The Prognostic Role of the IL-6 and RBP4 in Colorectal Cancer

1. Introduction

Colorectal cancer (CRC) is one of the most common cancers worldwide, ranking fourth in incidence and fifth in mortality. In 2022, 1 142 286 new CRC cases were diagnosed globally, and projections estimate that by 2040, this number will rise to 1 916 781 [1,2]. A major challenge in CRC management is the often-late onset of clinical symptoms, which hampers early diagnosis and reduces the effectiveness of treatment strategies. [3,4]

Chronic inflammation has emerged as a key driver in CRC development and progression. Immune cells infiltrating the tumor microenvironment (TME) release numerous pro-inflammatory cytokines, including interleukin-6 (IL-6), which amplify local inflammation.[5] IL-6, produced by immune cells (mainly macrophages), tumor cells, and intestinal epithelial cells, orchestrates immune responses, regulates lymphocyte maturation, and contributes to macrophage differentiation.[6,7] In the CRC microenvironment, IL-6 stimulates cancer-associated fibroblasts (CAFs), which in turn further secrete IL-6, creating a positive feedback loop that sustains chronic inflammation.[5,6] Moreover, IL-6 promotes angiogenesis, enhances cancer cell motility through hepatocyte growth factor (HGF) signaling, and impairs anti-tumor immunity by generating dysfunctional natural killer (NK) cells.[8,9,10,11,12,13] Elevated IL-6 levels have been associated with larger tumor size, advanced tumor stage, metastasis, poor prognosis, and increased recurrence risk in esophageal cancer, multiple myeloma, and renal cell carcinoma among others.[7,11,14]

Interestingly, growing evidence points to the role of metabolic dysregulation and adipokines in modulating the inflammatory TME and influencing cancer progression. One such molecule is retinol-binding protein 4 (RBP4), a serum transporter of retinol secreted primarily by the liver and adipocytes.[15,16,17] Recent studies have highlighted that RBP4 is not only a metabolic regulator but also a contributor to tumor biology. Mechanistically, RBP4 can activate macrophages via exosomal signaling, trigger STRA6-mediated Janus kinase (JAK)/STAT3/STAT5 activation, and promote pro-inflammatory responses that support tumor progression, migration, and metastasis.[17,18,19] In gastric cancer cells, RBP4 expression is upregulated, and its silencing inhibits tumor proliferation, migration and invasion.[20] RBP-4 also increases insulin resistance by reducing phosphatidylinositol 3-phosphate kinase, leading formation of pancreatic adenocarcinoma.[21] In CRC, elevated RBP4 expression in tissue correlates with metastasis and unfavorable prognosis.[15,16,22] Furthermore, RBP4-driven endothelial activation enhances vascular permeability and angiogenesis, contributing to tumor growth and dissemination.[19] Notably, RBP4 links adiposity and insulin resistance to cancer biology, with its role being more pronounced in individuals with lower BMI due to adipose tissue dysfunction in obesity.[19,21,23,24,25]

In this study, we aimed to investigate the expression patterns and potential clinical significance of IL-6 and RBP4 in colorectal cancer (CRC). The initial phase involved the immunohistochemical analysis of IL-6 and RBP4 expression in CRC tissues and adjacent non-cancerous tissues. Protein expression levels were evaluated based on our own immunohistochemical data, while associations with clinicopathological parameters and overall survival (OS) were systematically analyzed. To broaden our understanding, bioinformatic analyses were conducted using publicly available transcriptomic datasets to assess IL-6- and RBP4-related transcriptional landscapes and their potential regulatory networks. Functional enrichment and pathway analyses were performed to identify biological processes, molecular functions, and cellular components associated with IL-6 and RBP4 expression. By integrating experimental and bioinformatic approaches, this study aims to define the prognostic value of IL-6 and RBP4 expression in CRC and to provide novel insights into their roles in colorectal tumorigenesis and potential as therapeutic targets.

2. Materials and Methods

2.1. Patients and Tissue Material

The study was conducted on tumor and adjacent non-tumor tissue samples collected from patients undergoing colectomy in 2015–2021 at the Department of Surgery of Dr J. Biziel from the University Hospital No. 2 in Bydgoszcz. Patients diagnosed with colorectal adenocarcinoma were included in this investigation, while the exclusion criteria were as follows: hereditary colorectal cancer, neoadjuvant chemotherapy before resection, and reoperation due to recurrent cancer. The study group included 118 patients (47 females, 71 males) with an average age of 68 years at operation (median 68, range 27–89). Herein, pathologic TNM stage is based on the American Joint Committee on Cancer (AJCC) 8th edition. Histologically normal tissues adjacent to the tumor were available for 42 cases, and these constituted the control group. Clinical follow-up data, including patient survival status, were collected from hospital records. Overall survival (OS) was defined as the time from surgery until the last recorded follow-up or death. The end date for survival analysis was set at 25 March 2025.

This study was carried out in accordance with the guidelines of the Declaration of Helsinki. The protocol was approved by the Nicolaus Copernicus University Ethics Committee (no. 454/2021).

2.2. Tissue Microarrays and Immunohistochemical Staining

Immunohistochemical (IHC) staining was performed on tissue microarrays according to the previously described protocol by Antosik et al. IL-6 and RBP4 protein was stained in BenchMark® ULTRA (Roche Diagnostics/Ventana Medical Systems, Tucson, AZ, USA) using ultraView Universal DAB Detection Kit (Roche Diagnostics/Ventana, Tucson, AZ, USA) according to the previously validated protocol. [26] Tissue sections were incubated with primary rabbit polyclonal anti-Il-6 antibody (dilution 1:100, cat. no: ab6672, Abcam, Cambridge, UK) and rabbit polyclonal anti-RBP4 antibody (dilution 1:250, cat. no: 5950-RBM23-P1ABX, Thermo Fisher Scientific, Waltham, MA, USA). Sections of normal human testis were used as a positive control, while a negative control was obtained by omitting the primary antibody.

2.3. Evaluation of Immunohistochemical Staining

The protein expression was assessed at 20× objective magnification in a blinded manner by two independent researchers, including a senior pathologist (DG), using the DP600 scanner (Roche Diagnostics/Ventana Medical Systems, Tucson, AZ, USA). Immunohistochemical expression was evaluated using the H-score system, which combines staining intensity (scored 0–3) and the percentage of positively stained cells. The final score ranged from 0 to 300. For analytical purposes, protein expression was dichotomized into low (negative) and high (positive) categories based on discriminatory cut-off values determined using the Evaluate Cutpoints software (R version 3.4.1).[27] The cut-off values were set at <225 for low and ≥225 for high IL-6 expression, and <175 for low and ≥175 for high RBP4 expression, respectively.

Gene expression data were obtained from the UCSC Xena platform in the form of RSEM expected counts and subsequently normalized using the DESeq2 package. The initial RNA-seq transcriptomic data were aligned with the STAR aligner and quantified via RSEM. The study cohort comprised 350 CRC cases from The Cancer Genome Atlas (TCGA). Expression analyses of IL-6 and RBP4 were conducted based on RNA-seq data sourced from the UCSC Xena Browser (http://xena.ucsc.edu; accessed July 7, 2025). The DESeq2 median-of-ratios normalization method was employed to correct for differences in sequencing depth and RNA composition among samples. For IL-6, a cut-off value of 6.728 was determined, classifying samples into low (<6.728) and high (≥6.728) expression groups. For RBP4, a cut-off value of 7.213 was determined, classifying samples into low (<7.213) and high (≥7.213) expression groups.

To identify genes with positive correlation to IL-6 and RBP4, data from the TCGA cohort were analyzed using the cBioPortal platform (https://www.cbioportal.org; accessed July 7, 2025), extracting the top 50 co-expressed genes for each target. Pathway enrichment analyses were performed with the Reactome Pathway Database (https://reactome.org; accessed July 7, 2025) and the KEGG Pathway Database (https://www.genome.jp/kegg/pathway.html; accessed July 9, 2025) to explore molecular mechanisms potentially involved in CRC pathogenesis.

Protein–protein interaction (PPI) networks were constructed for the top IL-6- and RBP4-associated genes using the STRING database (https://string-db.org; accessed July 14, 2025), applying a medium confidence score threshold (0.700). Network visualizations were generated in Cytoscape software (version 3.10.3, Cytoscape Consortium, San Diego, CA, USA). Functional annotation of co-expressed genes, including classification into Gene Ontology (GO) categories—biological processes (BP), molecular functions (MF), and cellular components (CC)—was conducted using the Database for Annotation, Visualization and Integrated Discovery (DAVID; https://david.ncifcrf.gov; accessed July 9, 2025).

2.4. Statistical Analysis

All statistical analyses and visualizations were performed using GraphPad Prism (version 8.0; GraphPad Software, San Diego, CA, USA) and RStudio (version 1.3.1093). The Shapiro–Wilk test was applied to assess the normality of continuous variables. Based on the distribution of the data, appropriate parametric or non-parametric statistical tests were selected. Comparisons of IL-6 and RBP4 expression levels between tumor tissues and adjacent non-tumorous tissues were conducted using the Mann–Whitney U test. The relationships between categorized IL-6 and RBP4 expression and clinicopathological parameters were analyzed using the chi-square test or Fisher’s exact test, depending on the expected cell counts. Overall survival (OS) was estimated by Kaplan–Meier survival analysis, and differences between survival curves were evaluated with the log-rank test. All p-values were two-tailed, and a significance level of p < 0.05 was considered statistically significant.

3. Results

3.1. Immunoexpression of Il-6 and RBP4 in Tumor and Normal Adjacent Tissue and Its Clinicopathological Associations of Our Cohort

Cytoplasmic expression of IL-6 was observed in CRC tissues. The staining pattern was characterized by diffuse, non-uniform cytoplasmic labeling of tumor cells, with areas of more intense staining, particularly at the invasive front. Additionally, IL-6 expression was noted in stromal immune cells, including macrophages, within the tumor microenvironment. In adjacent non-cancerous tissues, cytoplasmic IL-6 expression was also detected but at markedly lower levels. Representative images illustrating IL-6 immunohistochemical staining in tumor and adjacent tissues are presented in Figure 1. Quantitative analysis revealed significantly higher IL-6 expression in CRC tissues compared to adjacent tissues (median: 183.0 vs. 129.0; p < 0.0001; Figure 2A).

Cytoplasmic expression of RBP4 was demonstrated in CRC tissues. The cytoplasmic staining pattern was characterized by ubiquitous, diffuse labeling of tumor cells, although with variable intensity across the tumor tissue. Additionally, RBP4 expression was observed in macrophages present within the tumor microenvironment. In the control group, consisting of adjacent non-cancerous tissues, cytoplasmic RBP4 expression was also detected. Representative images illustrating RBP4 immunohistochemical staining in tumor and adjacent tissues are presented in Figure 1. RBP4 expression was significantly elevated in colorectal cancer (CRC) tissues compared to adjacent non-tumorous tissues (p < 0.0001; Figure 2B).

The association between IL-6 protein expression and clinicopathological features was assessed in 118 colorectal cancer (CRC) patients (Table 1). High IL-6 expression was observed in 27 cases (22.9%), while low expression was present in 91 cases (77.1%). No significant associations were identified between IL-6 expression and gender, tumor grade, primary tumor (pT) status, lymph node involvement (pN), distant metastasis (pM), TNM stage, vascular invasion (VI), perineural invasion (PNI), resection margin status, or primary tumor location (all p > 0.05). However, significantly higher IL-6 expression was observed in patients aged over 65 years (p = 0.0254), in those with BMI ≥30 kg/m² (p < 0.0001), and in patients with elevated waist circumference (≥88 cm for females, ≥102 cm for males; p = 0.0127) (Table 1).

Applying the defined cut-off value, high RBP4 expression was identified in 50 cases (42.37%), while low expression was observed in 68 cases (57.63%). High RBP4 expression was significantly associated with advanced pT status (p = 0.0398), positive lymph node involvement (pN1–2; p = 0.0044), and advanced TNM stage (III–IV; p = 0.0119). Furthermore, patients with obesity (BMI ≥ 30 kg/m²) demonstrated significantly higher RBP4 expression compared to non-obese patients (p = 0.0264). Patients with elevated waist circumference (≥88 cm in females, ≥102 cm in males) showed a higher frequency of RBP4 overexpression (26 cases, 52%) compared to those with lower waist circumference (16 cases, 33%) (p = 0.0396). No significant associations were found between RBP4 expression and age, gender, tumor grade, distant metastasis (pM), vascular or perineural invasion, resection margin status, or primary tumor location (Table 1).

3.2. Association with the Clinical Outcome of Our Cohort

In Kaplan–Meier survival analysis of our cohort, IL-6 high expression was significantly associated with shortened median overall survival (OS) of 67 months compared with 105 months in the IL-6 low-expression group (p = 0.0081; HR = 2.136, 95% CI: 1.061–4.302; Figure 3A). Similarly, RBP4 high expression was significantly associated with reduced median OS of 59 months, whereas the median OS in the RBP4 low-expression group was not reached (p < 0.0001; HR = 4.360, 95% CI: 2.417–7.868; Figure 3B).

3.3. In Silico Investigation of Functional Pathways Connected to RBP4 and IL-6-Related Genes of the TCGA Cohort

In the TCGA cohort, IL-6 mRNA expression was significantly higher in CRC tissues compared to adjacent normal tissues (p = 0.0243; Figure 4A). In contrast, RBP4 mRNA expression did not show a significant difference between tumor and adjacent tissues (p = 0.6338; Figure 4B).

We evaluated whether IL-6 and RBP4 mRNA expression levels were associated with clinicopathological parameters of CRC patients (Table 2). Based on the defined cut-offs, 191 (54.57%) patients showed low IL-6 expression, while 159 (45.43%) demonstrated high IL-6 expression. Regarding RBP4 expression, 95 (27.14%) cases were classified as low, and 255 (72.86%) as high. In terms of age, 81 patients aged ≤65 years showed high IL-6 expression, compared to 78 (43.09%) patients aged >65 years. For RBP4, high expression was observed in 121 and 134 patients aged ≤65 and >65 years, respectively. No statistically significant associations were observed for either IL-6 or RBP4 expression with age. When examining gender, high IL-6 expression was observed in 79 males and 80 females. A significant association was found between RBP4 expression and gender (p = 0.0359), with high expression more prevalent among females. Regarding tumor invasion (pT status), high IL-6 expression was detected in 29 T1–T2, 113 T3, and 17 (36.17%) T4 patients. No significant associations were found for either IL-6 or RBP4 expression with pT status. For nodal status (pN), high IL-6 expression occurred in 94 N0 and 65 N1–N2 cases. In contrast, RBP4 expression was significantly associated with pN status (p = 0.0018). As for distant metastasis (pM status), high IL-6 expression was found in 110 M0 and 27 M1 cases. Notably, RBP4 expression was significantly associated with pM status (p = 0.0086), with a lower frequency of high RBP4 expression in M1 patients. Cases with undefined pM status (MX, n = 55) were excluded from the statistical analysis. Finally, when evaluating TNM stage, high IL-6 expression was observed in 25 stage I, 63 stage II, 48 stage III, and 23 stage IV patients. RBP4 expression showed a significant association with TNM stage (p = 0.0017), with the highest proportion of high expression seen in early-stage patients.

In Kaplan–Meier survival analysis of the TCGA cohort, patients were stratified into low- and high-expression groups for IL-6 (cut-off = 6.728) and RBP4 (cut-off = 7.213) using Evaluate Cutpoints software (R version 4.4.2). High IL-6 expression was significantly associated with poorer overall survival (median overall survival (OS) of 65.85 months compared with undefined) (p = 0.0215; HR = 1.654, 95% CI: 1.067–2.563; Figure 5A). For RBP4, high expression showed a trend toward reduced survival, though it did not reach statistical significance (median overall survival (OS) of 67.30 months compared with undefined) (p = 0.0513; HR = 1.757, 95% CI: 1.076–2.868; Figure 5B).

Spearman’s rank correlation analysis of CRC data obtained from cBioPortal (accessed July 7, 2025) identified genes whose expression levels were positively or negatively associated with IL-6 and RBP4. Positive correlations indicate genes with expression patterns that increase alongside IL-6 or RBP4, while negative correlations reflect genes whose expression decreases as IL-6 or RBP4 expression rises. Among the top 50 genes positively correlated with IL-6 (Table 3), IL11 (ρ = 0.765, p = 5.83 × 10⁻¹⁰²) and SOCS3 (ρ = 0.753, p = 9.05 × 10⁻⁹⁷) showed the strongest associations. For negatively correlated genes (Table 4), HNF1B (ρ = –0.356, p = 4.10 × 10⁻¹⁷) and IL17RE (ρ = –0.356, p = 4.39 × 10⁻¹⁷) demonstrated the highest inverse relationships. Similarly, for RBP4, COLEC11 (ρ = 0.369, p = 2.19 × 10⁻¹⁸) and CDIPT (ρ = 0.368, p = 2.95 × 10⁻¹⁸) were the most strongly positively correlated genes (Table 5), while HSPA4L and PPAT again emerged as the strongest negative correlations (ρ = – 0.398, p = 2.71 × 10⁻²¹ and ρ = –0.365, p = 6.35 × 10⁻¹⁸, respectively; Table 6).

Functional enrichment analysis of the top 50 genes most significantly positively correlated with IL-6 in CRC was performed using Reactome (accessed July 7, 2025). In (Figure 6A), the network visualization highlights strong enrichment for multiple immune-related pathways. All these pathways showed the same level of statistical significance (p < 1.1 × 10⁻¹⁶). In (Figure 6B), the heatmap illustrates the distribution of individual genes across these pathways, with CXCL8 (Interleukin-8) being the most frequently represented, emphasizing its central role in immune system regulation.

In contrast, analysis of the top 50 genes negatively correlated with IL-6 revealed enrichment for signaling and transcriptional pathways unrelated to immune processes (Figure 7A). These included the Nuclear Receptor transcription pathway and FGFR3 mutant receptor activation. As shown in (Figure 7B), ERBB2 was the most frequently involved gene among the negatively correlated set, appearing across multiple enriched pathways. All these pathways reached the same level of statistical significance (p < 5.37 × 10⁻⁸).

To further explore these associations, a similar analysis was performed for RBP4-correlated genes. For the top 50 positively correlated genes, the network visualization (Figure 8A) demonstrated enrichment predominantly in signal transduction and developmental biology pathways. The most significant pathway was GLI proteins bind promoters of Hh responsive genes to promote transcription with a p value of 2.38 × 10⁻⁸. The heatmap (Figure 8B) revealed ACTB as the most frequently represented gene across these enriched categories.

Meanwhile, the top 50 negatively correlated genes showed enrichment mainly in pathways related to cell cycle and immune system processes (Figure 9A). The most significant pathways included Cell Cycle Mitotic (p = 0.0012) and Mitotic Anaphase (p = 0.0016). The heatmap (Figure 9B) identified CDC27 as the most frequently represented gene within these pathways.

A parallel Gene Ontology enrichment analysis was carried out for genes whose expression correlates with IL-6 using the DAVID tool (accessed July 9, 2025). Genes positively correlated with IL-6 (Figure 10A, 11C, 11E) are most significantly enriched for the cellular response to lipopolysaccharide (GO:0071222), localize predominantly to the extracellular space (GO:0005615), and exhibit chemokine activity (GO:0008009). In contrast, genes negatively correlated with IL-6 (Figure 10B, Figure 10D, Figure 10F) are dominated by brush border assembly (GO:1904970) among biological processes, cytosolic localization (GO:0005829) as their principal cellular component, and growth factor binding (GO:0019838) as their top molecular function.

In Figure 11, Gene Ontology enrichment analysis of proteins whose abundance correlates positively (Figure 11A, 11C, 11E) or negatively (Figure 11B, Figure 11D, Figure 11F) with RBP4 gene expression was performed using the DAVID tool (accessed July 9, 2025). Among the proteins up-regulated in parallel with RBP4, the Wnt signaling pathway (GO:0016055) is the most significant biological process (Figure 11A), the extracellular space (GO:0005615) is the predominant cellular component (Figure 11C), and hedgehog family protein binding (GO:0097108) is the top molecular function (11E). By contrast, proteins whose levels decrease as RBP4 expression rises are most strongly enriched for cell division (GO:0051301) as a biological process (Figure 11B), localization to the mitochondrion (GO:0005739) as a cellular component (Figure 11D), and ATP binding (GO:0005524) as a molecular function (Figure 11F). Together, these data indicate that proteins co-regulated with RBP4 partition into an extracellular signaling/developmental module versus a proliferation/mitochondrial energetics module.

A protein-protein interaction (PPI) network was constructed for genes positively correlated with IL-6 using STRING (https://string-db.org; accessed 14 July 2025) and visualized in Cytoscape v3.10.3. Figure 12A shows the top ten hub genes ranked by connectivity in the IL-6 positive network, with IL1B, TNF and IL10 as the three most connected nodes. Figure 12B presents the complete interactome of 50 nodes and 859 edges. This network has an average of 34.36 neighbors per node, a clustering coefficient of 0.819, a network density of 0.701, a diameter of 3, a radius of 2 and a characteristic path length of 1.300. It exhibits highly significant protein interaction enrichment (p < 1 × 10^-16).

For genes negatively correlated with IL-6, Figure 12C highlights the top ten hub genes, led by CDH1, PPARG and CDH17, while Figure 12D shows the full network of 50 nodes connected by 144 edges. This network has an average of 5.76 neighbors per node, a clustering coefficient of 0.517, a network density of 0.118, a diameter of 5, a radius of 3 and a characteristic path length of 2.663, reflecting a more modular architecture centered on epithelial adhesion and differentiation pathways.

Similarly, PPI network was constructed for proteins positively correlated with RBP4 using STRING (https://string-db.org; accessed 14 July 2025) and visualized in Cytoscape v3.10.3. Figure 13A shows the top ten hub proteins ranked by connectivity in the RBP4 positive network, with ACTB, MAPK3 and ACTA2 as the three most connected nodes. Figure 13B presents the complete interactome of 50 nodes and 130 edges. This network has an average of 5.31 neighbors per node, a clustering coefficient of 0.410, a network density of 0.111, a diameter of 7, a radius of 4 and a characteristic path length of 2.771.

For proteins negatively correlated with RBP4, Figure 13C highlights the top ten hub proteins, led by HSPA4, HSPA9 and ABCE1, while Figure 13D shows the full network of 50 nodes connected by 353 edges. This network has an average of 14.12 neighbors per node, a clustering coefficient of 0.653, a network density of 0.288, a diameter of 4, a radius of 2 and a characteristic path length of 2.109.

4. Discussion

Colorectal cancer is a major public health concern due to its high rate of occurrence and the fact that its symptoms are often non-specific, making diagnosis challenging.[28,29,30] For this reason in recent years, numerous studies have been conducted to identify new biomarkers involved in the pathogenesis and progression of colorectal cancer. Recent discoveries in this area have drawn researchers' attention to the link between tumorigenesis, obesity, and inflammation as important drivers of colorectal cancer.[10,11,12,18,19,21,24,31,32,33,34] We believed that this line of research was necessary and should continue. Thus, we selected IL-6 as the factor connecting inflammation to the growth of colorectal cancer and RBP4 as the link between obesity and tumorigenesis.

IL-6 is a pro-inflammatory cytokine that has been shown to induce acute phase responses, act as a maturation factor for B lymphocytes, participate in T lymphocyte differentiation, and activate natural killer (NK) cells.[6,7] Elevated levels of IL-6 are indicative of inflammation, a common occurrence in cancerous diseases. The initiation of the JAK/STAT3, Ras/MAPK, and PI3K/Akt pathways results in the regulation of gene products that stimulate proliferation, angiogenesis, or metastasis. Its role has been described in a variety of neoplasms, including multiple myeloma, renal cell carcinoma, cervical cancer, and prostate cancer, among others.[7,11] The presence of CEA in colorectal cancer has been demonstrated to stimulate the secretion of IL-6 by Kupffer cells. This, in turn, promotes colorectal cancer metastasis to the liver.[8,9] IL-6 stimulates the secretion of HGF (hepatocyte growth factor), which increases cell motogenic activity and also promote metastasis formation.[9,10] Furthermore, a correlation with a worse prognosis has been observed, which may be attributable to the inhibition of the anti-tumor response by IL-6. [9,12]

The present study demonstrated that the expression of interleukin-6 (IL-6) in the tumor tissue was significantly higher than in the adjacent tissue. In both tumor and adjacent tissue, the sites of highest IL-6 expression were cancer cells and inflammatory cells infiltrating the tumor, but to a lesser extent, expression also appeared in fibroblasts in the lining. This finding aligns with most reports on IL-6 expression. [8,9,10,11,14,33,34,35,36,37] Notably, the results obtained from immunohistochemical evaluation were further corroborated by in silico analyses of transcriptomic data derived from The Cancer Genome Atlas (TCGA), which confirmed increased IL-6 mRNA expression in colorectal cancer tissue compared to adjacent normal mucosa. These concordant findings strengthen the evidence for IL-6 upregulation in CRC and underscore its potential involvement in tumor-related inflammation.

In our study, a statistically significant association was observed between IL-6 expression and patient age, with higher expression levels noted in individuals over 65 years. This finding aligns with the concept of "inflammaging," which refers to age-related chronic low-grade inflammation and has been widely associated with increased circulating IL-6 levels. Although previous reports have demonstrated a general trend of elevated IL-6 with advancing age, the relationship is not universally consistent and may be influenced by factors such as comorbidities, gender, or immune status. Our findings support the hypothesis that aging contributes to a pro-inflammatory tumor microenvironment, which may in turn influence cancer progression.[38] A comparable absence of significant associations between IL-6 expression and clinicopathological parameters—including gender, tumor location, histological subtype, and tumor stage—was also demonstrated by Lu et al.further supporting the heterogeneity of IL-6 involvement in colorectal cancer pathophysiology. [34] In turn, results from other authors demonstrated that have shown that higher IL-6 expression correlates with metastasis,[10,39] angiogenesis,[35] greater tumor cell proliferation[35] and larger tumor size.[9,36] This phenomenon can be attributed, at least in part, to the role of inflammation in stimulating angiogenesis by inducing the expression of angiogenesis-related proteins such as VEGFA.[35] In addition, IL-6 autocrine increases the motogenic activity of tumor cells by binding to the receptor for IL-6 on the surface of tumor cells and increases the secretion of HGF, which promotes metastasis. [10,11] In the in silico analysis assessing the relationship between IL-6 mRNA expression and clinicopathological variables such as age, sex, pT, pN, pM status, and TNM stage, no statistically significant associations were identified. Discrepancies between the findings of different studies may be attributed to variations in cohort size, characteristics of the studied populations, as well as the inherent heterogeneity of colorectal cancer. This observation aligns with the findings of Legrand-Poels et al., who reported inconsistent IL-6 expression due to promoter suppression. [40] Moreover, some studies suggest that IL-6 expression may decrease in advanced-stage tumors, potentially reflecting adaptive changes in the tumor microenvironment. [33] In our study, significantly higher IL-6 expression was observed in colorectal cancer (CRC) patients with obesity (BMI ≥ 30 kg/m²) and increased waist circumference. This finding is consistent with the report by Lara Kern et al., who described IL-6 as a pleiotropic cytokine playing a central role in the crosstalk between metabolic dysfunction and inflammation. Elevated IL-6 levels in obesity contribute to a chronic low-grade inflammatory state that supports tumor development and progression. IL-6 exerts its effects through the activation of several oncogenic pathways, including JAK/STAT3, PI3K/Akt, and MAPK, which promote tumor cell proliferation, inhibit apoptosis, and enhance metastatic potential. In the context of CRC, our findings suggest that obesity-related upregulation of IL-6 may contribute to the creation of a tumor-promoting microenvironment. These results highlight the importance of IL-6 as a molecular link between obesity and colorectal carcinogenesis, emphasizing its potential value as a prognostic biomarker or therapeutic target. [41]

In our Kaplan–Meier survival analysis, a significant association was observed between IL-6 expression and overall survival in patients with colorectal cancer (CRC). Patients with high IL-6 expression exhibited shorter survival times compared to those with low expression, both at the protein and mRNA levels. Similarly, Olsen et al. (2015) reported that elevated IL-6 transcript levels in colon carcinoma tissues were significantly associated with poorer overall survival. [41] This finding is further supported by other studies demonstrating that IL-6 contributes to worse prognosis through activation of STAT3 via the gp130 subunit. STAT3 promotes cell cycle progression through the G1 and G2/M phases, thereby stimulating tumor cell proliferation and prolonging their survival. These mechanisms position IL-6 as a tumor-promoting cytokine. Conversely, the absence of IL-6 receptor expression, which prevents cytokine-receptor binding, has been associated with longer survival and more favorable outcomes. This relationship has been documented not only in CRC but also in other malignancies, including esophageal cancer, multiple myeloma, and Kaposi’s sarcoma. [11,14]

RBP4 is a retinol-binding protein secreted by adipocytes that has been shown to transport retinol to extrahepatic tissue.[15,16,17] In addition to adipose tissue and liver, it is also synthesized to a lesser extent by the lungs, brain and kidneys.[16] RBP4 activates STRA6, leading to activation of Janus kinase and STAT3/STAT5, which results in pro-inflammatory response[17,42] and tumorogenesis.[19]This results in cancer progression, migrations and metastasis.[17,18,19] In addition elevated RBP4 expression has been mentioned in ovarian cancer [17,43], breast cancer[42,44] and pancreatic cancer. [21,23] RBP4 levels have been shown to increase with increasing body weight [16] and greater distribution of central adipose tissue[25] what may make them a link between obesity and tumorigenesis.[16]

The study revealed that RBP4 expression levels were higher in tumor tissues compared to adjacent non-cancerous tissues. A strong cytoplasmic staining pattern was observed in tumor cells, as well as in the epithelial cells of normal colon tissue. Notably, during the evaluation of RBP4 in colorectal cancer patients, positively stained macrophages were also observed within the tumor microenvironment. This observation adds an additional layer to our understanding of RBP4 distribution and aligns with previous reports highlighting the role of RBP4 in various malignancies. [15,17,18,19,32,42,45] Interestingly, Zhang et al. also reported strong RBP4 expression in immune cells, particularly macrophages. [18] According to our immunohistochemical staining results, Zhao et al. (2025) analyzed TCGA/GTEx RNA-seq (COAD/READ) and found significantly higher RBP4 mRNA in colorectal tumors than in normal mucosa, using log2-transformed counts and standard statistical testing. They complemented this with RT-qPCR/IHC validations, confirming increased intratumoral expression. Together, these independent results strengthen the interpretation that elevated, tumor-intrinsic RBP4 expression contributes to CRC progression.[46]

Our study demonstrated that high RBP4 expression in colorectal cancer (CRC) tissues was significantly associated with pT status, lymph node metastasis, and advanced clinical stage according to the TNM classification. Additionally, these findings were corroborated by in silico analysis, which revealed a significant association between RBP4 expression and lymph node involvement, distant metastases, as well as TNM staging. The association between RBP4 and tumor size suggests that RBP4 may promote tumor growth, as also described in ovarian cancer, where RBP4 enhances migration and proliferation by activating RhoA/Rock1 and ERK pathways and upregulating MMP2 and MMP9.[17] The link between RBP4 and lymph node metastasis aligns with reports from hepatocellular carcinoma, where elevated RBP4 levels were associated with increased cell proliferation, invasiveness, and worse prognosis and TNM staging.[47] The observed relationship between RBP4 and TNM staging in CRC is consistent with studies indicating that higher RBP4 serum levels are linked to advanced disease and poor survival also in breast cancer, where RBP4 was shown to correlate with ER- and PR- negative status what results with more unfavorable prognosis [48] Binding of RBP4 to STRA6 activates JAK, which phosphorylates STATs. STATs activate cell cycle regulatory genes in the nucleus. This results in tumor progression and an increased risk of metastasis. [15,16,18] Inflammation and consequent endothelial dysfunction induced by stimulation of the expression of pro-inflammatory molecules, such as TGF-beta, by RBP4 may also be involved in metastasis formation.[49] Reduction in RBP4 levels results in inactivation of N-cadherin, MMP2,-3,-9 and an increase in E-cadherin, which also proves the roles of RBP4 in cancer cell migration. [20]

Furthermore, the significant correlation between RBP4 expression and obesity-related measures in our study supports the well-established role of RBP4 as an adipokine reflecting adipose tissue mass and visceral fat distribution [25]. Because RBP4 is secreted by adipocytes, more adipose tissue results in overexpression of RBP4, and this may affect tumor progression.[17] In addition, RBP4 can activate macrophages through the adipose tissue exosome,[31] which increases inflammation that enhances tumorigenesis. RBP4 has been implicated in insulin resistance and chronic low-grade inflammation, both of which are recognized contributors to cancer development and progression [16,17,19,23,24] RBP4 has been demonstrated to decrease phosphoinositide-3-kinase activity in muscle and increase phosphoenolpyruvate carboxykinase activity. It increases the expression of GLUT4, which inhibits glucose uptake by muscle and adipocytes. [19] Moreover, STAT's target gene is SOCS3, which inhibits insulin receptor signaling, which also results in insulin resistance. [16] However, the role of RBP4 in the development of insulin resistance and tumorigenesis is less significant in patients with lower BMI, which may be attributable to adipose tissue dysfunction in obesity. [19,44]

Our Kaplan–Meier survival analysis in the TCGA cohort revealed a trend suggesting an association between RBP4 expression levels and overall survival in patients with colorectal cancer (CRC). In our own patient cohort, individuals with elevated RBP4 expression demonstrated a shorter survival time compared to those with lower expression levels of this protein. It was reflected in the other reports demonstrating that enhanced RBP4 correlates with unfavorable prognosis in gastric carcinoma, [20] hepatocellular carcinoma,[47] glioblastoma.[50] Also in breast cancer showed higher RBP4 expression.[44] In ovarian cancer, RBP4 overexpression stimulates the expression of MMP-2 and MMP-9, which increases the migration of tumor cells and affects the unfavorable prognosis.[17]

In the subsequent findings of our study, we demonstrated that genes IL11, SOCS3 have been identified as top positively correlated with IL-6 in CRC. This correlation is primarily linked to the activation of the IL-6/STAT3 signaling pathway, which plays a crucial role in tumor progression and prognosis in CRC. [51,52] When IL-6 increases in CRC, we observe a fall in HNF1B and IL17RE expression. This is biologically plausible: HNF1B is frequently reduced (and sometimes promoter-methylated) in colorectal carcinomas and low levels associate with recurrence and shorter disease-free survival, indicating a tumour-suppressive role. [53] There’s little published specifically on IL17RE in CRC. By contrast, studies on IL17 itself report that it can promote colorectal tumorigenesis through several pathways and is associated with metastasis and poor prognosis, although some data also suggest a protective role [54] Reactome analysis shows that genes positively correlated with IL-6 cluster in immune and inflammatory pathways, with CXCL8 most frequently represented, underscoring an IL-6/IL-8–driven inflammatory microenvironment in CRC. In contrast, negatively correlated genes are enriched in non-immune signaling and transcriptional pathways, with ERBB2 appearing most often, suggesting attenuation of epithelial growth–related programs when IL-6–mediated inflammation is dominant. Rising IL-6 aligns with parallel upregulation of immune genes and reciprocal downregulation of alternative signaling axes, highlighting a functional dichotomy in CRC biology. [55]

Within the set of genes correlated with RBP4, COLEC11 (CL-11) ranked among the top positive hits; given that CL-11 drives cancer cell proliferation and tumor growth and its elevated expression associates with advanced stage and poorer survival in CRC, this positive correlation in our data points to a shared pro-tumorigenic program. [56,57] For CDIPT, there are essentially no published data in colorectal cancer; its positive correlation with RBP4 in our set therefore represents a novel observation that may reflect shared lipid/inositol metabolism but remains to be experimentally validated. Reactome analysis shows that RBP4, genes positively correlated with its expression cluster in signal transduction and developmental pathways, dominated by the GLI/Hedgehog transcriptional module. Genes negatively correlated with RBP4 fall into cell cycle and immune system pathways. This divergence suggests that RBP4-high tumors prioritize developmental signaling programs, while proliferative and immune processes are comparatively reduced. This inverse relationship suggests that reduced RBP4 expression may release constraints on cell cycle progression, potentially facilitating uncontrolled proliferation and driving CRC progression. Such a mechanism highlights the relevance of RBP4 as a possible modulator of cell division dynamics within the tumor context.

5. Conclusions

This study provides comprehensive insights into the role of IL-6 and RBP4 in colorectal cancer (CRC). Both proteins were found to be significantly overexpressed in tumor tissues compared to adjacent non-cancerous mucosa. Elevated IL-6 expression was associated with older age and obesity-related parameters, while RBP4 overexpression correlated with tumor size, lymph node metastasis, advanced TNM stage, and markers of obesity, such as increased BMI and waist circumference. Kaplan–Meier survival analysis demonstrated that high expression of IL-6 and RBP4 tended to associate with shorter overall survival, supporting their potential prognostic relevance. In silico analyses confirmed the transcriptional upregulation of IL6 and RBP4 in CRC and revealed distinct molecular pathways—immune/inflammatory signaling for IL-6 and developmental signaling for RBP4—highlighting their possible contributions to tumor biology.

Taken together, our findings suggest that IL-6 and RBP4 may function as biomarkers of poor prognosis and may represent potential therapeutic targets, particularly in CRC cases linked to inflammation and obesity. Further research is warranted to explore their mechanistic roles and validate their clinical utility in larger, independent cohorts.