Chlamydia pneumoniae, heat shock proteins 60 and risk of secondary cardiovascular events in patients with coronary heart disease under special consideration of diabetes: a prospective study

Chlamydia pneumoniae (CP) is an obligate intracellular organism. Sero-positivity is low in early childhood (< 5 years), rising rapidly during school age such that at age 20 years approximately 50% of individuals will have detectable antibodies. Seroprevalence continues to rise, at a slower pace, in adulthood and reaches 75% in the elderly population [1]. The first description of the clinical presentation of C. pneumoniae infection appeared in 1985 [2].

Chlamydia pneumoniae is one infectious agent that has received particular attention as a potent atherogenic stimulus. Correlative studies suggest a possible relationship between atherosclerosis and CP infection [3, 4]. Pathological studies have revealed the localization of CP in atherosclerotic lesions [5]. Several animal models have suggested a causal role of CP in atherogenesis [6, 7], however, the precise mechanism(s) by which CP promote atherosclerosis still remains unclear.

Inflammation, infection and autoimmunity to human heat shock protein 60 (h-hsp60) have emerged as novel risk factors for coronary heart disease during the past decade. Heat shock proteins are a class of highly conserved proteins produced by all organisms as a result of stress and damage caused by, for instance, exposure to chemical and environmental factors, inflammation and infections [8]. Because of the high sequence homology between bacterial and human heat shock proteins, they have been postulated to be critical antigens in autoimmune diseases. In particular, it has been proposed that the immune response to microbial hsp60 may lead to autoimmunity to h-hsp60 and by that means nourish the development of atherosclerosis [9]. Because serum antibodies against chlamydial hsp60 from subjects with atherosclerosis may cross-react with human hsp60 and mediate endothelial cytotoxicity, [10] it is suggested that humoral immune reactions to hsp60s may play an important role in vascular endothelial injury, a key process in the early stages of atherosclerosis.

Two independent groups have shown that an elevated anti-human hsp60 antibody level is a risk factor for primary coronary atherosclerosis, especially when it is present with Chlamydia pneumoniae infection [11, 12]. Therefore, it is suggested that heat shock protein (h-hsp) 60 may serve as a possible link between Chlamydia pneumoniae infection and atherosclerosis [12, 13]. Elevated levels of h-hsp60 antibodies have also been reported in human sera obtained from patients with atherosclerosis [14]. In a recent study, Heltai and colleagues demonstrated associations of acute myocardial infarction with high levels of anti-h-hsp60 and anti-CP antibodies, these being independent risk factors for primary CHD [15].

The role of heat-shock proteins in patients with prevalent coronary heart disease on prognosis has been rarely investigated. Furthermore, despite the fact that patients with diabetes are immuno-compromised, remain persistently at risk for infectious complications and are prone to develop accelerated atherosclerosis [16], there are few epidemiological studies available investigating a potential interaction between chlamydial sero-markers of infection, heat shock proteins and secondary cardiovascular disease (CVD) events in diabetics.

In this study we are investigating prospectively the role of Chlamydia pneumoniae, chlamydial heat shock protein 60 and a possible intermediate role of human heat shock protein 60 sero-status in the development of secondary CVD events in patients with coronary heart disease (CHD) under special consideration of diabetes, also comparing the relationship between chlamydial sero-markers of infection, heat shock proteins and secondary cardiovascular disease (CVD) events in patients with diabetes with those without.

In addition, the association of CP sero-status and serum concentration of C-reactive protein (CRP) was explored, as CRP is considered a main determinant of secondary CVD events.

Furthermore, we found associations between CP IgA, CP IgG sero-positivity and CRP levels at baseline after adjusting for age and gender (geometric mean: CP IgA negative, 3.22 mg/L vs CP IgA positive, 3.98 mg/L, (p = 0.01)), (geometric mean: CP IgG negative, 3.31 mg/L vs CP IgG positive, 3.77 mg/L, (p = 0.07)). Our analyses also showed no association between Ch-hsp60 IgG and CRP levels at baseline after adjusting for age and gender.

In this prospective study, we investigated the role of and Chlamydia pneumoniae (CP), Chlamydia heat shock protein (Ch-hsp) 60 and a possible intermediate role of human heat shock protein (h-hsp) 60 antibodies in the development of secondary cardiovascular events in patients with coronary heart disease (CHD) under special consideration of diabetes mellitus. The results from this study showed that the occurrence of a secondary CVD event was more common among CP (IgA) and tentatively CP (IgG) sero-positive than among sero-negative patients in crude analysis. However, the associations and a possible interaction between the sero-status of these infections and history of diabetes were not statistically significant after multivariate adjustment for covariates. Although the study clearly demonstrated an increased risk in developing a secondary CVD event among sero-positive patients with a history of diabetes compared to sero-negative without and sero-positivity adding a hazard to diabetes, it is suggested that mainly diabetes was responsible for the increased risk. Furthermore, this study showed that human heat shock protein 60 (h-hsp60) is unlikely an intermediate factor for developing secondary CVD events in the population studied.

Chlamydia pneumoniae is one infectious agent that has received particular attention as a potent atherogenic stimulus. However, the question whether serological evidence of prior infection with CP and the presence of Ch-hsp60 and h-hsp60 is associated with MI and CHD death remains a source of controversy, as several prospective studies have failed to demonstrate consistent associations between the presence of IgG antibodies to CP and incident MI [17–19]. Examples of some studies that looked at primary CVD events include a nested case-control study conducted within the context of the Cardiovascular Health Study [20]. Results showed that high-titer CP IgG was associated with an increased risk of primary CHD (OR 2.2, 95% CI, 1.1–4.4), whereas, a low to moderate CP antibody titre showed a weak non-significant association of IgG antibodies to CP with the risk of primary CHD among the elderly (OR 1.1, 95% CI, 0.7 to 1.8). Furthermore, Haider and colleagues found CP infection, as evidenced by sero-positivity, not to be associated with increased risk for primary CVD in the Framingham Heart Study cohort: CP IgG (HR 0.91, 95% CI, 0.68–1.20) and CP IgA (HR 0.65, 95% CI, 0.39–1.07) [21].

Further examples of the limited number of studies that looked at CP sero-positivity and secondary CVD events include a prospective study in the context of the Heart Outcomes Prevention Evaluation (HOPE) Study. Their study showed that among patients with pre-existing CVD, neither CP IgG (adjusted HR, 0.87; 95% CI, 0.68–1.10) nor CP IgA (adjusted HR, 1.10; 95% CI, 0.90–1.34) predicted cardiovascular events during the 4.5 years of follow-up. [22]. Another prospective study evaluated the effect of pathogens and pathogen burden on the risk of MI or death among coronary artery disease (CAD) patients. This study demonstrated that baseline prevalence of antibodies directed against CP was not significantly associated with development of secondary MI or CVD death in an elderly cohort after a 3 years follow-up period. [19]. Furthermore, Haim and colleagues evaluated the association between previous exposure to CP and future coronary risk in patients with CHD. This was a prospective nested case-control design including patients from a trial study of bezafibrate for the treatment of CHD. Results showed that mean titers of anti-CP antibodies (IgG and IgA) in the baseline sera of 136 patients who had coronary events during a mean follow-up period of 6.2 years were similar with 136 age- and gender-matched controls from same trial without subsequent coronary events. [23].

In addition, another study investigated the role of infections, autoimmunity and inflammation in atherosclerosis. They studied the joint effect of chronic Chlamydia pneumoniae (CP) infection, persistently elevated human heat-shock protein 60 (h-hsp60) antibodies and C-reactive protein (CRP) on coronary risk. Results showed that, compared with persistently low levels, the risk of coronary events was 2-fold for persistently elevated IgA antibodies to CP (OR, 1.96; 95% CI, 1.14–3.36) and also for IgA antibodies to h-hsp60 (OR,2.11; 95% CI, 1.08–4.13). The risks associated with elevated antibodies were much higher when CRP was also elevated [24]. In a previous study, Huittinen and colleagues showed that an elevated level of IgA antibodies to h-hsp60 in baseline serum predicted the occurrence of coronary event several years later [11]. Although CP, a common respiratory pathogen throughout the world, has been associated with atherosclerosis, in the above mentioned study, this finding indicated that elevated levels of CP antibodies were not alone a significant risk factor for coronary events. Instead, a 7-fold risk emerged when an elevated level of IgA antibodies to CP in baseline serum was accompanied by factors related to auto-immunity and inflammation, such as elevated levels of h-hsp60 IgA antibodies and CRP. Although hsCRP levels were associated with C. pneumoniae antibodies in our study, elevated hsCRP levels of patients did not have any type of joint effect in predicting secondary cardiac event (data not shown). This is in accordance with observations of others [25, 26]. CRP may be a more modest predictor of CHD in elderly subjects, especially if already having CHD, than previously reported. The intake of aspirin and statins may further attenuate the predictive power for later events in patients with CHD.

Furthermore, all patients included in this study had an acute cardiovascular event within last 3 months before baseline. However, such an event may also reduce antibody titers, as has been shown for (h-hsp) 60 [27] and this may partly explain the negative results for secondary events in patients with prevalent CHD of our study.

The following limitations of our study should be considered. Although we had a large sample of patients with CHD (over 50% with a MI), fatal-CVD events were rare in this study population. This might be explained by the fact that mortality of MI is highest within the pre- and early hospital phase. As the acute events leading to diagnosis of CHD or MI had occurred at least 3 weeks before inclusion in this study, we were dealing with a selection of patients with a better prognosis compared to a patient population within the early phase of a newly diagnosed CHD. Furthermore, not all patients of course are willing or able to participate in a rehab program. This is another reason why severely ill-patients are probably underrepresented in our study sample. In addition, proof of chlamydial heat shock protein 60 immunoglobulin G is genus specific and not specific to C. pneumoniae, although the majority of antibodies against chlamdial LPS are due to C. pneumoniae.

Although inflammation represents an important feature in all stages of atherogenesis and its complications [28, 29], we could not demonstrate an association between infection markers, sero-status and levels of CRP, even after consideration of history of diabetes. This may indicate, however, that another pathway might be of more importance.

Despite the above mentioned limitations, results of our study cannot exclude a possible moderate increase in risk of secondary CVD events among patients with a positive infection sero-status, however, they clearly indicate an increased risk among diabetic patients. This study further showed that h-hsp60 is unlikely an intermediate factor for developing secondary CVD events considering mechanisms involved in autoimmunity. Larger studies or meta-analysis of multiple studies are needed to address the interaction between infection sero-status and diabetes with adequate power. Notwithstanding the need of further evidence from such studies, patients with diabetes are at special risk for developing secondary CVD events and should be a target group for preventive measures.

Results from this cohort of 1052 patients with pre-existing CHD cannot exclude a possible moderate increase in risk of secondary CVD events among patients with a positive infection sero-status. However, our study showed no intermediate role of human heat shock protein (h-hsp) 60 sero-status in the development of secondary CVD events in patients with CHD. Larger studies or meta-analysis of multiple studies are needed to address the interaction between infection sero-status and diabetes with adequate power.