INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Previous exposure to Chlamydia pneumoniae and cytomegalovirus (CMV) has been associated with heart disease (6). C. pneumoniae antigen and DNA have been detected in coronary and carotid atheroma and in aortic aneurysms, and culture of C. pneumoniae from atheroma has been reported previously (10, 19). However, recent large prospective studies have not confirmed an association between anti-C. pneumoniae immunoglobulin G (IgG) serology results and vascular events (16, 20, 21), and there was poor correlation between serology results and the presence of C. pneumoniae antigen or DNA in tissue (5).

The detection of C. pneumoniae DNA circulating in peripheral blood mononuclear cells (PBMC) has been reported, although estimates of prevalence varied widely. In one study, 59% of 101 heart disease patients and 46% of 52 blood donor controls were positive for C. pneumoniae DNA (3). Among 804 men undergoing coronary angiography, the prevalence of C. pneumoniae DNA was 8.8% in those with heart disease versus 2.9% in those without heart disease (23). In 41 aortic aneurysm patients, detection of C. pneumoniae DNA in PBMC correlated with the isolation of C. pneumoniae DNA from aortic aneurysms (1). Potentially, detection of C. pneumoniae DNA in PBMC could enable large-scale epidemiological studies to clarify the role of C. pneumoniae in atherosclerotic heart disease and its complications.

CMV is associated with accelerated atherosclerosis of cardiac transplants and may be associated with coronary artery restenosis or thrombosis after angioplasty or atherectomy (7, 15). In a rat model, rat CMV increases neointimal cell proliferation after balloon injury to the carotid artery (18). A role in human disease remains unproven, and prospective studies of CMV serology have not confirmed a relationship with vascular events (20).

This study had three primary objectives: first, to determine the prevalence of circulating C. pneumoniae DNA and CMV DNA in patients undergoing coronary angiography; second, to determine if C. pneumoniae DNA detection increased after coronary angioplasty, on the assumption that disrupted endothelium would release C. pneumoniae (but not CMV) DNA into the bloodstream; and third, to determine whether DNA isolation was prognostically important.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Patients. Consecutive elective outpatients were recruited from the Hamilton Regional Angiography Suite, Hamilton Health Sciences Corporation, Hamilton, Ontario, Canada, between February and October 1999. Information regarding age, gender, and a history of previous cardiac disease, smoking, diabetes mellitus, hyperlipidemia, and hypertension was obtained. Sample size calculations required 100 patients in the angioplasty stratum for an 80% probability of detecting an increase in DNA prevalence from 10 to 20%. Angiography and angioplasty patients were enrolled until predetermined strata of 100 patients each were filled, with recruitment of angiography patients being complete by April 1999 and recruitment of angioplasty patients continuing until October 1999. The angiogram report was scored by the presence of any arterial narrowing (>25%) and by the number of epicardial coronary arteries with at least 50% narrowing in two orthogonal views or at least 70% narrowing in one view by visual assessment. Six-month clinical outcomes (cardiac hospitalization, repeat angiogram or angioplasty, myocardial infarction, coronary artery bypass surgery, or death) were obtained by telephone calls to the patient and by hospital chart review. All clinical data were collected by study nurses blinded to laboratory data. All participating patients gave written consent, and the study protocol was approved by research ethics boards at St. Joseph's Hospital (Hamilton, Ontario, Canada), Hamilton Health Sciences Corporation, and McMaster University (Hamilton, Ontario, Canada).

Blood collection. Serum was collected prior to angiography or angioplasty. Circulating PBMC were obtained by venipuncture into an 8-ml Vacutainer CPT cell preparation tube (BD Vacutainer Systems, Franklin Lakes, N.J.) prior to, immediately after, and 4 h after the procedure for a total of three tubes. Specimens obtained prior to the procedure were obtained in a precatheterization outpatient clinic days to weeks before the procedure. CPT tubes contain a blood separation medium composed of a thixotropic polyester gel and a density gradient liquid solution. Laboratory personnel processed CPT tubes essentially according to the manufacturer's instructions, except for a second centrifugation. Briefly, CPT tubes were centrifuged in a Beckman GPR centrifuge at 1,500 × g for 30 min and refrigerated. After transport to the research laboratory (generally within 24 h), the specimens were mixed by inversion and recentrifuged, and the mononuclear cell layer (if visible) or 1 ml of plasma directly above the gel was aspirated and frozen at 70°C. In batches, mononuclear cell preparations were thawed and 200-µl aliquots were extracted using QIAamp DNA minikits (Qiagen, Mississauga, Ontario, Canada) into 100 µl of elution buffer.

Detection of DNA. Laboratory staff were blinded to all clinical data. A 2.5-µl aliquot was amplified by a nested PCR (22), consisting of 40 amplification cycles for a 333-bp product (external primers CP1 and CP2) and 30 cycles for a 207-bp product (internal primers CPC and CPD), followed by separation on a 2.0% (wt/vol) agarose gel containing ethidium bromide and UV light visualization. The 207-bp product was confirmed as C. pneumoniae by hybridization with an in-house specific fluorescein-labeled oligonucleotide probe (5' TAC GGA GAC TAT GTT TTC GA 3', GenBank accession no. AF131889, positions 196 to 215). The probe detected C. pneumoniae VR 1310 but not Chlamydia trachomatis (LGV 434) or Chlamydia psittaci (6BC). Six controls, consisting of one positive control (C. pneumoniae VR 1310), four negative water controls without DNA, and one additional tube with master mix open to the air throughout specimen addition, were run for every 48 specimens. PCR extraction and amplification were performed in separate rooms. In addition, all positive samples were confirmed by reextraction from the original patient sample, followed by amplification in triplicate and probing. C. pneumoniae DNA-positive status was defined as samples which were positive initially and in at least one of the replicates after reextraction. Twelve amplification products from different patients were sequenced. Oligonucleotide probe synthesis and sequencing of amplification products were carried out at the Institute of Molecular Biology, McMaster University. PCR-positive samples were also amplified by one or more of three different nonnested PCRs, targeting 23S ribosomal DNA (8), 16S ribosomal DNA (13), or a cloned PstI fragment (4), with confirmation of all positive results by specific oligonucleotide probes.

Serology. CMV and C. pneumoniae serology data were obtained by enzyme immunoassay: anti-C. pneumoniae IgG and IgA were measured by Sero-CP (Savyon Diagnostics, Kiryat Minrav, Israel), and anti-CMV IgG antibody was measured by ETI-CYTOK-G Plus (DiaSorin SRL, Saluggia, Italy). Anti-C. pneumoniae IgG or IgA seropositivity was defined as a cutoff index of 1.1 or greater, where the index was obtained by dividing the specimen optical density by twice the mean of two negative controls, in accordance with the manufacturer's product insert. Anti-CMV IgG seropositivity was defined as a value greater than 0.4 IU/ml, in accordance with the product insert. Serology and PCR assays were performed independently by different staff in separate laboratories, with blinding to other laboratory results.

Statistical analysis. Proportions were compared using chi-square or Fisher exact tests (unmatched data) or the McNemar test (matched data). Logistic regression modeling (SPSS for Windows 10.0; SPSS, Inc., Chicago, Ill.) was undertaken using DNA status as the response variable and the following explanatory variables: serological status, age, gender, clinical history (unstable angina, myocardial infarction, angioplasty, or coronary artery bypass grafting), and clinical risk factors (angiogram score, smoking, diabetes mellitus, hypercholesterolemia, or hypertension). Six-month outcome (as the response variable) was modeled by logistic regression with C. pneumoniae DNA, CMV DNA, age, gender, clinical risk factors, and angiogram score as explanatory variables. A P value of <0.05, two-tailed, was considered statistically significant.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Description of patients. One hundred eighteen patients undergoing angiography and 90 patients undergoing angioplasty were enrolled. The median age was 60.4 years (minimum to maximum, 21 to 85 years, respectively), and 75.4% of patients were male. Medical history included the following: diabetes mellitus, 21.6% of patients; hypertension, 56.8%; hypercholesterolemia, 66.1%; current smokers, 18.1%; former smokers, 59.9%; and lifelong nonsmokers, 22.1%. A history of myocardial infarction was present in 46.2%, one of unstable angina was present in 74.1%, one of previous angioplasty was present in 21.6%, and one of coronary artery bypass grafting was present in 10.1% of patients.

C. pneumoniae DNA detection. A total of 547 samples from 208 patients had DNA extracted and amplified for C. pneumoniae, including 207 samples from before, 182 samples taken immediately after, and 158 samples taken 4 h after the angiogram or angioplasty.

CMV DNA detection. Forty of 559 samples (7.2%) from 36 of 208 patients (17.3%) were positive for CMV DNA (Table 1). For all but four patients, a single sample from the three time periods was positive. Seven amplification products from different patients were sequenced, and all matched CMV sequences exactly. In a test of 49 plasma samples (31 of which had CMV in the corresponding mononuclear cell fraction), 9 were positive for CMV DNA. Five of these nine were CMV DNA positive in both plasma and mononuclear cell fractions, and eight of these nine were anti-CMV IgG positive.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

This study addressed several issues of direct relevance for the development and validation of molecular tests for cardiovascular studies. First, we examined the prevalence of C. pneumoniae DNA in PBMC from patients undergoing angiography, most of whom had documented atherosclerotic heart disease. The estimate of 11.5% for C. pneumoniae DNA prevalence was similar to the results of one recent study (23) and may be more representative than the higher prevalence of 59% reported in another study (3). The discrepancy may represent interlaboratory variation or a different population being sampled during a time of high C. pneumoniae activity in the community. Our laboratory used the same primers (22) as those used in the higher-prevalence study (3), and we have previously demonstrated that this PCR was more sensitive for the detection of C. pneumoniae in blood samples, despite an analytic sensitivity similar to those of four other PCRs (14). The higher prevalence in smokers and in the winter-spring months requires confirmation in a separate study. The finding that positive serology results were not associated with circulating C. pneumoniae DNA was similar to the lack of association between seropositivity and DNA in atherosclerotic lesions (5) and may explain why anti-C. pneumoniae IgG did not predict cardiovascular events in several prospective studies (16, 20, 21). Binding of antibody in immune complexes, which has been detected for patients with atherosclerotic heart disease, is a plausible explanation for this discrepancy (11, 12).

Second, we examined the reproducibility of PCR results. Despite use of a very sensitive PCR, samples were usually positive in only one of three sampled time periods and on repeat testing were positive in only one of three replicates. Potential explanations for these results include nonspecific amplification, contamination of the PCR, between-sample variation (biological sampling), and within-sample variation (statistical sampling). The amplification products were C. pneumoniae, as evidenced by hybridization with a specific oligonucleotide probe and by molecular sequencing of 12 amplification products. We believe that contamination is not a likely explanation, given that all positives had been extracted on two separate occasions, 75% were positive when tested a third time, and 40% of the samples were confirmed by one or more PCRs targeting a different part of the genome. Furthermore, among the plasma samples and initially negative mononuclear cell samples, we detected only a single positive sample in 237 PCR runs. Finally, of over 200 C. pneumoniae-negative controls run using this PCR by our laboratory in the last year, including 40 negative controls which were open during the entire specimen addition step, there was not a single false positive. Biological variation is plausible, with intermittent shedding of infected mononuclear cells and clearance by the reticuloendothelial system. The most likely explanation for intermittent positivity is statistical sampling. We have recently demonstrated that the proportion of replicates which are positive is directly related to the concentration of C. pneumoniae organisms in a sample and that a positive proportion of one in three replicates can be predicted using probit analysis for samples with low DNA concentrations (M. Smieja et al., unpublished data).

Third, this and a smaller study (1) have demonstrated that the Vacutainer CPT tube method is a simple and feasible method for obtaining mononuclear cell-associated DNA. The method eliminates the need for standard Ficoll-Hypaque gradient centrifugation and may facilitate the undertaking of similar studies by other laboratories. The finding that mononuclear cell fractions, but not the corresponding plasma fractions, were positive for C. pneumoniae DNA constitutes clear evidence for humans that circulating C. pneumoniae DNA is cell associated and found within the mononuclear cell fraction. This corroborates the recent finding of C. pneumoniae antigen in circulating mononuclear cells (2).

Fourth, we demonstrated the routine detection of CMV DNA in peripheral blood of nontransplant vascular patients. To our knowledge, this has not been previously reported. CMV was found in both the mononuclear cell fraction and plasma, although a greater number of positives was detected in the mononuclear cell fraction.

Fifth, we examined whether C. pneumoniae DNA or CMV DNA detection was augmented by coronary angioplasty. We reasoned that C. pneumoniae in macrophages within the atherosclerotic plaque would seed the bloodstream during and after the angioplasty, because of disruption of the endothelium and plaque. No such increase in C. pneumoniae DNA or CMV DNA was observed, either immediately after or 4 h after angioplasty. Potential explanations include too low a concentration of organism in coronary lesions, incorrect sampling time, incorrect sampling fraction (mononuclear cells), or lack of organisms in the atheroma of most patients. Further improvements in the methods of detection, and other methods of sampling such as those with intra-coronary-artery catheters or before and after coronary thrombolysis, may help to resolve this issue.

Last, we examined whether detection of C. pneumoniae or CMV DNA in PBMC had any prognostic significance. Neither CMV DNA nor C. pneumoniae DNA was associated prospectively with subsequent revascularization or other clinical events. To examine whether the detection of C. pneumoniae DNA and CMV DNA has prognostic importance, large, well-designed, prospective observational and treatment studies are needed. The measurement of DNA in PBMC using the methods described here may facilitate such studies and may be a more specific measure of recent exposure than serology.