Elevated carbonylated proteins are associated with major cardiovascular events in patients with chronic coronary syndrome: A cohort study

Abstract

Background: Protein carbonylation is reported in atherosclerosis, but its predictive value is unknown.
Aims: We evaluated plasma carbonylated protein (PC) levels in association with clinical outcomes in coronary artery disease (CAD) in long-term follow-up.
Methods: In patients with advanced stable CAD, we assessed plasma PC content along with fibrin clot properties, i.e., permeability (Ks) and clot lysis time, and its determinants: plasminogen activator inhibitor-1 (PAI-1) and thrombin-activatable fibrinolysis inhibitor. We recorded a composite of myocardial infarction, ischemic stroke, systemic embolism, and cardiovascular death during a follow-up of 8.3 (1.8) years.
Results: The analysis involved 178 patients aged 64.0 (57.0–70.0) years. The baseline PC content was 2.9 (2.2–3.7) nmol/mg protein and was elevated above the reference value obtained for a control group (2.03 nmol/mg protein) in 82.6% of patients. In linear regression models, high PC adjusted for age was associated with lower Ks, longer clot lysis time, and elevated PAI-1 and thrombin-activatable fibrinolysis inhibitor. Baseline PC was 48% higher in patients with the composite endpoint (n = 67, 37.6%) compared with others (P <0.001). Patients with PC in the highest quartile (3.7–5.1 nmol/mg protein) were more likely to develop the composite endpoint compared to the lowest quartile (hazard ratio [HR] 4.9; 95% confidence interval, 2.1–11.3; P <0.001).
Conclusions: This is the first study showing that in CAD the extent of protein carbonylation, in part via its antifibrinolytic effects, predisposes to cardiovascular events in long-term follow-up, highlighting the role of persistent oxidative protein modifications in atherosclerotic vascular disease.

WHAT’S NEW? This study shows, for the first time, that increased plasma carbonylated protein levels in patients with advanced coronary artery disease are independently associated with adverse cardiovascular events, including death, in long-term follow-up. We have shown that the harmful effects of enhanced protein carbonylation are, at least in part, linked to formation of more compact fibrin clot networks and impaired susceptibility to lysis. Our results provide new insights into the role of posttranslational oxidative modifications in atherothrombosis.

INTRODUCTION

Oxidative stress is a common denominator for major cardiovascular (CV) risk factors, i.e., hypertension, hypercholesterolemia, diabetes, and smoking [1] and it is involved in coronary microvascular angina [2]. Reactive oxygen species and oxidized polyphenols lead to protein carbonylation (PC), i.e., introduction of reactive carbonyl groups, i.e., ketone, aldehyde, and lactam into the side chains of amino acids, which might alter the protein structure and function [3]. Such a modification may also be secondary to lipid peroxidation [4]. A study of 350 subjects showed that advanced oxidation protein products were positively associated with mean and maximum carotid intima-media thickness [5]. Becatti et al. [6] have shown that in myocardial infarction (MI) survivors assessed 6 months following the index hospitalization, PC was associated with denser fibrin clots and resistance to plasmin-dependent fibrinolysis. In acute ischemic stroke (IS), baseline PC levels were associated with an unfavorable fibrin clot phenotype and higher risk of poststroke disability or death assessed 3 months following the acute event [7].

To our knowledge, there have been no studies on the prognostic value of PC in patients with stable coronary artery disease (CAD). We evaluated plasma PC levels and their determinants in stable CAD, and then the prognostic value of this modification during long-term follow-up.

METHODS

Between 2013 and 2015, we enrolled consecutive patients with stable CAD and a ≥50% stenosis in at least 1 major epicardial artery, i.e., >2 mm reference diameter on coronary angiography. The subgroup of that population was described previously [8]. The exclusion criteria included acute coronary syndrome 1 month before enrollment or percutaneous coronary intervention within 6 months before enrollment, prior coronary artery bypass grafting, neoplasms, end-stage kidney or liver disease, inflammation, or anticoagulation (Supplementary material, Figure S1). Definitions of comorbidities were presented in Supplementary materials. The study protocol conformed to the 1975 Declaration of Helsinki ethical guidelines and was approved by the Jagiellonian University Medical College Ethics Committee. Study participants provided informed written consent.

Laboratory measurements

In fasting venous blood sample complete blood counts, plasma lipid profile, glucose, and creatinine were assayed by standard laboratory techniques. High-sensitive C-reactive protein and fibrinogen were assessed by nephe­lometry (Siemens, Marburg, Germany). Enzyme-linked immunosorbent assays were used to determine plasminogen activator inhibitor-1 (PAI-1) antigen (Zymutest PAI-1 Antigen, Hyphen BioMed, Neuville-Sur-Oise, France) and thrombin-activatable fibrinolysis inhibitor (TAFI) zymogen (Zymutest Activable TAFI, Hyphen BioMed), which is presented as a percentage of pooled standard plasma. 8-iso-prostaglandin F2α (8-iso-PGF2α) was assessed immunoenzymatically (Cayman Chemicals, Ann Arbor, MI, US).

Fibrin clot analysis

Fibrin clot permeation was assessed as described previously [9]. In short, we mixed 60 µl of a coagulation trigger containing 1 IU/ml human thrombin, 20 mM CaCl2, and 60 µl of citrated plasma to generate a clot in a plastic cylinder (Sars­tedt, Nümbrecht, Germany). We measured the volume of the percolated buffer. We calculated the permeability coefficient (Ks), as a measure of the average pore size in the fiber network, using the equation: Ks (× 10–9 cm2) = Q × L × η/t × A × ∆P, where Q (cm3) is the flow rate at time t (s), L (cm) is the length of the fibrin gel, η (dyne × s/cm2) is the viscosity of the li­quid, A (cm2) is the cross-sectional area and ∆p (dyne/cm2) is differential pressure. In our laboratory, the upper reference Ks value in healthy patients is 7.4 × 10–9 cm2. The inter- and intra-assay variability were <7%.

Clot lysis time (CLT) was evaluated, as previously described [9]. Briefly, we added 15 mmol/l CaCl2, 0.6 pM human tissue factor (Innovin, Siemens), 12 µmol/l phospholipid vesicles, and 60 ng/ml recombined tissue plasminogen activator (rtPA, Boehringer Ingelheim, Ingelheim, Germany) to 100 µl of citrated plasma. We measured absorbance at 405 nm at 37°C (Tecan Sunrise). We defined CLT as the time from clot formation to lysis, i.e., from the midpoint of the clear-to-maximum-turbid transition to the midpoint of the maximum-turbid-to-clear transition. The upper refe­rence value for CLT in healthy subjects at our laboratory is 84 min. The inter- and intra-assay variability were <6%.

Carbonylation measurement

Carbonyl content was measured using a method by Becatti et al. [6]. Briefly, 400 µl DNPH was added to 100 µl of plasma. Following incubation, trichloracetic acid was added for precipitation. The pellet was washed with a 1:1 solution of ethanol/ethyl acetate and resuspended in 500 µl of guanidine hydrochloride. PC content was normalized for total protein concentration and expressed as nmol/mg of protein. The reference values for apparently healthy controls, aged 56 (39–64) years, 70% male, were 0.54–2.03 nmol/mg. The detailed characteristics of the control group can be found in Supplementary material, Table S1. A similar reference value was reported for healthy volunteers by Becatti et al. [6]. The inter- and intraassay variability of the results was <10%.

Follow-up

Data were censored in January 2023. The primary endpoint was a composite of MI, IS, systemic thromboembolism (SE), and CV death. Secondary endpoints were MI, IS/SE, and CV death analyzed separately. We did not record type 2 MI. CV mortality was coded when the cause of death was: MI, IS, thromboembolism of any other vascular bed, heart failure, arrhythmia, or a CV procedure. Follow-up was conducted during an ambulatory visit or a phone call. We asked patients or their families to provide medical records for confirmation. We also used the National Mortality Regi­stry maintained by the State Systems Department of the Ministry of Digital Affairs to assess patient status.

Statistical analysis

We assumed the rate of the composite endpoint at 15% for patients with PC in the bottom quartile and 60% for patients in the top quartile, based on data from large regi­stries [10, 11]. The estimated hazard ratio (HR) was 4.0, significance level was set at 0.05, and the power of the test at 90%, which led to the overall number of participants of 156, 39 patients per quartile.

The Shapiro–Wilk test was used for assessment of continuous variable distribution. Continuous data were presented as means (standard deviations) or as medians (Q1–Q3). Continuous variables were compared using Student’s t-test or Mann–Whitney U test, as appropriate in the case of 2 groups, and ANOVA or Kruskal–Wallis tests when more than 2 groups were compared. Post hoc Dunn’s or Tukey’s tests were used, as appropriate. Categorical data were presented as numbers (percentages), and Fisher’s exact test was used to compare them. Spearman correlation was used to check for associations between continuous variables. The effects of variables on the endpoints were evaluated using Cox proportional hazard models. Results were presented as HR with 95% confidence intervals (CI). CIs for area under the curve scores were calculated using DeLong’s method. Kaplan–Meier survival curves were plotted. Cox proportional hazard models were used to find endpoint predictors. Multivariable Cox proportional hazard models were used to evaluate HR adjusted for confounders. We chose possible confounders from parameters that differed between patients with and without the endpoints based on the results of ANOVA or Kruskal–Wallis tests; we included age and sex as additional confounders. The Akaike information criterion was used for selection of best models. We established optimal cut-off points for PC values predicting endpoints based Youden’s J statistic. Statistical analyses were performed using Python and R libraries. A level of significance of 0.01 was used for the normality of data distribution and assumptions for Cox analysis. Otherwise, a level of significance <0.05 was considered statistically significant.

RESULTS

We analyzed 178 patients, aged 64.0 (57.0–70.0) years, 75.8% male (Table 1). Two patients (1.1%) were excluded due to C-reactive protein >100 mg/l. Most of the patients (n = 132, 74.2%) had multivessel disease while 127 (71.3%) had a history of MI or percutaneous coronary intervention (Table 1).

Table 1. Patient characteristics according to quartiles of plasma carbonylated protein content

		Carbonylated protein content, nmol/mg				P-value
	Whole group (n = 178)	Q1 (1.3–2.2)	Q2 (2.2–2.9)	Q3 (2.9–3.1)	Q4 (3.1-5.1)
Age, years	64 (57–70)	57 (53–64)	63 (57–68)	65 (60–69)	73 (64–73)	<0.001a
Male, n (%)	135 (75.8)	35 (81.4)	30 (66.7)	37 (86.1)	31 (68.9)	0.10
BMI, kg/m2	26.9 (3.9)	26.3 (4.1)	27.1 (4.1)	26.8 (3.6)	27.5 (3.6)	0.54
Comorbidities, n (%)
Smoking	57 (32.0)	19 (44.2)	14 (31.1)	11 (25.6)	13 (28.9)	0.27
Diabetes	36 (20.2)	8 (18.6)	10 (22.2)	9 (20.9)	9 (20.0)	0.98
Hypertension	133 (74.7)	31 (72.1)	33 (73.3)	32 (74.4)	35 (77.8)	0.94
Prior MI/PCI	127 (71.3)	32 (74.4)	30 (66.7)	30 (69.8)	33 (73.3)	0.85
Medications, n (%)
ACE-I/ARB	125 (70.2)	32 (74.4)	30 (66.7)	26 (60.5)	35 (77.8)	0.29
Statins	156 (87.6)	41 (95.4)	40 (88.9)	36 (83.7)	37 (82.2)	0.24
Laboratory parameters
WBC, 103/μl	6.6 (5.5–8.3)	6.6 (5.3–8.1)	7.1 (6.4–8.6)	6.1 (4.9–7.4)	6.4 (5.5–7.9)	0.06
Hemoglobin, g/dl	13.6 (1.4)	13.5 (1.3)	13.8 (1.5)	13.7 (1.3)	13.6 (1.5)	0.77
Creatinine, μmol/l	79 (66–90)	82 (67–89)	84 (67–95)	73 (63–87)	76 (67–85)	0.25
TC, mmol/l	4.4 (3.7–5.3)	4.6 (3.7–5.7)	4.1 (3.6–4.7)	4.3 (3.6–5.3)	4.5 (3.8–5.2)	0.42
LDL-C, mmol/l	2.5 (1.9–3.4)	2.9 (1.9–3.6)	2.5 (2.1–3.1)	2.5 (2.0–3.4)	2.7 (1.9–3.4)	0.79
HDL-C, mmol/l	1.2 (1.0–1.4)	1.3 (1.1–1.4)	1.2 (1.1–1.3)	1.1 (1.0–1.3)	1.2 (1.0–1.3)	0.38
Glucose, mmol/l	5.3 (5.0–5.9)	5.2 (5.0–5.8)	5.5 (5.0–6.0)	5.4 (5.1–6.0)	5.2 (4.9–5.6)	0.42
Fibrinogen, g/l	3.3 (2.6–4.3)	3.1 (2.5–4.1)	3.2 (2.8–3.8)	3.5 (2.7–4.6)	3.5 (2.8–4.7)	0.57
8-iso-PGF2α, pg/ml	346 (279–423)	297 (251–330)	323 (280–367)	395 (321–458)	376 (279–455)	<0.001c
Fibrin clot properties and associated proteins
Ks, 10-9 cm2	6.6 (0.9)	7.1 (0.9)	6.5 (0.9)	6.4 (0.8)	6.4 (1.0)	<0.001a
CLT, min	103.3 (18.0)	92.0 (13.8)	103.3 (13.7)	107.45 (16.8)	111.1 (20.2)	<0.001a
TAFI, %	100 (91–111)	96 (86–103)	100 (91–107)	100 (94–113)	106 (96–118)	0.005b
PAI-1, ng/ml	51.1 (13.0)	43.3 (11.9)	51.4 (11.3)	54.6 (11.4)	55.4 (13.1)	<0.001a

Continuous data were shown as mean (SD) or median (Q1–Q3)

aPost hoc test: significant differences: Q1 vs. Q2, Q1 vs. Q3 and Q1 vs. Q4. bPost hoc analysis: significant differences: Q1 vs. Q3 and Q1 vs. Q4. cPost hoc analysis: significant differences: Q1 vs. Q3, Q1 vs. Q4 and Q2 vs. Q3

Abbreviations: 8-iso-PGF2α, 8-iso-prostaglandin F2α; ACE-I, angiotensin-converting enzyme inhibitor; ARB, angiotensin receptor blockers; BMI, body mass index; CLT, clot lysis time; HDL-C, high-density lipoprotein cholesterol; Ks, permeation coefficient; LDL-C, low-density lipoprotein cholesterol; MI, myocardial infarction; PAI-1, plasminogen activator inhibitor-1; PCI, percutaneous coronary intervention; TAFI, thrombin-activatable fibrinolysis inhibitor; TC, total cholesterol; WBC, white blood cells

The median PC content in the cohort was 2.9 (2.2–3.7) nmol/mg protein while 147 (82.6%) patients had PC above the upper reference limit (>2.03 nmol/mg protein). There were no differences in clinical, demographic, or routine laboratory parameters between patients with PC in different quartiles, except age, which was associated positively with PC (r = 0.50; P <0.001; Supplementary material, Figure S2). PC levels correlated positively with 8-iso-PGF2α (r = 0.33; P <0.001).

Patients with PC in the highest quartile had lower Ks and longer CLT, compared with those in the lowest one (Table 1). Subjects in the top PC quartile had higher PAI-1 and TAFI than those in the bottom quartile (Table 1). PC levels were inversely associated with Ks (r = –0.26; P <0.001; Figure 1A) and positively with CLT (r = 0.30; P <0.001; Figure 1B), along with its determinants, i.e., PAI-1 (r = 0.33; P <0.001; Figure 1C) and TAFI (r = 0.32; P <0.001; Figure 1D).

Figure 1. Correlation charts with regression lines, 95% confidence intervals. Association between carbonyl protein content and A. Permeation coefficient (Ks). B. Clot lysis time (CLT). C. Plasminogen activator inhibitor 1 antigen (PAI-1 Ag). D. Thrombin-activatable fibrinolysis inhibitor antigen (TAFIAg)

Follow-up

During a follow-up of 8.3 (1.8) years, the composite endpoint occurred in 67 (37.6%) patients, (0.43 per 100 patient-years), including 35 with MI, 25 with IS plus SE, and 30 with CV death. Eight patients experienced MI and CV death, 12 patients had IS and died of CV causes, 2 had SE and died of CV causes, and 1 patient experienced IS, SE, and CV death (Supplementary material, Figure S1). Patients with and without the composite endpoint were similar, except for higher body mass index values and hypertension prevalence in the former group (Table 2).

Table 2. Patient characteristics based on the occurrence of the composite endpoint during follow-up

	Composite endpoint		P-value
	Yes (n = 67)	No (n = 111)
Age, years	66 (57–73)	63 (56–69)	0.08
Male, n (%)	53 (79.1)	82 (73.9)	0.43
BMI, kg/m2	27.9 (3.5)	26.4 (4.0)	0.009
Smoking, n (%)	21 (31.3)	36 (32.4)	0.88
Comorbidities, n (%)
Diabetes	14 (20.9)	22 (19.8)	0.83
Hypertension	56 (83.6)	77 (69.4)	0.03
Prior MI or PCI	52 (77.6)	75 (67.6)	0.15
Medications, n (%)
ACE-I	51 (76.1)	74 (66.7)	0.18
Statins	55 (82.1)	101 (91.0)	0.08
Laboratory parameters
White blood cells, 103/μl	6.7 (5.3–8.5)	6.5 (5.5–8.1)	0.64
Hemoglobin, g/dl	13.8 (12.7–14.8)	13.8 (12.5–14.4)	0.37
Creatinine, μmol/l	76.2 (65.2–94.0)	79.1 (66.0–89.1)	0.80
CRP, mg/l	2.2 (1.4–3.6)	2.0 (1.2–3.5)	0.49
TC, mmol/l	4.8 (4.0–5.8)	4.2 (3.4–5.0)	0.001
LDL-C, mmol/l	2.9 (2.1–3.8)	2.4 (1.9–3.2)	0.013
HDL-C, mmol/l	1.2 (1.0–1.4)	1.2 (1.0–1.3)	0.29
Glucose, mmol/l	5.3 (4.9–6.0)	5.3 (5.0–5.8)	0.54
Fibrinogen, g/l	3.3 (2.5–4.5)	3.2 (2.7–4.3)	0.87
Fibrin clot properties and associated proteins
Ks, 10–9 cm2	6.4 (0.9)	6.8 (1.0)	0.009
CLT, min	108 (100–127)	99 (89–107)	<0.001
TAFI Ag, %	103.9 (96.7–113)	100 (87–110)	0.023
PAI-1, ng/ml	52.9 (13.6)	49.9 (12.5)	0.14
Total PC content, nmol/mg protein	3.7 (2.9–4.1)	2.5(2.1–3.2)	<0.001

Continuous data were shown as mean (SD) or median (Q1–Q3)

Abbreviations: CRP, C-reactive protein; thrombin-activatable fibrinolysis inhibitor antigen; other — see Table 1

Patients with the composite endpoint had lower Ks and longer CLT, along with higher TAFI. PC levels were 48% higher in the group with the composite endpoint (Table 2). After adjustment for potential confounders, PC in the highest quartile remained a predictor of the composite endpoint (HR, 4.89; 95% CI, 2.12–11.27; P <0.001; Figure 2A). An increase in PC by 1 nmol/mg was associated with a 2.2-fold higher risk of the composite endpoint on multivariable analysis (Supplementary material, Table S3).

Figure 2. Kaplan–Meier curves showing the probability of event-free survival with regard to carbonylated protein content, after adjustment for potential confounders, in patients with coronary artery disease

Abbreviations: CI, confidence interval; HR, hazard ratio

We recorded MI in 35 (19.7%) patients, 0.20 per 100 patient-years. The PC content in this group was 38.5% higher compared with other groups (Supplementary material, Table S2). After adjustment for age, sex, and LDL-C, PC in the highest quartile was associated with higher MI risk (HR, 6.25; 95% CI, 1.70–23.01; P = 0.006; Figure 2B). On multivariable analysis, PC remained an independent predictor of MI (Supplementary material, Table S3).

Twenty-five (14.0%) patients experienced IS/SE (0.14 per 100 patient-years). Baseline PC were 44.4% higher in patients with an IS/SE compared to others (Supplementary material, Table S2). After adjustment for age and sex, PC in the top quartile was associated with IS/SE risk compared to the bottom quartile (HR, 21.37; 95% CI, 2.67–170.96; P = 0.004; Figure 2C), and this association remained signi­ficant on multivariable analysis (Supplementary material, Table S3).

Thirty patients (16.9%) died of CV causes (0.17 per 100 patient-years). PC content and prolonged CLT distinguished these patients from others (Supplementary material, Table S2). After adjustment for age and sex, PC in the highest quartile was associated with the risk of death from CV causes (HR, 4.87; 95% CI, 1.46–26.21; P = 0.01; Figure 2D). On multivariable analysis, PC remained independently associated with CV mortality (Supplementary material, Table S3).

Based on ROC curves (Supplementary material, Figure S3), the optimal cut-off value for baseline PC was 3.03 nmol/mg protein with an area under the curve of 0.753; 95% CI, 0.678–0.827 for the composite endpoint.

DISCUSSION

This study is the first to show that in patients with stable CAD, high plasma PC content can predict MI, IS, SE, and CV death, analyzed separately and as a composite endpoint, in long-term follow-up. The composite endpoint was mostly driven by MI; however, elevated PC had the strongest impact on IS/SE risk, with a 3.8-fold risk increase per 1 nmol/mg of protein. Our results suggest that persistent protein carbonylation contributes to a prothrombotic state in CAD. A novel finding in stable CAD is that the extent of protein carbonylation is associated with low clot permeability and prolonged CLT, along with elevated concentrations of PAI-1 and TAFI, which indicates prothrombotic and antifibrinolytic effects of this protein modification. Our study provides further evidence that in advanced CAD complex mechanisms governing carbonylation persist and predispose to CV events, largely thromboembolic in nature.

Our population was similar to those reported in large registries [10, 11]. In the seminal study by Becatti et al. [6], patients aged 71 (59–77) years, 69% of men, 6 months after an MI, had PC levels of 2.87 (1.02) nmol/mg [5]. This mean value was almost identical to the median of 2.9 (2.2–3.7) nmol/mg protein in our cohort with comparable demographics. In our study, PC content correlated with age, which is consistent with previous data [12]. Redox imba­lance has been implicated in several atherosclerosis-related diseases such as hypertension [13], hypercholesterolemia [14], diabetes [15], and obstructive sleep apnea [16]. We found no such associations, which is most likely attributable to advanced CAD in our group, with over 74% of patients with multivessel disease, which implies (via the abundant atherosclerosis burden) substantial reactive oxygen species generation.

Mechanisms leading to the persistent presence of circulating PC in CAD are largely unknown. Free radicals generated in a low-grade inflammatory state typical of CAD can cause carbonylation. It might be speculated that in atherosclerosis, protein carbonylation is the most suitable marker of the detrimental impact of oxidative stress on circulating proteins contributing to cardiovascular events.

We have provided further evidence that elevated PC enhances the prothrombotic state in CAD. We observed altered fibrin clot properties including decreased permeability and resistance to lysis in association with PC, and such a prothrombotic clot phenotype was reported in chronic and acute coronary syndromes [17]. Since fibrinogen concentration and function are key determinants of fibrin clot properties [18], it might be speculated that enhanced fibrinogen carbonylation largely contributes to the prothrombotic clot phenotype in plasma-based assays. Paton et al. have shown that in MI patients with PC in the top quartile, fibrinogen polymerization was 1.4-fold faster and gave 1.4 times higher maximum turbidity compared with those in the bottom quartile, reflecting faster lateral fibrin aggregation [18]. It is likely that similar reactions, though of lesser intensity, could be observed in stable CAD since fibrinogen is particularly susceptible to oxidative modifications [19]. Fibrinogen carbonylation contributes to the formation of fibrin clots that are resistant to tissue plasminogen activator-induced lysis, an effect attributable to lysine carbonylation and subsequent modification of the binding sequence for plasminogen [20]. Such impaired plasminogen-fibrin interactions in enhanced PC have been demonstrated in patients after a venous thromboembolism patients [21]. As concluded by de Vries et al. [19], the effects of fibrinogen oxidation on clot properties in vitro vary depending on the type of oxidation trigger used and oxidant concentrations.

In our study, elevated PC was weakly associated with higher PAI-1 and TAFI, well-established modulators of clot lysability [22], and longer CLT, which represents impaired global fibrinolysis. Similar associations have been recently reported in acute IS [7]. Oxidative stress was linked to elevated PAI-1 expression [23]. Moreover, lysine carbo­nylation could decrease the number of cleavage sites for both plasmin and tissue-type plasminogen activators [24], leading to impaired fibrinolysis.

The reasons for a rise in TAFI with elevated PC are unclear. TAFI circulates in the plasma bound to plasminogen and is activated by the thrombin/thrombomodulin complex, which cleaves TAFI on Arg92 [25]. Since arginine is prone to carbonylation, carbonylation could alter the cleavage site for thrombin and facilitate its dissociation from plasminogen, which would lead to a detectable higher concentration. Our unexpected observations deserve further mechanistic studies.

A positive correlation between PC and 8-iso-PGF2α confirms that both markers reflect enhanced oxidative stress in advanced CAD. Protein carbonylation represents redox imbalance over longer periods, while isoprostanes reflect short-term effects.

The current study has several limitations. First, the sample size was relatively small; however, it was sufficient to show the assumed effect, according to power calculation. The patients in this study used similar pharmacotherapy, with almost 90% on statins, which did not allow for analysis of these drugs as potential confounders. Also, modifications of pharmacotherapy during the follow-up period could have affected the results. The mechanistic basis for the observed associations between PC levels and impaired fibrinolysis were not investigated in our study. We have, however, shown positive correlations between PC and PAI-1, as well as TAFI, which are 2 key antifibrinolytic proteins. In addition, an association between PC and impaired fibrinolysis was reported for MI and IS survivors [6, 7]. We did not determine carbonylated fibrinogen, however, a positive association between overall and fibrinogen carbonylation has already been documented [6].

To conclude, this study shows a novel, independent association between plasma PC content, a stable marker of oxidative stress, and the risk of atherothrombotic events in stable CAD, likely linked to prothrombotic alterations of the fibrin clot phenotype. The current management of advanced CAD may not substantially reduce these processes, therefore, plasma PC concentrations could be perceived as a specific type of “residual risk” to be treated differently.

Supplementary material

Supplementary material is available at https://journals.viamedica.pl/polish_heart_journal.

Article information

Conflict of interest: None declared.

Funding: This study was supported by the Jagiellonian University Medical College in Krakow (N41/DBS/000682 to AU).

Open access: This article is available in open access under Creative Common Attribution-Non-Commercial-No Derivatives 4.0 International (CC BY-NC-ND 4.0) license, which allows downloading and sharing articles with others as long as they credit the authors and the publisher, but without permission to change them in any way or use them commercially. For commercial use, please contact the journal office at polishheartjournal@ptkardio.pl