Introduction

Colorectal cancer (CRC) ranks as the third most common malignancy worldwide in terms of incidence and second most common in terms of mortality [1]. Despite the availability of effective and resource-sensitive preventive and therapeutic interventions for CRC within the healthcare system [2], the incidence of new CRC cases and related mortality rates have been steadily rising [1].

Recent research has highlighted the frequent association of coagulation disorders with the onset, progression, or treatment of many malignancies, including CRC [3]. Virtually all types of cancer are linked to a hypercoagulable state [4], which manifests as coagulation activation in cancer patients—a subclinical condition referred to as cancer-associated hypercoagulability [5]. The interplay between cancer and hemostasis is bidirectional [6]: cancer cells can activate components of the hemostatic system through multiple mechanisms [7], thereby creating a hypercoagulable microenvironment through the expression of procoagulants, tissue factors, or inflammatory cytokines [8]. Conversely, elements of the hemostatic system can actively contribute to cancer pathogenesis [7], promoting malignant proliferation, invasion, and metastasis [3]. In investigations into the correlation between CRC and coagulation indicators, platelet count, D-dimer, and fibrinogen have been identified as factors associated with CRC progression and prognosis [9–16].

Fibrinogen, an acute-phase protein, exhibits varying degrees of abnormalities in individuals experiencing surgery, infection, trauma, tumors, and coagulation-related disorders [17–22]. Extensive research has focused on the role of this glycoprotein. Fibrinogen, a key inflammatory regulator, has been implicated in processes such as the formation of an inflammatory microenvironment, angiogenesis, malignant cell proliferation, invasion, and metastasis [23]. Studies have demonstrated that CRC growth and metastasis are dependent on the involvement of fibrinogen [24], and elevated preoperative fibrinogen levels have been linked to poor CRC prognosis [25–29]. Nevertheless, limited research has explored the relationship between fibrinogen and CRC in the general populace [14,16], especially among Chinese individuals.

Exploring the risk factors for CRC can contribute to the development of more detailed, noninvasive colorectal screening methods [30]. The objective of this research was to analyze the relationship between fibrinogen levels and CRC in the Chinese population.

Materials and methods

This retrospective study was approved by the Medical Ethics Committee of Shijiazhuang Traditional Chinese Medicine Hospital (NO.20220919029), and the requirements for informed consent were waived due to the retrospective nature.

The study involved the enrollment of 3595 consecutive inpatients from April 2015 to June 2022. All individuals underwent colonoscopy, and each patient was included only once to ensure exclusivity. Histopathological reports confirmed the presence of colorectal cancer (initial diagnosis). Subjects with normal colonoscopy results were considered as controls (for details, see Fig 1). The data for this study were accessed on September 22, 2022.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. The flow chart.

https://doi.org/10.1371/journal.pone.0322789.g001

Clinical (including comorbidities and laboratory data) and demographic information were derived from clinical records. Laboratory data were selected from the first qualifying report prior to colonoscopy during hospitalization. The definition of liver disease was based on a previous study [31].

The statistical analyses were performed using R 3.3.2 (http://www.R-project.org, The R Foundation) and Free Statistics software version 1.7.1 (http://www.clinicalscientists.cn/freestatistics, Beijing, China). Significance was determined at a P value < 0.05 (2-sided).

The study data utilized Mean ± standard deviation or median (Q1–Q3) values based on the normality of the distribution. Statistical tests such as the Mann‒Whitney U test or Student’s t test were used for continuous variables. In contrast, Fisher’s exact test or the chi-squared test were used for categorical variables.

To analyze the relationship between fibrinogen and CRC, we used multivariate binary logistic regression and restricted cubic spline analysis. Three models were constructed in multivariate binary logistic regression: (a) unadjusted; (b) adjusted for age and sex; (c) additionally adjusted for weight, family history of CRC, drinking and smoking status, total bilirubin (TBIL), aspartate aminotransferase (AST), alanine aminotransferase (ALT), albumin (ALB), gamma-glutamyl transferase (GGT), cholinesterase (ChE), alkaline phosphatase (ALP), total bile acid (TBA), glucose (GLU), creatinine (CREA), urea, low-density lipoprotein (LDL), high-density lipoprotein (HDL), total protein (TP), lipoprotein(a) [Lp(a)], platelets (PLT), white blood cell (WBC), neutrophil count, activated partial thromboplastin time (APTT), PT-INR, thrombin time (TT), hypertension, and liver disease. These potential confounders were chosen based on a change of ≥ 10% in effect estimates or significant covariates in Table 1 (P < 0.05). Subgroup analysis was conducted using a stratified binary logistic regression followed by a test for interaction. Missing data were handled using multiple imputations by chained equations with five replications. In addition, a sensitivity analysis using complete cases was performed to assess the robustness of the results.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Participants characteristics.

https://doi.org/10.1371/journal.pone.0322789.t001

Results

Fibrinogen and colorectal cancer

Table 2 presents a significant relationship between fibrinogen and CRC in the multivariate logistic regression analyses. In the crude model, fibrinogen was strongly associated with CRC (odds ratio [OR] = 3.04 [2.66–3.46]); adjusted for age and sex, the OR was 2.48 (2.15–2.86); and adjusted for the full model, the OR was 1.83 (1.52–2.20). After transforming fibrinogen into a categorical variable, fibrinogen was positively associated with CRC in different models, adjusting for other potential confounding factors ( P < 0.001).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Multivariable logistic regression analyses of fibrinogen and CRC.

https://doi.org/10.1371/journal.pone.0322789.t002

The restricted cubic spline analysis (Fig 2, 99.5% data) revealed a non-linear relationship between fibrinogen levels and CRC risk, identifying a threshold at 3.794 (95% CI: 3.747–3.841) g/L. Specifically, below this threshold, the CRC risk significantly increases (OR 5.944, 95% CI 4.084–8.65, P < 0.001), whereas, above this threshold, the association is no longer statistically significant (Table 3, OR = 0.838, 95% CI: 0.431–1.632, P = 0.604).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. Association between fibrinogen and colorectal cancer.

Adjusted for age, sex, weight, family history of colorectal cancer, drinking status, smoking status, albumin, gamma-glutamyl transferase, alanine aminotransferase, alkaline phosphatase, aspartate aminotransferase, cholinesterase, total bilirubin, total bile acid, glucose, creatinine, urea, total protein, low-density lipoprotein, high-density lipoprotein, lipoprotein(a), platelets, white blood cell, neutrophil count, the international normalized ratio of prothrombin time, activated partial thromboplastin time, thrombin time, hypertension, and liver disease. Only 99.5% of the data is shown.

https://doi.org/10.1371/journal.pone.0322789.g002

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Threshold effect analysis of the relationship of fibrinogen and CRC.

https://doi.org/10.1371/journal.pone.0322789.t003

Sensitivity analysis

To explore the stability of the association between fibrinogen and CRC, we performed subgroup analyses. The relationship between fibrinogen and CRC was not significantly affected by subgroups such as sex, age, HLP, hypertension, ischemic cerebrovascular disease, and CHD (P for interaction >0.05, Fig 3).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Association between fibrinogen and colorectal cancer.

Except for the subgroup component itself, each subgroup factor was adjusted for age, sex, weight, family history of colorectal cancer, drinking status, smoking status, albumin, gamma-glutamyl transferase, alanine aminotransferase, alkaline phosphatase, aspartate aminotransferase, cholinesterase, total bilirubin, total bile acid, glucose, creatinine, urea, total protein, low-density lipoprotein, high-density lipoprotein, lipoprotein(a), platelets, white blood cell, neutrophil count, the international normalized ratio of prothrombin time, activated partial thromboplastin time, thrombin time, hypertension, and liver disease.

https://doi.org/10.1371/journal.pone.0322789.g003

In the sensitivity analysis, participants with missing data were not included, leaving 1695 individuals. S1 Table showed that the relationship between fibrinogen and CRC was stable. In the crude model, fibrinogen was associated with CRC (OR = 3.06, 95% CI: 2.55–3.68); in Model I (adjustment for age and sex), the OR was 2.54 (2.08–3.10); and in Model II (adjustment for full model), the OR was 1.67 (1.27–2.21). The significant relationship also existed in different adjustment models when fibrinogen was transformed into quartiles (P for trend <0.001).

Discussion

In this case-control study, we found that fibrinogen levels are positively correlated with CRC risk. Subgroup analyses further confirmed the stability of this association. Notably, we identified an inverse L-shaped relationship between fibrinogen levels and CRC risk, with a significant threshold at 3.794 g/L.

Our findings are in line with previous investigations exploring the relationship between fibrinogen and CRC in the general population. These studies have consistently suggested that elevated levels of fibrinogen are independently and positively associated with CRC risk [14,16]. For example, a prospective study involving 84,000 individuals from the general population in Denmark found that high levels of fibrinogen were associated with an increased risk of CRC. Similarly, research from the Moli-Sani cohort indicated that individuals. These studies suggest that fibrinogen may be an important biomarker for assessing CRC risk, potentially offering new approaches and methods for the early diagnosis and treatment of CRC.

Beyond the hemostatic factors, we meticulously explored the potential confounders, including biomarkers associated with CRC or other cancers identified in prior studies (e.g., UA, AST, ALT, GLU, HDL, and PLT) [32–36]. Additionally, we considered other potential confounders such as biochemical indicators, comorbidities, and family history. Remarkably, these factors had only minimal impact on the relationship between fibrinogen and CRC, underscoring the robustness of our findings.

Although this retrospective observational study cannot confirm a causal relationship between fibrinogen and CRC, evidence from both basic and clinical research indicates an interaction between cancer and the hemostatic system. Cancer may promote thromboembolism, while pathological activation of the hemostatic system can drive cancer progression. The thrombin-fibrinogen axis plays a crucial role in CRC [37]. Fibrinogen interacts with leukocyte integrin αMβ2, promoting inflammatory responses that lead to increased stress and phosphorylation of proliferation markers in intestinal epithelial cells, indicating a direct link between fibrinogen/inflammatory function and tumorigenic changes in colon epithelial cells [38]. Fibrinogen promotes angiogenesis by stimulating endothelial cell migration, proliferation, and tubular structure formation, and it binds growth factors such as bFGF and VEGF, protecting them from degradation and enhancing endothelial cell activation, thereby reinforcing angiogenesis [39]. Additionally, the fibrin matrix plays a key role in platelet stabilization and activation, and platelets promote the expansion and function of immunosuppressive myeloid cells during the early stages of inflammation-induced adenoma formation, and their infiltration into colon adenomas, further driving tumor cell metastasis [40]. These mechanisms collectively promote tumor growth, angiogenesis, and metastasis in CRC patients through the interplay of the inflammatory and coagulation axes. Conversely, tumors can induce hypoxia and hypercoagulable states in vivo [41,42], leading to elevated fibrinogen levels in CRC patients, which are a result of tissue damage and subsequent inflammatory responses [43].

We discovered a non-linear, inverse L-shaped relationship between fibrinogen levels and CRC risk, identifying a critical threshold at 3.794 g/L. This finding provides strong evidence supporting the use of fibrinogen levels in clinical interventions to reduce CRC risk. Notably, when fibrinogen levels are controlled below 3.794 g/L, CRC risk decreases significantly as fibrinogen levels decline, making this threshold crucial for managing fibrinogen levels. Additionally, the diagnostic utility of fibrinogen may be enhanced when combined with other tumor markers, such as carcinoembryonic antigen and prealbumin, potentially improving diagnostic performance in CRC [44]. Future clinical trials and studies are needed to further explore the role of fibrinogen in the adenoma-carcinoma sequence and validate its effectiveness as a diagnostic biomarker. Upon such validation, monitoring fibrinogen levels could become a valuable risk assessment tool prior to colonoscopy in CRC screening, owing to its efficiency, speed, and non-invasive nature. When elevated fibrinogen levels are detected, they should be assessed in conjunction with other risk factors to evaluate CRC risk, and colonoscopy should be employed for screening high-risk individuals.

Several limitations should be acknowledged in our study. Firstly, missing data is a common issue in retrospective studies. However, we addressed this by employing multiple imputations, and our results remained consistent. Secondly, we lacked long-term data on fibrinogen levels, which could offer valuable insights into its impact on the progression of CRC. Thirdly, despite controlling for various factors, unmeasured lifestyle factors (e.g., diet, physical activity), infections, and socioeconomic variables could still influence fibrinogen levels and CRC risk. Furthermore, our study design, being retrospective, primarily reveals correlation rather than causation. Finally, our study primarily involved Chinese participants, limiting the generalizability of our results to other demographic groups. For greater reliability and wider applicability, it is crucial that future research validates these results with independent cohorts.

Conclusion

Fibrinogen levels are positively correlated with CRC risk, and the relationship between them exhibits an inverse L-shape, with a critical turning point at 3.794 g/L. Future clinical trials and studies are necessary to further validate the role of fibrinogen in CRC clinical management.

Supporting information

Acknowledgments

We thank Free Statistics team for providing technical assistance and valuable tools for data analysis and visualization.