Study on the predictive effect of fibrinogen on vascular calcification

Abstract

Background

Fibrinogen, lipoprotein, and high-density lipoprotein levels were associated with vascular calcification, but their predictive capacity for a vascular calcification was not reported.

Aims

The purpose of this study was to evaluate the predictive efficacy of fibrinogen, lipoprotein, and high-density lipoprotein by retrospective analysis of fibrinogen, lipoprotein, and high-density lipoprotein levels in patients with vascular calcification, to explore the effective predictive indexes of vascular calcification, to predict the occurrence and development of vascular calcification, and to provide a simple and effective method for the diagnosis and prevention of vascular calcification.

Hypothesis: Fibrinogen is a good prediction of vascular calcification.

Methods

Univariate and multivariate analyses were used to assess the effects of fibrinogen, lipoprotein, and high-density lipoprotein on the CV, and the ROC curve of the predictive model was used to assess its predictive effectiveness. We collected the relevant indicators of 462 patients admitted to the Department of Vascular Surgery of the First Hospital of Hebei Medical University from August 2018 to July 2020, including 189 patients with vascular calcification (40.9%) and 273 patients without vascular calcification (59.1%); 75% of the collected data is used for modeling (modeling group) and 25% for verification (verification group).

Results

Results from the multivariate analysis showed fibrinogen, lipoprotein, and high-density lipoprotein to be independent predictors of vascular calcification. Next, the three-factor models are developed respectively. The area below the ROC curve in the fibrinogen, lipoprotein, and high-density lipoprotein forecast model was 0.8018, 0.7348, and 0.7019, respectively.

Conclusions

Fibrinogen is more predictive than high-density lipoprotein and lipoprotein in patients with arteriosclerosis.

Keywords

Fibrinogen vascular calcification high-density lipoprotein lipoprotein predict

Introduction

Vascular calcification (VC) is a common pathological feature of atherosclerosis, hypertension, diabetic vascular lesions, vascular damage, chronic kidney disease, and aging.^1,2 VC is characterized by increased vascular wall stiffness and decreased compliance, which can lead to serious clinical adverse reactions, including systolic hypertension, left ventricular hypertrophy, coronary artery ischemia, congestive heart failure, and possible plaque rupture, thrombosis, and myocardial infarction.^3,4 Once considered to be a passive process, it is now recognized that VC is a complex and highly regulated process involving a series of mediating factors, including activation of cellular signaling pathways, circulatory calcification inhibitors, genetic factors, and hormones. Different phenotypes may have different effects on plaque vulnerability and clinical outcomes.^5,6 For example, histological studies have shown that coronary artery calcification is mainly limited to the intima; however, intima and media calcification can occur in large arteries, including the aorta; media calcification is more closely related to aging, diabetes, and severe nephropathy.⁷ The structural characteristics of different vascular beds may also play a role. Because of its complex and multifaceted characteristics, there is no targeted treatment for VC at present.

The pathogenesis of intravascular calcification is not fully understood, but recent studies have shown that it is similar to neointimal calcification, which is considered to be a repeat of bone formation and ultimately depends on the nucleation and crystal growth of hydroxyapatite.^5,8 Related studies have shown that inflammation, vesicle secretion, oxidative stress in apoptotic bodies and plaques, as well as increased levels of VSMC and cholesterol, radiotoxicity, and adipogenesis contribute to the progression of calcification.⁶

VC is usually the result of errors in adaptive mechanisms. Inflammatory reaction and oxidative stress are important pathological processes of VC.⁹ In general, chronic inflammation seems to be the central factor for abnormal soft tissue calcification, and the site of chronic inflammation in the vascular system has been proved to be the site of atherosclerotic calcification in mice.⁴ There is strong evidence that inflammation plays an important, at least partially reversible, role in the development of VC, and inflammatory markers may be useful additional tools for assessing cardiovascular risk in clinical practice. The combined evaluation of VC and inflammatory markers can improve the non-invasive assessment of cardiovascular risk, so those high-risk patients can be selected for preventive treatment or more regular physical examination, and the possibility of de-targeting therapy for pro-inflammatory mechanism can be developed in the future.¹⁰ As an important biomarker of systemic inflammation, FIB is a glycoprotein synthesized and secreted by hepatocytes. FIB is the main coagulation protein in plasma, the determinant of blood viscosity, and the cofactor of platelet aggregation. Current studies have shown that serum inflammatory markers are significantly correlated with VC,¹¹ but the predictive efficiency of FIB in VC has not been reported.

According to related studies, lipoprotein can regulate the initiation of VC, which is an important determinant in the progression of VC.^12–16 Low-density lipoprotein is more likely to cause VC, especially, the elevated level of oxidized low-density lipoprotein is a risk factor for cardiovascular disease.^17–19 But high-density lipoprotein (HDL) has the inhibitory ability to VC.^6,20 As a controversial lipoprotein, the inhibitory effect of HDL on VC is mainly mediated by its role in cholesterol reverse transport,²¹ a biological process that promotes the transfer of free cholesterol from the arterial wall back to the liver for reuse or excretion into bile. HDL can inhibit the oxidation of low-density lipoprotein and its anti-inflammatory effect on cellular signals caused by oxidized low-density lipoprotein. HDL also inhibits macrophage toll-like receptor 4-mediated inflammation, which is dependent on cholesterol efflux through ATP binding cassette A1 and transporter G1. Besides, HDL has an anti-diabetic effect.²² Although previous studies have shown that there is a correlation between lipoprotein and HDL and VC, its ability to predict VC has not been reported.

There are no clear predictors of vascular calcification; the purpose of this study was to evaluate the predictive efficacy of FIB, lipoprotein, and HDL by retrospective analysis of FIB, lipoprotein, and HDL levels in patients with VC to explore the effective predictive indexes of vascular calcification, to predict the occurrence and development of vascular calcification, and to provide a simple and effective method for the diagnosis and prevention of vascular calcification.

Patients

This study was approved by the Ethics Committee of the First Hospital of Hebei Medical University (ethics code is 20200617). We collected the relevant indicators of 462 patients admitted to the Department of Vascular Surgery of the First Hospital of Hebei Medical University from August 2018 to July 2020, including 189 patients with vascular calcification (40.9%) and 273 patients without vascular calcification (59.1%). Coronary artery calcification score is a quantitative assessment of overall coronary artery calcification using CT. First, the CT score was assigned according to the changes of the condition, 130–199 Hu was 1 point, 200–299 Hu was 2 points, 300–399 Hu was 3 points, and 400 Hu and above were 4 points. Then, the calcification area (measured in square mm) was multiplied. Finally, the total calcification score was obtained by adding the scores of each coronary artery in all CT sections. A score of 0–10 makes the probability of significant coronary artery disease very unlikely; 11–100 indicates that mild disease is likely; 101–400 makes non-obstructive disease highly likely, and over 400 indicates a high likelihood of at least one significantly stenosed lesion.^23–25 The abdominal aortic calcification score was described by Pepe et al.²⁶ Lateral lumbar spine radiographs were obtained in the standing position to measure the severity of calcification in the aorta at the level of the first four lumbar vertebrae. The following scores were assigned for the presence of calcifications in the longitudinal aortic wall opposite each vertebra: 1, small calcified deposits occupying less than 1/3 of the aortic wall; 2, one-third or more but less than two-thirds of the wall of the aorta; and 3, calcification of two-thirds or more of the wall of the aorta. The anterior and posterior aortic walls were evaluated separately and the total was obtained to get a score out of 24.²⁷ In this study, vascular calcification was assessed using the abdominal aortic calcification score. The abdominal aortic calcification score of 0 is the patient without vascular calcification, and the abdominal aortic calcification score > 0 is the patient of vascular calcification.

Methods

Data collection

We collected indicators of age, height, weight, fibrinogen, fibrinogen degradation products, triglycerides, cholesterol, HDL, low-density lipoprotein, lipoprotein, glutamyl transferase, and C-reactive protein from the database, and collected related indicators of vascular-free calcification for 273 patients for comparison.

Statistical analysis

All statistical analyses and random allocation were performed by Empower Stats and R project version 3.3.3 (http://www.rproject.org/). The results were randomly selected by the system, and 75% of the data collected was used for modeling (modeling group) and 25% for validation (verification group). QQ plot was used to check the normal distribution of data. The measurement data that obey normal distribution are expressed by mean ± standard deviation, and the comparison between groups is expressed by independent sample t-test, count data use case (%), Chi-square test, or Fisher exact probability method. Univariate analysis and multivariate regression analysis were used to analyze the differences between groups by Kaplan–Meier method and Log-Rank method. The test level was α = 0.05 (p <0.05). The VC prediction ROC curve based on FIB, lipoprotein, and low-density lipoprotein was established by Empower Stats.

Results

The population analysis is shown in Table 1. We collected 462 patients, including 189 patients with vascular calcification (40.9%) and 273 patients without vascular calcification (59.1%). In this study, there were 325 males (70.3%) and 137 females (29.7%). In this data, the incidence of coronary atherosclerosis was 7.8%, peripheral atherosclerosis 40.9%, and hypertension 41.3%. In cases of vascular calcification, the incidence of hypertension was 66.7%.

Table 1.

Population analysis.

Variables	Modeling group	Verification group	p value
Normal	202	71
VC	139	50
Age	60.842 ± 12.504	61.777 ± 11.840	0.474
High	168.748 ± 27.059	167.851 ± 10.772	0.723
Kg	71.891 ± 11.373	71.296 ± 12.751	0.635
Neutrophiles	0.637 ± 0.102	0.630 ± 0.105	0.562
Lymphocyte	0.269 ± 0.093	0.276 ± 0.096	0.447
FIB	3.272 ± 1.085	3.160 ± 0.898	0.313
FDP	3.571 ± 26.759	1.407 ± 3.854	0.376
Triglycerides	1.500 ± 1.145	1.586 ± 1.427	0.519
Cholesterol	4.495 ± 0.996	4.452 ± 1.053	0.694
HDL	1.050 ± 0.322	1.075 ± 0.253	0.457
LDL	2.801 ± 0.705	2.758 ± 0.738	0.584
Lipoproteins	286.245 ± 283.271	275.690 ± 293.930	0.736
Glutamyl transferase	29.080 ± 32.553	32.244 ± 29.112	0.350
CRP	12.962 ± 37.953	11.323 ± 36.595	0.694

FIB: fibrinogen; FDP: fibrin degradation products; HDL: high-density lipoprotein； LDL: low-density lipoprotein；CRP: C-reactive protein.

The QQ plot of FIB data normal distribution is shown in Figure 1. The QQ plot of lipoprotein data normal distribution is shown in Figure 2. The QQ plot of HDL data normal distribution is shown in Figure 3.

Figure 1.

The QQ plot of FIB data normal distribution.

Figure 2.

The QQ plot of lipoprotein data normal distribution.

Figure 3.

The QQ plot of HDL data normal distribution.

All variables were analyzed by univariate analysis (Table 2) as follows. Result variable is vascular calcification. Exposure variables are neutrophil (NEUT), lymphocyte (LY), FIB (FIB), FIB degradation product (FDP), triglyceride (TG), total cholesterol (TC), HDL (HDL), low-density lipoprotein (LDL), lipoprotein (Lipoproteins), glutamyltransferase (GGT), and C-reactive protein (CRP), there is no adjustment variable. In univariate analysis, we can see FIB, lipoprotein, and HDL were significantly correlated with VC (p <0.05). These significant correlation variables were used in the multivariate analysis (Table 3). The results of multivariate analysis showed that FIB, lipoprotein, and HDL were independent predictors of VC (p <0.05).

Table 2.

Univariate analysis of vascular calcification.

Variables	Modeling group
Variables	HR(95% CI)	p value
Neutrophiles	2074.501 (99.313, 43333.169)	<0.00001
Lymphocyte	0.000 (0.000, 0.002)	<0.00001
FIB	2.628 (1.736, 3.978)	<0.00001
FDP	1.498 (1.123, 1.998)	0.00596
Triglycerides	0.974 (0.755, 1.257)	0.83900
Cholesterol	0.759 (0.572, 1.007)	0.05574
HDL	0.036 (0.009, 0.138)	<0.00001
LDL	0.774 (0.521, 1.150)	0.20525
Lipoproteins	1.003 (1.002, 1.004)	<0.00001
Glutamyl transferase	1.021 (1.008, 1.034)	0.00117
CRP	1.134 (1.070, 1.203)	0.00003

FIB: fibrinogen; FDP: fibrin degradation products; HDL: high-density lipoprotein；LDL: low-density lipoprotein；CRP: C-reactive protein.

Table 3.

Multivariate analysis of vascular calcification.

Variables	Modeling group
Variables	HR(95% CI)	P value
Neutrophiles	0.000 (0.000, 43.945)	0.17334
Lymphocyte	0.000 (0.000, 2.254)	0.06521
FIB	2.772 (1.307, 5.880)	0.00788
FDP	1.082 (0.702, 1.668)	0.72090
HDL	0.135 (0.023, 0.803)	0.02772
Lipoproteins	1.003 (1.001, 1.004)	0.00004
Glutamyl transferase	1.014 (0.997, 1.032)	0.09675
CRP	1.033 (0.970, 1.101)	0.30588

FIB: fibrinogen; FDP: fibrin degradation products; HDL: high-density lipoprotein；CRP: C-reactive protein.

Analysis of the ROC curve of VC prediction model by FIB, lipoprotein, and HDL. Univariate analysis and multivariate analysis showed that FIB, lipoprotein, and HDL had effects on VC. To further detect and compare the predictive efficacy of these indexes on VC, the ROC curves of FIB, lipoprotein, and HDL on the VC prediction models were established (Figure 4). The ROC curve analysis and optimal threshold analysis of FIB, lipoprotein, and HDL to VC prediction model are shown in Table 4. The area under the ROC curve of FIB was 0.8018, and the area under the ROC curve of lipoprotein was 0.7348 and the area under the ROC curve of HDL was 0.7019. The models with ROC curves ranging from 0.70 to 0.80 are considered to be good, while the models with areas under the curves ranging from 0.80 to 0.90 have excellent resolution. The area under the ROC curve of the prediction model of FIB, lipoprotein and HDL was 0.8018, 0.7348, and 0.7019, respectively. Our research shows that lipoprotein and HDL are effective in predicting VC, and FIB is strong in predicting VC.

Figure 4.

ROB curve of fibrinogen, lipoprotein and high-density lipoprotein on the vascular calcification prediction model.

Table 4.

ROC curve analysis and optimal threshold analysis of the prediction model.

Test	FIB	Lipoproteins	HDL
ROC area (AUC)	0.8018	0.7348	0.7019
95%CI low	0.7538	0.6785	0.6444
95%CI upp	0.8499	0.7912	0.7593
Best threshold	3.3050	182.2500	1.0850
Specificity	0.8543	0.6774	0.5376
Sensitivity	0.6403	0.7333	0.7926
Accuracy	0.7663	0.7009	0.6449
Positive-LR	4.3937	2.2733	1.7142
Negative-LR	0.4211	0.3937	0.3858
Diagnose-OR	10.4345	5.7750	4.4435
N-for-diagnose	2.0220	2.4346	3.0280
Postive-pv	0.7542	0.6226	0.5544
Negative-pv	0.7727	0.7778	0.7812
a	90	105	114
b	29	64	91
c	50	38	30
d	172	134	106

Discussion

Vascular calcification is a very common pathological phenomenon, and the prevalence rate of calcification increases with age. Relevant statistics show that VC exists in ≥ 90% of men and ≥ 67% of women over the age of 70.⁵ VC is not only an important cause of increased morbidity and mortality of cardiovascular disease but also a widespread pathological phenomenon harmful to the health of the middle-aged and elderly. VC is a marker of atherosclerosis and an independent predictor of cardiovascular events. It is closely related to all-cause mortality and has always been an important field of cardiovascular medical research.^5,6,12,28

Fibrinogen is a glycoprotein consisted of three pairs of distinct polypeptide chains, α, β, and γ. It is mainly synthesized by hepatocytes in the liver and can be degraded by fibrinolytic enzymes to form fibrinogen degradation products. The increase in plasma FIB concentration is an independent risk factor for cardiovascular disease.^29,30 Our results also show that FIB has an effect on vascular calcification, which is consistent with previous studies. Lipoprotein is a kind of hydrophobic core rich in sterolipid and triglyceride and spherical particles composed of protein, phospholipid, cholesterol, and so on. Lipoproteins play an important role in the packaging, storage, transport, and metabolism of extracellular lipids. High-density lipoprotein (HDL) is a complex lipoprotein composed of lipids and proteins and their regulatory factors, mainly produced by the liver and small intestine. Studies have shown that lipoprotein can regulate the start-up of VC, and HDL has an inhibitory effect on VC. Our results are consistent with previous studies.

In this study, using univariate and multivariate analysis, we determined that FIB, lipoprotein, and HDL were independent predictors of VC. The models with ROC curves ranging from 0.70 to 0.80 are considered to be good, while the models with areas under the curves ranging from 0.80 to 0.90 have excellent resolution. The ROC curve analysis of the prediction model established in our study shows that lipoprotein and HDL have good prediction efficiency for VC, and fibrinogen has better predictive ability than high-density lipoprotein and lipoprotein in patients with arteriosclerosis.

The present study had several limitations. The sample size was still small. In this study, there were 325 males and 137 females. Due to the small number of females, this study has not yet explored the influence of gender on VC. In the future, we will continue to increase the sample size to further study this issue. Although our study shows that FIB has a good predictive ability for VC, it still needs to be further determined in combination with clinical imaging examination.

To sum up, this study found that FIB has a strong ability to predict VC. Elevated levels of fibrinogen suggest an increased risk of calcification. The results of this study have good clinical practical value, which is a necessary item for hospitalized patients, and the collection of samples is simple. FIB can also be measured repeatedly, cheap, easily accepted by patients, and can be regulated by the use of drugs such as Bezafibrate. FIB is expected to become a new molecular marker for VC.

Footnotes

Authors’ contributions

Conceptualization: Lihua Zheng, Yaheng Zhao.

Data curation: Yan Yang, Pengju Gao, Zepeng Zhang, Wenxiu Fu.

Formal analysis: Jing Zhang, Yan Yang, Hongsong Zhang.

Investigation: Yan Yang, Hongsong Zhang.

Methodology: Yan Yang, Lihua Zheng, Yaheng Zhao.

Visualization: Pengju Gao, Zepeng Zhang, Wenxiu Fu.

Writing-original draft: Jing Zhang.

Writing-review & editing: Lihua Zheng, Yaheng Zhao.

Declaration of conflicting interests

The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The author(s) received no financial support for the research, authorship, and/or publication of this article.

ORCID iD

Jing Zhang

References

Vossen

Kroon

Schurgers

, et al. Pharmacological and nutritional modulation of vascular calcification. Nutrients 2019; 12: 100.

Smith

Hewitson

Holt

SG.

Diagnostic tests for vascular calcification. Adv Chronic Kidney Dis 2019; 26: 445–463.

Shao

Sierra

Cohen

, et al. Vascular calcification and aortic fibrosis: a bifunctional role for osteopontin in diabetic arteriosclerosis. Arterioscler Thromb Vasc Biol 2011; 31: 1821–1833.

Demer

Tintut

Inflammatory, metabolic, and genetic mechanisms of vascular calcification. Arterioscler Thromb Vasc Biol 2014; 34: 715–723.

Leopold

JA.

Vascular calcification: mechanisms of vascular smooth muscle cell calcification. Trends Cardiovasc Med 2015; 25: 267–274.

Akers

Nicholls

Di Bartolo

BA.

Plaque calcification: do lipoproteins have a role?

Arterioscler Thromb Vasc Biol 2019; 39: 1902–1910.

Ayhan

Kasapkara

Aslan

, et al. Relationship of neutrophil-to-lymphocyte ratio with aortic stiffness in type 1 diabetes mellitus. Can J Diabetes 2015; 39: 317–321.

Raggi

Bellasi

Bushinsky

, et al. Slowing progression of cardiovascular calcification with SNF472 in patients on hemodialysis: results of a randomized phase 2b study. Circulation 2020; 141: 728–739.

Arefin

Buchanan

Hobson

, et al. Nrf2 in early vascular ageing: calcification, senescence and therapy. Clin Chim Acta 2020; 505: 108–118.

10.

Mozos

Malainer

Horbańczuk

, et al. Inflammatory markers for arterial stiffness in cardiovascular diseases. Front Immunol 2017; 8: 31.

11.

Moe

Chen

NX.

Inflammation and vascular calcification. Blood Purif 2005; 23: 64–71.

12.

Yang

Zou

Hou

, et al. Extracellular vesicles in vascular calcification. Clin Chim Acta 2019; 499: 118–122.

13.

Lee

Kim

Park

, et al. Effects of low-density lipoprotein cholesterol on coronary artery calcification progression according to high-density lipoprotein cholesterol levels. Arch Med Res 2017; 48: 284–291.

14.

Pechlivanis

Mahabadi

Hoffmann

, et al. Association between lipoprotein(a) (Lp(a)) levels and Lp(a) genetic variants with coronary artery calcification. BMC Med Genet 2020; 21: 62.

15.

Fiuza

Lipoprotein(a) as a novel therapeutic target for preventing cardiovascular disease: a whiter shade of pale?

Rev Port Cardiol 2019; 38: 495–496.

16.

Kälsch

Lehmann

Moebus

, et al. Aortic calcification onset and progression: association with the development of coronary atherosclerosis. J Am Heart Assoc 2017; 6: 30.

17.

Taskinen

MR.

Diabetic dyslipidaemia: from basic research to clinical practice. Diabetologia 2003; 46: 733–749.

18.

Lewandowski

Płusajska

Horzelski

, et al. Prevalence of dyslipidaemia and pre-diabetes among women with polycystic ovary syndrome (PCOS): do we overestimate cardiovascular risk. Horm Metab Res 2019; 51: 539–545.

19.

Griffin

BA.

Lipoprotein atherogenicity: an overview of current mechanisms. Proc Nutr Soc 1999; 58: 163–169.

20.

Parhami

Basseri

Hwang

, et al. High-density lipoprotein regulates calcification of vascular cells. Circ Res 2002; 91: 570–576.

21.

Toth

PP.

When high is low: raising low levels of high-density lipoprotein cholesterol. Curr Cardiol Rep 2008; 10: 488–496.

22.

Kurano

Anti-thrombotic properties of high-density lipoprotein. Rinsho Byori 2016; 64: 66–71.

23.

Thompson

Partridge

Coronary calcification score: the coronary-risk impact factor. Lancet 2004; 363: 557–559.

24.

Mori

Torii

Kutyna

, et al.

Coronary artery calcification and its progression: what does it really mean?

JACC Cardiovasc Imaging 2018; 11: 127–142.

25.

Kaur

Rahimi

Razali

, et al. H. Association of coronary artery calcium score with calcification and degree of stenosis: an autopsy study. Malays J Pathol 2019; 41: 177–183.

26.

Pepe J, Diacinti D, Fratini E, et al. High prevalence of abdominal aortic calcification in patients with primary hyperparathyroidism as evaluated by Kauppila score. Eur J Endocrinol 2016; 175: 95--100.

27.

NasrAllah

Nassef

Elshaboni

, et al. Comparing different calcification scores to detect outcomes in chronic kidney disease patients with vascular calcification. Int J Cardiol 2016; 220: 884–889.

28.

Youssef

Guo

McClelland

, et al. Risk factors for the development and progression of thoracic aorta calcification: the multi-ethnic study of atherosclerosis. Acad Radiol 2015; 22: 1536–1545.

29.

Bielak

Klee

Sheedy

2nd , et al. Association of fibrinogen with quantity of coronary artery calcification measured by electron beam computed tomography. Arterioscler Thromb Vasc Biol 2000; 20: 2167–2171.

30.

Dellegrottaglie

Saran

Rajagopalan

Vascular calcification in patients with renal failure: culprit or innocent bystander. Cardiol Clin 2005; 23: 373–384.