Antibiotic Therapy for Coronary Artery Disease: Title and subTitle BreakCan a WIZARD Change It All?

The possible association between Chlamydia pneumoniae and atherosclerosis was first reported in 1988 by Saikku et al¹ who noted that antichlamydial antibodies were present more often among patients with acute myocardial infarction (MI) than among matched controls. Since then, several other serological studies have yielded conflicting results about this association, possibly due to lack of uniformity in methods used (ie, different classes of antibodies tested, different methods, different cutoff points) and by the high prevalence of C pneumoniae exposure in the population. On the other hand, data from basic laboratory studies have strongly suggested a possible role of C pneumoniae in the etiology of atherosclerosis. Cells involved in atherogenesis (vascular endothelial cells, smooth muscle cells, macrophages, and platelets) have been found to be infected by C pneumoniae.² In mouse and rabbit models, C pneumoniae has been detected in the vasculature after nasal inoculation and has been associated with various degrees of atherogenesis and progression.² Chlamydia pneumoniae also has been documented in atherosclerotic lesions from patients.²

Although Muhlestein et al³ were the first to demonstrate a protective effect of antibiotic therapy against accelerated atherosclerosis in a rabbit model, the "infection hypothesis" of atherosclerosis only gained momentum after Gurfinkel et al⁴ reported a 7% absolute reduction in the combined end point of cardiac ischemic death, MI, and severe recurrent angina after 30 days of treatment with roxithromycin (in the Randomised Trial of Roxithromycin in Non–Q-Wave Coronary Syndromes [ROXIS] trial). Since then, the number of published reports investigating the role of C pneumoniae in atherogenesis has more than tripled in the last 5 years compared with the preceding decade.

Yet after this positive initial result, data from laboratory studies and clinical trials seem to diverge. More evidence in favor of a role of C pneumoniae emerged from basic laboratory studies, whereas clinical trials of patients with stable and unstable coronary heart disease failed to demonstrate a significant benefit of antibiotic therapy.⁵

In addition to the ROXIS trial, only 3 other trials have suggested a benefit of antibiotic therapy. In the Clarithromycin in Acute Coronary Syndrome Patients in Finland (CLARIFY) trial,⁶ 148 patients with unstable angina or non–Q-wave MI were randomly assigned to receive 3 months of clarithromycin therapy or placebo. There was no significant effect of antibiotic therapy noted in the primary end point (death, MI, or unstable angina within the 3-month treatment period). There was an absolute risk reduction of 14.9% in the secondary end point (combination of death, MI, unstable angina, ischemic stroke, and critical limb ischemia); this was evident after 3 months of therapy and persisted throughout the median 555 days of follow-up. However, in the St George's Hospital trial,⁷ of 220 consecutive men who survived MI, 60 patients with persistently elevated antibody titers against C pneumoniae were randomly assigned to receive placebo or a single or double 3-day course of azithromycin. There was a significantly higher incidence of death, MI, or unstable angina during the 18-month follow-up among those receiving placebo than those with negative chlamydial serology (odds ratio [OR], 4.2; P = .03). On the other hand, patients receiving azithromycin had a similar outcome as those with negative serology (OR, 0.9; 95% CI, 0.2-4.6). In the South Thames Trial of Antibiotics in Myocardial Infarction and Unstable Angina (STAMINA),⁸ 325 patients admitted with acute MI or unstable angina were randomly assigned to receive placebo or triple therapy (azithromycin or amoxicillin plus metronidazole plus omeprazole). Patients receiving antibiotic therapy had a 36% reduction in the incidence of cardiac death, unstable angina, or MI noted at 12 weeks, and the reduction persisted throughout the 1-year follow-up.

In this issue of THE JOURNAL, O'Connor and colleagues⁹ report the results of the WIZARD (Weekly Intervention with Zithromax for Atherosclerosis and Its Related Disorders) study. This is by far the largest trial of antibiotic therapy for coronary artery disease, with 7747 patients enrolled at 271 centers in 9 countries. All patients had documented previous MI in a stable phase (ie, the MI had occurred >6 weeks before screening) and serological evidence of exposure to C pneumoniae. Patients were randomly assigned to receive azithromycin (600 mg/d for 3 days then weekly for the next 11 weeks) or placebo. After 14 months of follow-up, patients in the antibiotic group had a nonsignificant 7% reduction in the risk of the primary end point (death, MI, coronary revascularization, or hospitalization for angina).

Although the findings from the WIZARD trial show, once more, that antibiotic therapy failed to prove useful in the treatment of coronary artery disease, the question is then whether a true contradiction exists between laboratory data and clinical trials. The answer is impossible to give at present, but some points need to be highlighted. Most laboratory data come from in vitro experiments. Although such studies are unequaled in their power to identify mechanisms that may link C pneumoniae infection and atherogenesis, their true validity is difficult to ascertain in vivo. Questions that have been raised ask whether C pneumoniae detected in atheromatous lesions plays an active role or simply represents an innocent bystander in the process.² In addition, most in vivo experimental data are derived from small-animal models, which have inherent limitations.¹⁰ Data from large-animal models of chlamydial infection are practically nonexistent.¹¹ Studies on the effects of antibiotic therapy on atherogenesis in animal models are limited. Moreover, animal models are just what they are, experimental models. Indeed, the literature abounds with studies of interventions that are highly successful in animals but fail to have similar effects in patients. Yet despite these concerns, the bulk of experimental data favor a role for C pneumoniae in atherosclerosis. Although this pathogen is unlikely the unique cause of atherosclerosis, current data strongly suggest that C pneumoniae is a plausible and potentially modifiable risk factor in cardiovascular diseases.²

Several factors might help explain the negative results of clinical trials with antibiotic therapy. Since C pneumoniae is not uniformly present in atherosclerotic plaques, this organism probably does not play a role in all patients. The challenging task is to identify individuals who may benefit most from antibiotics. In the WIZARD trial, post hoc analysis showed trends toward a favorable effect of antibiotic therapy in men who smoke or who have diabetes or hypercholesterolemia; a significant 55% reduction in events was noted in individuals who have diabetes and who smoke. Another approach may be to select potential candidates for antibiotic therapy based on the detection of chlamydial DNA in peripheral leukocytes. Alveolar macrophages may take up C pneumoniae during the course of pulmonary infection and deliver the organism to the site of vascular inflammation or injury, where it can induce chronic inflammation, potentially leading to the formation of atherosclerotic lesions.¹² Identification of chlamydial DNA in these cells may be more revealing than positive serological studies.

In addition, the optimal timing, dose, and duration of antibiotic therapy needed for an antiatherogenic effect are not known. Some clinical studies seem to suggest an early benefit that is lost when antibiotic therapy is discontinued.⁵ Considering that late antibiotic treatment of C pneumoniae infection does not seem to prevent atherogenesis in animal models,¹³ it is possible that inclusion of stable patients at a median time of 31 months after the acute event in the WIZARD trial may have reduced the chance of detecting a beneficial effect of azithromycin.

In the end, the main question is whether antibiotics fail to improve outcomes in coronary artery diseases because C pneumoniae does not play a role in atherogenesis or because the proper antibiotic regimen has not been used in the proper clinical setting. Before completely rejecting the "infection hypothesis" of atherosclerosis, clinicians should remember lessons from the past. In the 1950s, streptokinase therapy for MI was considered only marginally effective.¹⁴ It took more than 20 years and large-scale clinical trials to realize that improper patient selection and inadequate dosage were responsible for the limited efficacy noted in initial studies; today thrombolytic therapy is established as one of the most beneficial interventions for patients with acute MI. Will antibiotic therapy follow the same pattern? It is hoped that ongoing large clinical trials with longer-term antibiotic therapy (such as the Azithromycin and Coronary Events Study [ACES]¹⁵ and the Pravastatin or Atorvastatin Evaluation and Infection Therapy [PROVE IT]¹⁶ study—a gatifloxacin trial) will help provide more answers. Until then, the evidence appears insufficient to support use of antibiotics for primary or secondary prevention of coronary heart disease.