*Article* **Management of High-Risk Atherosclerotic Patients by Statins May Be Supported by Logistic Model of Intima-Media Thickening**

**Dorota Formanowicz 1,\*,†, Jacek B. Krawczyk 2,†, Bartłomiej Perek 3,†, Dawid Lipski 4,‡ and Andrzej Tykarski 4,‡**


**Abstract:** While the use of statins in treating patients with atherosclerosis is an undisputed success, the questions regarding an optimal starting time for treatment and its strength remain open. We proposed in our earlier paper published in Int. J. Mol. Sci. (2019, 20) that the growth of intimamedia thickness of the carotid artery follows an S-shape (i.e., logistic) curve. In our subsequent paper in PLoS ONE (2020, 15), we incorporated this feature into a logistic control-theoretic model of atherosclerosis progression and showed that some combinations of patient age and intima-media thickness are better suited than others to start treatment. In this study, we perform a new and comprehensive calibration of our logistic model using a recent clinical database. This allows us to propose a procedure for inferring an optimal age to start statin treatment for a particular group of patients. We argue that a decrease in the slope of the IMT logistic growth curve, induced by statin treatment, is most efficient where the curve is at its steepest, whereby the efficiency means lowering the future IMT levels. Using the procedure on an aggregate group of severely sick men, 38 years of age is observed to correlate with the steepest point of the logistic curve, and, thus, it is the preferred time to start statin treatment. We believe that detecting the logistic curve's steepest fragment and commencing statin administration on that fragment are courses of action that agree with clinician intuition and may support decision-making processes.

**Keywords:** atherosclerosis; statins; control-theoretic model; logistic growth

#### **1. Introduction**

Knowledge of the intima-media (IMT) growth process is essential for decision making regarding statin therapy initiation and intensification. The purpose of our study is to assess whether the mathematical modeling of IMT growth, proposed in [1,2], can assist clinicians at crucial stages of the process.

Thickening of the intima-media complex, which is an undisputed symptom of atherosclerosis, is an inevitable consequence of the process of aging of the human vascular system. The age-related changes can be observed at both micro- and macrolevels. At the micro-level, cellular senescence manifests as reduced cell proliferation, an irreversible arrest of growth, apoptosis, DNA damage, etc. [3]. At the macro-level, atherosclerotic plaques with calcium deposits can be detected by imaging examinations

**Citation:** Formanowicz, D.; Krawczyk, J.B.; Perek, B.; Lipski, D.; Tykarski, A. Management of High-Risk Atherosclerotic Patients by Statins May Be Supported by Logistic Model of Intima-Media Thickening. *J. Clin. Med.* **2021**, *10*, 2876. https://doi. org/10.3390/jcm10132876

Academic Editor: Anna Kabłak-Ziembicka

Received: 18 May 2021 Accepted: 23 June 2021 Published: 29 June 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

applied routinely in clinical practice; see [4]. The deposits are of clinical significance, as they are the final stage of vascular degeneration.

Clearly, many factors contribute to atherosclerosis development; see e.g., [5]. In particular, there are some inherited predisposing factors and other factors that may be modified by our lifestyle, which include diet; physical activity; and adherence to recommendations of optimal management of many atherosclerosis-modifying diseases, such as arterial hypertension, diabetes, and hyperlipidemia—see [6].

According to our knowledge, it is very difficult, if at all possible, to regress atherosclerotic plaque development. However, in some cases, doctors are able to stabilize the plaque and, therefore, inhibit disease progression. This can happen if intensive treatment with high-dose statins is applied; see [7].

However, doubts surrounding the dosage of statins and the therapy timing remain. A few years ago, a cohort study [8] on the general UK population reported statin overuse in patients with low cardiovascular risk and underuse in patients with high cardiovascular risk. Furthermore, another study (see [9]) points to possible adverse events, such as muscle and liver injury, cognitive impairment, new-onset diabetes mellitus, and even hemorrhagic stroke, as a result of long-term statin therapy. Moreover, there is no strong clinical evidence that the elderly would benefit from statin therapy; see [10]. Statin therapy should be individualized and based on the patient's risk profile. This study uses a model that may alleviate the above concerns.

The measurement of the carotid artery IMT can be achieved by the simple and noninvasive technique of measuring atherosclerotic burden [11]. Consequently, IMT has been utilized as a reliable marker of drug efficiency and tested in clinical trials devoted to atherosclerosis treatment. Moreover, IMT is widely accepted as a screening tool that can be used together with the traditional risk factors assessment. The size and dynamics of IMT can help in determining optimal therapeutic interventions as well as in the application of other diagnostic tools; see [12]. Although there are studies that have presented a discrepancy between carotid IMT changes, prognosis, and the course of cardiovascular pathologies [13,14], the overwhelming clinical data (see [15–19]) strongly confirm that the IMT will continue to be used as a valuable tool in clinical research.

It was conjectured in [1] that the IMT growth process follows an S-shape (i.e., *logistic*) curve. An application of a mathematical model based on this proposition to atherosclerosis management by statins was developed in [2]. However, the number of observations in [1] of the logistic model being calibrated was low (27); therefore, the quantitative reasoning based on that model was mainly of conceptual value rather than being immediately applicable to management of atherosclerosis. The recent availability of the large Cardio Poznan Database (122 observations) [20] has created an opportunity for us to perform a new calibration of the logistic model from [1] and provide some managerial advice.

In the next section, we describe briefly the Cardio Poznan Database, the source of our new data. Then, Section 3 discusses the new data support for the S-shaped IMT growth. Subsequently, in Section 4, we propose a procedure for inferring an optimal age to start statin treatment for a particular group of patients. The paper ends with brief Concluding Remarks and an Appendix, which contains a few summary statistics concerning the observations gathered in the database.

#### **2. Cardio Poznan Data**

The data collection project, see [20], received a positive opinion (decision No. KB 341/21) of the Bioethics Committee of the Poznan University of Medical Sciences.

The data collection involved 122 consecutive patients: 78 males (63.9%) and 44 females (36.1%). Their mean age was 49.6 years, and the standard deviation was 15.6 years. These patients were treated in the Department of Hypertension and Angiology and Internal Medicine at the Poznan University of Medical Sciences in the first quarter of 2020. From this group, we selected the male subjects (*n* = 31) who had arterial hypertension-related cardiac disease, i.e., coronary artery disease (CAD) and/or vascular complications, such as peripheral vascular disease (PVD). This group of 31 *male* patients represents the observation sample for our study. These patients are referred to as *severely sick men*. They are split into two subgroups: (1) patients undergoing statin therapy, denoted as *statin(+)*, and (2) patients not treated with statins, denoted as *statin(-)*.

We provide a summary of the demographic and clinical data of the studied patients in Appendix A, Table A1. The findings of the laboratory tests and imaging examinations (carotid artery Doppler ultrasonography) are provided in Table A2.

#### **3. Support for S-Shaped Growth of the Atherosclerotic Plaque**

#### *3.1. Importance of S-Shaped Growth for Atherosclerosis Treatment*

Confirming both, the S-shape of the IMT growth process (see [1,2]) and its quantitative features is important for clinical reasons. Notably, knowledge of the S-shape model for the atherosclerosis process will enable us to indicate the patient age ranges of the disease's fast and slow growth. Specifically, we aim to identify when the atherosclerotic process is fast. In an attempt to prevent the IMT from thickening, the ideal place (here, the age range) to start statin treatment is when the curve is steep, i.e., before it eventually flattens at the patient's older age. Below, we present why we think this line of thought could help clinicians.

A slow IMT growth can occur when the patient is very young—too early to administer statins—or when the patient is very old (see [21])—too late to start treatment. Assuming statin treatment slows down the plaque growth by a certain amount, if this amount is subtracted from fast growth, the disease-slowing effect will be more substantial than if statins are administered when the disease progresses slowly. A mathematical explanation for this effect involves the model of nonlinearity. This emulates nonlinearity of the underlying medical process, as postulated in [1]. With the help of new data, we aim to confirm the logistic growth of IMT and establish the steepest fragment on an IMT growth model.

#### *3.2. The S-Shape Conjecture*

The S-shape of the atherosclerosis process was postulated in [1] (see also [2]). Regrettably, the number of data points available in [1] is low, and the proposed model's goodness-of-fit statistics are mixed. In an attempt to improve the statistical significance of the model, we now use a larger dataset [20] and carry out a new parameter identification procedure for the logistic model of atherosclerosis.

The starting point of our analysis of atherosclerosis in [1], continued in [2], is Figure 1 presented below (produced out of Figures A1 and A2 published in [1]). The figure panels show S-shaped curves calibrated ibidem using our previous data on IMT vs. age of severely sick men on dialysis.

The S-shaped curves in Figure 1 are of the following analytic form:

$$\mathbf{x}(t) = \frac{c\mathbf{x}\_0 e^{\mathbf{a}t}}{c + \mathbf{x}\_0(e^{\mathbf{a}t} - 1)} \,. \tag{1}$$

Each is a solution to the logistic differential equation *dx dt* <sup>=</sup> *ax*- <sup>1</sup> <sup>−</sup> *<sup>x</sup> c* , *x*(0) = *x*<sup>0</sup> where


The difference in the appearances of the S-shaped curves is attributed to the patients' medical conditions. Typically, patients whose medical conditions require dialysis will have a large *c* that will grow more steeply. This is why the left panel curve grows faster and reaches higher values. The model coefficients for the severely sick men on dialysis (see

the left panel) are provided in Table 1. (The model coefficients for the healthy patients' behavior are omitted, because healthy patients are not considered in this paper.)

**Figure 1.** Fitting of the IMT time profile for severely sick patients (on dialysis)—**left panel** (11 observations), and fitting of the IMT time profile for healthy patients—**right panel** (16 observations).

**Table 1.** Parameters of the IMT time profile model for severely sick patients (on dialysis); see [1]. SSE is the sum of squared estimate of errors, and RMSE is the root mean square error. *R*<sup>2</sup> is the coefficient of determination.


The evidence provided by the curves, together with the existing clinical literature cited in [1], led us to propose ibidem that the atherosclerotic plaque's growth over a patient's life span has an S-shape and can be represented mathematically by a logistic function.

As mentioned above, the number of data points in Figure 1 is low, and the curves' goodness-of-fit statistics, reported in [1] and cited in Table 1, are mixed. For example, while the coefficient of determination *R*<sup>2</sup> = 0.36 for the case of severely sick patients might be considered satisfactory, the model-corresponding values of SSE = 0.389 and RMSE = 0.1881 are ordinary for such a small data sample of 11 observation points. Nevertheless, we conjectured ibidem that in aggregate, the above goodness-of-fit statistics provide support for a logistic process of IMT formation. However, given the low number of observations, they do not carry sufficient weight for the obtained model to be relied upon in clinical diagnostics and treatment, which are our ultimate goals of the model usage (see [2]).

#### **4. New Data Support**

#### *4.1. Patient Aggregate*

Figure 2 shows the new 62 IMT measurement points from [20] for the left and right arteries of severely sick men (see the legend for points marked L and R). The 11 black circles are the same as those in the left panel of Figure 1 and represent the 2008–2011 sample ([22]) of the severely sick men on dialysis. Their IMT values point to accelerated atherosclerosis.

The data from [1] did not distinguish between the left and right arteries; thus, it is likely that both were represented. Furthermore, both datasets are *aggregates* of sick patients who take statins (for different period lengths) and sick patients who do not take statins. Hence, from the point of view of gravity of the sickness, the old and new data may be compatible.

We used the new data to obtain the logistic curve in Figure 2. The identified parameters of this curve are provided in Table 2.

**Figure 2.** Fitting of the IMT time profile for severely sick men from [20].

**Table 2.** Parameters of the IMT growth model for severely sick men; see [20] (infl. point stands for the inflation point, and max slope indicates the fastest growth of IMT in mm/y for this group of patients).


There are several comments to make concerning Figure 2.


**Proposition 1.** *We propose that, on balance, for small RMSE (good) versus small R*<sup>2</sup> *(bad), the logistic curve in Figure 2 supports the conjecture of the authors of [1] regarding the S-shaped process of the IMT growth.*

**Proposition 2.** *The steepest part of the logistic curve is around its inflection point (*≈*38, 0.6; see Table 2 and the beige circle on the curve in Figure 2). Therefore, propose that for the group of sick men, starting patient medication at around 38 years old may be the most beneficial approach.*

The new dataset (from [20]), which we use in this study, concerns severely sick patients. However, this set is an aggregate of many patient types. The data are *inhomogeneous* in (at least) two aspects. First, they contain statin-medicated and statin-non-medicated patients. Second, the medicated patients take various doses of different statins (atorvastatin and rosuvastatin) for a varying number of months. Of course, model coefficients crucially depend on the sample patients' conditions. For example, the 2008–2011 data [22] of the severely sick men *on dialysis* generated a steeper S curve than that of the new data. Arguably, models built for a homogeneous patient group should be more reliable than a model built for a patient aggregate. In the next sections, we *disaggregate* the severely sick patient group into non-medicated and medicated patient subgroups and propose an IMT growth model for each subgroup.

#### *4.2. Non-Medicated, Severely Sick Men*

The group analyzed here is composed of patients who are severely sick but remain non-medicated. We refer to this subset of patients classified in the database [20] as *nonmedicated*.

We now analyze the IMT growth process of the non-medicated patients.

Figure 3 shows 18 pairs (age and IMT) of the measurement points for severely sick, non-medicated men from [20]. Of course, these points are also represented in Figure 2. Now, they are analyzed alone.

The parameters of the logistic curve in Figure 3, which represents the IMT growth model, are provided in Table 3.

**Table 3.** Parameters of the IMT growth model for non-medicated, severely sick men from [20].


**Figure 3.** Fitting of the IMT time profile for non-medicated, severely sick men from [20].

Observing Figure 3, it can be noted that, in comparison to that of Figure 2, the logistic curve stabilizes here at a much lower IMT value. This may suggest that these patients' atherosclerotic plaque has been developing in a different manner to that of the other patient aggregate. This is discussed further in Section 4.1.

These patients are likely to have stable atherosclerotic plaque and be free from many of the clinical symptoms recorded in [20]. In particular, they may have developed collateral circulation, hence remaining in a clinically stable condition. Moreover, the IMT measurements, consequential for our study, are the IMT's thickness quantifications only. However, using the thickness measurement only, it is impossible to conclude the morphology of the atherosclerotic plaque. Nevertheless, it may be the morphology that is relevant to the classification of a patient as *severely sick*. For better insight into these statin-non-medicated patients, their characteristics, demographic data, and clinical and laboratory data are compared with those of severely sick but statin-medicated patients in Table A2.

While this group of non-statin-medicated patients might not profit from statin treatment, the observations of their IMT growth, as shown in Figure 3 and Table 3, assist in furthering the discussion of the growth's S-shape.


We draw the information on qualitative and quantitative properties of the IMT growth process in the non-medicated patients from Figure 3 and Table 3 as follows:


**Proposition 3.** *The above comments lead to the proposal that the model for the non-medicated, severely sick men (see Figure 3 and Table 3) supports the conjecture of the authors of [1] regarding the S-shaped process of the IMT growth.*

#### *4.3. Statin-Medicated, Severely Sick Men*

The complement to the non-medicated, severely sick men within the severely sick men aggregate in [20] is the group of severely sick men receiving statins or *statin-medicated* men. We now analyze the IMT growth process of these statin-medicated patients.

Figure 4 shows 44 pairs (age, IMT) of the measurement points for severely sick statinmedicated men from [20]. Of course, these points are also represented in Figure 2. Now, they are analyzed alone.

The parameters of the logistic curve in Figure 4, which represents the IMT growth model, are provided in Table 4.

**Table 4.** Parameters of the IMT growth model for statin-medicated, severely sick men; see [20].

*x***<sup>0</sup>** *c a* **SSE RMSE** *R***<sup>2</sup> Infl. Point Max Slope Sample Size** 0.325 1.25 0.02625 1.369 0.1784 −0.13 [≈40, 0.625] 0.008203 44

**Figure 4.** Fitting of the IMT time profile for statin-medicated, severely sick men from [20].

Below are our comments regarding the model for the statin-medicated, severely sick men.


15. The goodness-of-fit statistics of the statin-medicated patient model do not suggest that this model is better than the one proposed for the patient aggregate. Although SSE has improved (1.369 < 1.689), the expected distances between an actual measurement and the corresponding model's value (RMSE 0.1784 > 0.1664) and *R*<sup>2</sup> have worsened. We remind the reader that even *R*<sup>2</sup> < 0 should not alone disqualify a nonlinear model.

**Figure 5.** The slopes of the plaque formation processes for the severely sick men on statins, without statins, and their aggregate.

**Proposition 4.** *The logistic growth model (see Figure 4 and Table 4) for the statin-medicated, severely sick men has an explanatory value in that it helps to explain the IMT plaque formation process for this group of patients. However, the goodness-of-fit statistics for this model do not indicate that this model is an improvement on the patient aggregate's model referred to in Proposition 1.*

#### **5. How to Infer an Optimal Age for Starting Statin Treatment**

We have previously proposed that the optimal patient age for a specific group of patients to start statin treatment is when the curve is at its steepest. This seems the best locus on the S-curve to prevent it from rising or, in clinical terms, to prevent IMT from thickening. Figures 1–4 can help to find such loci for a specific group of patients.

The patterns of the speed of plaque formation differ between the aggregate of the severely sick patients and the non-medicated patients. We can see these speeds in Figure 5: the solid line represents the patient aggregate and the dash-dotted line the non-medicated patients. The dashed line indicates patients on statin; see the next section.

The maximum speed of the plaque formation for the non-medicated patients (a) is higher than the patients' aggregate i.e., 0.008915 > 0.008625 for the former and latter group and (b) occurs 17 years earlier for the former and latter groups.

We claimed in Proposition 2 that 38 years of age (see infl. point in Table 2) is the right age to start statin treatment in the aggregated group. Our argument is that a decrease induced by treatment—in the slope of the IMT logistic curve is most efficient when the curve is at its steepest, whereby the efficiency concerns lowering the future IMT levels. Beginning treatment of non-medicated patients at the young age of 21 years old (see infl. point in Table 3) appears to be early. However, as judged by the logistic curve in Figure 3, these patients do not need medication. Arguably, their own bodies manage to considerably slow down plaque growth at this age. Just after the inflection point, we can see in Figure 5 that the dash-dotted (red) line quickly drops far below the other lines. The corresponding graph of the plaque formation in Figure 3 flattens as if these non-medicated patients were submitted to treatment. In fact, one could claim that it is their own bodies that generate this plaque formation pattern.

Briefly, an analysis of IMT slopes can help in making decisions regarding the best patient age for commencing statin treatment. The steepest slope can be learned from the slope's first derivative graph (see Figure 5). The steepest slope is where the derivative attains a maximum.

#### **6. Limitations and Strength of This Study**

An important limitation of our study is that it concerns male patients only. The reason for this is that the female population was less represented in [20]. Furthermore, the female patient population is less homogeneous than the male patient population in that their symptoms associated with cardiovascular pathologies are more variable and therefore more difficult to calibrate in the model than those of male patients.

Another limitation is related to the patient sample size in [20]. Although significantly larger than that in [1], our sample size is modest when compared with that of international studies; see e.g., the JUPITER trial [23]. With more patients in the database, perhaps augmented by a population-based study, we would be able to attempt to model IMT growth in female patients.

Notwithstanding these limitations, we strongly believe that the male patient sample size we used in this study was sufficient to validate the model proposed and explained in [1,2]. The results of the biostatistical data analyzed in this paper should assure clinicians regarding our model's usefulness.

#### **7. Concluding Remarks**

Our study showed that logistic models of IMT growth can support clinician decisions concerning the use of statins in the treatment of atherosclerosis. Specifically, we suggest that the steepest segment of the IMT-growth's S-shape curve, obtained for a specific group of patients, can be recommended as the appropriate disease phase to commence treatment. In numerical terms, we identified some statistical evidence that 38 years old may be an appropriate age to start treatment for the group of severely sick men in [20].

**Author Contributions:** Conceptualization: D.F., J.B.K., and B.P.; methodology: D.F., J.B.K., and B.P.; validation: D.F., J.B.K. and B.P.; formal analysis: D.F., J.B.K. and B.P.; investigation: D.F., J.B.K. and B.P.; writing—original draft preparation: D.F., J.B.K. and B.P.; writing—review and editing: D.F., J.B.K. and B.P.; data collection and furnishing: D.L. and A.T. All authors have read and agreed to the published version of the manuscript.

**Funding:** D.F., B.P., D.L. and A.T. were supported by Poznan University of Medical Sciences' statutory founds.

**Institutional Review Board Statement:** Bioethics Committee at Poznan University of Medical Sciences has confirmed in decision No. KB 341/21 that this study was not a medical experiment and therefore according to Polish law and the GCP regulations this research did not require approval of the Bioethics Committee.

**Informed Consent Statement:** The IMT assessment used in this study to validate our atherosclerotic plaque build-up model is a routine non-invasive diagnostic procedure performed using Doppler ultrasound. Anonymous use of the IMT results and the routinely performed laboratory parameters did not require any informed consent of the subjects.

**Data Availability Statement:** Data are available at https://www.researchgate.net/publication/3513 55752\_CARDIO\_POZNAN\_DATA (accessed on 25 June 2021).

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **Appendix A. CARDIO POZNAN DATA—Summary Statistics**

We refer the reader to the Excel file at [20] for our field observations. Here, we present a few summary statistics concerning the observations.


**Table A1.** Summary of demographic and clinical CARDO POZNAN DATA severely sick male patients.

Variables are presented as either means (standard deviation (SD)) for continuous data or numbers (%) for categorical data. \* Included symptomatic (stroke/TIA) and asymptomatic with significant lesions (>80% of diameter) in the carotid artery (noted in Doppler ultrasound examination); \*\* Defined if eGFR calculated by means of simplified (short) MDRD formula was below 60 mL/min/1.73 m<sup>2</sup> BSA (body surface area). *Abbreviations* : ACS = acute coronary syndrome; CABG = coronary artery bypass grafting; CAD = coronary disease; CKD = chronic kidney disease; COPD = chronic obstructive; pulmonary disease; GI = gastrointestinal; PCI = percutaneous coronary intervention; PVD = peripheral vascular disease.

**Table A2.** Results of laboratory studies and Doppler ultrasonography of Poznan Cardio severely sick male patients.


Variables are presented as means (SD). *Abbreviations*: CRP = C-reactive protein; eGFR = estimated glomerular filtration rate; HDL = high-density lipids; HGB = hemoglobin concentration; IMT = intima-media thickness; NLR = neutrophil-to-lymphocyte ratio; PLR = platelet-to-lymphocyte ratio; RBC = red blood cell count; WBC = white blood cells count.

#### **Appendix B. Assessment of Carotid Intima-Media Thickness (IMT)**

The carotid IMT measures were performed with the use of high-quality sonography (System EPIQ 5, Philips N.V., The Netherlands) using the linear 12–3 MHz transducer. IMT was measured bilaterally at the distal wall of carotid arteries in length (cm) starting 1 cm below the carotid bulb. At least five thickness measurements were performed on each site, and its average value was accepted as the outcome.

#### **Appendix C. Data Management and Statistical Analysis**

Continuous variables were validated for normality by means of the Shapiro–Wilk W test. If they satisfied criteria of normal distribution, they were expressed as means (SD).

#### **References**


## *Review* **Fibrin Clot Properties in Atherosclerotic Vascular Disease: From Pathophysiology to Clinical Outcomes**

**Michał Z ˛abczyk 1,2,†, Joanna Natorska 1,2,† and Anetta Undas 1,2,\***


† Michał Z ˛abczyk and Joanna Natorska contributed equally.

**Abstract:** Fibrin is a major component of thrombi formed on the surface of atherosclerotic plaques. Fibrin accumulation as a consequence of local blood coagulation activation takes place inside atherosclerotic lesions and contributes to their growth. The imbalance between thrombin-mediated fibrin formation and fibrin degradation might enhance atherosclerosis in relation to inflammatory states reflected by increased fibrinogen concentrations, the key determinant of fibrin characteristics. There are large interindividual differences in fibrin clot structure and function measured in plasma-based assays and in purified fibrinogen-based systems. Several observational studies have demonstrated that subjects who tend to generate denser fibrin networks displaying impaired clot lysis are at an increased risk of developing advanced atherosclerosis and arterial thromboembolic events. Moreover, the majority of cardiovascular risk factors are also associated with unfavorably altered fibrin clot properties, with their improvement following effective therapy, in particular with aspirin, statins, and anticoagulant agents. The prothrombotic fibrin clot phenotype has been reported to have a predictive value in terms of myocardial infarction, ischemic stroke, and acute limb ischemia. This review article summarizes available data on the association of fibrin clot characteristics with atherosclerotic vascular disease and its potential practical implications.

**Keywords:** atherosclerosis; coronary artery disease; fibrin clot; fibrinolysis; thromboembolism

#### **1. Introduction**

Growing evidence indicates that the formation of denser fibrin networks, which are less susceptible to lysis, characterizes patients with atherosclerosis and arterial thromboembolic events. Several cardiovascular risk factors, such as hyperlipidemia, hypertension, smoking, or diabetes have also been shown to be associated with unfavorably altered fibrin clot properties in the general population. Low-dose aspirin, statins, better diabetes control, or smoking cessation have been shown to increase fibrin clot permeability and its susceptibility to lysis. Moreover, it has been shown that non-vitamin K antagonist oral anticoagulants (NOACs) are able to improve fibrin clot characteristics and contribute to the reduced risk of adverse clinical outcomes. The current review article summarizes available basic research and clinical papers deposited on PubMed over the last decade regarding associations between fibrin clot phenotype and atherosclerotic vascular disease, supported by the seminal papers from previous years. Moreover, data on novel therapeutic strategies, which can potentially influence fibrin clot characteristics, have been discussed.

#### **2. Atherosclerotic Plaque Formation**

Atherosclerosis is a major cause of cardiovascular disease that encompasses coronary artery disease, cerebrovascular disease, peripheral arterial disease, and aortic atherosclerosis. The current concept of the pathogenesis of atherosclerosis is based on chronic inflammation associated with modified lipid deposition and dysregulated immunity within the

**Citation:** Z ˛abczyk, M.; Natorska, J.; Undas, A. Fibrin Clot Properties in Atherosclerotic Vascular Disease: From Pathophysiology to Clinical Outcomes. *J. Clin. Med.* **2021**, *10*, 2999. https://doi.org/10.3390/jcm10132999

Academic Editors: Anna Kabłak-Ziembicka and Gregory Y. H. Lip

Received: 18 June 2021 Accepted: 29 June 2021 Published: 5 July 2021

**Publisher's Note:** MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations.

**Copyright:** © 2021 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https:// creativecommons.org/licenses/by/ 4.0/).

<sup>1</sup> John Paul II Hospital, 31-202 Kraków, Poland; michalzabczyk@op.pl (M.Z.); j.natorska@szpitaljp2.krakow.pl (J.N.)

arterial wall [1–3]. The key driver of atherosclerosis is elevated low-density lipoprotein (LDL) prone to undergoing oxidative modification. Following endothelial cell injury with the subsequent influx of monocytes transformed into heterogeneous macrophages and other inflammatory cells, modified LDLs are extracellularly accumulated below the endothelium, leading to fatty streaks, an initial stage of plaque formation [2]. The formation of fibroatheroma and, finally, advanced atherosclerotic plaque is associated with the secretion of multiple chemoattractants and growth factors by leukocytes and arterial smooth muscle cells (SMCs) [2]. The proliferation of SMCs is associated with the production of large amounts of extracellular connective tissue matrix, including collagen, elastin, and proteoglycans [2,3]. Oxidized LDLs (oxLDLs) are taken up by immune cells within the atherosclerotic lesion with their subsequent transformation into foam cells, leading to plaque growth [4].

Neovascularization within the advanced plaque contributes to its gradual growth in part due to intraplaque hemorrhages. An increased density of microvessels has been found in ruptured atherosclerotic plaques [5], suggesting an important link between neovascularization and plaque instability [6]. Recent studies strongly suggest that plaque healing occurs in the natural course of atherosclerosis, with higher prevalence of healed plaques in patients with chronic manifestations of atherosclerotic vascular disease compared to those with recurrent acute coronary syndromes [7,8]. Unstable or vulnerable atherosclerotic plaques that show such characteristics as a thin fibrous cap, high macrophage content, high amounts of proinflammatory factors or a large necrotic core composed of foam cells and extracellular cholesterol [9] are prone to rupture, accompanied by occlusive thrombosis.

#### **3. Blood Coagulation and Fibrin Formation in Atherosclerosis**

The role of blood coagulation in the development and progression of atherosclerotic vascular disease reaches beyond thromboembolic complications. Macrophages can produce tissue factor (TF) [4], expressed also on microvesicles and its expression is regulated by inflammatory mediators, demonstrating the link between inflammation and thrombosis [10]. TF is the high-affinity receptor and cofactor for factor (F) VII/VIIa and the resultant TF-FVIIa complex activates FIX and FX [10]. A prothrombinase complex formed on activated platelets, including FXa, its cofactor FVa converts prothrombin to thrombin, the key enzyme of blood coagulation [11]. Pro-atherogenic actions of thrombin are associated with the activation of protease-activated receptors (PARs), leading to increased endothelial permeability, SMC migration and proliferation, and the activation of platelets and leukocytes, which promote vascular calcification and plaque development [11]. Activated platelets interact with leukocytes and stimulate them to release proinflammatory cytokines, reactive oxygen species, and provide the surface for the formation of tenase and prothrombinase complexes to generate thrombin from circulating prothrombin [10]. Fibrin formation occurs when minimal amounts of prothrombin have been activated (less than 5% of thrombin generation capacity) [8]. Fibrin accumulation within atherosclerotic plaque is involved in the disease progression, especially at the late stage of plaque formation [12]. Presence of fibrin within the necrotic core of damaged plaques supports its role in plaque growth and rupture [12,13]. Intraplaque fibrin has been shown to be more common in symptomatic than in asymptomatic atherosclerotic plaques [12]. Thrombin-activated FXIII, which covalently crosslinks fibrin fibers, also catalyzes the formation of intermolecular bonds between α2-antiplasmin, fibronectin, vitronectin, thrombospondin, and collagen, which in part explains fibrin accumulation with impaired fibrinolytic degradation within the lesions. Borissoff et al. [14] showed that ApoE-/-mice, which are prone to atherosclerosis, with genetically imposed 50% reduction in prothrombin were characterized by diminished atherosclerotic lesion formation and increased plaque stability, which suggests that coagulation activation is implicated in plaque development and progression and could be a potential therapeutic target. Thrombin promotes the accumulation of neutrophils and the production of reactive oxygen species, enhancing vascular inflammation. Of pivotal importance are observations made in 2010 suggesting that enhanced blood coagulation can

be associated with plaque stability, given the fact that TF, FII, FX, and FXII activities are diminished along with plaque transformation to advanced stage [15]. Borissoff et al. [15] have also suggested that the loss of coagulation protein activity may contribute to the risk of plaque rupture.

Adventitial fibroblasts in normal arteries are able to express TF, while in atherosclerotic lesions, TF is also expressed by SMCs, foam cells, and macrophages, which can additionally release microparticle-derived TF [16]. TF was locally detected in 43% of patients with unstable coronary syndromes and in 12% of patients with stable coronary syndromes [17]. Moreover, about 40% higher blood levels of FVII have been reported in men with vulnerable atherosclerotic plaques in the coronary arteries, compared to those who had stable plaques [18]. The colocalization of several proteins involved in blood coagulation within the plaques, largely on macrophages, microvesicles, and SMCs [19], provides the rationale for the role of a local thrombin-mediated conversion of soluble fibrinogen into fibrin, the final product of blood coagulation, in the formation of atherosclerotic lesions.

Several studies have suggested that increased fibrinogen concentration, a key determinant of fibrin formation and its characteristics, is a risk factor for atherosclerotic vascular diseases, in particular coronary artery stenosis and myocardial infarction (MI) [20–22]. In the meta-analysis of 154,211 participants from 31 prospective studies, the hazard ratio (HR) for coronary heart disease and stroke was 1.78 (95% confidence interval [CI] 1.19–2.66) per 1 g/L increase in plasma fibrinogen concentrations [23]. The US National Health and Nutrition Examination Survey (NHANES) study showed that fibrinogen is associated with cardiovascular disease and about a 2.5-fold higher risk of all-cause and cardiovascular mortality during the 14 years of follow-up [24]. On the other hand, a Mendelian randomization study has shown no causal effect of fibrinogen on cardiovascular disease [25].

The localization of fibrin degradation products (i.e., D-dimer) within the human arterial wall suggests their potential atherogenic properties [26]. Higher D-dimer levels can be associated with atherosclerotic plaque remodeling or ongoing fibrinolysis [27]; however, both processes may trigger lipid deposition and modulate local inflammation within atherosclerotic plaques. Moreover, high levels of plasminogen activator inhibitor type 1 (PAI-1) have been identified both in the blood of coronary artery disease (CAD) patients and within unstable plaques [28]. Some genetically determined fibrinogen disorders, dysfibrinogenemias, have also been linked to atherosclerotic vascular disease and its thromboembolic manifestations, supporting the view that alterations to fibrin structure and function might be of greater importance than the fibrinogen concentration itself [29–31].

Fibrin acts as a scaffold for intravascular blood thrombi, enhancing platelet aggregation and thrombin generation, leading to a further increase in fibrin formation [29]. Fibrin(ogen) can interact with red blood cells through specific receptors, such as CD47, and with platelets (i.e., integrin αIIbβ3 or intercellular adhesion molecule 1) [29]. Cellular components embedded within the fibrin network after thrombus formation can modulate its properties. FXIII-dependent red blood cell retention in clots has been shown to impair the fibrin network structure, which delays thrombus degradation [32]. Scanning electron microscopic analysis of intracoronary thrombi obtained from patients with STsegment elevation MI (STEMI) showed that fibrin content increased from about 30% to 80%, while red blood cell content decreased from 31% to 2% over time after the onset of chest pain [33,34]. It suggests that thrombus formation and its major component, fibrin, is a dynamic process [35,36]. Moreover, intravascular thrombi rich in red blood cells contain more neutrophils, reflecting a high thrombus burden, which has been shown to be associated with impaired reperfusion assessed at six months after the index event among patients with STEMI [35]. The presence of polyhedral erythrocytes, polyhedrocytes, has been linked with higher erythrocyte content, higher fibrinogen, and more significant stenosis in the culprit artery [36].

A contribution of blood components, in particular key coagulation factors to atherothrombosis along with potential therapeutic targets, which can modulate fibrin clot structure and function are summarized in Figure 1.

**Figure 1.** The initiation of an atherosclerotic lesion is associated with retention of low-density lipoproteins (LDL) and their oxidation (oxLDLs). oxLDLs stimulate recruitment of blood monocytes and their differentiation into macrophages. An uptake of oxLDLs by macrophages results in the formation of foam cells. Upon stimulation, vascular smooth muscle cells (SMCs) migrate and proliferate. Tissue factor (TF), expressed on macrophages and SMCs, is involved in coagulation activation, resulting in prothrombin (factor II, FII) conversion to thrombin (FIIa), which, in a prothrombinase complex with active FV (FVa), converts fibrinogen to fibrin. Several drugs, including aspirin, statins, angiotensinconverting enzyme inhibitors (ACEI), or non-vitamin K antagonist oral anticoagulants (NOACs) have been shown to modulate fibrin clot phenotype by different mechanisms. Proprotein convertase subtilisin kexin (PCSK) type 9 inhibitors are able to attenuate interplaque inflammation, but their effect on fibrin clot properties has not been reported yet. ↑—up-regulation, ↓—down-regulation.

#### **4. Measures of Fibrin Clot Properties**

Several parameters have been used in human subjects to assess fibrin clot properties. The structure of a fibrin clot generated from plasma (or purified fibrinogen) can be described using clot permeability (Ks or Darcy's constant; reflecting volume of a buffer flowing through a fibrin gel during prespecified time) [37–39], turbidity (clot absorbance measured using a spectrophotometer at 405 or 340 nm) [40], or the direct measurement of fibrin fiber diameter, pore size, or fiber branching using microscopic techniques [41]. Fibrin clot susceptibility to lysis is measured by turbidimetry using several assays with either exogenous tissue plasminogen activator or plasmin added to clotted plasma [40,42,43]. A so-called prothrombotic fibrin clot phenotype encompasses reduced Ks associated with typical changes in fibrin structure, such as lower fibrin fibers diameter, lower pore size area between particular fibers, and an increased number of branch points, along with faster fibrin formation and prolonged lysis time (Figure 2).

**Figure 2.** Fibrin clot structure differs between healthy persons and patients with atherosclerosis. A key measure describing plasma fibrin clot structure is its permeability. Reduced fibrin clot permeability (Ks) is a typical feature of the prothrombotic fibrin clot phenotype, which is associated with lower fibrin fiber diameter, lower pore size area, and increased number of fibrin branch points. Faster clot formation results in denser fibrin network (indicated by higher clot turbidity), which is relatively resistant to lysis (prolonged clot lysis time; CLT). ↑—up-regulation, ↓—down-regulation.

#### **5. Cardiovascular Risk Factors**

The majority of well-established cardiovascular risk factors which have been reported to be associated with prothrombotic fibrin clot properties are presented in Table 1 [44–54]. Of note, there is controversy around the association of unfavorably altered clot properties and hypercholesterolemia. Low HDL cholesterol has been reported to be associated with more prothrombotic clot features [55].

**Table 1.** Cardiovascular risk factors in association with fibrin clot properties.



**Table 1.** *Cont*.

Myocardial infarction (MI), clot lysis time (CLT), fibrin clot permeability (Ks), ↑—up-regulation, ↓—down-regulation.

#### **6. Coronary Artery Disease**

#### *6.1. Acute MI*

Acute MI is a leading cause of mortality in high-income countries and is a major thrombotic manifestation of atherosclerotic lesions in coronary arteries [56]. Dense fibrin networks, as evidenced by reduced Ks and impaired clot susceptibility to lysis, have been reported in patients with acute MI, at least in part associated with increased oxidative stress and the extent of inflammation [57]. Prolonged clot lysis time has been confirmed as a risk factor for MI in men and women in a case–control study performed on 800 acute MI patients and 1123 controls [58]. It has been hypothesized that unfavorably modified fibrin clot properties observed in acute MI are also driven by increased thrombin generation and platelet activation, expressed by a release of large amounts of proteins affecting clot features, for instance beta-thromboglobulin and platelet factor 4 [34]. In a cohort of 421 men with acute MI compared to 642 controls, hypofibrinolysis has been associated with the risk of a first MI in young men, but not in subjects aged ≥ 50 years, and CLT strongly correlated with body mass index [44]. As expected, patients with acute coronary events compared to stable coronary artery disease were characterized by a more prothrombotic fibrin clot phenotype, as reflected by lower Ks and prolonged lysis time, related to a higher body mass index, higher blood pressure and higher C-reactive protein levels [57]. Plateletderived factors, such as P-selectin of platelet-factor 4, exert a similar effect, promoting prothrombotic fibrin clot features [59]. Serum levels of P-selectin and soluble CD40 ligand were also positively associated with thrombus fibrin content [34].

It has been demonstrated that fibrin is the main constituent (60%) of intracoronary thrombi obtained during thrombectomy in acute STEMI (within 12 h since the symptom onset), with amounts increasing with time. Increased intracoronary fibrin content has also been found to be positively associated with denser plasma fibrin networks, reflected by reduced Ks [36], which indicates that plasma clot characteristics have an impact on fibrin formed in intravascular thrombi. Recently, it has been suggested that intracoronary

thrombi may have another type of fibrin on the surface. We have identified a thin layer of a fibrin biofilm on the surface of 15% intracoronary thrombi from acute MI patients, which was solely associated with higher plasma fibrinogen levels [60]. Heparin infusion during coronary angiography and thrombectomy probably reduces this proportion. This observation provided additional ex vivo evidence supporting the findings of Macrae and colleagues [61], who have shown that fibrin forms a film connected to the clot network and covers whole blood clots, which may protect against infiltration by bacteria or viruses, with potential impact on wound healing and thrombus fragmentation. Despite that there are no available reports describing the presence of fibrin biofilm on thrombi obtained from other locations. The relevance of the biofilm formation in human thrombosis requires further investigation.

It is unclear whether fibrin clot composition differs among patients with the same condition. Proteomics data has shown that clot-bound protein composition can influence fibrin properties [62]. A preliminary shotgun proteomic analysis performed on plasma clots from four patients during acute MI and two months later revealed time-dependent changes in the clot structure, which may influence clot stability and its susceptibility for lysis [63]. Differences in fibrinolysis proteins, such as increased amounts of α2-antiplasmin, in acute MI may at least in part explain time-dependent changes in the clot structure following MI [63].

From a practical point of view, the issue of prognostication based on plasma fibrin clot characteristics appears to be of vital importance. Growing evidence indicates that the prothrombotic fibrin phenotype can predict cardiovascular events. A PLATO substudy performed on 4354 patients following acute MI has shown that a validated turbidimetric assay employed to assess plasma clot lysis time and clot maximum turbidity at hospital discharge, while on dual antiplatelet therapy, is able to predict adverse clinical outcomes during a 12 month follow-up period [64]. After adjustment for cardiovascular risk factors, each 50% increase in lysis time was associated with a 1.17-times higher risk of cardiovascular death or MI, and a 1.36-fold higher risk of cardiovascular death alone. A similar increase in plasma clot maximum turbidity was associated with a 1.24-fold increased risk of death (hazard ratio 1.24, 95% confidence interval 1.03–1.50) [64]. Fibrin clot density and resistance to lysis increased with higher levels of N-terminal pro B-type natriuretic peptide (NT-proBNP) and troponin T, which are known to be associated with a greater inflammatory response. The authors concluded that higher NT-proBNP levels can be associated with worse outcomes as a consequence of impaired fibrin clot features, which may lead to an increased risk of atherothrombosis [64]. Even after additional adjustment for leukocyte count, high-sensitivity C-reactive protein, high-sensitivity troponin T, cystatin C, NT-proBNP, and growth differentiation factor-15 levels, the association with death remained significant solely for lysis time [64]. It remains to be established whether clot density, or permeability, could have a similar prognostic value.

#### *6.2. Stable CAD*

Stable CAD, defined as angina pectoris, MI history, or presence of atherosclerotic plaques, has been associated with unfavorably modified fibrin clot properties [65]. Reduced fibrin clot lysability has been reported in asymptomatic women with present coronary plaque determined by computed tomography angiography, compared to both women without plaque and men, suggesting a sex-dependent link between coronary atherosclerosis and prolonged clot lysis time [66]. Patients with symptomatic CAD formed clots with more rigid structures and increased fibrin fiber mass-to-length ratio [67]. A history of MI in 33 young patients with documented CAD was associated with prothrombotic fibrin clot characteristics, including increased clot stiffness and a slower fibrinolysis rate, compared to healthy controls [68]. Increased lipoprotein(a) levels, a well-established risk factor for premature atherosclerosis, have been shown to alter fibrin clot properties and were associated with reduced Ks and prolonged clot lysis time in patients with a history of MI [69]. Moreover, in advanced CAD patients, lipoprotein(a), a genetically determined risk factor

for premature atherosclerosis, levels predicted Ks but not lysis time [70]. Similarly, type 2 diabetes concomitant to CAD was associated with prolonged clot lysis time and a more compact fibrin clot structure compared to non-diabetic CAD patients [71]. The rs495828 risk allele within the ABO locus, which is known to be associated with an increased risk of MI in CAD patients, has also been shown to be associated with a more compact fibrin network structure, as evidenced by higher clot maximum absorbance, but not lysis time, among 773 stable CAD patients [72]. In a long-term follow-up study, the area under the curve of turbidimetrically monitored clot formation and lysis predicted future cardiovascular events in stable CAD (HR = 2.4, 95%CI 1.2–4.6) [73]. Altogether, CAD is associated with the prothrombotic clot phenotype governed by several largely environmental factors.

#### **7. Peripheral Arterial Disease**

Prothrombotic fibrin clot properties were observed for the first time in 2009 in patients with intermittent claudication, a typical manifestation of peripheral arterial disease (PAD) [74,75] which occurs in 18–20% in individuals over 70 years of age, including the most serious presentation, limb ischemia, affecting up to 3% of patients with PAD. [76]. The group from Leeds reported that unfavorably altered fibrin clot structure and function are detectable in apparently healthy close relatives of patients with claudication [74,75]. In 106 PAD patients, there was a reduction by 20% in Ks when compared to controls, and it was associated with 31% prolonged CLT; the two alterations in fibrin properties predicted PAD progression during long-term follow-up [77]. Similarly, 13.4% reduced Ks with no difference in CLT was found in patients with a history of acute lower limb ischemia compared to individuals without any history of such event [77]. Interestingly, premature PAD has also been identified as the clinical condition associated with a less favorable fibrin clot phenotype, in particular 30% lower clot permeability, compared to normal conditions, and no difference in the phenotype is observed in typical older PAD patients [78]. Moreover, in critical limb ischemia patients, who represent up to 3% of patients with PAD, restenosis detected within one year following endovascular therapy was associated with slightly reduced Ks and prolonged CLT at baseline, accompanied by elevated thrombin generation and von Willebrand factor antigen; however, the fibrin clot variables cannot predict re-intervention, amputation, and death during further 3-year follow-up [79].

#### **8. Aortic Aneurysm**

Scott et al. [80] have shown that patients with abdominal aortic aneurysm (AAA) form denser fibrin clots with smaller pore sizes which are more resistant to lysis. Such prothrombotic clot phenotype was associated with the size of aneurysm and may play a role in its development. A further study performed on 169 AAA patients, including about 40% with a history of stable angina or MI, showed that plasma levels of D-dimer and thrombinantithrombin (TAT) complexes were independent predictors of AAA growth rate [81]. An increase in D-dimer level by 500 ng/mL, or TAT level by 1 μg/mL was associated with enlargement of the aneurysm size by 0.21 and 0.24 mm per year, respectively. However, it is not known to date whether prothrombotic fibrin clot phenotype in patients with aneurysm may contribute to clinical outcomes in particular its rupture or rapid enlargement.

#### **9. Pharmacological Treatment and Fibrin Clot Properties**

#### *9.1. Cholesterol-Lowering Agents*

Statins (3-hydroxy-methylglutaryl coenzyme A reductase inhibitors) may exert several cholesterol-independent antithrombotic effects, including the down-regulation of TF expression and enhanced protein C activation via increased endothelial thrombomodulin expression [82]. Although data linking hypercholesterolemia with prothrombotic clot characteristics are limited and unconvincing, statins (simvastatin or atorvastatin at a dose of 40 mg/day for 4 weeks) used in patients with stable CAD to effectively reduce LDL cholesterol have been shown to reduce plasma clot density, reflected by higher Ks and shortened

CLT, despite no impact on plasma fibrinogen levels [83]. Favorable effects of statins on fibrin clot structure and function were supported by the study in which a 3-month use of simvastatin (40 mg/d) led to a slight, though significant, increase in Ks and a shortened clot lysis time in patients with LDL cholesterol < 3.4 mmol/L free of clinically evident CAD, like in the JUPITER trial with rosuvastatin [84]. This effect appeared to be associated with a decrease in CRP concentrations, suggesting links between the antithrombotic and anti-inflammatory effects of statins.

Novel therapeutic strategies to lower LDL cholesterol based on fully humanized monoclonal antibodies that bind free plasma proprotein convertase subtilisin/kexin type 9 (PCSK9) [85], different cholesteryl ester transfer protein (CETP) inhibitors [86], and antisense oligonucleotides targeting apolipoprotein(a) [87,88] are currently under investigation. To our knowledge, their potential effects on fibrin clot properties have not been investigated yet.

#### *9.2. Aspirin*

Aspirin was the first effective antiplatelet therapy for the prevention of ischemic events in patients with atherosclerotic vascular disease. Aspirin treatment has been shown to be associated with the formation of thicker fibrin fibers and improved clot susceptibility to lysis in stable CAD patients [89]. The mechanism of aspirin action on fibrin clot structure is unclear; however, fibrinogen acetylation [90,91] has been postulated as a major contributor despite controversies as to whether aspirin at therapeutic doses (75–150 mg/day) might exert such effects observed largely in vitro. Interestingly, low-dose aspirin (75 mg/day) has been shown to exert a stronger effect on fibrin clot properties than 320 mg/day [92,93]. In ten stable CAD patients treated with aspirin at a dose of 75 mg/day, aspirin withdrawal was associated with 32% reduced Ks after one week and 41% reduced Ks after two weeks when compared to values observed during treatment [94]. It is unclear to what extent fibrin-related mechanisms might add to the well-known antithrombotic effect caused by cyclooxygenase-1 inhibition and antiaggregatory effects on platelets.

#### *9.3. Angiotensin-Converting Enzyme Inhibitors (ACEI)*

Antihypertensive therapy with ACEI has been found to modulate fibrin clot properties in association with reduced complement component C3 levels [54]; however, similar effects were noted for other agents lowering blood pressure. In a double-blind study performed in men aged < 70 years with a history of MI or hospitalization for unstable angina, a four-week treatment with quinapril was associated with a Ks increase by 13% and shortening of clot lysis time by 28% [83]. It has been suggested that the effect of ACEIs on increased fibrin clot porosity can be associated with reduced plasma fibrinogen levels [95].

#### *9.4. NOACs*

NOACs, including rivaroxaban and apixaban, which are selective and direct FXa inhibitors, and dabigatran as a direct thrombin inhibitor, are used to prevent and treat thromboembolic events [96]. Clots formed from normal plasma spiked with rivaroxaban (174 ng/mL) or apixaban (128 ng/mL) at therapeutic levels resulted in a less dense and more permeable clot structure with thicker fibers [97]. Varin et al. [98] have demonstrated more permeable fibrin networks composed of thicker fibrin fibers in plasma spiked with rivaroxaban at a concentration of 0.15 μg/mL, which was in line with the findings of Janion-Sadowska et al. [99] made in plasma-based assays in patients 2–6 h after rivaroxaban intake (20 mg/day). A seminal COMPASS trial has shown that rivaroxaban use at a dose of 2.5 mg twice daily combined with 100 mg of aspirin can prevent cardiovascular death in patients with advanced cardiovascular disease, mostly represented by those with stable CAD [100,101]. It has also been shown that pharmacological inhibition of FXa promotes the regression of advanced atherosclerotic plaques and enhances plaque stability in mice treated with rivaroxaban (1.2 mg/g) for 14 weeks [102], suggesting that the inhibition of FXa may be beneficial both in the prevention and regression of atherosclerotic vascular

disease by down-regulated activation of PARs. The influence of low-dose rivaroxaban on fibrin clot phenotype still remains to be elucidated; however, based on the current knowledge, it can be assumed that rivaroxaban 2.5 mg bid might improve fibrin properties like rivaroxaban 20 mg/day, though to a smaller extent, and thus contributing to reduced risk of adverse clinical outcomes largely thromboembolic by nature.

#### **10. Clinical Implications**

The formation of dense fibrin networks which are relatively resistant to lysis has been observed in patients with atherosclerotic vascular disease, in particular CAD and PAD, along with those who experience arterial thromboembolism. The prothrombotic features of fibrin clots that are largely determined by environmental factors can be improved by the control of cardiovascular risk factors, in particular the normalization of glycemia and statin use. Growing evidence suggests that measures of clot characteristics, such as clot permeability and clot lysis time, may predict arterial thromboembolic events. Therefore, fibrin clot measures could serve as prognostic markers in patients at risk of arterial thromboembolism. However, there is a need for large studies to validate the available observations and standardization of the assays used to characterize the clot phenotype, though the first step initiated by the scientific subcommittee of the International Society on Thrombosis and Haemostasis has been made to implement the measurement of Ks and CLT in clinical practice [43,103].

#### **11. Conclusions**

Taken together, several studies demonstrated that the formation of more compact fibrin networks displaying lower susceptibility to lysis are implicated in the progression of atherosclerosis and the occurrence of thromboembolic manifestations, in particular MI (see Table 2). The observations might support accumulating data on clinical benefits from the use of anticoagulant agents in the prevention of cardiovascular mortality. It remains to be established whether any specific modulators of fibrinolytic efficiency might be useful in the prevention of clinical outcomes in atherosclerotic vascular disease, still the major cause of morbidity and mortality worldwide.




**Table 2.** *Cont*.

Myocardial infarction (MI), coronary artery disease (CAD), peripheral arterial disease (PAD), abdominal aortic aneurysm (AAA).

**Author Contributions:** Conceptualization, A.U.; methodology, M.Z. and J.N.; writing—original draft preparation, M.Z. and A.U.; writing—review and editing, A.U.; visualization, M.Z. and J.N.; supervision, A.U.; project administration, A.U.; funding acquisition, A.U. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research was funded by the Jagiellonian University Medical College, grant number N41/DBS/000184.

**Conflicts of Interest:** The authors declare no conflict of interest.

#### **References**

