1. IntroductionThyroid cancer is the most common malignant tumor of the endocrine system, and its incidence has been increasing over the past decade [1]. Papillary thyroid carcinoma (PTC) is the most common subtype of all thyroid carcinomas. It is generally characterized by a high degree of differentiation, slow clinical course, and low specific mortality rate. However, a considerable number of patients still experience persistent or recurrent disease, with an incidence of 14–30% [2,3,4]. Stratification systems currently recommended by the American Thyroid Association, and European Thyroid Association, are organized mainly based on the aspects of primary tumor size, lymph node, and pathology. Unfortunately, about 10% of patients classified as low-risk still experience progression or recurrence [5]. For PTC patients, regular postoperative review and thyroid-stimulating hormone (TSH) suppression should be performed to reduce the risk of tumor recurrence [6]. However, long-term TSH suppression therapy also brings about a series of adverse reactions, such as osteoporosis, arrhythmia, and hyperthyroidism [7]. Therefore, a more accurate assessment of the recurrence risk is essential to avoid overtreatment in low-risk patients and to ensure that high-risk patients receive more aggressive treatment, such as higher TSH suppression and more frequent postoperative follow-ups.Programmed cell death, including apoptosis, autophagy, and pyroptosis, refers to the way in which cells die depending on the signal or activity encoded by specific genes. It is a special mechanism regulating cell proliferation, homeostasis, and tumor development [8]. Pyroptosis is a type of programmed cell death mediated by the gasdermin protein family. It is triggered by a specific inflammasome, leading to the activation of the canonical caspase-1 or the non-canonical caspase-4/5/11 inflammasome pathways. Caspase can cleave gasdermin D and form plasma membrane pores. Eventually, this results in cytoplasmic swelling, membrane rupture, and release of inflammatory factors, inducing an inflammatory response [9,10,11]. Pyroptosis was initially thought to play a key role in fighting infection, but related studies have shown that pyroptosis is closely related to cancer. For example, the programmed death ligand 1 (PD-L1) has been found to switch cell apoptosis into pyroptosis, leading to tumor necrosis [12]. It has also been reported that chemotherapy drugs can induce cell pyroptosis via caspase-3 cleavage of gasdermin E [13]. In general, the impact of pyroptosis on cancer may be two-sided. On one hand, pyroptosis can eliminate cancer cells and inhibit tumor growth. On the other hand, pyroptosis may also form a suitable microenvironment for tumor cell growth [14]. Several studies based on public databases have investigated the value of pyroptosis in predicting cancer prognosis, such as in gastric [15] and ovarian cancers [16]. The mechanism and potential role of pyroptosis in tumor regulation can provide new strategies for cancer treatment.At present, there are still few studies on the relationship between pyroptosis and thyroid cancer. Previous studies [17] reported the role of PRGs in predicting the overall survival of PTC. However, in view of the extremely low mortality rate of PTC, it is more clinically meaningful to focus on the recurrence and progression of PTC. The present study determined the expression level of pyroptosis-related genes (PRGs) in PTC, explored the value of these genes in predicting disease-free survival (DFS), established a nomogram, and discussed its relationship with immune characteristics and tumor mutations. In addition, the expression of PRGs was verified using quantitative real-time PCR in clinical tissue samples. 4. DiscussionThe incidence of thyroid cancer is increasing worldwide. Although surgery is an effective treatment, there is still a risk of recurrence, which introduces significant challenges to clinical treatment [3]. Therefore, it is necessary to accurately predict the recurrence of thyroid cancer. Pyroptosis has been confirmed to be associated with the development of many cancers [21]. However, its role and mechanism in PTC remain to be studied. Therefore, the present study aimed to explore the value of pyroptosis in predicting recurrence and treatment. We first obtained the transcriptome data of 510 PTC samples and 58 normal tissues from the TCGA database, and obtained 33 PRGs from the literature. Then we detected the expression of PRGs in PTC and normal tissues. Based on these differentially expressed genes, the risk score models of DFS were established, based on four PRGs, including CASP6, CASP9, NOD1, and IL-18, followed by in-group and qRT–PCR validation. Next, we divided all samples into high–low risk groups, based on the median of risk score. We enriched the differential genes between the two groups, compared the characteristics of immune infiltration and gene mutation between them, so as to explore the possible reasons for the influence of pyroptosis on prognosis. Finally, we combined the risk scores with other clinical features to create a prognostic model.The caspases are proteases at the heart of networks that govern programmed cell death and play vital roles in the induction, transduction, and amplification of intracellular signals [28,29]. Caspases can be divided into three types: initiator, executioner, and inflammatory [30]. Caspase-6 is the executor caspase. It has been implicated in a number of neurological diseases and can be used as a therapeutic target [31]. However, the role of caspase-6 in pyroptosis remains unclear. Caspase-6 is not present in typical inflammasome activation and is not required for classic trigger-induced apoptosis and necrosis [32]. However, caspase-6 can cleave STAT1 to inhibit tumor cells in leukemia [33]. Similarly, STAT1 acts as a tumor suppressor in thyroid cancer. The long non-coding RNA transcriptionally mediated by STAT1 regulates thyroid cancer cell growth, migration, and invasion [34,35]. In the present model, CASP6 was associated with a shorter DFS. Thus, it was speculated that a similar pathway exists in PTC recurrence. Inhibiting CASP6 may be a new strategy to reduce PTC recurrence, but further basic studies are needed to verify this hypothesis. Different from caspase-6, caspase-9 is an initiator caspase, which is the initiation protein of the innate apoptosis pathway. Caspase-9 is activated by binding to Apaf-1 apoptotic bodies, triggering the CASPs cascade, which then activates CASP3 and leads to cell death [36]. Since it is located upstream of the entire pathway, caspase-9 is considered a promising therapeutic target. For example, previous studies have found that tumor cells can hijack caspase-9 signaling to suppress radiation-induced immunity [37]. The present study showed that CASP9 was down-regulated in PTC and was a tumor suppressor gene. It might be involved in the process of killing tumor cells and can be inhibited in a tumor, although this mechanism remains unclear. In addition, the loss of CASP9 signaling can lead to tumor immune escape and tumor recurrence by up-regulating PD-L1, which also provides new insights for immunotherapy of thyroid cancer.NOD1 is one of the most important members of the NOD-like receptor family. NOD1 can detect conserved motifs in bacterial peptidoglycan and initiate nuclear factor-κB (NF-κB)-dependent and mitogen-activated protein kinase-dependent gene transcription, resulting in proinflammatory and antimicrobial responses [38,39]. Furthermore, NOD1 has been reported to be involved in the development of many cancers, such as breast and colon cancer [40]. NF-κB is one of the most prominent cascades involved in tumor development [41], which regulates the proliferation and apoptosis signaling pathways of cancer cells. It has recently been shown to play an important role in thyroid cancer, with related genes, such as RET/PTC, RAS, and BRAF, acting through NF-κB [42,43]. However, the specific regulatory mechanisms of NOD1 and NF-κB in PTC remain unclear. IL-18 is a member of the IL-1 cytokine super family and is expressed in inflammation, autoimmune diseases, various cancers, and many infectious diseases [44]. Activation of caspase-1 and -11 can drive the release of IL-18 in pyroptosis [9]. In general, IL-18 plays a complex role in cancer. On one hand, it can stimulate NK and T cells, enhance the T helper type-1 immune response, and secrete interferon-γ, which effectively destroy cancer cells. On the other hand, IL-18 can induce angiogenesis, which helps tumor metastasis and proliferation. The creation of a chronic inflammatory microenvironment also helps tumor cells evade the immune response of host cells [45]. Therefore, IL-18 may be considered to be a new tumor therapeutic target, and overcoming its avoidant immunophenotype may be a new cancer therapeutic strategy. In the present study, there was a difference in the expression of IL-18 in the TCGA database and clinical samples, which might be related to the presence of inflammation in the patients themselves. In the risk score model, the higher the expression of IL-18, the longer the DFS, suggesting that IL-18 might be involved in the process of inhibiting cancer cell growth, and increasing its expression could help improve the prognosis.The relationship among the risk score, immune characteristics, and tumor mutation burden was then explored further. There is a complex network of interactions between the immune system and cancer, and understanding these interactions may improve immunotherapy treatment for thyroid cancer. The immune system can specifically destroy tumor cells and control tumor development through immune monitoring [46]. However, tumor cells can also reduce their immunogenicity by recruiting immunosuppressive cells and overexpressing immune checkpoints to achieve immune evasion [47]. Several previous studies, based on the TCGA data for thyroid cancer, have found that immune scores were associated with BRAF mutations, immune checkpoint expression, and prognosis [48,49]. In the present study, there were also differences between the high- and low-risk groups. Generally, the low-risk group showed a higher immune infiltration level and lower tumor purity. In terms of the immune checkpoint, previous studies suggested a high expression of PD1, PD-L1, and CTLA4 in PTC. In the present study, higher expression of CTLA4 was present in the low-risk group, while PD1 and PD-L1 showed no differences. In addition, higher BRAF mutation rates were evident in the low-risk group. BRAF inhibitors combined with immune checkpoint inhibitors (ICIs) were reported to significantly inhibit thyroid cancer growth and prolong survival in mouse models [50]. Therefore, these results suggested that patients in low-risk groups were more likely to benefit from ICI treatment. Finally, TMB differences between the two groups were compared. For tumors, the higher the TMB, the more neo-antigens were expressed and the more sensitive they were to ICIs [51]. For patients not treated with ICIs, a higher TMB was often associated with poor survival [52], while in patients treated with ICIs, a higher TMB was associated with longer survival [53]. The present study results were similar. TBM in the high-risk group was significantly higher than that in the low-risk group and was associated with TMB and risk score. In this regard, higher TMB offers new possibilities for immunotherapy in high-risk patients.

There were several limitations in the present study. First, due to the lack of DFS data, this research was based on a single database and lacked external validation. Second, the PRGs were obtained from previous literature. With the development of research on pyroptosis, there may be more PRGs that can be incorporated into the model in order to improve its accuracy. Third, few studies on the correlation between pyroptosis and thyroid cancer have been performed, and the role of pyroptosis in tumor development and immune characteristics needs to be confirmed by additional in vitro and in vivo studies.