenvironment, while CD8+ T cells, CD4+ memory T cells, B cells, M1 macrophages, and NK cells portray a protective role [10, 11, 12] (Figure 1).
2.1
T Cells in TC
T cells are a class of lymphocyte that mature in the thymus and can be classified into helper T cells, cytotoxic T cells (CTLs), and Tregs. Studies have demonstrated that in PTC patients harboring BRAF mutations, the expression of tumor‐infiltrating lymphocytes (TILs), particularly CD4+ T cells and Tregs, is elevated [13]. Furthermore, Wang et al. [14] reported that in progressive PTC, CD8+ T cells exhibit increased levels of exhaustion compared to adjacent normal tissues. The co‐occurrence of these exhausted CD8+ T cells with elevated Treg levels suggests impaired immune surveillance. Another study found that, in comparison to multinodular goiter and peripheral blood samples, PTC tissues contain a higher abundance of Tregs, with recurrent patients showing an increased proportion of effector Tregs [15]. CXCL13+ T lymphocytes, enriched in ATC, may facilitate the formation of early tertiary lymphoid structures, playing a central role in ATC's response to immunotherapy [16]. γδ T cells also hold significant potential, as their infiltration inversely correlates with the differentiation status of TC [17]. One study proposed that the cell surface glycoprotein dipeptidyl peptidase 4 (DPP4) is positively correlated with the infiltration of various immune cells, including B cells, CD8+ T cells, neutrophils, macrophages, and dendritic cells [18]. Moreover, Jing et al. [19] discovered that inhibiting DPP4 alleviates CD8+ T cell exhaustion and reduces IL‐13 secretion, effectively disrupting the IL13‐IL13RA2 axis. This disruption promotes mesenchymal‐to‐epithelial transition in PTC cells, influencing both immune responses and tumor plasticity.
Current immunotherapy efforts largely rely on the targeted protection provided by tumor‐reactive T cells, primarily through immune checkpoint inhibitors (ICIs), such as PD‐1/PD‐L1 and CTL‐4, as well as adoptive cell therapies, such as chimeric antigen receptor (CAR) T cell therapy [9], which will be discussed in detail later in this review regarding their progress in TC.

2.2
Macrophages in TC
Macrophages are the most abundant cell population in the TME and are classified into M1, M2, and unstimulated M0 types based on their activation state and functional roles [20]. M1, classically activated macrophages, primarily produce pro‐inflammatory cytokines and exhibit anti‐tumor functions, whereas M2, alternatively activated macrophages, produce immunosuppressive cytokines that promote tumor progression [21]. Studies have shown that the infiltration of macrophages in PTC is significantly higher than in adjacent normal tissues or benign tumors [12, 22]. Tumor‐associated macrophages (TAMs), which often refer to both M1 and M2 types, contribute to tumor progression as both “immune suppressors” and “tumor promoters,” highlighting the importance of identifying macrophage‐specific targets to optimize current immunotherapies [23].
TAMs play a critical role in TC, especially in ATC, implicate to immune suppression [24, 25, 26], angiogenesis [16, 27], invasion [27] and therapy resistance [28]. In ATC, more than 50% of the infiltrating cells are TAMs, which are polarized into pro‐tumor M2 types by paracrine signals secreted by ATC cells [29]. ATC‐specific macrophage subpopulations, such as IL2RA+ VSIG4+ TAMs, co‐expressed M1 and M2 markers, are associated with BRAF and RAS signaling, as well as increased TILs like B cells, CD8+ T cells and Tregs. High infiltration of IL2RA+ VSIG4+ ATAMs correlates with a favorable patient prognosis [30]. The M2 macrophage subtype has garnered particular attention in the search for new therapeutic targets. In xenograft models of ATC, Palacios et al. [31] found that 24%–28% of CD45+ immune cells were macrophages, with 40% of these macrophages exhibiting an M2‐like phenotype. In human ATC, macrophage marker expression was positively correlated with T‐cell immunoglobulin and mucin‐domain containing protein‐3 (TIM‐3) expression, which may represent a newly identified immune checkpoint in macrophages. In addition to M1 and M2 macrophages, Li et al. [32] identified SPP1+ macrophages in TC with immunosuppressive functions, and CD14+ monocytes were found to contribute to tumor progression and angiogenesis. The SPP1‐CD44 and MIF‐CD74 axes, mediating communication between SPP1+ macrophages and T cells, could potentially reverse the immunosuppressive TME, enhancing the efficacy of immunotherapy.

2.3
Myeloid‐Derived Suppressor Cells in TC
Myeloid‐derived suppressor cells (MDSCs) are a heterogeneous population of immature myeloid cells [33], which have been identified to be one of the major contributors of immunosuppressive populations in TME [34]. Major subtypes include early, polymorphonuclear, and monocytic MDSCs, and although they differ in their phenotype and potency, they share a common ability to suppress effector T cell activity, similar to M2‐polarized TAMs [35]. Research has found that in the bone marrow, myeloid cells in TC patients undergo significant transcriptional and functional changes prior to tumor infiltration. These changes are already initiated in the bone marrow, suggesting that they play an active role in the formation of the TME [36]. Additionally, some scholars have proposed that BRAF V600E promotes the development of TC by facilitating the recruitment of MDSCs [37]. Studies have shown that the levels of circulating MDSCs detected in peripheral blood of TC patients are closely associated with tumor staging [35] and poor prognosis [38]. Specifically, for ATC, a study utilized multiplex immunofluorescence and immunohistochemistry techniques to retrospectively analyze surgical specimens from 26 ATC patients, proposing that the presence of MDSCs in the ATC TME may lead to resistance to anti‐PD therapy [39]. Additionally, long‐term survivors of ATC had lower tumor‐infiltrating MDSC counts, suggesting that MDSCs play a role in preparing the pre‐metastatic microenvironment before distant metastasis manifests clinically [40]. However, other studies have found that when mice tumors exhibit the greatest response to combination therapy with BRAF inhibitors and anti‐PD‐1/PD‐L1 antibodies, there is a significant reduction in the monocyte MDSC‐like cell component, but MDSC‐like cells remain low during tumor regeneration. This suggests that MDSCs may play a role in the response to combination therapy but not in resistance to combination therapy [41].

2.4
Other Immune Cells in TC
Beyond T cells, macrophages and MDSCs, other immune cells such as dendritic cells [14, 42], mast cells [43], natural killer (NK) cells, and eosinophils [44] also exhibit distinct patterns in TC. For instance, the infiltration of activated dendritic cells and M0 macrophages is increased in PTC compared to normal tissues, whereas the infiltration of activated NK cells and eosinophils is decreased. These patterns of immune cell infiltration are closely associated with the tumor's clinicopathological features and may influence patient prognosis [45]. Moreover, the cell surface glycoprotein DPP4, also known as CD26, positively correlates with the infiltration of B cells, CD8+ T cells, neutrophils, macrophages, and dendritic cells. Inhibiting DPP4 can alleviate CD8+ T cell exhaustion, impact immune responses, and influence tumor plasticity [46].

3
Immune Checkpoints in TC
In the TME, immune surveillance is often suppressed through two main mechanisms: signaling inhibition and metabolic suppression. Signaling inhibition involves tumor cells downregulating the activity of stimulatory immune receptors while upregulating the activity of inhibitory immune receptors. Over the past few decades, numerous inhibitory immune receptors have been identified and studied in cancer, including PD‐1, CTLA‐4, LAG3, TIM‐3, TIGIT, and BTLA. These receptors are collectively known as “immune checkpoints” [47]. Most immune checkpoint molecules are expressed on cells of the adaptive immune system as well as those of the innate immune system. However, the expression of immune checkpoint molecules on tumor cells plays a critical role in maintaining several malignant behaviors [48]. Blocking the activation of inhibitory immune receptors has been shown to re‐activate the anti‐tumor functions of immune cells. This concept has been experimentally validated and is now applied in the clinical treatment of various cancer types [49, 50]. By analyzing the expression of immune checkpoint molecules through tissue staining, researchers can gain deeper insights into the mechanisms of immune evasion in TC and identify patients who may benefit from immunotherapy.
3.1
PD‐1/PD‐L1 Axis
The PD‐1/PD‐L1 axis is one of the most studied immune checkpoint pathways. PD‐1 is typically expressed on TILs, while PD‐L1 is expressed on antigen‐presenting cells and tumor cells. The interaction between these immune checkpoint proteins plays a key role in regulating immune checkpoints, tumor immune evasion, and T‐cell function restoration, and serves as a useful but imperfect biomarker for predicting patient responses to anti‐PD‐1 or anti‐PD‐L1 antibody treatments across various tumor types [51]. PD‐L1 overexpression on tumor cells binds to PD‐1 on T cells, transmitting inhibitory signals that lead to T‐cell exhaustion (e.g., reduced proliferation and cytokine secretion), ultimately resulting in tumor immune evasion [52]. This interaction forms the biological basis of immune surveillance escape in tumors [53]. PD‐L1 can also modulate metabolic pathways (e.g., mTOR signaling) and participate in DNA damage response, regulating gene expression via intracellular signaling to promote tumor survival and progression [54, 55]. The expression of PD‐1/PD‐L1 is regulated by post‐translational modifications such as phosphorylation, ubiquitination (e.g., MDM2‐mediated PD‐1 degradation), and glycosylation [56, 57, 58], offering potential targets for developing new inhibitors. Interestingly, tumor‐targeted PEG‐PD1‐PDL1 fusion proteins can block the PD‐1/PD‐L1 interaction while reducing T‐cell exhaustion, achieving an 88.9% tumor suppression rate in animal models [59].
The expression of PD‐1/PD‐L1 in TC varies significantly across different cancer types and is influenced by factors such as sample size, antibody type, and positive threshold (Table 1). Over the past 5 years, the most research has been conducted on ATC, with positive rates ranging from 47% (7/15) [78] to 80% (8/10) [63]. Most large‐scale studies reported positive rates between 60%–80%, which is significantly higher than in other tumor types. For PDTC, reported PD‐L1 positivity rates are generally lower, with multiple teams reporting no PD‐L1 positivity in their tested PDTC samples [60, 61]. When the positivity threshold was set to tumor proportion score (TPS) ≥ 5%, the highest reported positivity rate was only 47% [68]. For FTC, reported PD‐L1 positivity rates ranged from 10% [66] to 59.72% [69]. For PTC, the reported PD‐L1 positivity rates vary significantly, with several studies of larger sample sizes reporting positivity rates around 33% [64, 70, 72]. Studies on PD‐L1 staining in MTC are also numerous, but the reported positive rates for MTC are lower, all below 20%.
The difference in PD‐L1 positivity rates is one of the factors that affect the efficacy of anti‐PD‐1/PD‐L1 immunotherapy. However, the efficacy among different TC types remains complex due to variations in subtypes and individual differences. There are many other possible causes of the development of PD‐1/PD‐L1 antibody resistance. For example, while PD‐L1 expression is relatively high in ATC, existing studies have shown that PD‐1 expression on CTLs in ATC is typically low or absent [39, 67, 79], and this “target‐missing” situation is detrimental to treatment efficacy [39]. In addition, TC typically has a low tumor mutational burden (TMB), such as in PTC and MTC [80, 81]. A low TMB implies fewer neoantigens produced by the tumor, and the absence of tumor‐specific antigens may hinder the effective activation of immune responses by anti‐PD‐1/PD‐L1 therapy [82, 83]. Additionally, changes or interactions among components of the TME [39, 84, 85, 86, 87], expression of other alternative immune checkpoints [88, 89], as well as abnormalities in cellular signaling [90, 91] or antigen processing and presentation [83, 92], which have been reported in other cancers [93], may also influence the efficacy of anti‐PD‐1/PD‐L1 immunotherapy in TC.

3.2
B7‐H3
B7‐H3, also known as CD276, is a member of the B7‐CD28 family and has both anti‐tumor and pro‐tumor effects [94]. It is rarely expressed in normal tissues but is highly expressed in several human cancers, including TC [95, 96]. Clinical trials targeting B7‐H3 have mainly focused on other solid tumors (e.g., prostate cancer, lung cancer), with less research on TC. Studies have found that B7‐H3 is significantly upregulated in high‐risk TC (e.g., poorly differentiated or undifferentiated types), and its expression level is positively correlated with tumor invasiveness, metastasis risk, and poor prognosis [64]. Similar to PD‐L1, B7‐H3 may promote tumor progression by inhibiting T‐cell function and facilitating immune evasion. Overexpression of B7‐H3 is also associated with the infiltration of immunosuppressive cells (e.g., TAMs) in the TME, further weakening anti‐tumor immune responses [94, 97]. Hińcza‐Nowak et al. [98] found that B7‐H3 expression in MTC cells was three times higher than in normal tissues. Song et al. [94] demonstrated that B7‐H3 expression was relatively weak in MTC and FTC, while PTC and ATC exhibited moderate to strong expression. Zhao's study [99] on 343 PTC patients revealed that 84.8% (291/343) had B7‐H3 positivity, with higher expression levels of B7‐H3 mRNA and protein observed in tumor tissues compared to adjacent non‐tumor tissues, suggesting its potential as an independent predictor of recurrence‐free survival.

3.3
CTLA‐4
CTLA‐4 is another key immune checkpoint that fine‐tunes T‐cell activation and immune tolerance. It has been used by tumors to induce an immune‐suppressive state and promote tumor growth [100]. CTLA‐4 primarily acts during the early phases of immune tolerance induction, whereas PD‐1 plays a role in maintaining long‐term tolerance [101]. Anand's study [62] showed CTLA‐4 positivity in 6.6% of PTC cases (6/90), with co‐expression of PD‐L1 and CTLA‐4 in three PTC cases, highlighting the potential of combination immunotherapy. Another study reported CTLA‐4 positivity in 12.5% (25/200) of MTC patients [102]. Moreover, CTLA‐4 expression was significantly higher in BRAF‐mutant cases compared to BRAF‐wild‐type cases [13]. Further research is needed to fully understand the mechanisms of CTLA‐4 in TC.

3.4
Other Immune Checkpoints
Beyond the commonly studied PD‐1/PD‐L1, B7‐H3, and CTLA‐4, other immune checkpoint molecules, including intercellular adhesion molecule 1 [103], T‐cell immunoglobulin and TIM‐3 [102], indoleamine 2,3‐dioxygenase 1 [104, 105], CD73, and CD200 [13], have also garnered attention in TC. A study with a human ATC cell line reported that the CD56hiCD16hi/lo NK cells exhibited higher PD‐1 and TIM‐3 expression while demonstrating lower cytotoxicity. Further analysis using CAL‐62 cells, which co‐expressed PD‐L1 and Gal‐9 (the ligand for TIM‐3), revealed that blocking PD‐1 and TIM‐3 significantly enhanced NK cell cytotoxicity [106]. Multi‐target immune checkpoint blockade based on personalized immune profiling could represent a promising future direction for TC immunotherapy [102]. Exploring multiple immune checkpoints is crucial for developing more effective monotherapy or combination immunotherapies, overcoming tumor resistance, and improving TC treatment by better understanding and manipulating immune responses.

4
Immunotherapy in Clinical Trials for TC
Immunotherapy, by restarting and sustaining the tumor immune cycle, has become one of the most promising treatment strategies in cancer therapy [107]. It has shown remarkable efficacy in reducing disease recurrence and prolonging patient survival in TC. Through various combination therapies, immunotherapy has also managed to synergistically suppress tumors while controlling treatment‐related side effects. However, challenges remain, particularly regarding resistance and adverse effects [108]. Looking ahead, the future direction of immunotherapy in TC is likely to involve combined approaches, integrating it with chemotherapy or targeted therapies. However, immune‐related adverse events (irAEs), such as hypothyroidism, Hashimoto's thyroiditis (HT), and even pituitary inflammation or adrenal insufficiency, remain unavoidable [109, 110].
4.1
Immune Checkpoint Inhibitors in TC
The clinical exploration of immunotherapy in TC began in 1997, with ICIs emerging as a key therapeutic approach [111]. These monoclonal antibodies target immune checkpoint proteins to block their interactions with partner proteins, effectively enhancing the immune system's ability to combat cancer. The FDA has approved several ICIs targeting different checkpoint proteins in various tumor types, including CTLA‐4 inhibitors, PD‐1 inhibitors, and PD‐L1 inhibitors [112, 113] (Figure 2). ICI monotherapy demonstrated moderate anti‐tumor activity and relatively controllable safety [77, 114, 115], while dual ICI therapy was more effective than monotherapy, with objective response rate (ORR) reaching 33% in some subtypes, but treatment‐related adverse events (AEs) increased significantly and required close monitoring [116] (Table 2).
In a phase Ib trial with PD‐L1‐positive, advanced PTC/FTC (NCT02054806) [114], pembrolizumab monotherapy demonstrated an ORR of 9%, with median overall survival (OS) not yet reached, suggesting it may offer durable survival benefits in a specific patient population, with overall manageable safety. Another multi‐cohort phase II study of pembrolizumab monotherapy (NCT02628067) [77], the ORR was 8.7% in the PD‐L1‐positive group and 5.7% in the PD‐L1‐negative group, with an overall ORR of 6.8%. Researchers suggested that future evaluations of ICI in patients with DTC should focus on biomarker‐driven patient selection or combination with other drugs to achieve higher response rates than those observed in this study [77]. However, in a phase II study (NCT02404441) [115], spartalizumab monotherapy demonstrated good clinical activity and safety in a patient population with aggressive, incurable TC and a short life expectancy. Targeted PD‐1/PD‐L1 therapy may offer a much‐needed treatment option for patients with PD‐L1‐positive advanced ATC [115]. In summary, monotherapy with ICIs can achieve some clinical response in patients with advanced or recurrent TC, but the ORR is relatively low. A moderate‐to‐high proportion of patients experience treatment‐related AEs, most of which are low‐grade, but there are still 3–5 grade reactions that require intervention. This strategy is more likely to be suitable for patients with PD‐L1‐positive expression or immune‐responsive subpopulations.
Sehgal Kartik et al. [116] investigate the efficacy and safety of ICIs in a Phase II clinical trial (NCT03246958). Key findings indicate that for the radioiodine refractory (RAIR) DTC cohort, the median OS was 24.6 months with an ORR of 9.4%. For the ATC cohort, the median OS was not reached, and the objective response rate was 30%. This demonstrates promising survival outcomes in both RAIR DTC and ATC, particularly with a notable response in the oncocytic subtype of RAIR DTC. However, the trial had a treatment‐related adverse event incidence rate of 81.6%. Grade 3 and grade 4 adverse events occurred in 24.5% and 8.2% of cases, respectively. Dual ICIs warrant investigation in larger ATC‐focused clinical trials and hold promise as a new treatment paradigm for these patients.
Case reports have also highlighted the potential benefits of ICIs in specific TC patient populations. For example, 13 patients with inoperable ATC treated with DTP showed a 1‐year survival rate of 38%, although nearly half (46%) experienced irAEs [120]. Additionally, patients with metastatic MTC who had developed resistance to tyrosine kinase inhibitors (TKIs) benefited from ICIs after alkylating agent treatment induced high tumor mutation burdens [121].

4.2
Combination Strategies With ICIs
Although ICIs have revolutionized cancer treatment, not all patients respond effectively to monotherapy. To overcome this challenge, researchers are actively exploring combination therapies to enhance efficacy and improve resistance, particularly in advanced TC or cases where other single‐agent treatments have failed [122]. These combinations are designed to enhance the ability of the immune system to combat cancer more effectively than monotherapies [123]. Combining ICI with other therapies, such as targeted therapies, radiotherapy, or surgery, has shown promising results (Figure 3).
Current clinical trial evidence indicates that, compared to monotherapy with ICIs, the combination of ICIs with targeted therapy provides significant survival benefits for advanced TC, particularly for RAIR‐DTC and ATC (Table 2). In two phase II clinical trials, camrelizumab (NCT04521348) [117] and atezolizumab (NCT03181100) [119] combined with targeted therapy increased the ORR in the ATC cohort to 62.5% and 31%, respectively, with a median OS of 43.24 months and an ORR of 50% in the BRAF V600E mutation ATC cohort [119]. The combination of camrelizumab and the multi‐kinase inhibitor famitinib (a kinase inhibitor) for RAIR‐DTC achieved an ORR of 33.3% [119]. While Jena D. French et al. (NCT02973997) [118] further revealed that lenvatinib combined with pembrolizumab can enhance durability in RAIR‐DTC patients who have not previously received lenvatinib treatment. Additionally, for patients who have progressed on lenvatinib treatment, adding pembrolizumab may be a viable salvage therapy. For DTC (ineligible for 131I) and MTC patients, the combination of camrelizumab and famitinib achieved an ORR of 44.00% and 40.00%, respectively. Future efforts should focus on patient stratification guided by biomarkers and mechanism‐driven combination regimens to further expand the population benefiting from ICIs combined with targeted therapy and improve the treatment window. Multiple retrospective studies have also confirmed the efficacy of combination therapy in achieving higher OS and treating patients who are resistant to traditional therapies [124, 125, 126, 127].
Combination therapy can also be used as neoadjuvant treatment. A single‐center retrospective study [128] assessing 18 patients with unresectable ATC found that combining kinase inhibitors and anti‐PD‐1 antibody as neoadjuvant therapy was both safe and effective. Approximately 38.9% (7/18) of patients underwent surgical resection after treatment, with a median overall survival of 14.0 months and a 12‐month survival rate of 55.6%.
For patients with specific genetic mutations, targeted therapy combined with ICIs has shown promising results. Research has found that PD‐L1 expression in PTC and ATC is significantly associated with BRAFV600E mutations, suggesting that combined BRAFV600E and PD‐1/PD‐L1 axis targeting may have clinical potential [66]. Studies in mice and single‐cell RNA‐sequencing analysis of tumor samples from treated patients indicated that famitinib combined with anti‐PD‐1 antibodies could enhance the development of early tertiary lymphoid structures in ATC, making it more responsive to immunotherapy [16]. The combination of targeted therapy and ICIs appears to be effective as neoadjuvant therapy for BRAFV600E‐mutated ATC [129]. The FAST Multidisciplinary Group Consensus Statement recommends combining dabrafenib (selective BRAF kinase inhibitors), trametinib (MEK1/2 inhibitors), and pembrolizumab for BRAFV600E‐variant ATC patients, which may offer significant survival benefits over using dabrafenib and trametinib alone [130].
Several studies have reported that radiotherapy can increase tumor cell PD‐L1 expression, promote the release of inflammatory and immune‐stimulating factors, and enhance the tumor's immunogenicity, thus improving the effects of immunotherapy [131, 132]. Therefore, numerous studies have investigated the combination of ICIs with radiotherapy. In vitro studies have shown that adding atezolizumab to radiation‐treated ATC cells significantly reduces cell proliferation [133]. Xing [134] reported a case in which radiation combined with tislelizumab successfully treated an ATC patient with good tolerance. Goh [135] described the combination of radiotherapy and pembrolizumab in a PD‐L1 high‐expressing ATC patient with a combined positive score of over 70%, resulting in a complete response of both local and distant disease. However, a phase I clinical trial using durvalumab combined with tremelimumab and image‐guided stereotactic body radiation therapy for metastatic ATC failed to improve overall survival, with only one out of 12 patients surviving after 1 year [136]. Further randomized controlled trials are needed to fully understand whether combining radiotherapy with immunotherapy offers potential benefits and to optimize treatment strategies for different types of TC.

4.3
Adoptive Immunotherapy
Adoptive immunotherapy involves extracting immune cells from the patient, expanding or genetically modifying them to enhance anti‐tumor activity, and then re‐infusing them into the patient to bolster the immune system. Several approaches to adoptive immunotherapy are currently being explored in TC (Figure 4).
4.3.1
Chimeric Antigen Receptor T‐Cell Therapy
CAR‐T therapy is a groundbreaking treatment that enables T cells to recognize tumor antigens in an HLA‐independent manner. While it has shown significant success in hematologic malignancies, its application in solid tumors like TC remains limited due to TME complexity and antigen heterogeneity [137, 138]. TSHR, expressed on the basal membrane of thyroid follicular cells, is a key target in CAR‐T research. Li's team [139] assessed the safety and efficacy of anti‐TSHR CAR‐T in vitro and in vivo. In another case report, TSHR + CD19 CAR‐T successfully treated refractory and relapsed TC patients, achieving partial remission [140]. Additionally, CEA is another potential target. CAR‐T targeting CEA has shown promise for treating metastatic MTC [141], though concerns about off‐target effects due to CEA expression in normal tissues remain.

4.3.2
T‐Cell Receptor (TCR) Therapy
TCR therapy involves modifying T cells to express receptors that recognize specific tumor antigens presented on the cancer cell surface by the major histocompatibility complex [142]. This approach expands the range of targeted tumor antigens but faces challenges related to complex production processes, non‐specific cytotoxicity, and improving T‐cell persistence [142, 143]. Cui et al. [144] using TCR high‐throughput sequencing with small sample size (n = 6) confirmed that intratumor heterogeneity of the T‐cell quantity and TCR repertoire truly existed in PTC, and the number of CD3+ TILs was negatively associated with TCR clonality in PTC.

4.3.3
NK Cell Therapy
NK cells can be isolated, expanded, and sometimes genetically modified to enhance their tumor‐targeting capacity. TSHR‐CAR‐modified NK‐92 cells exhibited enhanced cytotoxicity against TSHR‐positive DTC cell lines along with increased degranulation and cytokine release, providing a promising option for advancing immunotherapy in DTC [145].

4.3.4
Tumor‐Infiltrating Lymphocyte Therapy
TILs possess diverse TCR clonality, excellent tumor‐targeting ability, and low off‐target toxicity [146]. A study on PTC used artificial intelligence models to analyze TIL density in tumor tissues, categorizing tumors into three immune phenotypes: immune desert (48%), immune exclusion (34%), and inflammation (18%) [147]. TIL therapy has already demonstrated potential in treating other solid tumors, such as melanoma [148]. A phase II trial (NCT03449108), expected to be completed by June 2025, is currently investigating the efficacy of autologous TILs (LT‐145 or LN‐145‐S1) in patients with ATC and other cancers.

5
New Explorations in Immunotherapy for TC
Recent advancements in immunotherapy for TC have significantly expanded our understanding beyond the identification of specific tumor markers. Researchers have leveraged advanced experimental technologies and online databases to uncover promising biomarkers that not only shed light on the molecular mechanism driving TC progression but also suggest novel strategies for optimizing immunotherapy. These studies enhance our understanding of tumor immunity in TC, improve the efficacy of immune therapies, and provide potential biomarkers for predicting treatment responses, ultimately aiding in the development of personalized treatment strategies [149].
5.1
N6‐Methyladenosine
N6‐methyladenosine (m6A) is a prevalent RNA modification in eukaryotes, playing a dual role in cancer by influencing the immune microenvironment, genomic stability, and non‐coding RNA networks [150]. In TC, m6A modifications regulate RNA metabolism, impacting tumorigenesis, invasion, and immune evasion. Key regulatory factors (e.g., METTL3, YTHDF2) and downstream target genes (e.g., ACSM5) are of significant clinical relevance [151, 152, 153]. METTL3, by mediating the m6A modification of target genes like ACSM5, suppresses the proliferation, migration, and invasion of TC cells [152]. Moreover, the loss of METTL3 can promote TC dedifferentiation, leading to worse patient prognosis [154].
M6A modification may influence immune cell infiltration within the TC microenvironment by regulating the expression of immune‐related genes. A study revealed that patients with low m6A scores exhibited higher expression of immune checkpoints such as LAG3 and CTLA‐4, suggesting poorer responses to immunotherapy [155]. Similarly, Zhou et al. [156] found that TC patients with low m6A scores had higher PD‐L1 and CTLA‐4 expression. These findings suggest that m6A scores could serve as potential indicators of immunotherapy efficacy, with m6A modification patterns offering valuable insights for guiding personalized immunotherapy. Cai [157] identified four m6A patterns, with the m6A cluster‐mc3 subtype exhibiting low m6A scores, high copy number burden, and poor prognosis. Xia et al. [158] constructed a risk model based on 16 m6A‐related genes, revealing that high‐risk groups had lower immune infiltration, suggesting that m6A influences tumor progression via immune pathways.
The role of m6A‐related long non‐coding RNAs (lncRNAs) is also noteworthy. Su's team [159] classified TC patients into three clusters based on m6A‐related lncRNAs, with Cluster 2 showing higher expression of PD‐L1 and CTLA‐4 and better survival. A risk model based on 11 key lncRNAs accurately assessed immune status and risk, with experimental validation showing that reducing representative lncRNAs inhibited TC cell proliferation and migration. Chen et al. [160] identified two m6A‐related lncRNAs, LINC02471 and DOCK9‐DT, linked to immune cell infiltration, offering potential targets for immunotherapy.

5.2
Sialic Acid‐Binding Immunoglobulin‐Type Lectin (SIGLECs)
SIGLECs are receptors on leukocyte membranes that recognize sialylated glycoproteins, helping immune cells distinguish between “self” and “non‐self.” This function influences immune responses and the TME. In the TME, SIGLECs promote immune evasion through mechanisms similar to the PD‐1/PD‐L1 pathway, which has garnered significant attention. However, the exact role of SIGLECs in TC remains unclear [161, 162]. Among the SIGLEC family, SIGLEC10 and SIGLEC15 have emerged as promising immunotherapy targets.
SIGLEC15 expression may be epigenetically regulated, affecting immune cell infiltration and shaping the immune characteristics of TC. Hou's study [162] indicated that m6A methylation regulators influence SIGLEC15 expression, providing a potential mechanism for its aberrant expression in cancer. Additionally, high SIGLEC15 expression in ATC cells may foster an immunosuppressive TME by enhancing proteasome activity, while anti‐SIGLEC15 antibodies can inhibit tumor growth by boosting T‐cell‐mediated cytotoxicity [163]. In an ATC mouse model, these antibodies increased infiltration of M1 macrophages, NK cells, and CD8+ T cells, while reducing myeloid‐derived suppressor cells. The treatment also promoted the secretion of IFN‐γ and IL‐2, reversing the immunosuppressive TME and enhancing anti‐tumor immune responses [163].
The expression of SIGLEC15 is associated with malignant progression in TC. High SIGLEC15 expression correlates with increased TME cell interactions and is linked to extrathyroidal extension, lymph node metastasis, and immune exhaustion [162, 163]. Co‐expression of SIGLEC10 and SIGLEC15 in PTC tumor cells and stroma is a significant predictor of recurrence risk [164]. Therefore, blocking SIGLEC15 may benefit TC patients [162]. Additionally, SIGLEC15 expression is inversely correlated with key DNA damage repair deficiencies, suggesting that different SIGLEC15 expression subgroups may exhibit distinct sensitivities to drugs [162]. Collectively, the expression patterns of SIGLEC family members, particularly SIGLEC10 and SIGLEC15, and their associations with the immune microenvironment and prognosis, open new avenues for next‐generation immunotherapy in TC.

5.3
Thyroiditis and TC
The co‐occurrence of HT and PTC raises ongoing debate about whether HT provides a protective or promoting effect on PTC. The underlying biological mechanisms remain unclear [165, 166]. Several studies have explored the impact of thyroiditis on PTC immunotherapy outcomes. In a cohort study of 9210 PTC patients, 19% were diagnosed with HT [167]. Some studies suggest that HT may increase the risk of TC [168, 169], while others indicate that TC patients with HT have better prognostic outcomes compared to those without HT [167, 170].
Li et al. [170] reported that HT exhibited significantly higher immune scores and CD8+ T cell abundance compared to those without HT, with the high abundance of CD8+ T cells being positively correlated with disease‐free survival in PTC patients. Moreover, PD‐1 gene expression was notably higher in the HT group than in the non‐HT group [170]. Ma's team [171] used data from 140,456 cells across 11 patients to investigate the role of HT in shaping the PTC tumor immune microenvironment. They found that HT‐associated cell populations, enriched in thyroid hormone pathways such as mTE3, nTE0, and nTE2, created a TSH‐suppressive environment, positively influencing PTC progression. Furthermore, Pani's team [172] proposed that iodine‐induced thyroiditis (IET) could serve as an adjunct therapy to modulate the immune system and improve responses to ICIs. Their research showed that in mouse models with concurrent IET and PTC, ICI treatment reduced PTC incidence, while no effect was observed in models with pre‐existing IET or without IET.
The chronic inflammation associated with HT leads to lymphocyte infiltration, which influences TC development. Preliminary studies, using single‐cell RNA sequencing analysis of peritumoral and intra‐tumoral immune cells in TC that are coexistent with thyroiditis, could be of paramount importance to elucidate their functions [173].

6
Conclusions and Future Perspectives
Recent advances in immunotherapy have demonstrated significant potential in treating advanced TC, particularly for aggressive subtypes such as ATC and refractory DTC. PD‐1/PD‐L1 inhibitors, among the most extensively studied ICIs, not only modulate the TME and induce T‐cell exhaustion but also influence tumor development through metabolic reprogramming and protein modification. CTLA‐4, which acts at early stages of immune tolerance, attenuates T‐cell responses and induce T‐cell exhaustion. B7‐H3, frequently overexpressed in aggressive subtypes, has been widely reported to promote oncogenesis, angiogenesis, invasion, and metastasis through diverse mechanisms. Beyond conventional immune checkpoints, several novel biomolecules offer promising therapeutic avenues for TC. These include SIGLEC15, which remodels the immunosuppressive tumor microenvironment and is targetable with blocking antibodies; DPP4 (CD26), whose inhibition reverses CD8+ T‐cell exhaustion and disrupts the EMT‐promoting IL13‐IL13RA2 axis; and SPP1+ macrophage‐driven axes (SPP1‐CD44/MIF‐CD74) that mediate T cell suppression. Additionally, m6A RNA modification regulators (e.g., METTL3, YTHDF2) influence immune checkpoint expression and predict immunotherapy response, while adoptive therapy targets like TSHR and CEA enable CAR‐T/NK cell strategies for refractory disease. Together, these targets may address key resistance mechanisms in ATC and warrant biomarker‐driven clinical validation, particularly for anaplastic and poorly differentiated subtypes. Furthermore, combination strategies that integrate ICIs with targeted therapies (e.g., BRAF/MEK inhibitors) or adoptive cell therapies (e.g., CAR‐T) are under investigation to overcome resistance and enhance efficacy [174].
Despite progress, challenges persist, including tumor heterogeneity, immune evasion mechanisms, and limited durable responses observed in clinical trials. Future research may focus on identifying predictive biomarkers (e.g., PD‐L1 expression, immune risk scores) to better stratify patients, optimizing combination regimens to mitigate irAEs, and exploring novel immunotherapeutic approaches, such as targeting TAMs heterogeneity (e.g., tissue‐resident TRM‐TAMs vs. monocytic‐derived MDM‐TAMs) [175]. Additionally, leveraging preclinical models and multi‐omics data will be crucial for unraveling the complex interactions within the immune microenvironment and understanding the epigenetic regulation of TC. As clinical trials progress and our understanding of tumor‐immune interactions deepens, immunotherapy is poised to redefine the therapeutic landscape for advanced TC.

Author Contributions
Q.Z. and F.W. conducted the literature research and drafted the manuscript. W.M. and N.L. drew the pictures according to the literature data. Y.L. and P.Z. revised the paper and made critical interpretations of the literature data. All authors have read and agreed to the published version of the manuscript.

Funding
Yihan Lu was supported by the National Nature Science Foundation of China (grant number 82200878) and Dalian Science and Technology Bureau (grant number 2023RQ013).

Conflicts of Interest
The authors declare no conflicts of interest.