The effect of azithromycin and Chlamydophilia pneumonia infection on expansion of small abdominal aortic aneurysms - A prospective randomized double-blind trial
Article Outline
Objective
The aim of the study was to evaluate the effect of azithromycin on the expansion rate of small abdominal aortic aneurysms (AAAs), and to determine whether or not a correlation exists between serological markers for Chlamydophilia pneumonia (Cpn) infection and AAA expansion.
Methods
Nine vascular centers were included and 259 patients were invited to participate. Ten patients declined and 2 patients had chronic kidney failure, leaving a total of 247 patients. Inclusion criteria were: AAA 35-49 mm and age <80 years. Patients were randomized to receive either azithromycin (Azithromax, Pfizer Inc, New York, NY) 600 mg once daily for 3 days and then 600 mg once weekly for 15 weeks, or placebo in identical tablets. The ultrasound scans were performed in a standardized way within a month before inclusion and every 6 months for a minimum follow-up time of 18 months. Cpn serology was analyzed in blood samples taken at inclusion and 6 months later. Serum was analyzed for Cpn IgA and IgG antibodies by microimmunofluorescence (MIF). Computed tomography (CT) scans were done in 66 patients at inclusion and at 1 year for volume calculations.
Results
Thirty-four patients were excluded, ie, could not be followed for 18 months, 20 in the placebo group and 16 in the active treatment group. A total of 211 patients had at least two measurements and all were analyzed in an intention-to-treat analysis. Detectable IgA against Cpn was found in 115 patients and detectable IgG against Cpn in 160 patients. No statistically significant differences were found between the groups regarding median expansion rate measured by ultrasound scan (0.22 cm/year, interquartile range [IQR]: 0.09 to 0.34 in the placebo group vs 0.22, IQR: 0.12 to 0.36 in the treatment group, P = .85). Volume calculation did not change that outcome (10.4 cm3/year in the placebo group vs 15.9 cm3/year in the treatment group, P = .61). No correlation was found between serological markers for Cpn infection and the expansion rate. Patients taking statins in combination with acetylsalicylic acid (ASA) had significantly reduced expansion rate compared to patients who did not take statins or ASA, 0.14 cm/year vs 0.27 cm/year, P < .001.
Conclusion
Azithromycin did not have any effect on AAA expansion. No correlation was found between serological markers for Cpn and AAA expansion, indicating no clinical relevance for Cpn testing in AAA surveillance. However, a significant reduction in AAA expansion rate was found in patients treated with a combination of ASA and statins.
Small abdominal aortic aneurysms (AAAs) are characterized by a massive inflammatory infiltrate in the adventitia and media of the aneurysm wall.1 It has been shown that there is a correlation between this infiltrate and the expansion of AAAs.2 This process seems to have a late onset, as AAA has a prevalence of 4-8% in men 65 years of age and older. Chlamydophilia pneumonia (Cpn) is a widespread pathogen and about 80% of men and 70% of women have been infected by the age of 65-years-old.3 The Cpn antigen has been found in the majority of abdominal aortic tissues using immunohistochemical (IHC) techniques.4, 5 Moreover, Vink et al6 reported the finding of Cpn by IHC at multiple locations in the arterial system within the same individual. IgA is an antibody with a short half-life,7 and high titers of this isotype have been associated with persistent infections.8 Lindholt et al9 have suggested that expansion of AAA is associated with increased titers of Cpn IgA antibody. Studies on rabbits have shown that Cpn can induce atherosclerosis and aortic dilatation, and that these lesions can be reversed by antibiotic therapy.10, 11 Antibiotics have been reported to temporarily reduce AAA expansion.12, 13 In other studies, aspirin and statins have been reported to inhibit the in vitro growth of Cpn in cell cultures.14, 15
We have earlier reported the viability of Cpn from AAA tissue, and higher antibody titers against Cpn in patients operated for AAA as compared to age and gender matched controls.5 Here we present a prospective randomized double-blind trial where azithromycin is compared to placebo for effects on the expansion of small AAAs. An ultrasound scan was used for all randomized patients and computed tomography (CT) scans for a subgroup in order to calculate aneurysmal volume. A possible correlation between Cpn antibodies and AAA expansion, as well as the role of acetylsalicylic acid (ASA) and statins were also studied.
Patients and methods
The study was a prospective randomized double-blind intention-to-treat trial, running between 2002 and 2005. Inclusion criteria were: AAA 35-49 mm, age ≤80 years. Exclusion criteria: AAA <35 mm or AAA ≥50 mm, age ≥81 years, intolerance to macrolide antibiotics, creatinine clearance <40 mL/minute, and medication with ergotamine. Included patients were both those who had previously been diagnosed with AAA and newly diagnosed patients meeting the inclusion criteria. Patients were randomized in a 1:1 proportion in blocks of four to receive either azithromycin (Azithromax, Pfizer Inc) 600 mg once daily for 3 days and then 600 mg once weekly for 15 weeks, or placebo in identical tablets. Pfizer AB (Sollentuna, Sweden), generously provided the study medication and matching placebo. Randomization was done in a standardized way at Kronans Droghandel (Moelnlycke, Sweden) and the medication code was kept at the Hospital Pharmacy (Akademiska Hospital, Uppsala, Sweden). The code was not revealed until all analyses and follow-up scans were completed.
Nine vascular centers were included and 259 patients were invited to participate. Ten patients declined and 2 patients could not be included because of chronic kidney failure, leaving a total of 247 patients (Fig 1). Patients were recruited from surveillance programs in each center and all participants gave written consent. Medical history, smoking habits at inclusion and after 18 months (by oral confession), heredity for AAA, and current medication were also recorded. Blood was sampled for control of kidney function, and serum taken at inclusion and 6 months later was stored at −70°C until analyzed for antibodies to Cpn. Each patient kept a record of their study medication, and these forms were collected at 6 months. Surgery was performed according to clinical practice at each center. An AAA was “allowed” to grow to the size where the surgeon in charge decided for surgical intervention using standard clinical practices. The decision for intervention was based upon the “last” ultrasound scan in the study; no additional ultrasound scan measurement was done. An operation was considered as an endpoint.
Fig 1.
Trial profile.
All participating centers performed ultrasonography scans in a standardized way. The widest anterior-posterior diameter was measured in both axial and transverse angles. An ultrasound scan was done within a month before inclusion and every 6 months for a minimum follow-up time of 18 months.
CT scans were done in 66 patients at inclusion and at 1 year for volume calculations, using a Siemens Somatom Volume Zoom (Siemens Medical Solutions, Erlangen, Germany). Eight centers used Acuson Sequoia (Universal Diagnostic Solutions, Oceanside, Calif), one used Siemens Elegra and all used a 4 MHz probe. Each patient analysis was run using either 3 or 5 mm slices. One experienced person did the volume calculations using a Siemens Somatom Workstation on one occasion.
Chlamydophilia pneumonia serology
All blood samples were taken at the department of clinical chemistry at each hospital. Sera were separated from erythrocytes immediately and frozen at −70°C until tested. All sera from each patient were analyzed at the same testing occasion. Cpn-antibodies were analyzed by microimmunofluorescence (MIF) using a previously standardized and optimized technique and reagents.16, 17 Sera were diluted 1:32 with phosphate buffered saline (PBS), pH 7.4 and tested for IgG and IgA antibodies using 21-well antigen slides with elementary bodies of C. psittaci, C. trachomatis, and C. pneumoniae in each test well (Thermobio Oy, Helsinki, Finland). Sera found to be positive in screening dilutions for IgG were re-diluted and tested in doubling dilutions. Sera positive for IgA were first absorbed with GullSorb (Gull Laboratories, Salt Lake City, Utah) to remove IgG, and then tested in doubling dilutions. Serum dilutions were incubated with antigen for 14 to 16 hours at +4°C. Slides were washed thoroughly in three changes of PBS pH 7.4, then incubated with fluorescein isothiocyanate (FITC) conjugated rabbit anti-human IgG, IgA or IgM (Dakopatts, BioPortfolio, Glostrup, Denmark) at +37°C for 30 minutes. Control sera with specified high and low titers were used on each testing occasion and tests were accepted only if the titers were within one dilution step of the predetermined mean titers for the respective quality control sera. All tests were read using an Olympus model BX40F4 UV microscope (Olympus Optical Co, LTD, Tokyo, Japan) with an ×50 oil immersion lens and an ×10 ocular lens (total magnification ×400) by an experienced microbiologist. All titers were reported as reciprocals, and all sera from the same individuals were tested on the same antigen slide on the same testing occasion.
Statistical analysis
The primary endpoint was the expansion rate of the AAA. For the purpose of sample size calculation, the standard error of the variable diameter in mm was estimated to 3 mm and the measuring uncertainty +/− 2 mm. This means that the root mean square error (RMSE) would be approximately √13. The number of measuring occasions would be a minimum of four. The expected annual expansion rate was assumed to be 0.26 cm/year for the placebo group. With these assumptions, 100 patients in each group would be sufficient to have a power of 80% to detect an annual difference of 0.5 mm in expansion rate between the groups at a significance level of 0.05. Categorical variables were summarized by percentages and counts and continuous variables by medians and inter-quartile ranges (IQR). For each patient with at least two measurements, the average expansion rate of the AAA diameter was estimated through a linear regression model. The distribution of the estimated slopes was subsequently compared between groups using the Wilcoxon-Mann-Whitney test. For IgA and IgG, an average change of the Log2-level of the concentration was estimated in a similar way with a linear regression model for each patient. The number of patients with decreasing, unchanged, or increasing levels was compared between groups using the χ2 tests. The primary objective was to compare azithromycin treatment with placebo. All other analyses are to be seen as hypothesis generating. All statistical analyses were done using R version 2.5.1.18 There were no interim analyses.
Ethics
All patients gave written consent and the study was approved by the Committee of Ethics of Uppsala University as well as local Ethic Committees, Dno: 01-349, Health Sciences Ethics Board, University of Alberta, Edmonton, Canada, and by the Swedish Medical Products Agency, Dno: 151:1254/01. The trial was registered according to the Law of Biological Samples Record, no. 543.
Results
Two hundred thirteen patients could be followed for a minimum of 18 months and 211 patients were measured at least twice. Five patients stopped taking the study medicine (2 in placebo and 3 in the active group). As this was an intention-to-treat study, these patients were followed and analyzed as they were randomized. Detectable IgA against Cpn was found in 115 and detectable IgG against Cpn in 160 patients with sera taken at inclusion and 6 months. As seen in Table I, the two groups were well balanced except for the history of stroke where 18.7% had stroke in the placebo group compared with 9.4% in the treatment group. There were 101 patients (52 in the placebo group and 49 in the treatment group) on low dose (75 mg) ASA, and 84 patients (43 in the placebo group and 41 in the treatment group) were on statins (Table I).
Table I. Baseline characteristics, 121 patients had detectable IgA at one occasion and 172 patients had detectable IgG on one occasion
| N | Control (n = 107) | Treatment (n = 106) | |
|---|---|---|---|
| N median (IQR) | N median (IQR) | ||
| Age; years | 213 | 71 (67,76) | 71(67,74) |
| Weight; kg | 210 | 76.0(70.0,85.0) | 79.5(69.0,89.0) |
| Height; cm | 207 | 174(169,178) | 174(169,181) |
| Gender: male | 213 | 73.8(79) | 79.2(84) |
| Previous MI | 213 | 32.7(35) | 29.2(31) |
| Previous stroke | 213 | 18.7(20) | 9.4(10) |
| Asthma | 213 | 3.7(4) | 3.8(4) |
| COPD | 213 | 7.5(8) | 3.8(4) |
| Current smoker | 213 | 37.4(40) | 37.7(40) |
| Smoking at 18 months | 68 | 89.7(35) | 94.3(33) |
| Package year | 187 | 40(30,50) | 40(30,50) |
| Heredity | 211 | 13.3(14) | 13.2(14) |
| Previous vasc. op | 213 | 27.1(29) | 28.3(30) |
| Adverse effect | 212 | 7.5(8) | 12.4(13) |
| Diabetes | 213 | 3.7(4) | 4.7(5) |
| Hypertension | 213 | 57.0(61) | 60.4(64) |
| Statin medication: yes | 213 | 40.2(43) | 39.6(42) |
| ASA medication: yes | 212 | 48.6(52) | 46.7(49) |
| Ultrasound: median size | 213 | 4.00(3.60,4.35) | 4.00(3.70,4.40) |
| IgA:16 | 121 | 15.0(9) | 13.1(8) |
| 32 | 11.7(7) | 19.7(12) | |
| 64 | 16.7(10) | 18.0(11) | |
| 128 | 11.7(7) | 9.8(6) | |
| 256 | 21.7(13) | 23.0(14) | |
| 512 | 13.3(8) | 9.8(6) | |
| 1024 | 5.0(3) | 4.9(3) | |
| 2048 | 1.7(1) | 1.6(1) | |
| 4096 | 3.3(2) | 0.0(0) | |
| IgG:32 | 172 | 11.4(10) | 4.8(4) |
| 64 | 5.7(5) | 10.7(9) | |
| 128 | 19.3(17) | 11.9(10) | |
| 256 | 13.6(12) | 22.6(19) | |
| 512 | 29.5(26) | 27.4(23) | |
| 1024 | 10.2(9) | 14.3(12) | |
| 2048 | 2.3(2) | 3.6(3) | |
| 4096 | 6.8(6) | 2.4(2) | |
| 8192 | 1.1(1) | 2.4(2) |
As seen in Table II and Fig 2, no statistically significant differences were found between the groups regarding median expansion rate measured by ultrasound scan (0.22 cm/year; IQR, 0.09 to 0.34 in the placebo group vs 0.22; IQR, 0.12 to 0.36 in the treatment group, P = .85). Volume calculation, ie, the difference in median growth of the volume, between 32 patients in the placebo group vs 34 in the treatment group, did not change that outcome: (10.4 cm3/year in the placebo group vs 15.9 cm3/year in the treatment group, P = .61). Individual growth rates regarding gender, age, smoking habits, history of myocardial infarction, previous stroke, asthma, chronic obstructive pulmonary disease (COPD), heredity for AAA, previous vascular operations, and hypertension are shown in Table II. Patients on statins had a lower expansion rate than those who did not: (0.16 vs 0.25 cm/year; P value, .008). In addition, patients on ASA had lower expansion rates than those not taking ASA (0.18 vs 0.26 cm/year; P value, .004). Further analysis demonstrated that patients taking statins in combination with ASA had significantly reduced expansion rate compared to patients who did not take statins or ASA, 0.14 cm/year vs 0.27 cm/year, P < .001 (Table III).
Table II. Summary of crude individual growth rates in cm/year. N is the number of non-missing observations, Med is the median, and Q1/Q3 is the lower and upper quartiles, P values from the Wilcoxon-Mann-Whitney test
| Subgroup | N | Q1 | Med | Q3 | P value |
|---|---|---|---|---|---|
| Treatment | |||||
| Treatment | 105 | 0.12 | 0.22 | 0.36 | .85 |
| Control | 106 | 0.01 | 0.22 | 0.34 | |
| Gender | |||||
| Women | 50 | 0.09 | 0.19 | 0.28 | .216 |
| Men | 161 | 0.12 | 0.23 | 0.37 | |
| Age | |||||
| ≤ 71 yrs | 110 | 0.08 | 0.20 | 0.32 | .118 |
| > 71 yrs | 101 | 0.13 | 0.23 | 0.43 | |
| Smoking | |||||
| No | 132 | 0.08 | 0.22 | 0.36 | .857 |
| Yes | 79 | 0.12 | 0.23 | 0.32 | |
| Prev MI | |||||
| No | 146 | 0.01 | 0.22 | 0.33 | .826 |
| Yes | 65 | 0.10 | 0.22 | 0.38 | |
| Perestroika | |||||
| No | 182 | 0.09 | 0.22 | 0.36 | .638 |
| Yes | 29 | 0.14 | 0.22 | 0.35 | |
| Asthma | |||||
| No | 203 | 0.10 | 0.22 | 0.35 | .781 |
| Yes | 8 | 0.00 | 0.21 | 0.36 | |
| COPD | |||||
| No | 199 | 0.10 | 0.22 | 0.36 | .337 |
| Yes | 12 | 0.08 | 0.20 | 0.28 | |
| Heredity | |||||
| No | 182 | 0.10 | 0.23 | 0.34 | .831 |
| Yes | 27 | 0.08 | 0.21 | 0.40 | |
| Prev Vasc Surg | |||||
| No | 153 | 0.01 | 0.22 | 0.33 | .370 |
| Yes | 58 | 0.10 | 0.26 | 0.38 | |
| Hypertension | |||||
| No | 87 | 0.09 | 0.22 | 0.34 | .720 |
| Yes | 124 | 0.10 | 0.22 | 0.35 | |
| Statin | |||||
| No | 127 | 0.13 | 0.25 | 0.39 | .008 |
| Yes | 84 | 0.06 | 0.16 | 0.32 | |
| ASA | |||||
| No | 110 | 0.13 | 0.26 | 0.40 | .004 |
| Yes | 101 | 0.061 | 0.18 | 0.30 |
Fig 2.
Diameter by time, by treatment – all patients. Lines show smoothed (based on the Lowess method) estimates of the relation.
Table III. Growth rates in cm/year, for patients not taking statins or ASA, statins or ASA separately, or in combination. N is the number of patients, Med is the median, and Q1/Q3 is the lower and upper quartiles
| N | Q1 | Med | Q3 | |
|---|---|---|---|---|
| Neither statins nor ASA | 84 | 0.17 | 0.27 | 0.40 |
| Only statins | 27 | 0.08 | 0.23 | 0.38 |
| Only ASA | 44 | 0.10 | 0.19 | 0.31 |
| Satins and ASA | 57 | 0.04 | 0.14 | 0.30 |
No correlation was found between serological markers for Cpn infection and the expansion rate (Table IV). When comparing the two groups, the distributions were similar indicating that there was no difference in changes of IgA levels over time. The majority of the patients had stable IgA levels at subsequent visits. There was a trend towards more patients who had a reduction in IgG levels in the treatment group compared with patients in the placebo group (Table IV).
Table IV. Changes in Cpn antibody-titers correlated to placebo vs active drug and ASA medication
| Subgroup | Decreasing | Unchanged | Increasing | |||
|---|---|---|---|---|---|---|
| N | (%) | N | (%) | N | (%) | |
| IgA | ||||||
| Control | 15 | (27.3) | 32 | (58.2) | 8 | (14.5) |
| Treatment | 17 | (28.3) | 34 | (56.7) | 9 | (15) |
| ASA No | 13 | (21.3) | 36 | (59) | 12 | (19.7) |
| ASA Yes | 19 | (35.2) | 30 | (55.6) | 5 | (9.3) |
| IgG | ||||||
| Control | 15 | (19.2) | 55 | (70.5) | 8 | (10.3) |
| Treatment | 22 | (26.8) | 54 | (65.9) | 6 | (7.3) |
| ASA No | 17 | (21.8) | 56 | (71.8) | 5 | (6.4) |
| ASA Yes | 20 | (24.4) | 53 | (64.6) | 9 | (11) |
Thirty-four patients were excluded, ie, could not be followed for 18 months for the following reasons: 13 deaths, none of which were AAA-related (death certificates were obtained from The Swedish National Board of Health and Welfare regarding all patients that died during the study period). All deaths outside a hospital and deaths from unknown causes were sent for autopsy. All expected in-hospital deaths were based on clinical information, 14 were operated on for AAA (all but 1 with symptoms went to surgery because of AAA size change), 2 could not be reached for follow-up, 2 had AAA >49 mm and should not have been included, 1 had an iliac aneurysm but no AAA, 1 had a disabling stroke, and 1 had a myocardial infarction and did not want further follow-up (Fig 1). The total study period was 36 months, and 29 patients were operated on for AAA in the 18-36 month surveillance period; 13 in the control group (1 with symptoms, 13 due to size change) and 16 in the active treatment group (1 with symptoms, 1 with rupture, and 14 due to size change).
Twenty-one patients, 8 in the control group and 13 in the active group reported side effects. All patients in the control group had gastrointestinal side effects and 2 stopped taking the medication. In the group with active substance, 3 stopped taking the study medication; 1 due to diarrhea, 1 due to arthralgia, and 1 who had an allergic reaction which initially was thought to be caused by study medication but turned out to be a reaction to antihypertensive medication. No patient reported taking other antibiotics while taking the study medication.
Discussion
The aim of the present study was to evaluate the effect of azithromycin on the expansion rate of small AAAs, and to determine whether or not a correlation exists between serological markers for Cpn infection and AAA expansion.
We have previously reported the presence of viable Cpn in the majority of tissue samples taken at their operation from the anterior wall of AAAs; a high proportion of these patients were found to have elevated specific IgG and/or IgA antibody titers, which could indicate chronic infection as compared to a control population.5 Lindholt et al9 reported that the expansion of AAA was related to the Cpn IgA antibody titer. Moreover, Tambiah et al11, 19 reported that Cpn caused dilatation of the aorta in an animal model and azithromycin prevented that process. Azithromycin is a macrolide with efficacy against Cpn, it achieves high tissue concentrations, has a long half-life, and is well tolerated by patients.20 Azithromycin may also exert anti-inflammatory activities,21 and it is reported to reduce of Cpn IgA titers after treatment with azithromycin.8
Cpn is a common cause of acute and chronic respiratory tract infections, although most of the infections result only in mild clinical symptoms,22 and might, therefore, not be diagnosed or treated with antibiotic, leaving Cpn infection “free” to initiate chronic/persistent damage. Furthermore, Cpn can infect and propagate in endothelial cells and smooth muscle cells of aorta.23 A chronic persistent infection in macrophages, endothelial cells, and smooth muscle cells can result in increased cytokine levels, and probably also in immune complex deposition, leading to a collection of lipids, macrophages, and thrombus. Human heat shock protein 60 (HSP 60) is associated with atherosclerosis and can provoke an autoimmune reaction.24 Chlamydial HSP 60 has close homology with and co-localizes with human HSP 60 in macrophages, stimulating enhanced production of MMP 9, an enzyme of vital importance in AAA pathogenesis.2
Although antibiotics have been reported to temporarily reduce AAA expansion12, 13 we could not find any difference in expansion rate between the groups. Vammen et al13 used roxithromycin, also a macrolide that has the same effect on Cpn as azithromycin. There are, however, some differences between that study and the present; we included patients from surveillance programs, ie, included both genders and all ages under 81 years rather than a selected screening cohort of men only, and we have twice as many patients in each group. Vammen et al13 reported that 1/3 of the eligible patients did not give informed consent, in the present study only 4% refused participation. Although Vammen et al13 used only two persons to perform the ultrasound scan, it was a strength of their study. The fact that the two study groups are different when it comes to gender and age distribution makes direct comparisons somewhat difficult.
Several studies indicate that when Cpn comes into contact with different agents such as tobacco smoke, antibiotics, or cytokines, a persistent phase is entered and the organism is not detectable by culture. The inflammatory reaction and subsequent aneurysmal formation might already be initiated despite Cpn entering a dormant, persistent phase.25, 26 We do not have data about penetration of azithromycin in AAA tissue nor the optimal dose or treatment length for this indication, however, we did use a higher dose of antibiotics and 12 times longer treatment periods compared to recommendations for community-acquired pneumonia. According to this, we believe that our choice of treatment regime is reasonable.
Smoking is a risk factor for AAA development and expansion.27, 28 Despite this, we could not detect any statistical difference in growth rate between smokers and non-smokers. The percentage of smokers (37% in both groups) is similar to the numbers reported in UK Small Aneurysm Trial (UKSAT).29 MacSweeney et al28 reports a difference in expansion rate between smokers vs non-smokers. Forty-three patients in all were followed in that study, and the difference in annual expansion rate was calculated to be 0.07 cm per annum. Our result is based on 40 patients/group, and is at odds with prior studies.30, 31 We did not use urine-cotinine levels to confirm smoking, which is a drawback in our study. The effect of smoking should perhaps be addressed in a prospective study to clarify doubts.
Statins in combination with ASA are commonly used for cardiovascular patients, and patients with an AAA have a high prevalence of coronary heart disease.32 Statins suppress the expansion of AAA33, 34, 35 via inhibition of MMP-9, and may also suppress the development of experimental AAA.15 It is, however, not clear whether this is a class effect of statins. Subgroup analysis demonstrated that patients taking statins had lower expansion rate than those who did not take statins, and are in line with what Schlösser et al33 and Schouten et al34 report.
It was found that patients on ASA had lower expansion rates compared with those not on ASA, indicating that ASA might have a role in slowing AAA progression. High dose ASA might also inhibit chlamydial growth.14 The results were observed also after adjustment between those groups. Statins and ASA have different anti-inflammatory properties, which might explain the complementary effect shown in this study, although ASA seems to be more effective than statins (Table III). These results are remarkable but as they are not primary endpoints they should be considered hypothesis-generating only. However, it might indicate that long-term medication with substances with anti-inflammatory properties could slow expansion of AAAs,36 supporting inflammation as part of AAA pathogenesis. The observation that statins and ASA may slow AAA expansion warrants further study.
The fact that the placebo group had a higher percentage with earlier stroke than the antibiotic group may be a confounding factor, as Cpn antibodies may be associated with risk of stroke.37
The strength of this study is that it is a prospective randomized double-blind investigation and we used not only ultrasound scans for AAA measurement, but also volume calculation with CT in a subgroup at inclusion and after 1 year. As no modality could detect any difference between the groups, we are comfortable with the results, especially as volume calculation might reflect the “true” expansion compared to ultrasound scan. This study can not exclude Cpn as a participant in AAA pathogenesis, because the damage could already be present, or Cpn might have entered a persistent phase and, therefore, not be accessible for antibiotic treatment.
We conclude that azithromycin treatment did not have any effect on AAA expansion, and there was no correlation between serological markers for Cpn activity and AAA expansion, indicating no clinical relevance for C. pneumoniae testing in AAA surveillance.
Author contributions
The authors want to thank the following colleagues for their recruiting and surveillance of study patients: Anders Henriksson and Marita Nyberg in Sundsvall, Torbjörn Tuveson and Marie Nätterlund in Gävle, Lotta Marquard in Uppsala, Bo Westman, Helena Danielsson, and Carina Andersson in Falun, Tobias Kjellberg and Irené Jensen in Karlstad, Ken Eliasson in örebro, Hans Ravn, Karin Day, and Britt Lantz in Eksjö, and Ingvar Jansson and Ylva Nilsson in Eskilstuna.
References
- . Pathogenesis of abdominal aortic aneurysms - cellular and biochemical mechanisms. Eur J Vasc Endovasc Surg. 1996;12:391–400
- . Metalloproteinases in degenerative aortic disease. Clin Sci (Lond). 1991;81:233–239
- . Chlamydia pneumoniae (TWAR) in coronary arteries of young adults (15-34 years old). Proc Natl Acad Sci USA. 1995;92:6911–6914
- Demonstration of Chlamydia pneumoniae in the walls of abdominal aortic aneurysms. J Vasc Surg. 1997;25:499–505
- . Detection of viable Chlamydia pneumoniae in abdominal aortic aneurysms. Eur J Vasc Endovasc Surg. 2000;19:630–635
- . Distribution of Chlamydia pneumoniae in the human arterial system and its relation to the local amount of atherosclerosis within the individual. Circulation. 2001;103:1613–1617
- . Structure and function of immunoglobulin A. Prog Allergy. 1972;16:181–213
- . Persistent Chlamydia pneumoniae infection in a Swedish family. Scand J Infect Dis. 1996;28:271–273
- . Indicators of infection with Chlamydia pneumoniae are associated with expansion of abdominal aortic aneurysms. J Vasc Surg. 2001;34:212–215
- . Can an antibiotic (macrolide) prevent Chlamydia pneumoniae- induced atherosclerosis in a rabbit model?. Clin Diagn Lab Immunol. 1999;6:891–894
- . Chlamydia pneumoniae antigens facilitate experimental aortic dilatation: prevention with azithromycin. J Vasc Surg. 2002;36:1011–1017
- . Use of doxycycline to decrease the growth rate of abdominal aortic aneurysms: a randomized, double-blind, placebo-controlled pilot study. J Vasc Surg. 2001;34:606–610
- . Randomized double-blind controlled trial of roxithromycin for prevention of abdominal aortic aneurysm expansion. Br J Surg. 2001;88:1066–1072
- . Aspirin inhibits Chlamydia pneumoniae-induced nuclear factor-kappa B activation, cytokine expression, and bacterial development in human endothelial cells. Arterioscler Thromb Vasc Biol. 2002;22:1075–1080
- Simvastatin suppresses experimental aortic aneurysm expansion. J Vasc Surg. 2006;43:117–124
- . Serological analysis of specific IgA to Chlamydia pneumoniae: increased sensitivity of IgA antibody detection using prolonged incubation and high antigen concentration. APMIS. 2000;108:357–362
- . Comparison of a new commercial EIA kit and the microimmunofluorescence technique for determination of IgG and IgA antibodies to Chlamydia pneumoniae. APMIS. 2000;108:819–824
- R Development Core Team (2008). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria. ISBN 3-900051-07-0, URL http://www.R-project.org.
- . Provocation of experimental aortic inflammation and dilatation by inflammatory mediators and Chlamydia pneumoniae. Br J Surg. 2001;88:935–940
- . Meta-analysis of randomized controlled trials on the comparative efficacy and safety of azithromycin against other antibiotics for lower respiratory tract infections. J Antimicrob Chemother. 2001;48:691–703
- . Effect of azithromycin treatment on endothelial function in patients with coronary artery disease and evidence of Chlamydia pneumoniae infection. Circulation. 2002;105:1298–1303
- . A new respiratory tract pathogen: Chlamydia pneumoniae strain TWAR. J Infect Dis. 1990;161:618–625
- . Replication of Chlamydia pneumoniae in vitro in human macrophages, endothelial cells, and aortic artery smooth muscle cells. Infect Immun. 1996;64:1614–1620
- . Autoantibodies against heat shock protein 60 mediate endothelial cytotoxicity. J Clin Invest. 1995;96:2569–2577
- . Tobacco smoke induces persistent infection of Chlamydia pneumoniae in HEp-2 cells. Microb Pathog. 2004;37:141–148
- . First-choice antibiotics at subinhibitory concentrations induce persistence of Chlamydia pneumoniae. Antimicrob Agents Chemother. 2004;48:1402–1405
- . Smokers‘ relative risk for aortic aneurysm compared with other smoking-related diseases: a systematic review. J Vasc Surg. 2003;38:329–334
- . Smoking and growth rate of small aortic aneurysms. Lancet. 1994;344:651–652
- . Mortality results for randomised controlled trial of early elective surgery or ultrasonographic surveillance for small abdominal aortic aneurysms. Lancet. 1998;352:1649–1655
- . Abdominal aortic aneurysm expansion: risk factors and time intervals for surveillance. Circulation. 2004;110:16–21
- . Smoking, but not lipids, lipoprotein(a) and antibodies against oxidised LDL, is correlated to the expansion of abdominal aortic aneurysms. Eur J Vasc Endovasc Surg. 2001;21:51–56
- . Prevalence of coronary artery disease, lower extremity peripheral arterial disease, and cerebrovascular disease in 110 men with an abdominal aortic aneurysm. Am J Cardiol. 2004;94:1358–1359
- . J Vasc Surg. 2008;47:1127–1133
- Statins are associated with a reduced infrarenal abdominal aortic aneurysm growth. Eur J Vasc Endovasc Surg. 2006;32:21–26
- . Simvastatin attenuates the activity of matrix metalloproteinase-9 in aneurysmal aortic tissue. Eur J Vasc Endovasc Surg. 2007;34:302–303
- . Vascular surgical society of Great Britain and Ireland: non steroidal anti-inflammatory drugs to treat abdominal aortic aneurysm. Br J Surg. 1999;86:707
- . Chlamydia pneumoniae but not cytomegalovirus antibodies are associated with future risk of stroke and cardiovascular disease: a prospective study in middle-aged to elderly men with treated hypertension. Stroke. 1999;30:299–305
The Research was supported by grants from County of Gävleborg Research and Development Center (FoU), Gore Swedish Research Foundation, Pfizer AB Sweden, Schyberg medical research fund, and Zoega medical research fund. We thank Prof Martin Björck for initial funding for the CT scans.
Competition of interest: none.
PII: S0741-5214(08)02287-8
doi:10.1016/j.jvs.2008.12.048
© 2009 Society for Vascular Surgery. Published by Elsevier Inc. All rights reserved.
