Predisposition to Atherosclerosis by Infections
Role of Endothelial Dysfunction
Background— Several microorganisms have been implicated in the pathogenesis of atherosclerosis. We hypothesized that infections may predispose to atherosclerosis by inflicting endothelial injury.
Methods and Results— Of 375 patients undergoing coronary angiography, 218 had assessment of endothelial function using intracoronary acetylcholine (ACH) and of endothelium-independent function with sodium nitroprusside and adenosine. Immunoglobulin-G antibody titers to cytomegalovirus, Chlamydia pneumoniae, Helicobacter pylori, hepatitis A virus, and herpes simplex virus-1 were measured. Pathogen burden was defined as the number of positive antibodies. Although positive serology to individual pathogens tended to be associated with increased incidence of coronary arteriosclerosis (CAD), the pathogen burden correlated with the presence of CAD, even after adjustment for risk factors (OR 1.3; 95% CI, 1.05 to 1.6, P=0.018). Moreover, the severity of CAD was independently associated with the pathogen burden (P=0.001). Pathogen burden was an independent predictor of endothelial dysfunction, determined as the percent change in coronary vascular resistance in response to ACH (P=0.009) but not the responses to sodium nitroprusside or adenosine. Pathogen burden was also an independent determinant of endothelial function in the subgroup with angiographically normal coronary arteries.
Conclusions— The immunoglobulin-G antibody response to multiple pathogens (pathogen burden) is an independent risk factor for endothelial dysfunction and the presence and severity of CAD. Endothelial dysfunction provides the crucial link by which pathogens may contribute to atherogenesis.
Received February 22, 2002; revision received April 22, 2002; accepted April 23, 2002.
Many patients with coronary atherosclerosis (CAD) lack conventional risk factors, suggesting that there are additional unidentified factors that contribute to vascular injury. Epidemiological studies indicate that infectious agents may predispose patients to atherosclerosis and its adverse clinical events. Organisms implicated are intracellular pathogens, including viruses such as cytomegalovirus (CMV), herpes simplex virus-1 (HSV-1), and hepatitis A virus (HAV) and bacteria such as Chlamydia pneumoniae and Helicobacter pylori.1–11⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓ Because some pathogens have been identified in atherosclerotic plaques, it has been hypothesized that they may precipitate vascular inflammation by either persistent infection12 or by immune-mediated injury.13,14⇓
See p 164
Endothelial dysfunction is one of the earliest manifestations of atherosclerosis, and it is noteworthy that certain pathogens can infect endothelial cells and increase synthesis of tissue factor, cell-surface thrombin expression, platelet adherence, and expression of adhesion molecules, cytokines, and growth factors and decrease prostacyclin release.15–19⇓⇓⇓⇓ Based on these observations and our preliminary findings that the risk of CAD correlates not only with the presence of individual infections (as determined by seropositivity) but more strongly with the aggregate number of infections (pathogen burden), we hypothesized that infection with multiple pathogens may contribute to human CAD by inflicting endothelial cell injury.7 In this investigation, we determined whether there is a relationship between the pathogen burden and severity of CAD and hypothesized that such a predisposition may be secondary to coronary vascular endothelial dysfunction.
We studied 375 patients with either CAD defined as angiographic presence of plaquing or more severe disease or angiographically smooth, normal coronary arteries (NCAs) undergoing diagnostic cardiac catheterization. Preliminary experience from a smaller cohort of 233 patients has been previously reported, and this study examines for the first time the relation between pathogen burden and disease severity and endothelial function.7 Disease severity was scored as 0 for those with smooth normal epicardial coronary arteries, 0.5 for plaquing (<50% severity), and 1, 2, or 3 for those with single-, double-, or triple-vessel epicardial coronary stenosis of >50%, respectively.
Of the 375 patients, 218 underwent measurement of vascular function. Inclusion criteria included presence of at least 1 unobstructed coronary artery (or bypassed artery). Patients with multivessel disease, recent myocardial infarction, or severe valvular heart disease were excluded. Cardiac medications were withdrawn for at least 48 hours. The protocol was approved by the Institutional Review Board of the National Heart, Lung and Blood Institute, and informed written consent was obtained.
Vascular Function Studies
A 6-French guide catheter was introduced into an unobstructed (<30% stenosis) coronary artery, and coronary blood flow (CBF) velocity was measured using a 0.014- or 0.018-inch Doppler flow wire (Cardiometrics Flowire, Endosonics Corp).20 Endothelium-dependent vasodilation was estimated by measuring CBF and epicardial responses to an infusion of intracoronary acetylcholine (ACH) at a rate of 30 μg/min for 2 minutes to obtain an estimated 10−6 mol/L intracoronary concentration. Endothelium-independent function was estimated with intracoronary sodium nitroprusside (SNP) (40 μg/min) infusion for 3 minutes and coronary flow reserve with adenosine infused at 2.2 mg/min for 2 minutes. When drugs were infused into a single coronary artery, the infusion rate was halved.
CBF was derived from the CBF velocity and diameter measurements using the formula (π×average peak velocity×0.125× diameter2).20 Coronary vascular resistance (CVR) was calculated as mean arterial pressure divided by CBF. For calculating CBF, diameter was measured in a 0.25- to 0.5-cm segment of vessel beginning 0.25 cm beyond the tip of the flow wire. Quantitative angiography was performed with the ARTREK software (Quantim 2001, Statview, ImageComm Systems, Inc) or PIE medical CAAS system.20 Additionally, mid and distal segments of the epicardial coronary arteries were also measured after each intervention.
Serum IgG Antibodies to CMV, HSV-1, HAV, C pneumoniae, and H pylori
Serum samples obtained from all study subjects were frozen at −80°C.7,21⇓ Enzyme-linked immunosorbent assay kits were used to determine serum immunoglobulin-G (IgG) antibodies to CMV (Wampole, Cranbury, NJ) and H pylori (Meridian Diagnostics, Cincinnati, Ohio), respectively. Serum IgG antibodies for HAV (Abbott HAVAB EIA, Abbott Park, Ill) and HSV-1 (American Medical Laboratories, Inc, Chantilly, Va) were determined using an enzyme immunoassay. Serum IgG and IgA (n=228) antibodies for C pneumoniae were assayed using a microimmunofluorescence test (University of Washington). Seropositivity was determined according to manufacturer’s instructions, as described before.7 The prevalence of IgG antibodies in the general population (age >50 years) has been reported as follows: 40% to 100% for CMV, 60% to 100% for HSV1, 50% to 100% for HAV, 70% to 95% for C pneumoniae, and >40% for H pylori based on socioeconomic class. Serum CRP was measured by fluorescence polarization immunoassay technology (TDxFLEx analyzer, Abbott Labs).7
Data are expressed as mean±SEM. Differences between means were compared by paired or unpaired Student’s t test, as appropriate. All probability values are 2-tailed, and a value <0.05 was considered statistically significant. Univariate correlations were performed using the Pearson’s correlation coefficient. Multivariate analyses included the use of stepwise regression and the General Linear Models procedure in SAS to identify significant relationships. This procedure was used to test whether the presence of CAD, the severity of CAD (0-, 1-, 2-, or 3-vessel disease) in the total population, or percent change in CVR with ACH (in patients undergoing vascular function studies) were related to age, sex, body mass index, hypertension (blood pressure >140/90 mm Hg), diabetes, cigarette use, history of hypercholesterolemia (cholesterol level >240 mg/dL or on lipid-lowering therapy), CRP level, antibody status to each of the 5 pathogens, or the total number of infections (pathogen burden).
Patient characteristics are summarized in Table 1. The prevalence of IgG antibodies directed against the 5 microorganisms was as follows: CMV, 64%; HSV-1, 83%; HAV, 52%; C pneumoniae, 76%; and H pylori, 42%. Thus, 11.5% had previous exposure to 0 or 1 pathogen, 44.5% to 2 or 3 pathogens, and 44% to 4 or 5 pathogens. Among conventional risk factors, univariate correlates of CAD prevalence were age (P<0.0001), male sex (P<0.0001), hypercholesterolemia (P<0.0001), hypertension (P=0.01), diabetes (P<0.0001), and smoking (P<0.0001). After multivariate adjustment, age (OR 1.09), male gender (OR 3.6), hypercholesterolemia (OR 4.1), smoking (OR 3.4), and diabetes (OR 5.1) remained independent predictors of CAD prevalence (all P<0.001).
CAD Prevalence and Seropositivity to Individual Pathogens
The frequency of past infection with each organism was higher in patients with CAD compared with those with NCA, with the difference being statistically significant in those with seropositivity for CMV, HAV, and H pylori after univariate analysis (Table 2). After adjustment for conventional risk factors, both CMV and HAV remained significant predictors of CAD. When additionally adjusted for CAD risk factors and presence or absence of all individual infections, only HAV remained independently predictive of CAD (Table 2). The prevalence of CAD in patients seropositive and seronegative for IgA antibody against C pneumoniae (n=228) was similar (67.1% versus 60.8%, P=NS, respectively).
CAD Prevalence and Seropositivity to Multiple Pathogens
Because most individuals possessed IgG antibodies to >1 pathogen, we investigated the relationship between the pathogen burden (an aggregate measure of number of past infections) and CAD prevalence. A highly significant correlation was observed between the pathogen burden and presence of CAD (r=−0.23, P<0.0001). The mean pathogen burden was higher in those with CAD compared with those without (3.4±0.1 versus 2.8±0.1, P=0.0001, respectively). When grouped into 3 subsets according to their antibody status to the 5 pathogens, the prevalence of CAD was higher in patients with multiple infections compared with those with few or no infections (Table 1, Figure 1). The pathogen burden remained an independent predictor for the presence of CAD even after adjustment for other risk factors (OR 1.3, P=0.018, Figure 1). Thus, the OR for presence of CAD was 5.0 (95% CI, 2.2 to 11.2) for patients seropositive for 5 antibodies compared with an OR of 1.0 for those seropositive for 0 pathogens or 1 pathogen.
Because of the strong association between seropositivity for HAV and CAD, we reanalyzed the relationship between pathogen burden and CAD after excluding the HAV data. The mean pathogen burden remained significantly higher in those with CAD compared those without CAD (2.8±0.1 versus 2.4±0.1, P=0.0007, respectively).
Pathogen Burden and Severity of CAD
The severity of CAD was also related to pathogen burden; the prevalence of multivessel disease (2- and 3-vessel disease) was significantly higher in patients with greater pathogen burden (P=0.001) (Figure 1). Pathogen burden remained an independent predictor of the severity of CAD after multivariate adjustment for conventional risk factors and CRP level (P=0.029).
Interaction With CRP
The CRP level was higher in patients with CAD (0.92±0.04 versus 0.72±0.06, P=0.03). After adjustment for conventional risk factors, a high CRP level (>0.5 mg/dL) remained an independent predictor of CAD (Table 2). There was also a correlation between pathogen burden and CRP levels (r=0.1, P=0.04), with a higher CRP level in patients with a greater pathogen burden (Table 1). To investigate whether the pathogen burden was an independent determinant of the presence of CAD, a multivariate analysis was repeated adjusting for conventional risk factors and CRP level. This demonstrated that both pathogen burden (OR 1.27, P=0.036) and high CRP level (OR 2.7, P=0.004) were independent predictors of CAD, together with the conventional risk factors described above (Figure 2).
Individual Pathogens and Vascular Function
The coronary vasodilator response to ACH was used as an index of endothelial function, and the responses to adenosine and SNP were used as indices of endothelium-independent function.20
Microvascular and epicardial dilation with ACH tended to be lower in subjects seropositive to any 1 of the pathogens compared with seronegative patients (Table 3), reaching statistical significance for patients seropositive for H pylori (P=0.002), HAV (P=0.03), and HSV1 (P=0.03) for the microcirculation and for HSV1 (P=0.02) for the epicardial circulation. No differences in the vasomotor responses to SNP or adenosine were observed between seropositive and seronegative patients (Table 3).
Pathogen Burden and Endothelial Function
There was a significant correlation between pathogen burden and the percent increase in flow (P=0.01) or the percent decrease in CVR with ACH (P=0.009) (Figure 3). Thus, compared with patients with antibodies to ≤1 pathogen, ACH-mediated microvascular dilation was diminished in patients seropositive to ≥4 pathogens (−50±4% versus −34±3% decrease in CVR) (Figure 3). The pathogen burden (P=0.009) and hypertension (P=0.014) were both predictors of the percent change in CVR with ACH, independent of risk factors, presence or absence of atherosclerosis, or CRP level.
Because the prevalence of CAD was higher in patients with a greater pathogen burden (Table 1), and because atherosclerosis is known to be associated with endothelial dysfunction, we examined this relationship in patients with NCA. Progressive impairment of ACH-dependent microvascular dilation was also observed with increasing numbers of infections in the subgroup of 86 patients with NCA (Figure 4), and the pathogen burden was in independent predictor of the percent change CVR after mutivariate analysis (P=0.03).
Serum CRP levels did not correlate with the ACH response. Multivariate adjustment for CRP levels did not alter the correlation between pathogen burden and endothelial function.
Patients with a higher pathogen burden (≥4 past infections) had a constrictor response with ACH, consistent with greater endothelial dysfunction compared with those with a lower pathogen burden (≤3 infections) in all patients (−2.7±1.2% versus 0.6±1.2%, P=0.05, respectively) and in the subset with NCA (−3.5±2% versus 2.2±1.4%, P=0.02).
Pathogen Burden and Responses to SNP and Adenosine
Vasodilation in response to SNP did not correlate with pathogen burden (P=0.9) (Figure 3). Similarly, there was no difference in SNP-mediated epicardial dilation in those with low (≤3 infections) compared with high-pathogen burden (≥4 past infections) (18±2% and 16±2%, respectively, P=0.3). Furthermore, flow reserve with adenosine was also similar in the 3 pathogen subsets (P=0.6, Figure 3) in all patients and those with NCA (Figure 4).
Increasing seroepidemiologic and mechanistic evidence suggests that infectious agents are implicated in the genesis of atherosclerosis.1–10,15–19⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓ The present investigation importantly extends the evidence that infections and, particularly, pathogen burden contribute to atherogenesis. Sixty-nine percent of our patients were infected by ≥3 of the common pathogens studied; most importantly, the pathogen burden was an independent risk factor for the presence of CAD, the severity of CAD, and coronary vascular endothelial dysfunction. The odds of having CAD were 4.1-fold higher in patients with 4 or 5 previous infections compared with those with 0 or 1 previous infection. This confirms our previous findings and extends these observations by illustrating that the severity of CAD was related to the pathogen burden.7 There was a progressive decline in the dilator response to ACH as pathogen burden increased, providing a mechanistic explanation of how infections might predispose to atherogenesis by causing endothelial dysfunction. Together with conventional risk factors, pathogen burden imposed an additional independent risk for the presence and severity of CAD and of endothelial dysfunction.
Increasing pathogen burden was also associated with a higher CRP level, indicating an increased systemic inflammatory response, and, as observed previously, a higher CRP level was an independent predictor of CAD. Combined together, these findings raise the intriguing possibility that the aggregate effect of multiple pathogens is to precipitate endothelial dysfunction and inflammation, both processes that predispose to atherosclerosis.
Many studies have demonstrated associations between seropositivity to individual pathogens and atherosclerosis.1–10,21⇓⇓⇓⇓⇓⇓⇓⇓⇓⇓ Importantly, we found that the pathogen burden predicted the presence and severity of CAD more strongly than any single pathogen, and recent studies examining multiple infections have confirmed our observations.22–25⇓⇓⇓ Nevertheless, other investigations have failed to demonstrate such associations,26–30⇓⇓⇓⇓ a disparity that may be because analysis was limited to single pathogens, because of the use of cardiac events as end points, or because the negative studies have been performed in relatively healthy populations.28–30⇓⇓
Our study suggests that the composite effect of multiple pathogens, rather than any single organism, is the development of endothelial dysfunction. The fact that endothelium-independent responses were preserved in all patient subsets indicates that the impairment of the endothelial layer does not affect smooth muscle function. Importantly, pathogen burden correlated with endothelial function even in those with NCA. It seems that endothelial dysfunction may be one of the underlying mechanisms by which multiple intracellular pathogens may actually contribute to the earliest processes leading to the development of atherosclerosis and to its more rapid progression to multivessel disease.
Endothelial dysfunction unequivocally predisposes to atherosclerosis, and patients with atherosclerosis and its risk factors exhibit both microvascular and epicardial coronary vascular endothelial dysfunction.20,31,32⇓⇓ Studies now demonstrate more rapid progression to adverse atherosclerotic events in individuals with endothelial dysfunction and also in those with multiple past infections.22,24,25,33⇓⇓⇓ Our observation that pathogen burden is an independent determinant of endothelial dysfunction additionally strengthens the epidemiologic, mechanistic, and prognostic link between infection with multiple pathogens and atherosclerosis.
It is likely that despite major differences in the biologic and clinical consequences that result from human infection with these pathogens, there may be shared pathways by which diverse organisms produce vascular endothelial injury. These are likely to involve the autoimmune and the coagulation pathways and, as demonstrated in this study, also lead to an integrated injury of the vascular endothelium and inflammation.13,14⇓
This is a cross-sectional investigation that establishes an association between pathogen burden and endothelial dysfunction but not causality. Thus, increasing pathogen burden may not be the direct cause of the endothelial dysfunction or atherosclerosis, but merely a surrogate marker of some other presently undiscovered risk factor. Thus, the conclusions of this study must be regarded as preliminary and hypothesis-generating. In addition, we considered a positive IgG antibody titer against these pathogens as evidence of past infection. It is not known whether seropositivity indicates recent, past, or reactivated latent infection or whether it represents reinfection.
The number of positive IgG antibody responses to multiple pathogens (pathogen burden) is an independent risk factor for the presence and severity of CAD and of endothelial dysfunction in a population of patients undergoing coronary angiography. As such, the study provides a crucial link by which pathogens may contribute to atherogenesis.25,34,35⇓⇓
We are grateful to Gloria Zalos, RN, for quantitative angiography and William Schenke and Rita Mincemoyer for their assistance.
Speir E, Modali R, Huang ES, et al. Potential role of human cytomegalovirus and p53 interaction in coronary restenosis. Science. 1994; 265: 391–394.
Nieto FJ, Adam E, Sorlie P, et al. Cohort study of cytomegalovirus infection as a risk factor for carotid intimal-medial thickening, a measure of subclinical atherosclerosis. Circulation. 1996; 94: 922–927.
Muhlestein JB, Horne BD, Carlquist JF, et al. Cytomegalovirus seropositivity and C-reactive protein have independent and combined predictive value for mortality in patients with angiographically demonstrated coronary artery disease. Circulation. 2000; 102: 1917–1923.
Zhu J, Quyyumi AA, Norman JE, et al. The possible role of hepatitis A virus in the pathogenesis of atherosclerosis. J Infect Dis. 2000; 182: 1583–1587.
Pasceri V, Cammarota G, Patti G, et al. Association of virulent Helicobacter pylori strains with ischemic heart disease. Circulation. 1998; 97: 1675–1679.
Epstein SE, Zhu J, Burnett MS, et al. Infection and atherosclerosis: potential roles of pathogen burden and molecular mimicry. Arterioscler Thromb Vasc Biol. 2000; 20: 1417–1420.
Visser MR, Tracy PB, Vercellotti GM, et al. Enhanced thrombin generation and platelet binding on herpes simplex virus-infected endothelium. Proc Natl Acad Sci U S A. 1988; 85: 8227–8230.
Key NS, Vercellotti GM, Winkelmann JC, et al. Infection of vascular endothelial cells with herpes simplex virus enhances tissue factor activity and reduces thrombomodulin expression. Proc Natl Acad Sci U S A. 1990; 87: 7095–7099.
Kol A, Sukhova GK, Lichtman AH, et al. Chlamydial heat shock protein 60 localizes in human atheroma and regulates macrophage tumor necrosis factor-α and matrix metalloproteinase expression. Circulation. 1998; 98: 300–307.
Dechend R, Maass M, Gieffers J, et al. Chlamydia pneumoniae infection of vascular smooth muscle and endothelial cells activates NF-κB and induces tissue factor and PAI-1 expression: a potential link to accelerated arteriosclerosis. Circulation. 1999; 100: 1369–1373.
Zhu J, Shearer GM, Norman JE, et al. Host response to cytomegalovirus infection as a determinant of susceptibility to coronary artery disease: sex-based differences in inflammation and type of immune response. Circulation. 2000; 102: 2491–2496.
Mayr M, Kiechl S, Willeit J, et al. Infections, immunity, and atherosclerosis: associations of antibodies to Chlamydia pneumoniae, Helicobacter pylori, and cytomegalovirus with immune reactions to heat-shock protein 60 and carotid or femoral atherosclerosis. Circulation. 2000; 102: 833–839.
Zhu J, Nieto FJ, Horne BD, et al. Prospective study of pathogen burden and risk of myocardial infarction or death. Circulation. 2001; 103: 45–51.
Rupprecht HJ, Blankenberg S, Bickel C, et al. Impact of viral and bacterial infectious burden on long-term prognosis in patients with coronary artery disease. Circulation. 2001; 104: 25–31.
Adler SP, Hur JK, Wang JB, et al. Prior infection with cytomegalovirus is not a major risk factor for angiographically demonstrated coronary artery atherosclerosis. J Infect Dis. 1998; 177: 209–212.
Havlik RJ, Blackwelder WC, Kaslow R, et al. Unlikely association between clinically apparent herpesvirus infection and coronary incidence at older ages: the Framingham Heart Study. Arteriosclerosis. 1989; 9: 877–880.
Ridker PM, Hennekens CH, Stampfer MJ, et al. Prospective study of herpes simplex virus, cytomegalovirus, and the risk of future myocardial infarction and stroke. Circulation. 1998; 98: 2796–2799.
Ridker PM, Kundsin RB, Stampfer MJ, et al. Prospective study of Chlamydia pneumoniae IgG seropositivity and risks of future myocardial infarction. Circulation. 1999; 99: 1161–1164.
Quyyumi AA, Dakak N, Andrews NP, et al. Contribution of nitric oxide to metabolic coronary vasodilation in the human heart. Circulation. 1995; 92: 320–326.
Schachinger V, Britten MB, Zeiher AM. Prognostic impact of coronary vasodilator dysfunction on adverse long-term outcome of coronary heart disease. Circulation. 2000; 101: 1899–1906.
Roivainen M, Viik-Kajander M, Palosuo T, et al. Infections, inflammation, and the risk of coronary heart disease. Circulation. 2000; 101: 252–257.
Siscovick DS, Schwartz SM, Corey L, et al. Chlamydia pneumoniae, herpes simplex virus type 1, and cytomegalovirus and incident myocardial infarction and coronary heart disease death in older adults: the Cardiovascular Health Study. Circulation. 2000; 102: 2335–2340.