Rectal chlamydial infection is a common bacterial sexually transmitted infection among men who have sex with men (MSM). In 2011, 11% of HIV-negative MSM and 21% of HIV-positive MSM attending a sexually transmitted disease (STD) clinic in the US tested positive for Chlamydia trachomatis in the rectum.1 More than half of all chlamydial infections among MSM are rectal infections,2,3 which are also more prevalent than urethral and rectal gonorrhea.3–5
Although rectal infections with gonorrhea and/or chlamydia are predominately asymptomatic,3 they are associated with an increased risk of HIV acquisition6–9 and may play a key role in sustaining high levels of chlamydia transmission among MSM. Consequently, screening MSM for these infections may be a cost-effective HIV prevention strategy10 and could be a critical component in the control of chlamydial infections in MSM.
These benefits of screening depend on the success of treatment. Currently, the Centers for Disease Control and Prevention (CDC) recommends doxycycline (100 mg orally twice daily for 7 days) or azithromycin (1 g orally once) for treatment of rectal chlamydia.11 However, few data on the efficacy of these regimens in the treatment of rectal chlamydia exist, and current guidelines primarily reflect an extrapolation of findings from randomized trials demonstrating that the 2 therapies are equally efficacious in the treatment of urogenital chlamydial infections.12
To date, no randomized trials have directly compared azithromycin and doxycycline for the treatment of rectal chlamydia. The sole published observational study13 comparing the 2 drugs found that azithromycin was associated with a significantly higher rate of treatment failure than doxycycline.
We conducted a retrospective analysis to evaluate the risk of persistent/recurrent rectal chlamydial infection associated with treatment with azithromycin compared with doxycycline.
MATERIALS AND METHODS
Study Design and Population
We created a retrospective cohort of all male patients who tested positive for rectal chlamydial infections in an STD clinic in Seattle, Washington, between 1993 and 2012. Women were excluded because they were not routinely screened for rectal chlamydia during the study period.
Data Collection and Measures
All data were obtained as part of routine medical care and recorded in the clinic’s electronic database. Clinicians used standardized instruments to record information about patient symptoms, concurrent sexually transmitted infections, and physical examination findings. Until October 2010, information on patient demographics (e.g., age, race), behavior (e.g., self-reported number of sex partners, self-reported contact to a sex partner with chlamydia), and STD history (e.g., self-reported prior chlamydial infection) was also collected by clinicians on standardized forms; thereafter, these data were routinely obtained from English-speaking patients using a computer-assisted self-interview.
Throughout the study period, clinic protocol was to obtain rectal specimens from all male patients who reported receptive anal sex in the prior year. Between 1993 and September 2010, the clinic tested rectal specimens for chlamydia using McCoy cell culture.14 Beginning in October 2010, the clinic tested all rectal specimens using nucleic acid amplification testing (NAAT) (APTIMA Combo 2; GenProbe Diagnostics, San Diego, CA).
Definition of Treatment and Outcome
Per CDC treatment guidelines,11 men who tested positive for rectal chlamydia were treated with azithromycin (1 g orally once) or doxycycline (100 mg orally twice daily for 7 days), at the clinician’s discretion. The treatment regimen was recorded by clinic staff on a standardized form. Clinicians treated men presumptively at the time of the initial evaluation if they reported recent sexual contact to a partner diagnosed as having gonorrhea, chlamydia, or nongonococcal urethritis.
We limited the analytic sample to men who were treated within 60 days of diagnosis to minimize the effect of spontaneous resolution of infection15 on our study outcome. We further excluded men who were cotreated with a quinolone or amoxicillin because both drugs have some activity against C. trachomatis. We also excluded men from the analysis if they were treated with a doxycycline or azithromycin regimen other than those recommended in the 2010 CDC STD treatment guidelines.11
We defined the outcome of interest, persistent/recurrent rectal chlamydial infection, as having a positive result on the first repeat test for rectal chlamydia obtained in the 14 to 180 days after the date of treatment. We only included repeat test outcomes if they were obtained using the same testing method (culture or NAAT) as that used for the initial diagnosis. Per CDC guidelines,11 men with rectal chlamydial infection were not asked to return to the clinic for a test of cure; thus, there was no active follow-up of the study population. We compared treatment outcomes in 4 repeat testing time intervals (defined as the number of days from the date of treatment): 14 to 30, 14 to 60, 14 to 90, and 14 to 180 days. We excluded tests occurring less than 14 days after treatment to diminish the risk of false-positive tests.16–18 Evaluating multiple testing intervals allowed us to maximize the number of follow-up tests included in the analysis, while also permitting an assessment of outcomes among men tested early after treatment (when the risk of reinfection is presumably low) and among men tested over a longer period (when the risk of reinfection as opposed to treatment failure is higher).
We used Pearson χ2 tests (for categorical variables) and t tests assuming unequal variance (for continuous variables) to compare characteristics of men who did and did not return for a repeat test and, among those who returned for a repeat test, to compare characteristics of men who received azithromycin versus doxycycline.
Our primary analysis aimed to compare the risk of persistent/recurrent rectal infection among men treated with azithromycin compared with men treated with doxycycline, in each repeat test interval defined above. We used Fisher exact test to analyze statistically significant differences in the proportion of persistent/recurrent rectal chlamydial infection among men treated with the 2 therapies. We used log binomial models to estimate the unadjusted and adjusted relative risk (RR) and corresponding 95% confidence interval (CI) of persistent/recurrent infection associated with treatment with azithromycin versus doxycycline. We conducted 2 multivariate analyses to control for potential confounding of the association of treatment regimen with persistent/recurrent infection. First, we developed a multivariate model that included factors associated with receipt of azithromycin or doxycycline in our population and that we defined a priori as being plausibly associated with persistent/recurrent infection. These adjustment factors included the following: anorectal symptoms, diagnosis of proctitis, contact to a chlamydial-infected partner, testing method, and year of diagnosis. We also included number of male sex partners in the 2 months before initial diagnosis (modeled as a continuous variable) to serve as a surrogate for sexual risk behaviors between initial and follow-up testing because our clinic did not collect standardized data on sexual behavior between treatment and follow-up testing. A second multivariate model included a larger array of factors, including demographic (e.g., age, race), clinical (e.g. proctitis), and behavioral (e.g., methamphetamine use; number of sex partners, modeled categorically). This analysis used log binomial models to estimate the association between each characteristic and persistent/recurrent infection. We performed a backward, stepwise elimination to identify factors significantly associated (P < 0.10) with persistent/recurrent infection. Because testing method is a known confounder, we maintained that variable in the model during the backward elimination procedure. We report the bivariate and multivariate RR with 95% CI.
We conducted 2 separate sensitivity analyses. First, we defined an additional repeat test time interval of 21 to 42 days after treatment. We chose this lower bound because repeat testing for rectal chlamydia before 21 days may detect residual nucleic acid from nonviable organisms11; the upper bound of 42 days decreases the probability that repeat test positivity is due to reinfection and conforms to the timing of testing used in randomized trials of urethral chlamydia and observational studies of rectal chlamydia.12,13,19,20 Second, we stratified our analytic sample by test method (culture vs. NAAT), to control for confounding by test method. We examined the outcomes of these men in the 14 to 90 and 14 to 180 days after treatment. For each sensitivity analysis, we used Fisher exact tests to compare the proportion of persistent/recurrent infection by therapy received.
All analyses were performed using Stata statistical software, version 12.2 (StataCorp, College Station, TX). Unless stated otherwise, tests were performed at a significance level of 0.05. Study procedures and analyses were approved by the University of Washington Human Subjects Division.
Clinicians diagnosed a total of 1461 unique MSM with 1835 cases of rectal chlamydia infection over the course of the study period. Of these, 1545 (84%) were treated in our clinic with azithromycin or doxycycline within 60 days of diagnosis, of whom 1505 (97%) received a standard regimen (1 g orally once or 100 mg twice daily for 7 days, respectively). Of these, 1480 (98%) did not receive a cotreatment with a quinolone or amoxicillin and were included in the analysis. The majority (81%; n = 1199) were treated within 1 week of diagnosis; only 8 (<1%) men were treated 30 to 60 days after testing.
The mean age of all treated cases was 32 years, approximately two-thirds of patients were white and more than 70% were HIV negative (Table 1). Approximately 34% of treated cases (502/1480) returned for a repeat rectal chlamydia test 14 to 180 days after treatment. Demographic and behavioral characteristics were comparable among men who did and did not return for a repeat test. However, men with a repeat test were significantly more likely to have had a chlamydial infection in the 12 months before their diagnosis (P = 0.005), more likely to have a symptomatic infection (P = 0.03), and less likely to report having a sex partner with chlamydia (P = 0.03) compared with men who did not return for a repeat test. Repeat testing did not significantly differ by therapy received: 407 (33%) of the 1231 men treated with azithromycin and 95 (38%) of the 249 men treated with doxycycline had a repeat test within 14 to 180 days of treatment (P = 0.12).
Of the 502 men who returned for a repeat test in the 14 to 180 days after treatment, 81% were treated with azithromycin and 19% were treated with doxycycline (Table 1). Among patients who were retested, there were no statistically significant differences in demographic or behavioral characteristics between men treated with azithromycin or doxycycline. Treatment with azithromycin was more common after 2009; consequently, testing by NAAT was significantly more likely among azithromycin-treated cases compared with doxycycline-treated cases (P = 0.003). Men treated with doxycycline were more likely than those treated with azithromycin to have a symptomatic infection (P = 0.005) and a diagnosis of proctitis (P < 0.001), but slightly less likely to report having a sex partner with C. trachomatis infection (P = 0.06). The median length of time between treatment and repeat testing was 84 days (interquartile range [IQR], 49–118) for azithromycin-treated patients and 83 days (IQR, 32–131) for doxycycline-treated patients.
At each repeat test interval, persistent/recurrent infection was higher among men treated with azithromycin compared with those treated with doxycycline (Table 2). Adjusting for potential confounders, treatment with azithromycin (compared with treatment with doxycycline) was significantly associated with a 5-fold higher risk of persistent/recurrent infection 14 to 90 days after treatment and a 2-fold higher risk of persistent/recurrent infection 14 to 180 days after treatment.
The significant association of treatment regimen with persistent/recurrent infection persisted in the second multivariate analysis (Table 3). Other factors independently associated with persistent/recurrent infection included having 5 or more sex partners in the 2 months before the initial diagnosis and being a known contact to a chlamydia-infected partner at the time of initial diagnosis.
Using an alternate time interval of 21 to 42 days from treatment to repeat testing produced results similar to our primary analysis: 12 (17%) of 69 azithromycin-treated men and 0 (0%) of 20 doxycycline-treated men had persistent/recurrent infection in the 21 to 42 days after treatment (P = 0.06).
Among men tested by culture, those treated with azithromycin were significantly more likely to have persistent/recurrent infection compared with those treated with doxycycline (Table 4). Persistent/recurrent infection among patients tested by NAAT was also higher for patients treated with azithromycin compared with those treated with doxycycline, although there were few men treated with doxycycline in the period after our clinic adopted NAAT testing of rectal specimens, and this difference was not statistically significant.
In this population of male STD clinic patients diagnosed as having rectal chlamydia between 1993 and 2012, we found that men treated with azithromycin had a significantly higher risk of persistent/recurrent rectal chlamydial infection compared with men treated with doxycycline. This association was consistent across several repeat test time intervals and was observed when all testing was done using culture and after our clinic switched to routinely test rectal specimens using NAAT. Together, these data suggest that a 7-day course of doxycycline may be superior to a single 1-g dose of azithromycin in the treatment of rectal chlamydial infection.
Our findings are largely consistent with prior observational studies. In the only previously published study directly comparing the 2 regimens, Hathorn and colleagues reported that 11 (26.2%) of 42 azithromycin-treated patients and 2 (5.0%) of 40 of doxycycline-treated patients had repeat positive NAAT test results for C. trachomatis 6 weeks after treatment (P = 0.014).13 Although Hathorn et al. did not account for differences in sexual behavior between the 2 treatment groups, other noncomparative studies of cohorts treated either with azithromycin or doxycycline also suggest that doxycycline may be the superior therapy. De Vries et al.21 observed persistent/recurrent rectal infection in only 1 (3.8%) of 26 doxycycline-treated men at 6 weeks, and Elgalib et al.22 found persistent/recurrent infections in 2 (1%) of 165 doxycycline-treated men tested a median of 45 days after treatment. In contrast, although Elgalib et al. excluded azithromycin-treated men from their primary analysis, 4 (15.4%) of 26 men receiving the drug had a persistent/recurrent infection at their test-of-cure visit. A second UK observational study of rectal chlamydial infection found that 9 (13%) of 68 azithromycin-treated men who retested at 6 weeks were infected,23 whereas a study in Australia reported persistent/recurrent rectal infections in 7 (15%) of 47 azithromycin-treated men retested within 3 months of treatment.24 Together, these studies present a consistent picture suggesting that doxycycline confers a lower risk of persistent/recurrent rectal chlamydial infection than azithromycin.
We postulate a number of reasons for the decreased effectiveness of azithromycin relative to doxycycline that we observed. First, although the high tissue concentration and long half-life of azithromycin25–27 make it an attractive therapy for rectal infection, we are not aware of any studies that have measured the concentration of azithromycin in the rectum. It is possible that rectal tissue penetration of azithromycin is poor or inferior to that of doxycycline. Second, and related to this, there may be differences in the pharmacokinetic properties of the 2 drugs that have not been elucidated but could explain our findings. Third, the high bacterial load of C. trachomatis in the rectum28 may foster the development of subpopulations of C. trachomatis that are less susceptible to certain antimicrobials (i.e., heterotypic resistance).29–31 Finally, doxycycline may decrease bacterial load more quickly than azithromycin and, consequently, might seem to be more effective. However, this seems unlikely to explain our findings among men retested many weeks after treatment.
Our study has several strengths. To our knowledge, this study is the largest to date to compare the efficacy of azithromycin and doxycycline for the treatment of rectal chlamydial infection and the first undertaken in the United States. Our clinic systematically obtains data on sexual behavior, and this allowed us to adjust for several possible confounding factors, strengthening the likelihood that the observed differences were real. Indeed, our primary finding that azithromycin seems to be inferior to doxycycline in the treatment of rectal chlamydial infections was extremely robust across time intervals and diagnostic testing modalities.
Our findings are also subject to several important limitations. First and foremost, this was an observational study, not a randomized trial, and we cannot eliminate the possibility that our results are due to confounding. Patients may have been treated with 1 of the 2 therapies based on certain factors that are prognostic of persistent/recurrent infection (i.e., confounding by indication). We attempted to address this issue through multivariable analysis, and this adjustment for confounders did not alter the significant association of azithromycin with persistent/recurrent infection. However, we did not measure several potential confounders (e.g., sex with anonymous partners, number of receptive anal sex partners), we assessed others using only a binary global measurement (e.g., ever used methamphetamine), and we could not measure unknown confounders. We were also concerned that our findings might reflect a change in testing methods from culture to NAATs because our clinic transitioned toward greater use of azithromycin during the same period that it switched from culture to NAAT testing of rectal chlamydial infection in 2010 to 2011. However, adjusting for testing method did not affect our principal finding, and the strong association of azithromycin with persistent/recurrent infection remained evident when we stratified our analysis by test type. Second, owing to the retrospective study design, there was no active follow-up of this study population and men included in this analysis were only those who returned for a repeat test 14 to 180 days after treatment. Third, we cannot differentiate between persistent infection and reinfection. Clinicians in our clinic do not systematically collect information from patients returning for follow-up testing about reexposure to infected or untreated partners; thus, it is likely that at least some of the persistent/recurrent infections are the result of reinfection. Men treated with azithromycin may have experienced a higher risk of persistent/recurrent infection because of earlier resumption of posttreatment sexual activity, although we believe that this is unlikely to explain the large difference in risk we observed. Fourth, our clinic does not routinely test specimens for lymphogranuloma venereum, so we cannot assess differences in outcomes among men with and without lymphogranuloma venereum. Finally, our study population was predominantly white and relatively old compared with some STD clinic populations, and our findings may not be generalizable to populations that are substantially different from the one studied here.
In conclusion, we found that azithromycin was associated with a significantly higher risk of persistent/recurrent rectal chlamydial infection than doxycycline. These findings add to the limited but growing body of evidence suggesting that azithromycin may be inferior to doxycycline in the treatment of rectal chlamydia. A randomized controlled trial is needed to definitively determine the best therapy for patients with rectal chlamydial infection.
1. Centers for Disease Control and Prevention. Sexually Transmitted Disease Surveillance, 2011. Atlanta, GA: US Department of Health and Human Services, 2012.
3. Kent CK, Chaw JK, Wong W, et al. Prevalence of rectal, urethral, and pharyngeal chlamydia and gonorrhea detected in 2 clinical settings among men who have sex with men: San Francisco, California, 2003. Clin Infect Dis 2005; 41: 67–74.
4. Cook RL, St George K, Silvestre AJ, et al. Prevalence of chlamydia and gonorrhoea among a population of men who have sex with men. Sex Transm Infect 2002; 78: 190–193.
5. Lister NA, Smith A, Tabrizi S, et al. Screening for Neisseria gonorrhoeae
and Chlamydia trachomatis
in men who have sex with men at male-only saunas. Sex Transm Dis 2003; 30: 886–889.
6. Craib KJ, Meddings DR, Strathdee SA, et al. Rectal gonorrhoea as an independent risk factor for HIV infection in a cohort of homosexual men. Genit Med 1995; 71: 150–154.
7. Bernstein KT, Marcus JL, Nieri G, et al. Rectal gonorrhea and chlamydia reinfection is associated with increased risk of HIV seroconversion. J Acquir Immune Defic Syndr 2010; 53: 537–543.
8. Zetola NM, Bernstein KT, Wong E, et al. Exploring the relationship between sexually transmitted diseases and HIV acquisition by using different study designs. J Acquir Immune Defic Syndr 2009; 50: 546–551.
9. Pathela P, Braunstein SL, Blank S, et al. HIV incidence among men with and those without sexually transmitted rectal infections: Estimates from matching against an HIV case registry. Clin Infect Dis 2013; 57: 1203–1209.
10. Chesson HW, Bernstein KT, Gift TL, et al. The cost-effectiveness of screening men who have sex with men for rectal chlamydial and gonococcal infection to prevent hiv infection. Sex Transm Dis 2013; 40: 366–371.
11. Centers for Disease Control and Prevention. Sexually transmitted diseases treatment guidelines, 2010. MMWR Recomm Rep 2010; 59: 1–110.
12. Lau CY, Qureshi AK. Azithromycin versus doxycycline for genital chlamydial infections: A meta-analysis of randomized clinical trials. Sex Transm Dis 2002; 29: 497–502.
13. Hathorn E, Opie C, Goold P. What is the appropriate treatment for the management of rectal Chlamydia trachomatis
in men and women? Sex Transm Infect 2012; 88: 352–354.
14. Stamm WE, Tam M, Koester M, et al. Detection of Chlamydia trachomatis
inclusions in Mccoy cell cultures with fluorescein-conjugated monoclonal antibodies. J Clin Microbiol 1983; 17: 666–668.
15. Apewokin SK, Geisler WM, Bachmann LH. Spontaneous resolution of extragenital chlamydial and gonococcal infections prior to therapy. Sex Transm Dis 2010; 37: 343–344.
16. Workowski KA, Lampe MF, Wong KG, et al. Long-term eradication of Chlamydia trachomatis
genital infection after antimicrobial therapy. Evidence against persistent infection. JAMA 1993; 270: 2071–2075.
17. Gaydos CA, Crotchfelt KA, Howell MR, et al. Molecular amplification assays to detect chlamydial infections in urine specimens from high school female students and to monitor the persistence of chlamydial DNA after therapy. J Infect Dis 1998; 177: 417–424.
18. Bachmann LH, Desmond RA, Stephens J, et al. Duration of persistence of gonococcal DNA detected by ligase chain reaction in men and women following recommended therapy for uncomplicated gonorrhea. J Clin Microbiol 2002; 40: 3596–3601.
19. Manhart LE, Gillespie CW, Lowens MS, et al. Standard treatment regimens for nongonococcal urethritis have similar but declining cure rates: A randomized controlled trial. Clin Infect Dis 2013; 56: 934–942.
20. Schwebke JR, Rompalo A, Taylor S, et al. Re-evaluating the treatment of nongonococcal urethritis: Emphasizing emerging pathogens—A randomized clinical trial. Clin Infect Dis 2011; 52: 163–170.
21. de Vries HJ, Smelov V, Middelburg JG, et al. Delayed microbial cure of lymphogranuloma venereum proctitis with doxycycline treatment. Clin Infect Dis 2009; 48: e53–e56.
22. Elgalib A, Alexander S, Tong CY, et al. Seven days of doxycycline is an effective treatment for asymptomatic rectal Chlamydia trachomatis
infection. Int J STD AIDS 2011; 22: 474–477.
23. Steedman NM, McMillan A. Treatment of asymptomatic rectal Chlamydia trachomatis
: Is single-dose azithromycin effective? Int J STD AIDS 2009; 20: 16–18.
24. Drummond F, Ryder N, Wand H, et al. Is azithromycin adequate treatment for asymptomatic rectal chlamydia? Int J STD AIDS 2011; 22: 478–480.
25. Girard AE, Girard D, English AR, et al. Pharmacokinetic and in vivo studies with azithromycin (CP-62,993), a new macrolide with an extended half-life and excellent tissue distribution. Antimicrob Agents Chemother 1987; 31: 1948–1954.
26. Lode H, Borner K, Koeppe P, et al. Azithromycin–review of key chemical, pharmacokinetic and microbiological features. J Antimicrob Chemother 1996; 37 (suppl C): 1–8.
27. Amsden GW. Advanced-generation macrolides: Tissue-directed antibiotics. Int J Antimicrob Agents 2001; 18 (suppl 1): S11–S15.
28. Twin J, Moore EE, Garland SM, et al. Chlamydia trachomatis
genotypes among men who have sex with men in Australia. Sex Transm Dis 2011; 38: 279–285.
29. Horner PJ. Azithromycin antimicrobial resistance and genital Chlamydia trachomatis
infection: Duration of therapy may be the key to improving efficacy. Sex Transm Infect 2012; 88: 154–156.
30. Suchland RJ, Geisler WM, Stamm WE. Methodologies and cell lines used for antimicrobial susceptibility testing of Chlamydia
spp. Antimicrob Agents Chemother 2003; 47: 636–642.
31. Wang SA, Papp JR, Stamm WE, et al. Evaluation of antimicrobial resistance and treatment failures for Chlamydia trachomatis
: A meeting report. J Infect Dis 2005; 191: 917–923.