- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
Atherosclerosis is an inflammatory disease characterized by vascular inflammation and a systemic inflammatory state (1, 2). Inflammation is involved in all stages of atherothrombosis, from the induction of endothelial dysfunction to plaque formation, plaque destabilization, and subsequent thrombosis.
The levels of systemic inflammatory biomarkers, including C-reactive protein (CRP) and tumor necrosis factor α (TNFα), are related to the severity of cardiovascular disease. These biomarkers predict future cardiovascular events in patients with cardiovascular disease, as well as in apparently healthy individuals (3). The role of systemic inflammation in atherosclerosis has not yet fully been elucidated. Systemic inflammation may be secondary to vascular inflammation and/or underlying cardiovascular risk factors (2). Systemic inflammatory biomarkers may also promote atherothrombosis (2, 4). Because cardiovascular morbidity is elevated in various disorders associated with chronic inflammation, including inflammatory rheumatic diseases (IRDs), chronic infections, diabetes mellitus, and obesity, a causal relationship between chronic inflammation and cardiovascular disease has been suggested (4).
Pentraxin 3 (PTX3) is a newly discovered marker of the acute-phase inflammatory response, and plays an important role in innate immunity (5, 6). Similar to CRP, PTX3 belongs to the PTX protein family (7). PTX3 is considered the prototype of the long PTX subfamily due to its long N-terminal domain and a C-terminal domain homologous to CRP. Unlike CRP, which is produced in the liver in response to interleukin-6 (IL-6), PTX3 is mainly produced at extrahepatic sites by several cell types, including cells of the myelomonocyte lineage (monocytes, macrophages, dendritic cells), endothelial cells, smooth muscle cells, fibroblasts, and adipocytes. PTX3 is also produced during neutrophil differentiation and stored in specific granules of mature neutrophils, ready to be released upon microbial recognition (8). Basal levels of PTX3 are generally undetectable, but protein levels are rapidly induced in response to proinflammatory signals (IL-1, TNFα, modified lipoproteins) and Toll-like receptor engagement. PTX3 has multiple complex nonredundant functions, ranging from assembly of a hyaluronic acid–rich extracellular matrix and female fertility to protection against pathogens (i.e., Aspergillus fumigatus, influenza viruses) (9–13). PTX3 also regulates the clearance of apoptotic cells and may participate in maintenance of immunologic tolerance.
Similar to CRP, increased levels of circulating PTX3 have been found in different human pathologies, including autoimmune, infectious, and degenerative disorders. One major difference between PTX3 and CRP is the rapidity of the PTX3 plasma level increase, likely reflecting the expression of PTX3 protein by different cell types compared with the liver-restricted production of CRP. This rapid induction is particularly evident in patients with acute myocardial infarction (MI), in which PTX3 plasma levels reach a peak within 6–8 hours after the onset of symptoms (14). Compared with established cardiovascular biomarkers, including CRP, PTX3 is the only independent predictor of mortality in acute MI (15). Circulating PTX3 also predicts cardiovascular and all-cause mortality independently of CRP and cardiovascular risk factors in older adults free of cardiovascular disease (16). PTX3 levels are higher in unstable compared with stable angina, possibly reflecting arterial inflammation (17). Compared with CRP, PTX3 was also found to be a more powerful predictor of restenosis after coronary artery bypass graft (CABG) and a more powerful predictor of poor outcome in heart failure (18, 19). Elevated PTX3 levels were observed in patients with rheumatoid arthritis (RA) and small-vessel vasculitis (20, 21), and in the latter case, PTX3 levels correlated with disease activity and response to therapy (21).
Data thus far strongly suggest that circulating PTX3 could function as a fast responding biomarker of cardiovascular pathologies. Therefore, we wanted to compare the levels of serum PTX3 in patients with coronary artery disease (CAD) and IRD, in patients with CAD without IRD, and in healthy controls.
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
Patient characteristics are shown in Table 1. The CAD/IRD group consisted of patients with RA (n = 24), psoriatic arthritis (n = 9), systemic lupus erythematosus (SLE; n = 3), ankylosing spondylitis (n = 6), reactive arthritis (n = 1), giant cell arteritis (GCA; n = 6), polymyalgia rheumatica (PMR; n = 15), primary Sjögren's syndrome (n = 1), and undifferentiated connective tissue disease (n = 1). Most patients with IRD included in our study had low to moderate disease activity. The distribution of traditional cardiovascular risk factors was similar between the CAD/IRD and CAD/non-IRD groups, but the CAD/IRD patients had higher levels of inflammatory biomarkers. Use of medications was mostly similar in both groups, except for more frequent use of disease-modifying antirheumatic drugs (DMARDs) and steroidal and nonsteroidal antiinflammatory drugs (NSAIDs) in the CAD/IRD group (Table 1).
Table 1. Patient characteristics*
| ||CAD/IRD (n = 66)||CAD/non-IRD (n = 52)||Healthy controls (n = 30)||P†|
|Age, mean ± SD years||67 ± 10||68 ± 10||57 ± 9||< 0.001‡|
|Men||41 (62)||34 (65)||17 (57)||0.735|
|Duration of CAD, mean ± SD years||6 ± 7||6 ± 6||–||0.848|
|History of myocardial infarction||38 (58)||23 (44)||0 (0)||< 0.001|
|Acute coronary syndrome||18 (27)||10 (19)||0 (0)||0.007|
|Time from angiography to CABG, mean ± SD days||13 ± 20||27 ± 44||–||0.023§|
|Left ventricular ejection fraction, mean ± SD||64 ± 12||66 ± 11||–||0.389|
|C-reactive protein level, mean ± SD mg/liter||11.5 ± 22.6||3.7 ± 4.8||1.1 ± 0.5||0.003¶|
|ESR, mean ± SD mm/hour||31 ± 24||16 ± 10||4 ± 3||< 0.001#|
|Body mass index, mean ± SD kg/m2||25.5 ± 4.3||25.7 ± 3.3||–||0.832|
|Hypertension||39 (60)||30 (58)||0 (0)||< 0.001|
|Family history of CAD||48 (74)||42 (81)||–||0.377|
|Hyperlipidemia||55 (83)||46 (89)||0 (0)||< 0.001|
|Diabetes mellitus**||8 (12)||6 (12)||0 (0)||0.139|
|Previous smoker||26 (39)||25 (48)||–||0.345|
|Current smoker||15 (23)||7 (14)||–||0.199|
|Current use|| || || || |
| Oral glucocorticoids**||27 (41)||0 (0)||0 (0)||< 0.001¶|
| DMARDs††||20 (31)||0 (0)||0 (0)||< 0.001¶|
| Cyclooxygenase 2 selective inhibitors**||11 (17)||0 (0)||0 (0)||0.001¶|
| Traditional NSAIDs**||8 (12)||1 (2)||0 (0)||0.029¶|
| Lipid-lowering drugs**||49 (75)||42 (81)||0 (0)||< 0.001|
| Acetylsalicylic acid**||56 (85)||47 (90)||0 (0)||< 0.001|
| Beta-blockers**||49 (74)||42 (81)||0 (0)||< 0.001|
| ACE inhibitors||21 (32)||18 (35)||0 (0)||0.001|
|Duration of IRD, mean ± SD years||16 ± 13||–||–||–|
|Patient global assessment of IRD (VAS), mean ± SD cm||3.1 ± 2.6||–||–||–|
The CAD/IRD group had significantly higher serum PTX3 levels than the CAD/non-IRD group (95% confidence interval [95% CI] 0.2, 0.9; P = 0.001) and the healthy controls group (95% CI 0.3, 1.2; P < 0.001). The difference in serum PTX3 levels between the CAD/ non-IRD and healthy controls groups was not statistically significant (P = 0.92) (Table 2).
Table 2. Mean values of serum PTX3 level in the diagnostic subgroups*
| ||Serum PTX3 level, ng/ml|
|Healthy controls||1.21 ± 0.59|
|CAD/non-IRD||1.41 ± 0.74|
|CAD/IRD||1.96 ± 0.98|
| RA||2.08 ± 0.99|
| Ankylosing spondylitis||2.48 ± 1.07|
| Psoriatic arthritis||1.79 ± 0.80|
| Reactive arthritis, mean||0.33|
| PMR||2.08 ± 0.95|
| GCA||1.98 ± 1.05|
| SLE||1.03 ± 0.84|
| Primary Sjögren's syndrome, mean||1.03|
| Undifferentiated connective tissue disease, mean||0.70|
We did not find any statistically significant differences in the levels of serum PTX3 between the respective rheumatic diagnoses (Table 2 and Figure 1). Except for patients with SLE, Sjögren's syndrome, undifferentiated connective tissue disease, and reactive arthritis, who had even lower serum PTX3 levels than the healthy controls group, patients with IRD had higher mean and median serum PTX3 levels than CAD/non-IRD patients. We compared RA, GCA/PMR, spondylarthritis, and connective tissue disease subgroups with the healthy controls and CAD/non-IRD groups and found significant differences between the RA subgroup and the healthy controls and CAD/ non-IRD groups (mean difference 0.87; P = 0.002, and mean difference 0.68; P = 0.016, respectively), and between the GCA/PMR subgroup and the healthy controls and CAD/non-IRD groups (mean difference 0.65; P = 0.042, and mean difference 0.84; P = 0.006, respectively).
Figure 1. Serum pentraxin 3 (PTX3) in inflammatory rheumatic disease (IRD) subgroups compared with healthy controls (HC) and CAD-nonIRD groups. The lines inside of the boxes show the median; the edges of the boxes show the upper and lower quartiles. CAD = coronary artery disease; RA = rheumatoid arthritis; PMR = polymyalgia rheumatica; GCA = giant cell arteritis; SLE = systemic lupus erythematosus.
Download figure to PowerPoint
We next analyzed the relationship between serum PTX3 levels and the following factors: history of IRD, CAD, diabetes mellitus, hypertension, hyperlipidemia, age, sex, education, marital status, body mass index, family history of CAD, exercise, current cigarette smoking, and alcohol consumption; length of time between angiography and CABG surgery; duration of CAD; high-density lipoprotein (HDL), low-density lipoprotein (LDL), and total cholesterol levels; current use of acetylsalicylic acid, statins, angiotensin-converting enzyme inhibitors, NSAIDs, coxibs, systemic corticosteroids, and DMARDs; and the presence of acute coronary syndrome. Of these variables, IRD, CAD, age, diabetes mellitus, acute coronary syndrome, and alcohol intake were significant at P < 0.005, and exercise ≥1 hour/week was significant at P < 0.01 for the effect on serum PTX3 levels in simple regression analyses. We performed 2 multiple regression models: model I was adjusted for age, sex, IRD, CAD, current acute coronary syndrome, and diabetes mellitus, and model II was adjusted for the same variables as model I, as well as exercise and alcohol consumption. IRD and acute coronary syndrome had consistent and significant effects on serum PTX3 levels in both models. Alcohol consumption showed a protective effect on serum PTX3 levels (Table 3). Age was related to serum PTX3 levels in model I but not in model II.
Table 3. Clinical predictors of serum PTX3 levels*
| ||Unadjusted||Adjusted: model I†||Adjusted: model II‡|
|β||95% CI||P||β||95% CI||P||β||95% CI||P|
|Age||0.02||0.009, 0.035||0.001||0.02||0.003, 0.03||0.020||0.01||−0.01, 0.03||0.206|
|Sex||0.04||−0.26, 0.34||0.78||−0.03||−0.31, 0.24||0.810||−0.20||−0.59, 0.19||0.304|
|IRD||0.62||0.35, 0.90||< 0.001||0.53||0.23, 0.82||0.001||0.43||0.09, 0.77||0.014|
|CAD||0.51||0.16, 0.86||0.005||−0.11||−0.51, 0.29||0.573||–||–||–|
|Acute coronary syndrome||0.33||0.13, 053||0.001||0.48||0.13, 0.83||0.008||0.44||0.02, 0.87||0.043|
|Diabetes mellitus||0.54||0.05, 1.03||0.030||0.31||−0.13, 0.77||0.177||0.16||−0.37, 0.68||0.56|
|Exercise ≥1 hour/week||−0.34||−0.68, 0.01||0.054||–||–||–||−0.20||−0.56, 0.16||0.274|
|Alcoholic drinks, unit/week||−0.18||−0.33, −0.04||0.016||–||–||–||−0.07||−0.12, −0.01||0.018|
IRD predicted serum PTX3 levels independently of CRP level and ESR (determined in analyses adjusted for IRD and CRP level or ESR, as well as in analyses adjusted for the same variables used in model I and CRP level or ESR). In simple regression analyses, ESR and CRP level correlated with serum PTX3 levels; however, after adjustment for CRP level, the effect of ESR was not statistically significant. In contrast, the effect of CRP level remained statistically significant after adjustment for the variables used in model I. In subgroup analyses, CRP level correlated with serum PTX3 level in the CAD/IRD group (β = 1.80, P = 0.012) but not in the healthy controls (P = 0.14) or CAD/non-IRD (P = 0.35) groups. We also performed analyses using combined variables of CRP for CRP levels ≥1 mg/ml and high-sensitivity CRP for CRP levels <1 mg/ml, but the results were similar to those for CRP only.
Among the patients with CAD, those with acute coronary syndromes had significantly higher serum PTX3 levels than those with stable CAD (2.2 versus 1.6 ng/ml; P = 0.002). In separate analyses of CAD/IRD and CAD/non-IRD groups, the difference reached statistical significance only in the CAD/non-IRD group (Figure 2).
Figure 2. Serum pentraxin 3 (PTX3) levels in patients with and without acute coronary syndromes. The lines inside of the boxes show the median; the edges of the boxes show the upper and lower quartiles. CAD = coronary artery disease; IRD = inflammatory rheumatic disease; * = P < 0.05.
Download figure to PowerPoint
Current use of DMARDs and systemic glucocorticoids and the dose of glucocorticoids positively correlated with serum PTX3 levels in simple analyses, but no correlations were detected after adjustment for IRD.
We examined associations between serum PTX3 levels and disease characteristics and treatment in the following subgroups: RA, spondylarthritis, GCA/PMR, and connective tissue diseases. We did not find any statistically significant associations between serum PTX3 levels and current use of all DMARDs, number of DMARDs used, systemic glucocorticoid treatment, patient and physician global assessment, visual analog scale (VAS) pain, numbers of swollen and tender joints, extraarticular manifestations (in chronic arthritides), or IRD duration. Levels of serum PTX3 positively correlated with VAS fatigue in the GCA/PMR subgroup (β = 0.15, P = 0.027) and with the modified Health Assessment Questionnaire (M-HAQ) score in the RA subgroup (β = 0.67, P = 0.025). In age-, sex-, and CAD-adjusted analyses, RA, spondylarthritis, and GCA/PMR diagnosis predicted serum PTX3 levels (β = 0.70, P < 0.001; β = 0.71, P = 0.003; and β = 0.58, P = 0.006, respectively).
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
In our study, patients with CAD and IRD had higher mean serum PTX3 levels than CAD patients without IRD or healthy controls. This effect was independent of age, sex, and other examined variables. CAD patients without IRD had higher serum PTX3 levels than healthy controls, but the difference was not statistically significant. Nevertheless, current acute coronary syndromes were independently related to higher serum PTX3 levels.
Because serum PTX3 is a marker of cardiovascular outcome in general, elevated serum PTX3 levels in patients with IRD may indicate adverse cardiovascular prognosis. This is consistent with previous observations, in which patients with RA have both a higher occurrence and severity of acute coronary syndromes compared with patients without RA (29, 30).
Patients with CAD and RA, ankylosing spondylitis, psoriatic arthritis, or GCA/PMR had higher mean and median serum PTX3 levels than CAD patients without IRD or healthy controls (the differences were statistically significant for the RA and GCA/PMR subgroups). In RA, and especially in GCA/PMR and ankylosing spondylitis, vasculitis is likely a relatively common, and often unrecognized, complication (22, 31–33). Therefore, the higher serum PTX3 levels observed with these diseases may be associated with the subclinical vascular inflammation (22, 34). PTX3 depositions have been observed in vessels of patients with leukocytoclastic small-vessel vasculitis (35).
Patients with CAD and SLE, Sjögren's syndrome, undifferentiated systemic tissue diseases, or reactive arthritis had lower mean serum PTX3 levels than the healthy controls and CAD/IRD groups. The patient with reactive arthritis did not have any disease flare during the 20 years prior to the CABG. The low serum PTX3 levels in patients with SLE may reflect a phenomenon similar to that observed for CRP level, which also tends to remain low in patients with active SLE. Low PTX3 levels in SLE patients were also observed in the study by Fazzini et al (21). The low serum PTX3 levels in SLE patients could be due to reduced protein production or increased elimination (e.g., by binding to apoptotic cells). In addition, PTX3 may be deregulated in the presence of anti-PTX3 antibodies. Dysregulation of PTX3 may contribute to breaking of immunologic tolerance, and therefore to the pathogenesis of SLE (36). However, the number of patients with systemic connective tissue diseases or reactive arthritis in our study was low (5 patients in total), and thus not suitable for drawing definitive conclusions.
Most IRD patients included in our study had low to moderate disease activity. Compared with the patients with vasculitis from the study by Fazzini et al, our patients with RA, PMR/GCA, psoriatic arthritis, or ankylosing spondylitis had slightly higher mean serum PTX3 levels than patients with quiescent systemic small-vessel vasculitis, but lower than those with active systemic small-vessel vasculitis (21). The mean level of serum PTX3 in non-IRD patients with CAD was similar to that in patients with quiescent vasculitis from the study by Fazzini et al. Our serum PTX3 level measurements and those by Fazzini et al were performed using the same technique in the same laboratory.
Except for a positive correlation with fatigue in the GCA subgroup, and with the M-HAQ score in the RA subgroup, serum PTX3 levels were not statistically significantly associated with IRD duration, treatment, or markers of IRD activity and severity.
In theory, serum PTX3 levels may be influenced by tissue injury during angiography. However, we did not find any association between serum PTX3 levels and the mean time between angiography and CABG, when the blood samples in the CABG groups were collected.
We found an inverse dose-related relationship between alcohol intake and serum PTX3 levels. Alcohol consumption has been observed to be protective against cardiovascular disease as well as RA and other conditions (37, 38). Thus, the lower serum PTX3 levels in patients with a higher intake of alcohol may be related to their lower cardiovascular risk.
Levels of serum PTX3 were statistically significantly related to age in simple regression analysis and in multiple regression model I, but not in model II. The significant effects of IRD, acute coronary syndrome, and alcohol consumption on serum PTX3 were independent of age.
Although PTX3 synthesis may be induced by oxidized LDL and HDL, we did not observe any correlation between serum PTX3 levels and serum levels of total cholesterol, LDL, or HDL. This lack of association may be due to several factors, such as a Type II error or the importance of the structure and function of the lipoproteins on their effects on serum PTX3 levels.
IRD predicted serum PTX3 levels independently of ESR and CRP level. CRP level correlated with the PTX3 levels, but in subgroup analyses, the association was apparent only in the CAD/IRD group.
PTX3 may have contrasting effects on inflammation: data obtained with genetically modified animal models have shown both an enhancement and inhibition of the inflammatory process (10, 39, 40). In humans, data collected in different pathologies thus far indicate a correlation between levels of circulating PTX3 and severity of disease, suggesting a possible role for PTX3 as a marker of pathology. Whether PTX3 is only a marker of the severity of cardiovascular disease (e.g., secondary to atherosclerotic process or to a cardiovascular risk factor) or if it has a pathogenic role is still unknown. As it up-regulates tissue factor, an important trigger of coagulation, and inactivates fibroblast growth factor 2, an important regulator of angiogenesis, PTX3 may potentially be involved in atherothrombosis (41, 42). In addition, PTX3 has been observed in atherosclerotic lesions (43, 44). In a murine model of acute MI, PTX3 depletion resulted in exacerbated heart damage, suggesting that PTX3 could play a cardioprotective role (45). Therefore, PTX3 produced during the inflammatory response may function in the restriction of cardiovascular damage, whereas the cardiovascular damage itself impairs the patient prognosis.
The cause of the increased serum PTX3 levels is unknown. PTX3 is predominantly released by inflamed tissue in response to proinflammatory cytokines (TNFα and IL-1β), microbial components, or biochemical substances (such as HDL and oxidized LDL); therefore, the elevated PTX3 serum levels may reflect local inflammation, e.g., in the joints or in the vessels and heart (46). Furthermore, PTX3 may also be released from circulating neutrophils. Since PTX3 levels might be modified by treatment (e.g., dexamethasone), prospective studies should examine the effect of the PTX3 reduction on the cardiovascular outcome (47, 48).
Circulating PTX3 is a more powerful predictor of cardiovascular outcome in the general population than CRP level (49), and might also have similar importance in IRD. However, similar to CRP, the role of PTX3 might differ among various IRDs, particularly in SLE and related disorders compared with other IRDs.
This study has several potential limitations. The number of patients was too small for detailed explorations of associations and subgroup analyses; therefore, the lack of statistical significance in some results might be caused by Type II errors that may have masked true differences. Nevertheless, our sample size is still relatively large compared with other published studies of PTX3 in IRD. Furthermore, to our knowledge, this is the only study comparing circulating PTX3 levels in rheumatic and nonrheumatic patients with CAD. Another limitation of our study is the heterogeneity of the IRD group, including various rheumatic diseases. On the other hand, this heterogeneity allowed us to provide information on the trends in the entire IRD group as well as in the respective diagnostic subsets. The healthy controls group was not age matched to the CAD groups because the inclusion of completely healthy elderly volunteers was not feasible. Nevertheless, the ages of the healthy individuals were within the age range of the CAD groups. In addition, we corrected our analyses for age using statistical methods. PTX3 levels are elevated in chronic kidney disease and related to the severity of the kidney disease, as well as to the presence of cardiovascular disease and endothelial dysfunction (50). However, our results may not be generalized to patients with advanced renal failure or severe multiorgan complications, as these patients were not typically accepted for CABG at the Feiring Heart Clinic. This study is cross-sectional and therefore does not confirm any cause and effect relationships.
In conclusion, this novel study revealed higher serum PTX3 levels in patients with IRD and CAD than in patients with CAD only and in healthy controls. The higher serum PTX3 levels may reflect the high cardiovascular risk in IRDs. However, because there were statistically nonsignificant variations in the serum PTX3 levels between diagnostic subsets, PTX3 may have different clinical significance in the respective IRDs, particularly in SLE and related disorders. PTX3 levels were higher in patients with acute coronary syndromes than in those with stable CAD. Alcohol consumption had a protective effect on serum PTX3 levels. Our findings show that serum PTX3 levels could potentially be used as an early predictor of the cardiovascular prognosis.
- Top of page
- PATIENTS AND METHODS
- AUTHOR CONTRIBUTIONS
All authors were involved in drafting the article or revising it critically for important intellectual content, and all authors approved the final version to be submitted for publication. Dr. Hollan had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Study conception and design. Hollan, Bottazzi, Førre, Mikkelsen, Saatvedt, Almdahl, Mantovani, Meroni.
Acquisition of data. Hollan, Bottazzi, Cuccovillo, Førre, Almdahl, Mantovani, Meroni.
Analysis and interpretation of data. Hollan, Bottazzi, Førre, Almdahl, Mantovani, Meroni.