Interleukin-33 (IL-33), a member of the IL-1 cytokine family, is expressed in endothelial cells, fibroblasts, and epithelial cells both during homeostasis and inflammation (1,2). The human suppression of tumorigenicity 2 (ST2) gene encodes for three splice variants; soluble ST2 (sST2), transmembrane ST2 (ST2L), and variant ST2 (ST2V) (1). The IL-33 receptor is a heterodimeric complex consisting of ST2L and IL-1 receptor accessory protein (2). ST2 is expressed on the membrane of a variety of immune cell types, such as T helper (Th)2 lymphocytes, group 2 innate lymphoid cells, macrophages, mast cells, basophils, eosinophils, dendritic cells, and natural killer (NK) cells (3). IL-33/ST2 signaling is crucial for tissue repair, type 2 immunity, allergic and non-allergic inflammation, and viral infection (3).

Recent studies have shown associations between the biological role of the IL-33/ST2 axis and progression of malignant tumors. However, the significance of tumoral ST2 expression remains controversial in some malignancies (4). In this study, we investigated the clinicopathological and prognostic relevance of tumoral ST2 expression in patients with resected pancreatic carcinoma after neoadjuvant chemoradiotherapy.

Patients. This retrospective study was approved by the Institutional Review Board of Kagawa University (approval number: 2020-090) and Shimane University (approval number: 20221012-1). All experiments were performed in accordance with relevant guidelines and regulations.

We reviewed records from patients with invasive pancreatic ductal adenocarcinoma who had undergone pancreatic resection after chemoradiotherapy at Kagawa University between 2009 and 2018 (n=81). After excluding patients with positive surgical margin (n=3) or ≤10% residual tumor after chemoradiotherapy (n=2), 76 patients were eventually included in our analysis. Clinical data were collected from a prospectively maintained pancreatic carcinoma database. Tumor, node, metastasis (TNM) stages were assigned based on the eighth edition of the American Joint Committee on Cancer TNM Staging Manual (5).

Immunohistochemistry using tissue microarrays and scoring. Formalin-fixed, paraffin-embedded tumor specimens from patients who met the inclusion criteria were used for tissue microarray construction. One representative main tumor site was marked on Hematoxylin & Eosin-stained slides. Using a tissue array (Tissue Microprocessor KIN-2, Azumaya, Japan), cylindrical 3-mm tissue cores from the corresponding paraffin block were arrayed on the recipient block.

Four-μm sections from stored formalin-fixed, paraffin-embedded tumor blocks were stained with an anti-ST2 rabbit polyclonal antibody (Proteintech, Rosemont, IL, USA; diluted at 1:100) using a BOND-III automated immunohistochemical slide staining system (Leica Biosystems). We used diaminobenzidine as chromogen and hematoxylin as nuclear counterstain.

Using an Olympus BX53 upright microscope (Olympus Corporation, Tokyo, Japan) with a standard 22-mm-diameter eyepiece, anti-ST2 stained tumor slides were reviewed by a pathologist (M.N.), who was blinded to patient outcomes. Immunohistochemistry scoring was based on the distribution and intensity of the staining in the main tumor. Distribution was scored based on a scale of 0 (0-25%), 1 (26-50%), 2 (51-75%), or 3 (76-100%) according to the percentage of positive cells in each core. Staining intensity was scored as 0 (no expression), 1 (mild expression), 2 (intermediate expression), or 3 (strong expression). Distribution and intensity scores were summed in a total score (0-6) for each patient (Figure 1). The score of ST-2 expression was dichotomized as low or high according to the median value (median: 2, range=0-6). When the score was equal or greater than the median, expression was classified as high.

Statistical analysis. Associations between variables were analyzed using chi-square tests for categorical variables. Disease-specific survival (DSS) was defined as the time from resection to the date of death related to pancreatic carcinoma or last follow-up. Disease-free survival (DFS) was defined as the time from resection to disease recurrence. DSS and DFS were estimated using the Kaplan–Meier method, and non-parametric group comparisons were performed using the log-rank test. All statistical tests were two-sided, with the significance level set at 5%. Statistical analyses were conducted using IBM SPSS Statistics for Windows (version 23.0; IBM, Armonk, NY, USA).

Patient characteristics and association of clinicopathological factors with patient outcomes. The clinicopathological characteristics of all patients who had undergone pancreatic resection after chemoradiotherapy (n=76) are summarized in Table I. During the study period, 52 patients recurred and 33 died from pancreatic carcinoma-related causes. The median duration of follow-up for patients who were alive at the time of the last follow-up was 27 months (range=5-113 months).

Univariate analyses of all patient outcomes (DSS and DFS) with clinicopathological factors are shown in Table II. None of the clinicopathological factors were associated with DSS. High T status (p=0.041), lymph node metastasis (p=0.008), pathological stage (p=0.030), and perineural invasion (p=0.001) were significantly associated with worse DFS. Univariate analyses of DFS with clinicopathological factors in stage I patients showed that perineural invasion (p=0.004) was significantly associated with worse DFS (Table III).

Association between ST2 expression and patient characteristics. High expression of ST2 was observed in 43 patients (57%). Associations between patient characteristics and ST2 expression are summarized in Table I. High expression of ST2 was more frequent in patients with high pathological T status (T1 vs. T2-3; p=0.002), in those with lymphatic invasion (p=0.049), and in those with ≤50% of tumor cells destroyed by chemoradiotherapy compared to those with >50% of tumor cells destroyed (Evans grade I-IIA vs. IIB; p=0.043) (Figure 2).

Association between ST2 expression and patient outcome. ST2 expression was not associated with DSS (p=0.70) (Table II). In all patients, DFS was lower in patients with high ST2 expression (median 11.7 months) than in those with low ST2 expression (median 23.7 months) (Figure 3A), although this difference was not statistically significant (p=0.067). In stage I patients, DFS was significantly lower in patients with high ST2 expression (median 10.6 months) than in those with low ST2 expression (median 43.4 months; p=0.046) (Figure 3B and Table III). In stage II-III patients, ST2 expression was not associated with DFS (p=0.49) (Figure 3C).

In this study, immunohistochemical analyses of tumoral ST2 expression were performed with the aim to interrogate its clinicopathological and prognostic relevance in patients with resected invasive pancreatic ductal adenocarcinoma who were submitted to neoadjuvant chemoradiotherapy. We have demonstrated that tumoral ST2 expression was associated with high T status, lymphatic invasion, and lower histopathological response grade after neoadjuvant chemoradiotherapy in patients with pancreatic carcinoma, while it was also associated with lower DFS in stage I patients.

The IL-33/ST2 signaling pathway promotes cancer development and remodels the tumor microenvironment. Therefore, the IL-33/ST2 pathway may be a potential prognostic marker and a possible immunotherapeutic target. However, how the IL-33/ ST2 axis is involved in the development of carcinomas remains unclear. ST2 is expressed in various types of immune and tumor cells. IL-33 is secreted from stromal cells, endothelial and epithelial cells, and tumor cells including tumor-initiating cells (4,6). In some studies, the IL-33/ST2 axis has been reported as a pro-tumoral effector pathway, by promoting expansion of immune suppressive cells, such as myeloid-derived suppressor cells or regulatory T cells. In contrast, the IL-33/ST2 axis may also promote formation of an antitumoral microenvironment by recruiting immune cells such as CD8+ T and NK cells (4).

Previously, the IL-33/ST2 axis was shown to facilitate proliferation of cancer cells in vivo, which was dependent on up-regulation of cyclooxygenase 2/prostaglandin E2 expression through NFĸB signaling (7), and to increase expression of IL-6, CXCR4, MMP-2, and MMP-9 in tumor cells, which were essential for enhanced tumor metastasis in rectal carcinoma (8). In contrast, Akimoto et al. have reported that sST2 negatively regulated tumor growth and metastatic spread of colorectal carcinoma through modulation of the tumor microenvironment (9). Moreover, Zhang et al. suggested that tumor-derived IL-33 modulated the tumor microenvironment, by enhancing recruitment of CD11b+ myeloid cells, to potently promote colon carcinogenesis and liver metastasis (10). In pancreatic carcinoma, clinicopathological and prognostic associations with IL-33/ST2 axis, based on immunohistochemical analyses in human cancer samples, have not yet been well investigated. Takenaga et al. have demonstrated that the cancer cell-derived interleukin-33 decoy receptor sST2 enhanced orthotopic tumor growth through up-regulation of CXCL3 via inhibition of IL-33/ST2L signaling in the tumor microenvironment in a murine pancreatic cancer model; however, they could not find a consistent association between IL-33 or ST2 expression and prognosis, based on analysis of public prognosis databases (11). Induction of IL-33/ST2 signaling results in activation of immune effector cells that leads to recruitment of pro- or anti-oncogenic cells into the tumor microenvironment (12). IL-33/ST2 signaling is negatively regulated by sST2, which acts as a decoy receptor and sequesters IL-33 to block its interaction with ST2L (13). Thus, IL-33 bioactivity is controlled by various factors, such as the balance between ST2L and sST2, and heterogeneous expression in various cell types and immunological conditions, indicating that this complexity may account for the inconsistent literature regarding the prognostic value of ST2 expression in carcinomas.

The incidence of pancreatic carcinoma has been increasing for years. Pancreatic carcinoma is the seventh leading cause of cancer death globally (14). and remains as one of the most lethal malignancies, with a 5-year survival rate of approximately 11% (15). Although radical resection is necessary for patients with pancreatic carcinoma, 80-85% of patients present with advanced unresectable disease (16). Neoadjuvant therapy added to surgery appears to be effective for pancreatic carcinoma. A systematic review and meta-analysis performed by Zhan et al. suggested that patients with borderline resectable or locally advanced pancreatic carcinoma may benefit from neoadjuvant therapy (17). Recently, Suto et al. have reported that the reduction rate of fluorodeoxyglucose uptake on positron emission tomography and carbohydrate antigen 19-9 levels may serve as pre-operative predictors of therapeutic response to neoadjuvant chemoradiotherapy in patients with resected pancreatic carcinoma, and were associated with better recurrence-free survival (18). However, prediction factors, based on pathological parameters using carcinoma tissues, of therapeutic response and recurrence after neoadjuvant chemoradiotherapy have not been thoroughly investigated in patients with pancreatic carcinoma. In our study, we have demonstrated that tumoral ST2 expression was significantly associated with lower histopathological response grade after neoadjuvant chemoradiotherapy in patients with pancreatic carcinoma and lower DFS in stage I patients. We suggest that ST2 expression may potentially predict response to neoadjuvant chemoradiotherapy and disease recurrence in patients with pancreatic carcinoma. However, we evaluated ST2 expression based on cancer tissues after chemoradiotherapy, but not biopsy specimens before neoadjuvant chemoradiotherapy and surgery, which may be considered as a limitation of our study; thus, further investigation would be required based on therapy-naïve cancer tissues using pre-operative biopsy. Interestingly, Kieler et al. have demonstrated a significant association between high sST2 plasma levels and inferior survival in patients with advanced non-resectable, pancreatic carcinoma undergoing chemotherapy (19), suggesting that sST2 may be a possible predictor of response to chemotherapy, although this finding was not based on surgically resected cases with neoadjuvant treatment.

In conclusion, we have identified tumoral ST2 expression as a marker of biological aggressiveness, such as disease recurrence, higher T status and lymphatic invasion, and as a possible predictor of recurrence and lower histopathological response grade in invasive pancreatic ductal adenocarcinoma after neoadjuvant chemoradiotherapy. These findings suggest that ST2 may serve as a predictor of response to current neoadjuvant therapy or as a therapeutic target in tumor immunotherapy in pancreatic carcinoma. However, the biological role of the IL-33/ST2 pathway in tumor progression is still unclear, although the controversial effects of IL-33/ST2 axis may be attributed to the disparate immunohistochemical detection of IL-33 receptor variants (sST2, ST2L, or other variant ST2), or the heterogeneous expression of ST2 in various cell types (tumor cells vs. immune cells in tumor microenvironment) that can be activated by IL-33. Our study was limited by the fact that IL-33 immunohistochemistry was not performed in this cohort, and that interactions between IL-33 and ST2 in tumor cells or tumor immune microenvironment were not investigated. Therefore, future in-depth investigation is warranted based on double staining with anti-IL-33 and anti-ST2 antibodies to confirm the role of IL-33/ST2 axis in the tumor microenvironment.

All Authors declare no conflicts of interest in relation to this study.

NM, EI, and KK designed the experiments; NM, EI, and KK performed the experiments; NM and KK analyzed the data; NM and KK helped with the discussion; NM and KK wrote the manuscripts; and RH and KO supervised the project. All Authors have reviewed the manuscripts.