Introduction

Colorectal cancer (CRC) is the third most common malignant tumor and the second leading cause of tumor-related deaths over the world. According to the latest global cancer statistics, more than 1.9 million people were diagnosed with CRC in 2020, with more than 900,000 cancer-related deaths1. During the early stages, the symptoms of CRC are often mild and non-specific, which is conducive to delaying diagnosis and treatment. Most CRC patients are diagnosed with late-stage disease, with significant organ infiltration and distant metastases, leading to loss of opportunity for radical resection and poor prognosis2.

Over the years, researchers have made significant strides with the discovery of molecular markers for the clinical diagnosis of CRC, such as carcinoembryonic antigen (CEA), carbohydrate antigen 199 (CA199), platelet to lymphocyte ratio (PLR), and neutrophil to lymphocyte ratio (NLR), etc. However, these markers are limited by their relatively poor specificity and sensitivity, which explains their use to monitor recurrence rather than disease diagnosis3. Therefore, it is necessary to further investigate the pathogenesis of CRC and identify more stable and specific molecular markers to improve the clinical diagnosis and treatment of CRC and the prognosis of this patient population.

TIPE2 (TNFAIP8L2) is a member of the TIPE family, discovered in a mice model of experimental autoimmune encephalomyelitis4. TIPE2 is highly expressed in immune cells such as macrophages, B-lymphocytes, and T-lymphocytes. There is growing evidence that TIPE2 is a novel negative regulator for maintaining immune homeostasis5. Moreover, it regulates both the intrinsic and adaptive host immune responses. An increasing body of evidence from recently published studies suggests that TIPE2 is involved in the pathophysiology of various diseases, such as autoimmune diseases, diabetic nephropathy, atherosclerosis, and tumors6,7,8,9.

It has long been recognized that tumor microenvironment (TME) plays a crucial role in the biological behavior of cancer. Based on previous studies, 3 prognostic immune biomarkers were selected for IHC staining: cytotoxic T cells (CD8), B cells (CD20), and neutrophils (CD66b)10,11. It is widely acknowledged that CD8 + T cells play an important role in innate and adaptive immune defense mechanisms. Interestingly, it has been shown that TIPE-2 inhibits allogeneic CD4 + and CD8 + T cell proliferation, activation, and cell cytokine production12. CD20 has been widely used in assessing the degree of inflammatory infiltration in different tumors. CD20 + tumor-infiltrating B cells showed a negative prognostic association in CRC13. Moreover, CD66b + tumor-associated neutrophils (TANs) have a close relationship with the development and prognosis of some tumors14. Furthermore, TIPE2 can reportedly promote leading-edge formation in neutrophils through phosphoinositide-dependent signaling and cytoskeleton remodeling15.

However, the role of TIPE2 in the pathophysiology of colorectal tumors and its relationship with tumor immunity is not fully understood. In this research, we explored the expression of TIPE2 in colorectal tumor tissues and adjacent normal tissues, combining the prognostic value of TIPE2 and immune infiltration to explore their relationships.

Materials and methods

Dataset information

The integrated pan-cancer dataset TCGA Pan-Cancer was downloaded from the UCSC database (https://xenabrowser.net/). We extracted the expression data of the TNFAIP8L2 gene in each sample and screened the sample sources. The samples with an expression level of 0 were also filtered. Subsequently, each expression value was log2(x + 1) transformed. In addition, we extracted the gene expression profile of the tumor separately that were mapped to GeneSymbol. In addition, the stromal, immune, and ESTIMATE scores were calculated for each patient in each tumor based on gene expression using the R package ESTIMATE. B cell, CD4 + T cell, CD8 + T cell, Neutrophil, Macrophage, and Dendritic cell (DC) infiltration scores were re-evaluated for each patient in each tumor based on gene expression using the TIMER method of R package IOBR.

Patients

Patients with different stages of CRC admitted to the First Affiliated Hospital of Anhui Medical University from January 2015 to December 2015 were included in this study. All patients underwent radical surgery with a confirmed postoperative pathological diagnosis of primary colorectal adenocarcinoma. All the subjects involved have been informed of the purpose and significance of this study, signed the informed consent, and provided the necessary biological samples for testing and analysis after being fully informed of the study details. All experiments were performed in accordance with relevant guidelines and regulations. The study has been approved by the ethics committee of the first affiliated hospital of Anhui Medical University(No. LLSC2021010). The exclusion criteria included: (1) Patients with unavailable complete clinical and pathological data. (2) Patients with a history of other malignant tumors. (3) Serious complications or death in the perioperative period. (4) Follow-up data unavailable.

A total of 110 patients with CRC were included in the study based on the above inclusion and exclusion criteria (Supplementary Fig. 1). This study was approved by the Ethics Committee of the First Affiliated Hospital of Anhui Medical University (No. LLSC2021010).

Follow-up

Long-term follow-up was conducted until June 31, 2021, ranging from 1 to 64 months, by phone call and outpatient review. The endpoints of this study were overall survival (OS), defined as the time from radical surgery to death.

Immunohistochemical staining

Surgically resected cancer tissue specimens (both cancer cells and the surrounding stroma, including infiltrating immune cells) and the corresponding paracancerous normal mucosal tissue specimens (colorectal mucosal tissue more than 5 cm from the edge of the tumor tissue) were collected from the First Affiliated Hospital of Anhui Medical University from January 2015 to December 2015. Tissues were immediately frozen in liquid nitrogen and paraffin-embedded for use in the following experiments. All samples were cut into 4-µm sections using a microtome (Leica, Germany).

The expression of TIPE2, CD8, CD20, and CD66b in cancer tissues was detected using immunohistochemical staining. Following incubation with an antibody against human TIPE2 (AB110389, Abcam, America), CD8 (AF5126, Affinity, China), CD20 (DF13319, Affinity, China), CD66b (DF10151, Affinity, China), PD-1 (18106-1-AP, Proteintech, China), and TIM-3 (11872-1-AP, Proteintech, China). The sections were stained using Envision System (PV6001, ZSGB-BIO, China). Two experienced pathologists interpreted the immunohistochemical staining results. First, using an electron microscope (Olympus, Japan), the general condition of the tissue section was assessed at a low power field (100×). Then, five representative high-power fields (200×) of each tissue section were selected to photograph and score. Then the average of 5 field scores was taken as the positive cell score. The immunoreactive scores were determined by a double-blind reading method, and discrepancies were resolved by consensus.

Different scoring criteria were used for different proteins. The scoring criteria of TIPE2 was obtained by the product of the staining intensity of TIPE2 expression (0, no staining; 1, weak staining; 2, moderate staining; 3, strong staining) with the extent of TIPE2 staining (0, < 1%; 1, 1–20%; 2, 21–40%; 3, 41–60%; 4, 61–80%; 5, > 81%). The two scores for each slide were combined to obtain a final TIPE2 expression score. The scoring criteria of CD8, CD20, CD66b, PD-1, and TIM-3 consisted of quantifying the number of positively stained cells. Based on the median of their scores, the expression of TIPE2, CD8, CD20, and CD66b in cancer tissues and adjacent normal tissues was grouped into low expression group and high expression group, respectively.

Clinicopathological parameters

And then, independent risk factors for CRC prognosis were identified by analyzing the association between different clinical and pathological features and CRC prognosis. The following clinical parameters were enrolled into subsequent research: grade, T-stage, N-stage, WBC, neutrophil, and CA199. CEA and CA199 above 5 ng/mL and 27 µg/mL, respectively, were considered abnormal. Besides, hemoglobin, WBC, neutrophil, lymphocytes, and monocyte within 130–175 g/L, 3.5–9.5 *10^9/L, 1.8–6.3 *10^9/L, 1.1–3.2 *10^9/L, and 0.1–0.6 *10^9/L were considered normal respectively. Based on the normal ranges, these parameters were divided into several subgroups, respectively.

Statistical analysis

Data were analyzed by SPSS software (version 22.0; IBM Corporation) with default parameters and R software (version 3.6.4). Continuous variables were tested for normality and the results showed that all continuous variables did not conform to a normal distribution. So continuous variables were subjected to Wilcoxon rank-sum test and expressed as median (interquartile range [IQR]). Frequency and percentage were used to express the categorical variables. During univariate analysis, Pearson’s χ2 or Fisher’s exact test was performed to analyze categorical variables, and the Student’s t-test or rank-sum test was used to examine continuous variables. Based on TIPE2, CD8, CD20, and CD66b expression groups and clinicopathological groups, the survival curves were plotted and log-rank test were used to analyze P-value (expressed as total survival time [OS]), respectively. Lasso regression analysis was used to screen the variables that may be meaningful. Univariate and multivariate analysis was performed using Cox proportional risk regression analysis to identify potential prognostic factors for OS, and independent risk factors for CRC were screened. Based on the above independent risk factors screened by Cox proportional risk regression analysis, the nomogram prediction models were constructed to predict 1-year, 3-year and 5-year OS in CRC patients. Receiver operating characteristic (ROC) curves were performed to examine outcomes. Calibration curves were used to detect consistency between the actual OS and the predicted OS probability from the nomogram. The clinical usefulness of the nomogram was evaluated using decision curve analysis (DCA) by quantifying the net benefits at different threshold probabilities. P < 0.05 was statistically significant.

Results

Correlation between expression of TIPE2 and immune infiltration scores in CRC

Immune infiltration scores for 9554 tumor samples from 39 tumor types were obtained. Spearman’s correlation analysis between gene and immune infiltration scores was conducted for each tumor using the corr. test function of the R package psych (version 2.1.6) to identify significantly correlated immune infiltration scores (Supplementary Fig. 2). Ultimately, 37 cancer species exhibited a positive correlation between gene expression and immune infiltration, including TCGA-COAD (N = 282, R = 0.77, P = 4.0e-57), TCGA-COADREAD (N = 373, R = 0.78, P = 2.1e-76) and TCGA-READ (N = 91, R = 0.79, P = 1.7e-20). The results indicated a significant positive correlation between TIPE2 expression in CRC and immune infiltration. Besides, TIPE2 expression was positively related to 5 immune cells, including B cell, CD4 + T cell, CD8 + T cell, Neutrophil, Macrophage, and DC (Supplementary Fig. 3).

The clinicopathological features of 110 CRC patients and the expression of TIPE2, CD8, CD20, and CD66b

The corresponding clinicopathological data of 110 CRC patients were collected, including rectal cancer (n = 45) and colon cancer (n = 65). The survival rate at the last follow-up was 68.2% (n = 75). The mean survival time was 52.63 months (range 1 to 64 months). The Reverse Kaplan-Meier method was used to calculate the median time to last patient follow-up. The median time to last follow-up of patients was 64 months. IHC staining was used to detect the expression of TIPE2, CD8, CD20, and CD66b in cancer and adjacent normal tissues of CRC patients (Fig. 1). Generally, TIPE2, CD8, and CD20 expression in cancer tissues was higher than in adjacent normal tissues. CD66b expression in cancer tissues was lower than in adjacent normal tissues. Univariate analysis was conducted to assess the relationship between TIPE2 expression levels in cancerous tissues, adjacent normal tissues, and each variable (Table 1). The differences in cancer tissues and adjacent normal tissues targeting high and low CD8 expression and patients’ clinicopathological indicators were analyzed, respectively (Supplementary Table 1). The results showed that there was a significant difference in survival status between patients with high and low CD8 expression. The remaining indicators did not show significant differences. The cut-off values of those are as follow. In cancer tissues, TIPE2 (median = 6, n = 78 and 32, respectively), CD8 (median = 120, n = 60 and 50, respectively), CD20 (median = 200, n = 58 and 52, respectively), and CD66b (median = 15, n = 60 and 50, respectively). In adjacent normal tissues, TIPE2 (median = 5, n = 79 and 31, respectively), CD8 (median = 80, n = 60 and 50, respectively), CD20 (median = 91.5, n = 57 and 53, respectively), and CD66b (median = 19, n = 58 and 52, respectively).

Fig. 1
Fig. 1The alternative text for this image may have been generated using AI.
Full size image

The expression of TIPE2, CD8, CD20, and CD66b in colorectal cancer tissues and adjacent normal tissues (200×, Scale bar = 100 μm). (A-D) High expression of TIPE2, CD8, CD20, and CD66b. (E-H) Low expression of TIPE2, CD8, CD20, and CD66b.

Table 1 The relationship between TIPE2 expression and clinicopathological features.
Full size table

Kaplan-Meier survival analysis

The relationships between IHC scores and survival were evaluated by Kaplan-Meier survival analysis (Fig. 2). High TIPE2 expression in CRC and normal tissues correlated with a better OS than low TIPE2 expression (P = 0.00048 and P = 0.01, respectively). In contrast, high CD8 expression in CRC and normal tissues correlated with a worse OS (P = 0.0033 and P < 0.0001, respectively). Moreover, high CD20 expression in CRC and normal tissues were associated with a worse OS (P = 0.041 and P = 0.025, respectively). Finally, high CD66b expression in normal tissues was associated with a worse OS (P < 0.0001). However, there was no significant difference of OS between high and low CD66b expression in CRC tissues (P = 0.086). Furthermore, the relationships between relevant clinicopathological features and survival were presented in Fig. 3. Poor differentiation was associated with a worse OS than moderate and well differentiation (P = 0.0099). T4 was associated with a worse OS than T1, T2, and T3 (P = 0.0059). N2 was associated with a worse OS than N0 and N1 (P < 0.0001). WBC count < 3.5 *10^9/L was associated with a worse OS than 3.5–9.5 *10^9/L and > 9.5 *10^9/L (P = 0.027). Neutrophil count < 1.8 *10^9/L was associated with a worse OS than 1.8–6.3 *10^9/L and > 6.3 *10^9/L (P = 0.029). CEA > 5 ng/mL was associated with a worse OS than < 5 ng/mL (P = 0.035). CA199 > 27 µg/mL was associated with a worse OS than < 27 µg/mL (P = 0.00076). However, there were no significant survival difference of OS in hemoglobin, lymphocytes, and monocyte (P = 0.46, 0.71 and 0.5, respectively). Besides, patients were divided into four subgroups based on TIPE2-C and CD8-A expression levels: (1) TIPE2-C high/CD8-A high; (2) TIPE2-C high/CD8-A low; (3) TIPE2-C low/CD8-A high; (4) TIPE2-C low/CD8-A low. Kaplan-Meier survival analysis showed that patients in the TIPE2-C low/CD8-A high group had the worst prognosis, while those in the TIPE2-C high/CD8-A low group had the best prognosis (Supplementary Fig. 4 A), which was consistent with earlier results. In addition, patients were stratified by TNM stage (Stage I, II, and III) and performed separate survival analyses for each subgroup. The results showed that high TIPE2 expression was significantly associated with poorer prognosis in Stage III colorectal cancer (Supplementary Fig. 4B, C, D). This suggests that TIPE2 may serve as a potential prognostic biomarker and could help guide treatment decisions, particularly in Stage III disease. In addition, patients with high and low CD8 + infiltration in cancer tissues were stratified and survival analyzed based on TIPE2 expression (Supplementary Fig. 4E, F). The results showed a significant difference in survival between patients with high and low expression of TIPE2, which was not only present in all patients, but also when stratified by the degree of CD8 infiltration. This phenomenon may indicate that TIPE2 has a very significant ability to differentiate patients’ prognosis, regardless of the degree of CD8 infiltration. Exploring the prognostic impact of TIPE2 or CD8 on colorectal cancer patients individually may not be as meaningful as a combined analysis, as these factors may influence each other.

Fig. 2
Fig. 2The alternative text for this image may have been generated using AI.
Full size image

Kaplan-Meier survival curves were generated for TIPE2, CD8, CD20, CD66b expression levels in colorectal cancer tissues and adjacent normal tissues. (A) TIPE2 expression level in colorectal cancer tissues. (B) CD8 expression level in colorectal cancer tissues. (C) CD20 expression level in colorectal cancer tissues. (D) CD66b expression level in colorectal cancer tissues. (E) TIPE2 expression level in adjacent normal tissues. (F) CD8 expression level in adjacent normal tissues. (G) CD20 expression level in adjacent normal tissues. (H) CD66b expression level in adjacent normal tissues.

Fig. 3
Fig. 3The alternative text for this image may have been generated using AI.
Full size image

Kaplan-Meier survival curves were generated for clinicopathological features. (A)Grade (B) T-stage of colorectal cancer patients. (C) N-stage of colorectal cancer patients. (D) WBC levels of colorectal cancer patients. (E) Hb levels of colorectal cancer patients. (F) CEA levels of colorectal cancer patients. (G) CA199 levels of colorectal cancer patients. (H) Neutrophil levels of colorectal cancer patients. (I) Lymphocyte levels of colorectal cancer patients. (J) Monocyte levels of colorectal cancer patients.

Cox proportional risk regression analysis for univariate, multivariate analysis

11 variables were selected by Lasso regression analysis and included into following Cox proportional risk regression analysis (Fig. 4), including grade, T-stage, N-stage, WBC, neutrophil, CA199, TIPE2-C, TIPE2-A, CD20-C, CD8-C, and CD8-A. Univariate analysis showed that all of above 11 variables were important factors of OS in CRC patients (Table 2). All variables were included to establish a Cox proportional risk model by multivariate analysis (Fig. 5). The multivariate analysis results for prognosis in the form of a table were presented (Supplementary Table 2). The results showed that N-stage, CA199, and CD8 in adjacent normal tissues were independent risk factors for OS. However, grade and TIPE2 expression in cancer tissues were independent protective factors for OS in CRC patients.

Fig. 4
Fig. 4The alternative text for this image may have been generated using AI.
Full size image

Lasso regression analysis for filtering variables. (A) LASSO coefficient profiles of the IHC scores and clinicopathological features. (B) A coefficient profile plot was generated against the log (λ) sequence.

Table 2 Cox univariate proportional hazard regression analysis of CRC patients.
Full size table
Fig. 5
Fig. 5The alternative text for this image may have been generated using AI.
Full size image

Forest plot for Cox proportional risk regression multivariate analysis.

Construction and validation of the prediction model for OS in CRC patients

Independent risk factors screened by Cox proportional risk regression analysis were incorporated into the construction of nomogram to obtain a risk prediction model for CRC patients (Fig. 6). In the nomogram, total points ranged from 0 to 200 for CRC patients. For example, when the differentiation was poor, patients with CRC scored 100 points. ROC curves showed that the1-, 3-, and 5-year survival area under the curve (AUC) were 0.642, 0.691 and 0.690, respectively (Fig. 7A). The calibration curves showed that the constructed prediction model can be used as an effective model for 1-, 3-, and 5-year OS (Fig. 7B, C, D). DCA was widely used to evaluate the clinical value of nomogram. The 1-, 3- and 5-year net benefits of the nomogram model are presented in Fig. 8. The DCA curves demonstrated a significant positive net benefit from the risk of death, indicating its great clinical practical value in predicting OS of CRC.

Fig. 6
Fig. 6The alternative text for this image may have been generated using AI.
Full size image

Nomogram prediction model for prognosis in CRC patients.

Fig. 7
Fig. 7The alternative text for this image may have been generated using AI.
Full size image

Validations for the nomogram prediction model in CRC patients. (A) ROC curve of the 1-, 3- and 5-year OS nomogram prediction model of CRC patients; AUCs were 0.642, 0.691 and 0.690, respectively. (B) Calibration curve of the 1-year OS nomogram prediction model of CRC patients. (C) Calibration curve of the 3-year OS nomogram prediction model of CRC patients. (D) Calibration curve of the 5-year OS nomogram prediction model of CRC patients. CRC, colorectal cancer; ROC, receiver operating characteristic; AUC, area under the ROC curve; OS, overall survival; TP, True Positive; FP, False Positive.

Fig. 8
Fig. 8The alternative text for this image may have been generated using AI.
Full size image

Decision curve analysis of the nomogram for 1-, 3- and 5-year survival prediction of CRC patients.

Detection of T cell exhaustion markers and correlation with TIPE2 expression and CD8 + T cell infiltration

To further investigate the potential mechanism underlying the unexpected association between high CD8⁺ T cell infiltration and poor prognosis, the hypothesis of T cell exhaustion was explored. IHC staining was performed to assess the expression of two well-established exhaustion markers, PD-1 and TIM-3, on 50 paired CRC tumor and adjacent normal tissue samples. The results showed that both PD-1 and TIM-3 were markedly upregulated in tumor tissues compared to matched adjacent normal tissues (Supplementary Fig. 5 A, B, D, E). To validate these findings, we analyzed gene expression data from the GEPIA database (http://gepia.cancer-pku.cn/). Consistently, PD-1 and TIM-3 mRNA levels were significantly higher in CRC tissues than in normal controls (Supplementary Fig. 5 C, F). Subsequently, correlation analysis was performed to examine the relationship between TIPE2, CD8, and the exhaustion markers. TIPE2 expression was positively correlated with PD-1 and TIM-3 protein levels (Fig. 9A, B), suggesting a potential role for TIPE2 in modulating immune exhaustion or facilitating immune escape. CD8⁺ T cell infiltration was also positively associated with PD-1 and TIM-3 expression (Fig. 9C, D), supporting the hypothesis that T cell exhaustion may contribute to the adverse prognostic impact of CD8⁺ T cell enrichment in CRC. Further validation using the GEPIA database revealed consistent results at the transcriptomic level, with PD-1 and TIM-3 showing positive correlations with both TIPE2 and CD8 expression (Fig. 9E–H).

Fig. 9
Fig. 9The alternative text for this image may have been generated using AI.
Full size image

Correlation analysis between TIPE2, CD8, and T cell exhaustion markers PD-1 and TIM-3. (A, B) TIPE2 protein expression was positively correlated with PD-1 (A) and TIM-3 (B) in CRC tissues, as determined by IHC staining and semi-quantitative scoring. (C, D) CD8⁺ T cell infiltration was also positively associated with PD-1 (C) and TIM-3 (D) expression. (E–H) Transcriptomic correlation analyses using the GEPIA database confirmed that PD-1 and TIM-3 mRNA levels were positively correlated with both TIPE2 (E, F) and CD8 (G, H) expression in CRC.

Discussion

The occurrence and development of CRC is a multi-stage process, starting with the normal mucous membrane to colon adenoma and progressing to carcinoma in situ and invasive carcinoma16. The pathogenesis of CRC is related to many aspects, including patient genetic factors, age, lifestyle, living environment, and inflammation17. Indeed, several treatment options are available for CRC. However, it is difficult to predict the prognosis for CRC patients, even if they have undergone radical surgery. In case of tumor metastasis and recurrence, fewer treatment options are available for this patient population and are often ineffective which leading to the poor prognosis18. Therefore, the current prognostic prediction approach for CRC warrants further improvements to guide the appropriate follow-up treatment such as radiotherapy and chemotherapy, which could decrease the mortality rate of CRC.

TIPE2 is a negative regulator of inflammation and immune homeostasis involved in various cellular immune responses and is preferentially expressed by lymphoid and myeloid cells5. However, TIPE2 is also expressed in various non-hematopoietic cell types, including hepatocytes, squamous epithelial cells, transitional epithelial cells, and glandular epithelial cells19. Moreover, the germline deletion of TIPE2 has been reported to cause fatal inflammation and hypersensitivity to Toll-like receptor and T-cell receptor signaling20. Importantly, TIPE2 expression is closely related to the occurrence and metastasis of tumors21. When challenged by noxious stimuli, endogenous Tipe2 paradoxically facilitates the early stages of carcinogenesis, particularly during AOM/DSS-induced inflammatory cascades. But, exogenous Tipe2 overexpression exerts inhibitory effects on neoplastic cell proliferation and survival22. A previous study reported that TIPE2 was strongly expressed in colon cancer tissues, and its expression was related to lymph node metastasis and Dukes stage. Besides, TIPE2 regulates the inflammatory effects of TLR4 by inhibiting caspase-8 activity in colon cancer cells and suppresses further amplification of the cascade23.

The TME consists of a heterogeneous population of tumor cells surrounded by endogenous stromal cells recruited by tumors. Immune cells are recruited from the local host matrix and promote extracellular matrix remodeling, cell migration, angiogenesis, invasion, and resistance generation by producing various growth factors, chemokines, and cytokines, and even evade immune surveillance24. Among these, CD8 + T cells protect against exogenous invasion (e.g., pathogens, viruses, and bacteria) and play an irreplaceable role against endogenous dangers (e.g., malignant cells)25. CD8 + T cells can recognize and destroy tumor cells through intracellular antigens and inhibit proliferation and metastasis of tumor cells26. Nevertheless, other studies indicated that CD8 + T cell pools deviate from tumor reactivity due to their low immunogenicity27 and poor activation in the TME28 which is conducive to cell proliferation29. Furthermore, some researches showed that, in the tumor microenvironment, the CD8 + T cell exhaustion is characterized by the loss of properties of the functional T cells which is the most studied type of T cell dysfunction and holds broad landscape30. In our study, the presence of large numbers of CD8 + T cells was associated with poor prognosis, possibly due to the dysfunction of these cells. For instance, due to continuous exposure to antigens and inflammatory signals, CD8 + T cells progressively lose their normal function and enter a state of exhaustion. These exhausted T-cells typically express multiple inhibitory receptors such as PD-1 and LAG3, lose their killing function and produce cytokines which promote tumors31. The presence of activated CD8 + T cells is reportedly associated with a positive prognosis in early-stage CRC32. The unexpected association between high CD8⁺ T cell infiltration and poor prognosis observed in our cohort may be explained by the presence of T cell exhaustion. IHC analysis of PD-1 and TIM-3 demonstrated that both exhaustion markers were highly expressed in colorectal cancer tissues compared to adjacent normal tissues. Furthermore, CD8 expression was positively correlated with PD-1 and TIM-3 levels, suggesting that increased CD8⁺ T cell density may not reflect effective anti-tumor immunity but rather an exhausted immune state. This phenomenon could account for the paradoxical link between high CD8⁺ T cell infiltration and unfavorable outcomes.

Most previous studies have compared adjacent normal tissue with cancer tissue, thereby highlighting the role of tumor markers in cancer tissue. Few studies have performed prognostic analyses of tumor patients only for adjacent normal tissue. By reviewing additional literature, there is evidence that the level of infiltration of adjacent non-tumorigenic immune cells was significantly higher in non-small cell lung cancer than in tumor samples. The difference in infiltration between the tumor region and the paraneoplastic region showed stronger prognostic predictive power than the tumor region alone, which indicating that the immune cell infiltration levels in the tumor and adjacent non-neoplastic regions provided complementary information33. One other study demonstrated that a high level of peritumorous zone infiltration by CD8 + T-lymphocytes has been identified as an unfavorable prognostic factor in renal cell carcinoma cases. The assessment of peritumorous zone infiltration by CD8 + T-lymphocytes has the potential to serve as an additional prognostic factor34. Besides, meta-analysis across various malignancies has demonstrated that PD‑1⁺ exhausted CD8⁺ T cells, which may accumulate in peritumoral sites, are associated with worse survival, including pancreatic cancer, gastric cancer and head and neck cancer35.

CD20 is a non-glycosylated protein expressed on the surface of B lymphocytes36 which has been widely used to assess the degree of inflammatory infiltration of different tumors37. It has been shown that tumor-infiltrating B lymphocytes (TIBs) are expressed in many tumors, secreting immunoglobulins to enhance T-cell responses and direct destruction of cancer cells, hindering tumor growth process and thus improving patient prognosis38. However, other studies have indicated that regulatory B cells (Bregs) promote tumor activity by secreting immunosuppressive substances, including IL10 and/or TGF-β39. Several studies have shown that TIL-Bs are associated with poor prognosis, such as clear cell Renal Cell Carcinoma40 and lower-grade glioma41. This situation may reflect the difficulty of distinguishing effector B cells from Bregs. Indeed, Bregs have consistently been associated with decreased patient survival when using IL-10, PD-1, or TIM-1 as CD19 complementary markers42. How to distinguish effector B cells from Bregs is a problem that needs to be addressed urgently.

CD66b + TAN is strongly associated with poor prognoses in tumors, such as hepatocellular carcinoma43 intrahepatic cholangiocarcinoma44 and gastric cancer45. Most importantly, it has been shown that CD66b + TAN is an independent risk factor for the prognosis of CRC patients46.

TIPE2 is closely related to many immune cells in the TME. Ample evidence suggests that the expression of TIPE2 may increase the percentage and absolute cell numbers of T cells in the spleen and lungs and their subtype CD8 + T cells in tumor-bearing mice by decreasing myeloid-derived suppressor cells (MDSCs)47,48. Interestingly, it has been shown that TIPE2-deficient leukocytes are defective in polarization and chemotaxis, and TIPE2-deficient mice are resistant to leukocyte-mediated neuroinflammation15.

In this study, the expression of TIPE2 with CD8, CD20, and CD66b in TME in cancer tissues and adjacent normal tissues was assessed by IHC staining. Data such as clinicopathological information and prognostic outcomes of the patients were retrospectively collected and analyzed. Importantly, we provided compelling evidence that TIPE2 is expressed in cancer and adjacent normal tissues, especially in the nucleus and cytoplasm of tumor cells. It is widely thought that differential expression of TIPE2 in tumor cells may be associated with different clinical features, which needs further validation. IHC staining was used to observe the varying degrees of immune cell infiltration in cancerous and adjacent normal tissues. Based on the previous literature, immune scoring was conducted for TIPE2, CD8, CD20, and CD66b. Based on IHC scores, we found that TIPE2, CD8 and CD20 expression in CRC tissues was higher generally than in adjacent tissues. Conversely, CD66b expression in CRC tissues was lower generally than in adjacent tissues. Survival analysis showed that overexpression of CD8 and CD20 in cancer tissues led to a poor overall prognosis for patients, while TIPE2 overexpression was associated with a better prognosis. However, CD8, CD20 and CD66b overexpression in adjacent normal tissues resulted in a poor overall prognosis, while TIPE2 overexpression was associated with a better prognosis. 11 variables selected by Lasso analysis were included into Cox univariate and multivariate analyses. It is showed that grade and TIPE2 expression in cancer tissues was the independent protective factors for OS. On the contrary, N-stage, CA199 and CD8 in adjacent normal tissue were independent risk factors for OS in CRC patients. Furthermore, the nomogram was construction to predict the risk of 1-, 3- and 5-year OS for CRC patients. The ROC curves and calibration curves were plotted to validate the prognostic model. In the present study, TIPE2 was correlated with CD8, CD20, and CD66b in the TME. Meanwhile, clinicopathological data were integrated with immunohistochemical scores to comprehensively analyze and predict the prognosis of CRC patients. Therefore, the analysis of TIPE2 and immune cell infiltration in tissues can improve the accuracy of CRC prognosis prediction to some extent.

Combined with our study and previous studies49,50 TIPE2 is indeed a tumor suppressor that plays an inhibitory role in the proliferation, migration and EMT of tumor cells. Deficiency of TIPE2 may accelerate the progression of colon cancer by enhancing IL-22 levels and loss of tumor suppressor function51. Colon cancer therapies targeting TIPE2 require cell- and pathway-specific approaches and provide a cautionary tale for colon cancer immunotherapy, as intestinal inflammation can both promote and suppress cancer. Additionally, although our study revealed that TIPE2 expression was positively correlated with immune checkpoint molecules PD-1 and TIM-3, which are associated with immune suppression and T cell exhaustion, we found that low TIPE2 expression was linked to poor prognosis. This apparent contradiction suggests that TIPE2 may play a more complex, context-dependent role in the tumor immune microenvironment. One possible explanation is that low TIPE2 expression reflects impaired immune regulation. In this context, both effective anti-tumor immune responses and immune homeostasis may be compromised, allowing tumor progression and immune evasion to occur simultaneously. Alternatively, TIPE2 might participate in non-immune pathways—such as those related to cell proliferation, apoptosis, or metastasis—which could independently contribute to disease progression when TIPE2 is downregulated. These findings imply that TIPE2 may function as a dual regulator in the tumor microenvironment, with its prognostic impact varying according to the immune context and cellular composition. Further functional studies are needed to clarify the cell-type specific roles of TIPE2 in cancer immunity and progression. Correlation analyses using GEPIA database further validated the co-expression patterns of TIPE2, CD8, and exhaustion markers at the transcriptomic level. Together, these data suggest that TIPE2 may contribute to shaping an immunosuppressive tumor microenvironment, partially through modulation of immune checkpoint pathways, thereby influencing clinical outcomes in colorectal cancer.

There were limitations in this research. First, this was a single-center retrospective study, and selection bias could not be avoided. Moreover, although different pathologists analyzed the IHC samples independently and a double-blind scoring was used, in addition to the limitations of the IHC method and tumor sampling, the field of view selection may have been biased, which may affect our results to a certain extent. Besides, the sample size included in the study was small. To address concerns about statistical power, we conducted a power analysis using R. Assuming a power of 80% and a significance level of 0.05, the minimum required sample size was 56 for TIPE2-C and 34 for CD8-A. Our actual sample size of 110 exceeded these thresholds. The calculated power was 88.3% for TIPE2-C and 99.9% for CD8-A, indicating adequate statistical power to support our findings. Non-parametric tests (Mann–Whitney U) was used to reduce the impact of sample size on statistical validity. Finally, only the relationship between TIPE2 and immune cells in TME was analyzed, while an in-depth analysis of the mechanism of TIPE2 and TME on CRC was not conducted. Due to technical constraints and the nature of available samples (formalin-fixed, paraffin-embedded tissue blocks only), we were unable to perform multiplex IHC or flow cytometry to identify specific immune cell subsets expressing TIPE2. We used single-marker IHC to assess PD-1 and TIM-3 in colorectal cancer tissues and conducted correlation analyses with TIPE2 and CD8 to explore their associations.

Conclusion

This study systematically analyzed the expression characteristics of TIPE2 in CRC. Besides, the prognosis of patients in the context of clinicopathological features and TME were explored. We provide hitherto undocumented evidence that TIPE2 combined with immune markers such as CD8, CD20, and CD66b can be used to assess disease progression and prognosis in CRC patients and contribute to the idea that TIPE2 may become a new immunotherapy-related target in the future.