Two Regimens of Azithromycin Prophylaxis against Community-Acquired Respiratory and Skin/Soft-Tissue Infections among Military Trainees
- 1Respiratory Disease Unit of the Military Hospital, Smolensk
- 2Medical Force of Moscow Military Region, Moscow, Russia
- 3Department of Epidemiology, College of Public Health University of Iowa, Iowa City, Iowa
- Reprints or correspondence: Dr. Igor A. Guchev, Respiratory Disease Unit of the Military Hospital, 214004, Smolensk, Russia (igouchev{at}mail.ru).
Abstract
Epidemics of community-acquired pneumonia (CAP) are a frequent cause of morbidity among Russian military trainees. We evaluated azithromycin prophylaxis against CAP. In 2001–2002, incoming military trainees were randomized to 1 of 3 trial arms by training group: azithromycin, 500 mg per week for 8 weeks (R1); azithromycin, 1500 mg once at enrollment (R2); or no therapy (R3). During the 22 weeks of training, CAP was diagnosed in 20.2% of 678 subjects in the R3 group, 8.6% of 508 subjects in the R1 group, and 10.3% of 507 subjects in the R2 group. Throat carriage cultures revealed that the proportion of Streptococcus pneumoniae isolates with resistance to macrolides correspondingly increased during the study, from 0% (all) to 40% (R1) and 22.6% (R2) by week 20. Azithromycin prophylaxis is effective against CAP in a healthy population of young men at transient high risk of disease; however, azithromycin use must be tempered with the possible concomitant risk of selection for resistant endemic pathogens.
As a result of crowded living conditions, physical and psychological stressors, and the periodic introduction of numerous pathogens, military training centers have frequently been the foci of epidemics of respiratory disease. Occasionally, these outbreaks have affected more than two-thirds of a training population [1, 2]. Community-acquired pneumonia (CAP) is the most frequent cause of morbidity in training centers of the Russian armed forces. CAP attack rates may be as high as 5.4% during the second month of training and 20% by the sixth month of training (I.A.G., unpublished data). Preventing these epidemics is a very difficult task. Although hygienic interventions may reduce morbidity [3], good hygiene alone is not entirely effective. Although some vaccines are available to prevent such respiratory infections (e.g., influenza and Streptococcus pneumoniae infection), no available vaccines exist for a number of important respiratory pathogens [4].
With the heterogeneous bacterial etiology of respiratory infections (S. pneumoniae, Mycoplasma pneumoniae, Chlamydia pneumoniae, or Streptococcus pyogenes) [1, 5–8] in the military training setting, the use of prophylactic antibacterial agents is the single most effective preventive measure. Another reason to use antibacterial therapy is the temporary suppression of host immune response in new military recruits, which predisposes them to nasopharyngeal content microaspiration. In addition, our studies have shown that as many as 50%–70% of Russian military trainees have nasopharyngeal carriage of S. pneumoniae.
Numerous antimicrobial prophylaxis treatments have been evaluated [1, 4, 7, 9–11], and the most promising has been a weekly dose of azithromycin [9, 10, 12]. At the same time, increased use of antibiotics is associated with the selection of resistant microorganisms and their spread in populations. This spread is particularly problematic in military populations, where the subjects are mobile and the potential for transmission is high [13]. Therefore, public health officials who serve military populations must weigh the benefit of the use of antimicrobial prophylaxis with the potential risk of increased trainee morbidity due to antimicrobial-resistant strains.
We sought to assess the safety and effectiveness of 2 regimens of azithromycin chemoprophylaxis on the incidence of acute respiratory tract infections (ARIs), CAP, and skin/soft-tissue infections (SSTIs) in a population at very high risk of infection. Although our study was not specifically designed for this purpose, as a secondary goal, we sought to learn whether chemoprophylaxis influences the spread of drug-resistant pneumococci in a crowded population.
Patients and Methods
Study design. This was an open-label, prospective, comparative clinical trial.
Study population. The study was conducted at a regional training camp in central Russia. This camp provides 22 weeks of basic training to military trainees from November until May. Trainees are boys or men aged 8–24 years who meet military physical examination standards, which prohibit the enrollment of young men with chronic diseases. They train in 100–120-person groups that have little contact with other groups. Training groups are generally housed on the same floors of the same barracks. Beds are positioned close together, with ∼12 m2 of floor space allotted per trainee.
Treatment groups. The previous 3 years of CAP incidence data were use to select a seasonal time period with the greatest likelihood of CAP. From November 2001 to late December, new conscripts were randomly aggregated to 100–120-person training groups. We used a random-number generator to assign each of these training groups to 1 of the 3 treatment arms: azithromycin, given orally in a 500-mg tablet once per week for 8 weeks (R1); azithromycin, given orally at 1500 mg (3 tablets) once at enrollment (R2); and no therapy (R3). Trainees who reported an allergy to macrolides, who were already taking any other antibacterial drugs, or who had evidence of moderate/severe ARIs were not permitted to participate.
Drug safety assessment. All patients who took ⩾1 dose of study medication were identified for drug safety analysis. A safety assessment was performed by medical personnel during medical examinations during the 4 or 11 weeks (R2 and R1, respectively) of observation.
Safety guaranties for the study population. Because of the possibility of selection and spread of macrolide-resistant strains of S. pneumoniae, most study subjects who developed respiratory infections received penicillin or amoxicillin. Less frequently, lincomycin (for SSTIs) or doxycycline were used. To reduce confounding, macrolides were excluded from the list of antibacterials available for medical care.
Data and specimen collection. At enrollment (visit 0), all study subjects were asked about signs and symptoms of ARIs and SSTIs. Three hundred subjects (culture cohort) from groups R1, R2, and R3 were further asked to permit a throat culture at enrollment. On the next day (visit 1), groups R1 and R2 received oral azithromycin under the supervision of medical staff. The other 7 doses in the R1 group were administered weekly during visits 2–8. During training week 10 (visit 9), the culture cohort provided a second swab sample for culture. Twenty-one weeks after enrollment (visit 10), the culture cohort provided a third throat swab sample. They were additionally asked about the signs and symptoms of nonreported ARIs and SSTIs.
During the 22-week period, trainees were continually monitored and closely examined once per day by regimental medical personal. Those who had signs of ARI or SSTIs were referred to the field medical hospital, where clinical data were collected and diagnoses were confirmed or rejected.
Clinical trial outcomes were identified as (1) ARIs (common cold, acute bronchitis, and/or acute bacterial sinusitis), (2) CAP, and (3) SSTIs (boils [a painful sore with a hard, pus-filled core] and/or other signs of mild soft-tissue infection). A case of common cold was defined as a mild respiratory illness presenting with an oral temperature of >37.2°C for no less than 24 h, accompanied by ⩾1 of the following symptoms: rhinorrhea and nasal congestion, sore throat, cough, sinus tenderness, rales, rhonchi, or wheezing on auscultation. A case of acute sinusitis was defined as ⩾7 days of local facial pain (that may radiate to maxillary teeth), purulent nasal and/or retronasal discharge, and pain on sinus percussion. CAP was defined as acute respiratory signs and symptoms with radiographic evidence of a new chest infiltrate, as read by 2 independent radiologists. If temperature, cough, diffuse rhonchi, rales, and/or breathlessness were observed in the absence of radiographic evidence of chest infiltration, this condition was classified as acute bronchitis. In questionable cases, additional laboratory signs of CAP were taken into account, such as WBC counts, differential WBC counts, and C-reactive protein level.
Laboratory studies. All laboratory personnel were blinded to the study treatment groups. Determination of complete blood cell counts and C-reactive protein level, isolation of organisms, and culturing of nasopharyngeal swabs for S. pneumoniae carriage were performed at the local military laboratory. Culture swabs were plated on Columbia blood agar (bioMérieux), incubated overnight at 35°C in a 5% CO2 environment, and sent to a central laboratory for reading. Bacterial identification was performed with Optochin (bioMérieux) susceptibility and bile (10% Natrii desoxycholate; Sigma) solubility tests. Isolates were frozen at -70° C in glycerol-containing broth. Susceptibility testing was performed by broth microdilution in cation adjusted Mueller-Hinton broth (Becton Dickinson) supplemented with 5% lysed horse blood. Inocula were prepared by suspending growth from overnight cultures in sterile Mueller-Hinton broth to a turbidity equal to a 0.5 McFarland standard. A 1 : 10 dilution was made in Mueller-Hinton broth so that the final inocula contained 5–104 cfu/spot. Plates were inoculated with a Steers replicator with 3-mm-diameter inoculating pins, delivering 1 µL of inoculum, and incubated overnight at 35° C in air. The lowest concentration of antimicrobial showing no growth was read as the MIC. Standard quality-control strains, including S. pneumoniae (ATCC 49619), were included with each run. All isolates were tested for susceptibility to erythromycin, azithromycin, clarithromycin, clindamycin, tetracycline, amoxicillin, and penicillin G. The laboratory used the antimicrobial breakpoints defined by the NCCLS [14].
Statistical analysis. Statistical analyses were performed with SAS software, version 8.2 (SAS Institute). General descriptive statistics (absolute and relative frequencies, means, medians, and ranges) were used to describe the study data. Fisher's exact test, analyses of variance, and risks ratios were used for comparison of data and for examination of data associations. All persons who finished the study (per-protocol population) were of primary interest, and all but safety results were calculated for per-protocol population. Assuming a untreated CAP incidence of 10%, we could be 95% confident to detect a 50% reduction in CAP with 80% power by assigning 474 subjects to each study arm.
Results
From 27 November 2001 to 10 January 2002, a total of 1798 recruits were enrolled onto the study by their 100–120-person training group. Among the enrolled, data sufficient for analysis (complete data) were obtained from 1733 subjects (per-protocol population). The rest (n = 65) withdrew from the study prematurely (before visit 4) without any signs of ARIs, CAP, or SSTIs. All 3 study groups had very similar characteristics (table 1). The groups were similar with regard to age, race, body mass index, duration of smoking, and living and military service conditions.
Table 1
Subject characteristics of Russian military trainees who received azithromycin prophylaxis against community-acquired pneumonia.
There were no significant differences in time to the start of prophylactic intervention between the R1 and R2 groups. Seventy-five percent of subjects from the R1 and R2 groups received the first dose of the study drug within the first 3 days after their arrival to the unit.
During the 22 weeks of follow-up, the incidence of pneumonia in the control group (R3) significantly exceeded the incidences in the treatment groups (P < .001) (table 2). There was no statistically important difference in CAP incidence between the R1 and R2 groups (P = .19). Comparisons of the course of pneumonia in the prophylactic and control groups did not demonstrate important differences either in their severity or in their subsequent response to antibacterial chemotherapy. There was no difference in the mean duration of hospitalization and convalescence (24 days), and there were no cases of severe CAP requiring intensive care unit admission. The number of patients needed to treat to prevent 1 case of CAP was 8.2 and 8.1 for the R2 and R1 regimens, respectively.
Table 2
Incidences of and risk ratios for clinical outcomes of Russian military trainees who received azithromycin prophylaxis against community-acquired pneumonia over 22 weeks of military training, by study group.
A protective influence of prophylaxis was also noted for the outcomes of ARI and specifically for common cold and acute bronchitis. In addition, there was a reduction in the number of cases of SSTIs (specifically for boils) in the R2 group. A similar protective effect was not observed for acute sinusitis or for SSTIs other than boils (table 2).
Both prophylactic regimens had high patient compliance (no subjects refused to take the prophylaxis). Only 1 case of mild diarrhea (R1 group) was reported as definitely associated with azithromycin. Azithromycin therapy for this subject was discontinued after 3 doses. Other mild reactions that were presumptively associated with azithromycin included abdominal pain lasting <2 days (4 cases in the R1 group and 8 cases in the R2 group) and nausea (3 cases in the R2 group). In addition, 2 unexplained cases of mild diarrhea and 2 cases of nausea were reported in the R3 group.
During assessment of nasopharyngeal carriage at visit 0, S. pneumoniae was isolated in 34.6%–43.3% of subjects (intergroup difference, P > .05). All pneumococci were susceptible to oxacillin (table 3). At subsequent visits, as a result of the temporary absence of the subjects who had been initially chosen for bacteriologic analysis (culture cohort), new subjects were added to each bacteriological study groups (table 1). It was noted that pneumococcal carriage was significantly more frequent in the intervention groups. At week 10, carriage S. pneumoniae rates in the R1, R2, and R3 groups were 75.6% (95% CI, 65.1%–84.2%), 66.2% (95% CI, 58.1%–73.7%), and 50.9% (95% CI, 37.1%–64.6%), respectively (P < .05). At week 21, S. pneumoniae carriage rates were 69.2% (95% CI, 60.6%–76.9%), 56.9% (95% CI, 46.7%–66.6%), and 35.6% (95% CI, 21.9%–51.2%) in the same respective groups (P < .05). However, these findings must be tempered with the observation that there was essentially no intergroup difference in S. pneumoniae carriage during visits 9 and 10, suggesting that the differences noted at week 21 could be influenced by chance.
Table 3
Nasopharyngeal Streptococcus pneumoniae sensitivity patterns in Russian military trainees who received azithromycin prophylaxis against community-acquired pneumonia and who were evaluated at least twice, by treatment group.
At visit 0, no macrolide resistance was detected among the 40 S. pneumoniae isolates tested. The background level of intermediate penicillin resistance was estimated to be 0%–14% among all strains tested at day 0 (P > .05). A dramatic increase in the prevalence of macrolide resistance was observed from week 0 to week 10 in the R1 group (44 resistant strains [95.7%]; azithromycin-clindamycin resistance was present in 37% of them) and the R2 group (34 resistant strains [89.5%]; azithromycin-clindamycin resistance was present in 11.9% of them). By week 21, the prevalence of macrolide resistance in the R1 group decreased to 40% (n = 16; azithromycin-clindamycin resistance was present in 10%) and to 22.6% (n = 7; azithromycin-clindamycin resistance was present in 5.4%) in the R2 group. No concomitant increase in resistance to penicillin G or other β-lactams was observed.
It is of special note that we isolated pneumococci with simultaneous resistance to macrolides and lincosamides. This phenotype was most frequently found in the R1 study group. Its prevalence was 37% at visit 9. At the same visit, the described resistance phenotype was found in only 11.9% of isolates from the R2 group. Two macrolide-resistant strains were isolated from the control subjects, who did not receive any antibacterial chemotherapy during the observational period. We speculate that these isolates might represent resistant isolates acquired the subjects in the other treatment arms; however, molecular fingerprinting studies of isolates were not performed to investigate this.
Discussion
Preventing outbreaks of respiratory infections in crowded, high-risk populations is an extremely complicated task. Our data suggest that oral azithromycin prophylaxis is an effective and safe method for reducing acute respiratory infections and acute bronchitis, and for controlling CAP epidemics. It seems likely that the protective effect of azithromycin may impact not only the incidence of S. pneumoniae but also reduce M. pneumoniae and C. pneumoniae respiratory disease. An important finding in our study is the similar (absence of a statistically significant difference) protective effect of the 2 prophylactic regimens.
Because of numerous epidemics of acute respiratory infections among Russian military trainees, the evaluation of azithromycin as a prophylactic agent became a major focus of our attention. This antibiotic has been used to successfully control the spread of M. pneumoniae in high-risk civilian groups [15]. Good prophylaxis results have also been previously been documented among US military forces [4]. It is interesting that in this US study, azithromycin-resistant S. pneumoniae isolates were similarly detected after receipt of treatment.
Although our study and previous studies [4, 16, 17] have documented the increased risk of selecting macrolide-resistant pneumococci in populations where azithromycin is used prophylactically, others have not revealed this unfavorable trend [18–21]. Moreover, no one has reported increased morbidity and mortality as a sequelae of prophylaxis.
Our study has a number of limitations. Resource constraints restricted our ability to use a placebo control or to randomize subjects individually into treatment arms. Thus, it is possible that our study may have had a number of negative influences, such as case ascertainment bias and herd effects. However, we conducted daily active surveillance for respiratory morbidity to maximize illness detection and believe we captured most cases of respiratory illness. Our study is also limited in that a small proportion of subjects (1.4%, 9.3%, and 0.7% in the R1, R2, and R3 groups, respectively) transferred from one treatment group to another and to unstudied groups during the first month of study. This too may have slightly influenced the outcome measurements. Finally, as a result of combat field training for 1–1.5 months, ∼30% of the initial culture cohort was unavailable for subsequent microbiological evaluation, and they were substituted by other representatives of the same groups for carriage testing. However, despite these study limitations, the magnitude of the protective effect of azithromycin suggests that its impact was real and worth reporting.
The study has a number of strengths. Our expected capture of clinical disease is thought to be high because of the very close monitoring of study subjects. To validate outcomes, we additionally reviewed hospital case histories and discharge resumes. In addition, an independent radiologist performed repeated retrospective assessment of the radiographic films, helping us reduce false-positive CAP outcomes.
Both regimens of azithromycin prophylaxis were effective and safe interventions among populations of healthy young men at transient high risk of respiratory disease. Although the sampling strategies were less than ideal, our data also suggest that macrolide resistance increased over time among asymptomatically carried S. pneumoniae isolates recovered from the azithromycin treatment groups. Accordingly, we caution that if azithromycin prophylaxis is used among similar high-risk populations, care must be taken to monitor macrolide resistance.
Acknowledgments
We thank the line commanders and medical staff of the Kovrov Training Center, Vladimir Region, Russia. We thank the following professionals for their assistance in conducting the trial: Pervov Yuri and Ivanitsa Gregory (clinical coinvestigators; State Institute for Postgraduate Medical Training of the Ministry of Defense, Moscow, Russia), Shturmina Svetlana (microbiology; State Center for Disease Control and Prevention, Kovrov, Russia), Rosman Serge (microbiology; State Center for Disease Control and Prevention, Tver, Russia), and Krechikova Olga (microbiology; State Center for Disease Control and Prevention, Smolensk, Russia).
Footnotes
-
The views expressed in this article are those of the authors and do not reflect the official policy or position of the Ministry of Defense of the Russian Federation. This research has been conducted in compliance with all applicable federal regulations governing the protection of human subjects in research.
-
Financial support: Medical Service of Moscow Military Region.
- Received August 6, 2003.
- Accepted December 3, 2003.
- © 2004 by the Infectious Diseases Society of America
References
This Article
- AbstractFree
- » Full Text (HTML)Free
- Full Text (PDF)Free
Classifications
Services
Citing Articles
Google Scholar
PubMed
Related Content
Share
- Email this article
-
CiteULike
Connotea
Delicious
Facebook
Google+
Mendeley
Twitter
Published on behalf of
IDSA Members: For your free access to this journal, log in via the IDSA members area.
The Journal
Impact Factor: 8.186
5-Yr impact factor: 7.898
For The Media
For Authors
- Services for authors
- Self-archiving Policy
- IDSA Conflict of Interest policy
- Instructions for Authors
- ICMJE Form
- Submit Now!
Open access options for authors visit Oxford Open
This journal enables compliance with the NIH Public Access Policy
Corporate Services
Alerting Services
Related IDSA Publications
