Abstract
Background/Aim: A three-dimensional network constructed using glycocalyx (GCX) extends throughout the cancer cell nest in human colorectal cancer (CRC). GCX was found to be closely related to cancer. We examined the prognostic correlation and potential of syndecan-1 (SDC1), a representative proteoglycan of GCX, as a biomarker. Patients and Methods: We analyzed SDC1 in the transcriptomic profiles of a major publicly available CRC cohort from The Cancer Genome Atlas (TCGA) using a computational algorithm. We investigated serum SDC1 levels preoperatively and on postoperative day seven in 48 patients with stage I-III CRC who underwent surgery during July-December 2019 at Gifu University Hospital. Results: For TCGA, no significant differences existed between the high and low SDC1 expression groups regarding disease-free, disease-specific, and overall survival for stage I-III, and only overall survival for stage IV was significantly different. In our study, among the 48 patients, 17 (no recurrence), 13 (1 recurrence), and 18 (10 recurrences) had stage I-III, respectively. Preoperative and postoperative day 7 SDC1 levels for patients with stage I-III were 10.7±2.3 and 9.9±3.1 ng/ml (p=0.40), 11.1±1.7 and 10.1±0.8 ng/ml (p=0.07), and 10.3±2.0 and 9.5±1.4 ng/ml (p=0.15), respectively. In stage II and III, patients were divided into two groups according to differences between preoperative and postoperative SDC1 levels (SDC1pre-pro). SDC1pre-pro ≤0 group significantly prolonged disease-free survival compared with SDC1pre-pro >0 group (p=0.048). Conclusion: Dynamic change in serum SDC1 levels serves as a prognostic biomarker for stage II and III colorectal cancer.
Cancer cells have higher levels of glycoproteins, including mucin-1, than healthy cells, and individual glycoproteins can transduce environmental signals that directly promote malignant transformation (1, 2). However, glycoproteins can also form glycocalyces in groups. Physical properties of the glycocalyx (GCX) can affect intracellular signaling by regulating the accumulation of cell surface receptors, thereby contributing to cancer progression and metastasis (3). Syndecans are involved in regulating cell-cell and cell-extracellular matrix adhesion and cell migration. In normal tissues, they mediate these processes by binding heparan sulfate chains to extracellular matrix molecules and effectors, including growth factors, cytokines, proteinases, and proteinase inhibitors (4, 5).
Syndecan-1 (SDC1) influences tumorigenesis by interacting with molecular mediators of tumor cell survival, proliferation, angiogenesis, and metastasis (6). SDC1 expression is reduced in colorectal adenocarcinomas compared with that in adenomas and normal tissues, and reduced expression is correlated with local metastasis (7). Elevated SDC1 levels are also present in the local stroma (8). However, the prognostic relevance of changes in SDC1 expression in colorectal cancer (CRC) remains unclear.
Cardiac oncology has attracted attention in recent years owing to the super-aging society and westernization of life in Japan. Risk factors for cancer mortality, heart failure, and atherosclerosis are considered similar and include increased body mass index, smoking, diabetes, and hypertension (9, 10). Endothelial damage is closely associated with several diseases via atherosclerosis (11). Endothelial glycoconjugates cover the inner surface of the vascular endothelium and regulate leukocyte adhesion (12). Therefore, leukocytes cannot adhere to glycan-covered endothelial cells, and damage to endothelial glycoconjugates may occur before atherosclerotic changes (13). We previously reported the microenvironment of the glycoprotein GCX based on scanning electron microscopy and transmission electron microscopy. GCX acts as a protective barrier against bacteria, regulates local enzyme levels, and inhibits thrombus formation, especially in sepsis and heart failure (14, 15).
General classes of glycans present in GCX include glycoproteins and proteoglycans (16). Glycoproteins are glycan complexes, in which one or more sugar chains are covalently attached to a polypeptide backbone. Cells are often covered with a surface layer of multifunctional glycans called GCX (17). The chemical composition, mechanical properties, and multiple physiological functions of this layer in vascular endothelial cells have been extensively described (18).
Cancer cells also possess glycans in the form of GCXs and produce abundant GCXs themselves. GCX can alter cancer cell phenotypes to become invasive due to physical stress in their surroundings (3). However, the role of GCX in human cancer tissues and gastrointestinal cancers remains unknown, making it an extremely important glycoprotein for studying the development of cancer neovasculature, metastatic mechanisms, and transport functions of drugs from the perspective of intrinsic cancer morphology.
This study analyzed the prognostic relevance of SDC1 in tissue samples of CRC with clinical stage I-IV from The Cancer Genome Atlas (TCGA) data, suggesting that SDC1 is a valid indicator of unresectable CRC, although no significant difference was found in stage I-III. We measured serum SDC1 levels before and seven days after surgery in patients with stage I-III CRC and examined the differences to determine whether it could be used as a biomarker or prognostic indicator.
Patients and Methods
The study included 48 patients who underwent colorectal resection for CRC at Gifu University Hospital between July and December 2019. The inclusion criteria were as follows: no previous history of cancer-related disease, no diagnosis of synchronous double cancer, no radiotherapy or chemotherapy before surgery, pathological diagnosis of CRC following surgery, no clinical diagnosis of familial adenomatous polyposis or hereditary nonpolyposis, and no clinical diagnosis of inflammatory bowel disease.
The study was conducted according to the World Medical Association Declaration of Helsinki and approved by the Ethics Committee of Gifu University (approval numbers:2019-074, 2021-C162). Because this was a prospective study that included potentially identifiable patient data, informed consent was obtained from all enrolled patients.
Laboratory data. The data were obtained by collecting blood samples at the time of enrollment with >12 h of fasting. Serum SDC1 levels were measured using an enzyme-linked immunosorbent assay (Diaclone, Besancon, Cedex, France).
Data acquisition from CRC cohorts. Expression data and clinical information of 592 CRCs (TCGA COADREAD) were obtained from UCSC University of California Santa Cruz Xena (19).
Statistical analysis. Statistical analyses were performed using the publicly available software R (version 4.02). For continuous variables, data are presented as median with interquartile range and compared using the non-parametric Mann–Whitney U-test. Fisher’s exact test was used to assess differences between groups. For the survival analysis, we estimated the disease-free survival (DFS), disease-specific survival (DSS), and overall survival (OS) by using Kaplan–Meier analysis and determined intergroup survival differences through the log-rank test. We used the Cox proportional hazards model to identify independent DFS risk factor. Statistical significance was set at p<0.05.
Results
Clinical and gene expression data of 589 patients were obtained from the TCGA and used to investigate whether intratumoral SDC1 was associated with survival. Patient characteristics are presented in Table I. The 589 patients were divided into the SDC1 High (n=295) and SDC1 Low (n=294) groups according to SDC1 expression levels in the TCGA cohort. The final clinical stage of the patients according to the Union for International Cancer Control classification in the SDC1 High group was stage I, II, III, and IV in 57 (19.3%), 101 (34.2%), 88 (29.8%), and 42 (14.2%) patients, respectively, with 7 (2.4%) patients of unknown stage, and that in the SDC1 Low group was stage I, II, III, and IV in 46 (15.6%), 117 (40.0%), 81 (27.6%), and 41 (13.9%) patients, respectively, with 9 (3.1%) patients of unknown stage (p=0.453).
Clinicopathological demographics of the SDC1 High and SDC1 Low groups in The Cancer Genome Atlas (TCGA).
The T-factor classification was T1, T2, T3, and T4 in 10 (3.4%), 57 (19.3%), 193 (65.4%), and 35 (11.9%) patients, respectively, in the SDC1 High group, and T1, T2, T3, and T4 in 10 (3.4%), 45 (15.3%), 206 (70.1%), and 32 (10.9%) patients, respectively, in the SDC1 Low group (p=0.583). The N-factor classification was N0, N1, and N2 in 168 (56.9%), 73 (24.5%), and 54 (18.3%) patients, respectively, in the SDC1 High group, and N0, N1, N2, and Nx in 171 (58.2%), 68 (23.1%), 54 (18.4%), and 1 (0.3%) patient, respectively, in the SDC1 Low group (p=0.916). The M-factor classification was M0, M1, and Mx in 215 (72.9%), 41 (13.9%), and 39 (13.2%) patients, respectively, in the SDC1 High group and M0, M1, and Mx in 223 (75.9%), 41 (13.9%), and 30 (10.2%) patients, respectively, in the SDC1 Low group (p=0.905).
The histological types were colon adenocarcinoma in 199 (67.5%), colon mucinous adenocarcinoma in 28 (9.5%), rectal adenocarcinoma in 58 (19.7%), rectal mucinous adenocarcinoma in 3 (1.0%), and unknown in 7 (2.4%) patients in the SDC1 High group, and colon adenocarcinoma in 172 (58.5%), colon mucinous adenocarcinoma in 33 (11.2%), rectal adenocarcinoma in 76 (25.9%), rectal mucinous adenocarcinoma in 10 (3.4%), and unknown in 3 (1.0%) patients in the SDC1 Low group (p=0.036).
Lymphatic invasion was positive in 108 (36.6%), negative in 151 (51.2%), and unknown in 36 (12.2%) patients in the SDC1 High group, and positive in 108 (36.7%), negative in 163 (55.4%), and unknown in 23 (7.8%) patients in the SDC1 Low group (p=0.724). Venous invasion was positive in 59 (20.0%), negative in 192 (65.1%), and unknown in 44 (14.9%) patients in the SDC1 High group, whereas it was positive in 65 (22.1%), negative in 195 (66.3%), and unknown in 34 (11.6%) patients in the SDC1 Low group (p=0.757). Perineural invasion was positive in 27 (9.2%), negative in 80 (27.1%), and unknown in 188 (63.7%) patients in the SDC1 High group and positive in 31 (10.5%), negative in 85 (28.9%), and unknown in 178 (60.5%) patients in the SDC1 Low group (p=0.879).
Microsatellite instability was MSI-High in 24 (8.1%), MSI-Low in 54 (18.3%), microsatellite stable (MSS) in 200 (67.8%), and unknown in 8 (2.7%) patients in the SDC1 High group and MSI-High in 55 (18.7%), MSI-Low in 40 (13.6%), MSS in 192 (65.3%), and unknown in 7 (2.4%) patients in the SDC1 Low group (p<0.001). The Kirsten rat sarcoma viral oncogene homolog (KRAS) status was wild type in 145 (49.2%), mutant type in 115 (39.0%), and unknown in 35 (11.7%) patients in the SDC1 High group, and wild type in 167 (56.8%), mutant type in 102 (34.7%), and unknown in 25 (8.5%) patients in the SDC1 Low group (p=0.157).
TCGA data showed no significant differences between the SDC1 High and SDC1 Low groups in terms of DFS, DSS, and OS for stage I (DFS: p=0.34, DSS: p=0.95, OS: p=0.74), II (DFS: p=0.19, DSS: p=0.36, OS: p=0.31), and III (DFS: p=0.65; DSS: p=0.35, OS: p=0.99). However, in stage IV, TCGA data showed a significant difference between the two groups only for OS (p=0.04) (Figure 1 and Figure 2).
Disease-free survival (DFS), disease-specific survival (DSS), and overall survival (OS) in relation to SDC1 expression in stage I-II colorectal cancer and each group in TCGA.
Disease-free survival (DFS), disease-specific survival (DSS), and overall survival (OS) in relation to SDC1 expression in stage III-IV colorectal cancer and each group in TCGA.
Patient characteristics are shown in Table II. Among the 48 patients, 17, 13, and 18 had stage I, II, and III CRC, respectively in our study. The mean age of the patients was 66.9 (42-86) years, the male-to-female ratio was 26:22, and the body mass index was 22.7 (15.9-40.3) kg/m2. The preoperative CEA level was 4.0 (0.4-18.8) ng/ml, and the CA19-9 level was 24.1 (0.4-206.9) U/ml. The number of patients who relapsed was 0, 1, and 10 for stage I, II, and III, respectively. The number of patients who died was 1, 1, and 4 for stage I, II, and III, respectively.
Patient characteristics.
Preoperative and postoperative day seven SDC1 levels for patients with stage I-III were 10.7±2.3 and 9.9±3.1 ng/ml (p=0.40), 11.1±1.7 and 10.1±0.8 ng/ml (p=0.07), and 10.3±2.0 and 9.5±1.4 ng/ml (p=0.15), respectively (Figure 3, Figure 4, Figure 5).
Respective preoperative and postoperative day seven values of SDC1 in patients with stage I colorectal cancer.
Respective preoperative and postoperative day seven values of SDC1 in patients with stage II colorectal cancer.
Respective preoperative and postoperative day seven values of SDC1 in patients with stage III colorectal cancer.
In patients with resectable advanced CRC (stage II and III), when divided into two groups according to the difference between pre-and postoperative SDC1 (SDC1pre-pro), SDC1pre-pro ≤0 group significantly prolonged DFS more than SDC1pre-pro >0 group (p=0.048) (Figure 6). In stage III advanced CRC, SDC1pre-pro ≤0 group tended to prolong DFS more than SDC1pre-pro >0 group (p=0.076) (Figure 7).
Disease-free survival by differences in SDC1 preoperatively and postoperatively in patients with stage II-III colorectal cancer.
Disease-free survival by differences in SDC1 preoperatively and postoperatively in patients with stage III colorectal cancer.
Discussion
The concept of “vascular mimicry” in cancer angiogenesis has been reported (20), but has not been validated in human samples. In human CRC tissues (21, 22), we reported that a 3-dimensional network constructed by GCX extends throughout the cancer cell nest. This structure appears to be the ultimate “cancer infrastructure”, enabling nutrients to reach cancer cells from distant blood vessels (23).
Higher glycoprotein levels are present in cancer cells than in healthy cells. Individual glycoproteins can convert environmental signals that directly promote malignant transformation (24) and collectively organize glycans. Therefore, this study was focused on SDC1, a representative proteoglycan that makes up GCX. SDC1 is a glycan component and its degradation suggests endothelial injury (25).
A relationship exists between SDC1 and serious diseases, including acute and chronic renal failure, cardiac arrest, cardiovascular disease, and sepsis (26, 27). SDC1 is a member of the four-member family of cell surface heparan sulfate proteoglycans. SDC1 plays an important role in regulating epithelial homeostasis, proliferation, and migration (4, 28) and suppresses the transformation and migration of several cancer cell lines. SDC1 expression is reduced in various cancer cell lines and tissues. Several mechanisms may explain the relationship between the loss of epithelial SDC1 expression and cancer progression (29).
Patients with CRC have higher serum soluble SDC1 levels than healthy adults, which correlates with decreased survival (30). Patients with high serum SDC1 levels are also less responsive to 5-fluorouracil, oxaliplatin, irinotecan, cisplatin, or paclitaxel chemotherapy (29). Furthermore, SDC1 is expressed at the basolateral border of normal colonic epithelial cells. However, in adenocarcinoma cells, SDC1 is present around the epithelial cell membrane and cytoplasm (30). In approximately 90% of adenocarcinomas studied, SDC1 expression was absent and correlated with lymph node metastasis. Stromal SDC1 is expressed in only a small percentage of these tumors. These findings highlight the possibility that loss of tumor SDC1 is a prognostic biomarker for human colon adenocarcinoma (7). In another study, epithelial SDC1 expression was found in >90% of colon cancer specimens and was associated with lower histological grade and less advanced clinical stage (8). Interstitial SDC1 expression was observed in 58% of specimens, but its expression did not significantly correlate with clinical outcomes (8).
Using TCGA data, we analyzed SDC1 levels in tissues of patients with stage I-IV CRC and found no prognostic significance of SDC1 in tissues in stage I-III. However, in stage IV disease, higher SDC1 levels were clearly associated with poorer prognosis in terms of OS. In the present study, serum SDC1 was measured pre- and postoperatively in stage I-III, and postoperative SDC1 levels tended to be lower than preoperative SDC1 levels. On the other hand, the prognostic value of serum SDC1 showed that DFS was prolonged in patients with resectable advanced CRC with increased postoperative SDC1 levels compared with preoperative SDC1 levels. The reason for this is that high SDC1 levels in the early postoperative period are an indicator of early recovery from surgical invasion and may be involved in homeostasis. The results also suggest that SDC1 is not directly involved in cancer cells, but is related to the environment surrounding cancer cells. Dynamic changes in SDC1 levels may be prognostic indicators of recurrence in patients with resectable advanced CRC.
Our study had a few limitations. First, this was a single-center study with a small sample size. Second, we could not identify any changes in SDC1 levels after postoperative day seven. Therefore, a multicenter study with a large sample size over a longer period is needed.
Conclusion
This study found that prolonged DFS was observed in patients with resectable advanced CRC when postoperative SDC1 levels were elevated compared with preoperative SDC1 levels. We suggest that dynamic changes in serum SDC1 levels serve as prognostic biomarkers for resectable advanced CRC.
Footnotes
Authors’ Contributions
Conceptualization, HT; methodology, HT, HH, HT, YT, MF; formal analysis, HT, YT, MF; resources, JYT, RY, HT, MK, YS, IY, RA, RM, YT, NO, MF, YT, NO, MF; writing – original draft preparation, HT; writing – review and editing, NM; supervision, NM; project administration, NM. HT and NM confirm the authenticity of all the raw data. All Authors read and approved the final manuscript.
Conflicts of Interest
The Authors declare that they have no competing interests in relation to this study.
- Received October 1, 2023.
- Revision received October 29, 2023.
- Accepted October 31, 2023.
- Copyright © 2024 International Institute of Anticancer Research (Dr. George J. Delinasios), All rights reserved.
This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY-NC-ND) 4.0 international license (https://creativecommons.org/licenses/by-nc-nd/4.0).