Treatment of Neisseria gonorrhoeae is complicated by the bacterium’s ability to develop antimicrobial resistance. Prompted by the emergence of fluoroquinolone resistance in the United States in 2007, the US Centers for Disease Control and Prevention (CDC) removed fluoroquinolones from the list of recommended antimicrobial agents for treating gonorrhea, leaving cephalosporins as the only class of antimicrobials recommended as first-line therapy.1 With reports of decreased cephalosporin susceptibility, treatment guidelines were later revised in 2010 to recommend dual therapy with a cephalosporin agent (250 mg ceftriaxone intramuscularly or 400 mg cefixime orally as a single dose) plus either 1 g azithromycin orally as a single dose or 100 mg doxycycline orally twice per day for 7 days,2 in part, as an effort to mitigate the emergence of cephalosporin resistance in the United States.
Documented cefixime treatment failures in both Asia3 and Europe4–6 and a 15-fold increase in the proportion of isolates collected through the Gonococcal Isolate Surveillance Project (GISP) with decreased susceptibility to cefixime prompted the CDC to remove oral cephalosporins as first-line antimicrobial agents for treating uncomplicated N. gonorrhoeae infection in August 2012.7 The current recommended first-line regimen is 250 mg ceftriaxone in combination with either 1 g azithromycin or 100 mg doxycycline twice daily for 7 days, with azithromycin preferred as the secondary antimicrobial.7,8 Although the preference for azithromycin is supported by the fact that it can be administered as a single dose and that a higher proportion of GISP isolates are resistant to tetracycline than to azithromycin,8 there are no data showing that ceftriaxone plus azithromycin is more efficacious than ceftriaxone plus doxycycline for the treatment of uncomplicated gonorrhea infection. Although nationally, the preferred recommended regimen is azithromycin, doxycycline is less expensive and, because of cost-savings, may be a more attractive option in many clinics.
Using data from 2 Baltimore sexually transmitted disease (STD) clinics, we sought to retrospectively measure the overall rate of gonorrhea retreatment and to compare retreatment rates between patients receiving ceftriaxone plus azithromycin and those receiving ceftriaxone plus doxycycline.
MATERIALS AND METHODS
The Baltimore City Health Department (BCHD) runs 2 public STD clinics, which provide STD testing and treatment services free of charge for more than 30,000 clients annually. Since 2004, BCHD has maintained an electronic health record database, which stores data on patient demographics, physical examinations, behavioral risk assessments, laboratory tests, and treatment from each patient visit to either clinic.
Using data extracted from the BCHD electronic health record database, we constructed a retrospective cohort of all persons diagnosed as having uncomplicated N. gonorrhoeae infection who were treated with (1) ceftriaxone plus azithromycin or (2) ceftriaxone plus doxycycline between January 1, 2004, and December 31, 2011. Entry into the cohort was defined as the date either ceftriaxone regimen was received as treatment of gonorrhea infection. For patients with multiple episodes of gonorrhea during this period, date of entry was defined as the first date on which either ceftriaxone regimen was received as treatment of gonorrhea. All patients were followed for 2 years from date of treatment or until March 31, 2012, whichever occurred first; all patients were followed for at least 90 days.
We defined date of gonorrhea retreatment as the first patient visit when that patient had positive Gram stain, culture, or nucleic acid amplification test at least 2 or 7 days, respectively, after the initial treatment. To identify any retreatments that may have been prescribed by a private provider within the state of Maryland, clinic records were matched to BCHD’s STD gonorrhea morbidity registry using a probabilistic record matching module, Reclink,9 in Stata Version 12.1 (StataCorp LP, College Station, TX). Matching on first and last name, sex, and date of birth, we linked only those records from registry with at least 90% probability of matching. We measured time to retreatment for each patient, defined as the number of days between the initial ceftriaxone regimen and the date on which a new gonorrhea infection was detected. Censoring was assumed to be noninformative; patients with no record of gonorrhea retreatment after 2 years or as of March 31, 2012, were assumed to be disease-free.
Because our assumption that patients with no record of retreatment are disease-free may underestimate the true retreatment rate, we also analyzed 2 subcohorts, which included all patients receiving a ceftriaxone regimen for uncomplicated gonorrhea who later returned to one of the STD clinics and were retested for gonorrhea within (1) 90 days and (2) 30 days of the initial treatment. Follow-up time was measured between the date treatment was received and the date on which a second Gram stain, culture, or nucleic acid amplification test was ordered.
We investigated a number of demographic and behavioral risk factors previously found to be associated with gonorrhea reinfection or treatment failure as potential confounders including the following: year of initial ceftriaxone treatment, sex, sexual orientation, age, race/ethnicity, anatomical site of infection, symptoms, HIV, history of gonorrhea or chlamydia infection, receipt of expedited partner therapy (EPT), and number of sex partners in the past 6 months.10–14 All variables were treated as categorical; data collected as continuous variables (e.g., age and number of sex partners) were categorized as above and below the median. For sexual orientation, we defined men who have sex with men (MSM) as males who self-identified as gay/bisexual or who reported receptive anal sex during their physical examination, regardless of whether or not they also reported having had sex with women. All other men were assumed to be men who have sex with women (MSW). Because the risk of retreatment is likely to be different among MSM than among MSW and among women, we collapsed sex and sexual orientation into one variable.
To assess differences in time to retreatment across treatment groups, we created Kaplan-Meier curves and tested statistical significance using log-rank tests of equality. Cox proportional hazards models were used to compare the hazards of retreatment across treatment groups, and multivariate Cox proportional hazard models were constructed to control for factors that may be associated with either reinfection or treatment failure. We tested the proportional hazards assumption by testing the significance of the interaction of each variable and survival time. To build the multivariate models, we included and retained all variables found to be significant during bivariate analysis at α = 0.10. All analyses were performed using SAS software, version 9.3 (SAS Institute, Inc, Cary, NC).
Ethical approval for this analysis was obtained from the Johns Hopkins Hospital institutional review board and BCHD.
Between January 1, 2004, and December 31, 2011, Baltimore’s public STD clinics diagnosed 9181 persons with uncomplicated gonorrhea. During this period, 4848 (52.8%) received ceftriaxone as treatment of at least 1 episode of gonorrhea, 4457 of whom received either ceftriaxone plus azithromycin (n = 1417; 29.2%) or ceftriaxone plus doxycycline (n = 3040; 62.7%). Of the 4457 patients retained in the cohort, most was male (67.0%), African American (94.3%), and younger than 25 years; the median age at the time the ceftriaxone regimen was administered was 24 years. Fifteen percent (n = 462) of male patients were MSM. Approximately 4% (3.5%) of patients in the cohort were known to be HIV positive upon entry into the cohort.
The demographic composition of the 485 patients who were retested for gonorrhea within 90 days was similar to that of the full cohort. However, a larger proportion of patients included in the 90-day cohort had a history of gonorrhea (29.2% full, 39.2% 90-day) or a history of chlamydia (21.4% full, 35.2% 90-day). The 140 patients who were retested within 30 days in the third cohort were more likely to be male (80.8%), older than 25 years (56.0%), and more likely to have a history of gonorrhea (48.2%).
Gonorrhea Retreatment Rates in the Full Cohort
In the full cohort, the median time under observation was 2 years (range, 3 days–2 years), during which 441 (9.9%) patients were retreated for gonorrhea. Patients who were treated with ceftriaxone plus doxycycline were slightly more likely to be retreated (n = 311; 10.2%) compared with those who were treated with ceftriaxone plus azithromycin (n = 130; 9.2%), although this difference was not statistically significant (χ 2 test, P = 0.27). The average time to retreatment (Fig. 1) was shorter among those who received ceftriaxone plus doxycycline (mean, 257.1 days; median, 207 days) compared with time to retreatment among patients who received ceftriaxone plus azithromycin (mean, 273.9 days; median, 238 days), although the difference in time to retreatment was not statistically significant (log-rank test, P = 0.26).
In the multivariate analysis, when controlling for year of initial treatment, sex/sexual preference, age, anatomical site of infection, symptomatic infection, HIV status, history of either gonorrhea or chlamydia, receipt of EPT, and the number of reported sex partners (Table 1), we did not observe a significant association between treatment regimen and the hazard of retreatment. However, older age (≥25 years) was significantly associated with a 42% reduction in the risk of retreatment (adjusted hazard ratio [aHR], 0.58 [0.47–0.72]), whereas having a history of gonorrhea or chlamydia and reporting 2 or more sexual partners within the past 6 months remained statistically significantly associated with an increased hazard of retreatment. Expedited partner therapy, which first became available in BCHD clinics on July 1, 2007, was statistically significantly associated with a decreased hazard of retreatment. Compared with patients diagnosed before EPT’s availability, patients who received EPT were 45% less likely to be retreated (aHR, 0.55 [0.31–0.96]); those diagnosed after July 1, 2007, who did not receive EPT were 40% less likely to be retreated (aHR, 0.60 [0.34–1.03]). A sensitivity analysis, restricted to all patients for whom 2 years of follow-up was possible, showed that our results were robust to our assumption that censoring was noninformative.
Gonorrhea Retreatment Rates in the 90-Day Cohort
Among patients who were retested for gonorrhea within 90 days of their initial ceftriaxone treatment, the median time between treatment and being retested was 47 days (range, 2–90 days), and nearly one fifth (n = 91; 18.8%) were retreated, double the rate observed in the full cohort. Similarly to the full cohort, those who were treated with ceftriaxone plus doxycycline were more likely to be retreated (n = 71; 21.1%) than those who received ceftriaxone plus azithromycin (n = 20; 13.5%); this difference was borderline significant (χ 2 test, P = 0.05). However, survival curves (Fig. 2) show that, on average, time to retreatment was statistically significantly shorter among those receiving doxycycline compared with those receiving azithromycin (median time to retreatment: 48 vs. 55 days, respectively; log-rank test, P = 0.02).
Multivariate analyses showed no association with treatment regimen when controlling for year of initial treatment, sex/sexual orientation, symptomatic infection, and number of reported sex partners (aHR, 0.95 [0.55–1.65]). Women were significantly less likely to be retreated than MSW; for all other variables, the observed associations with retreatment were similar to those observed in the full cohort, but failed to reach statistical significance. We observed neither an association between retreatment and reporting of sexual activity between clinic visits nor an association between retreatment and testing at any extragenital site at the second visit (Table 2).
Gonorrhea Retreatment Rates in the 30-Day Cohort
Among the 140 patients who were retested within 30 days, 19 (13.5%) were retreated, all of whom had received ceftriaxone plus doxycycline (Fisher exact test, P < 0.01). Average time to retreatment was 23.9 days (median, 25 days). Because no retreatments occurred among patients receiving the azithromycin regimen, we were unable to plot survival curves or build multivariate Cox proportional hazards models for the 30-day cohort. However, a descriptive analysis shows similar patterns in the 30-day cohort compared with the 90-day and full cohorts. A larger proportion of patients retreated were MSW (94.7% vs. 69.4% in the untreated group), were symptomatic (89.5% vs. 76.0%), had a history of gonorrhea (52.6% vs. 47.1%), and reported sexual activity between clinic visits (89.5% vs. 71.1%). None of these differences were statistically significant (Table 3).
We found that a substantial proportion of STD clinic patients were retreated for gonorrhea. One tenth of all STD clinic patients treated for gonorrhea were retreated within 2 years, and nearly a fifth of those who were retested at either clinic within 90 days of the initial treatment were retreated. Within the full and the 90-day cohorts, the relatively long median time to retreatment in both treatment groups, and the high proportion of retreated patients reporting sexual activity between infections (full cohort: 93%, data not shown) suggests that most retreatments may be reinfection rather than treatment failure. When controlling for factors related to reinfection, we found no overall association between treatment regimen and time to retreatment.
The finding that all retreatments within 30 days were limited to the ceftriaxone and doxycycline group suggests that the CDC’s preference for the ceftriaxone/azithromycin combination may be warranted. Azithromycin is given as a single oral dose, whereas the patient is still in clinic. In addition, azithromycin resistance has not been documented by the GISP in Baltimore (GISP collects Baltimore specimens from the 2 STD clinics that were included in this analysis), whereas resistance to doxycycline has averaged 7% among gonococcal isolates. Of note, through November 2011, all minimum inhibitory concentration to ceftriaxone and cefixime in Baltimore had been 0.125 µg/mL or less. In December of 2011, the first high-level (minimum inhibitory concentration 32 µg/mL) cefixime-resistant isolate was detected—that isolate, however, was sensitive to ceftriaxone.
In all 3 cohorts, women were significantly less likely than MSW to be retreated, which is unsurprising given that men are more likely to be symptomatic and thus more likely to seek care for a new, symptomatic infection or if symptoms persisted. Indeed, symptomatic infection was independently correlated with retreatment in both the full and 90-day cohorts, although the association was not significant in the 90-day cohort. Women also were more likely to have received the azithromycin regimen, which may explain why, when controlling for sex/sexual orientation and symptomatic infection, azithromycin was no longer significantly associated with a reduced hazard of retreatment in the 90-day cohort. In the full cohort, MSM were more likely than MSW to be retreated; however, the opposite was true in the 90-day cohort.
The most striking finding in our analyses was the protective effect of EPT in the full cohort; patients treated after EPT became available in Baltimore City were less likely to be retreated than patients treated for gonorrhea before July 1, 2007. Although the finding that patients treated after EPT became available but who did not receive EPT also had a decreased hazard of retreatment was of borderline significance (P = 0.06), we believe that this suggests that this intervention may have population-level effects. This finding should be interpreted with some caution. Although we included year of treatment in the model to control for any secular trends in gonorrhea epidemiology, there may be residual confounding. However, in a separate analysis performed for the yearly EPT report required by the Maryland legislature, when comparing retreatment rates within 365 days for persons given EPT versus those not given EPT during the same period, those who were given EPT had an 18% reduction in retreatment (suggesting that the EPT finding is not a period effect).15 These results further support the growing evidence that EPT is a successful intervention.16,17 However, with the demotion of oral cephalosporins from first-line to second-line agents for the treatment of gonorrhea coupled with the recommendation for tests of cure when oral cephalosporins are used, the future fate of this intervention for gonorrhea is unclear and greatly highlights the need for novel effective oral therapies.
Our analyses have several important limitations. This was a retrospective observational study and not a randomized trial. As such, selection and information biases may have influenced our results. The dose of ceftriaxone used was 125 mg (the recommended dose at that time) instead of the currently recommended 250 mg dose. Whether the enhanced dose of ceftriaxone would abrogate any potential benefit of using azithromycin instead of doxycycline is not known. Although we matched clinic patients to gonorrhea morbidity records, patients with asymptomatic infections who were not tested for gonorrhea by any medical provider would be misclassified as not retreated. However, this misclassification should have been nondifferential across both groups. Sexually transmitted disease clinic patients generally report higher levels of sexual risk than the general population and therefore are at higher risk for reinfection. In addition, given the regional variation in antimicrobial susceptibility trends18 and the relatively small proportion of MSM patients attending Baltimore’s STD clinic, 2 factors strongly associated with decreased cephalosporin, tetracycline, and azithromycin susceptibility, our results may not be generalizable to other populations.
Our analysis shows that gonorrhea retreatment is common, underscoring the need for rescreening and better partner treatment—our finding that EPT was associated with decreased risk of retreatment further highlights the need for effective oral therapies. Our analysis was conducted at a time when decreased cephalosporin susceptibility is relatively rare, and with CDC now requesting that state and local health departments enhance surveillance for cephalosporin resistant gonorrhea, in part by “monitoring and investigating patient with multiple laboratory or clinical reports of N. gonorrhoeae infections documented within the preceding 30 or 60 days,”19 our study suggests that such investigations are likely new infections rather than treatment failures. Finally, although over longer periods, there was no difference in retreatment across therapy groups, that none of the patients in the 30-day cohort who received the azithromycin regimen were retreated, our findings suggest that azithromycin may be preferable to doxycycline; however, doxycycline remains a reasonable alternative.
1. Centers for Disease Control and Prevention (CDC). Update to CDC’s sexually transmitted diseases treatment guidelines, 2006: Fluoroquinolones no longer recommended for treatment of gonococcal infections. MMWR Morb Mortal Wkly Rep 2007; 56: 332–336.
2. Workowski KA, Berman S, Centers for Disease Control and Prevention (CDC). Sexually transmitted diseases treatment guidelines, 2010. MMWR Recomm Rep 2010; 59: 1–110.
3. Yokoi S, Deguchi T, Ozawa T, et al. Threat to cefixime treatment for gonorrhea. Emerg Infect Dis 2007; 13: 1275–1277.
4. Unemo M, Golparian D, Syversen G, et al. Two cases of verified clinical failures using internationally recommended first-line cefixime for gonorrhoea treatment, Norway, 2010. Eur Surveill 2010; 15: 19721.
5. Ison CA, Hussey J, Sankar KN, et al. Gonorrhoea treatment failures to cefixime and azithromycin in England, 2010. Eur Surveill 2011; 16: 19833.
6. Unemo M, Golparian D, Stary A, et al. First Neisseria gonorrhoeae
strain with resistance to cefixime causing gonorrhoea treatment failure in Austria, 2011. Eur Surveill 2011; 16: 19998.
7. Centers for Disease Control and Prevention (CDC). Update to CDC’s sexually transmitted diseases treatment guidelines, 2010: Oral cephalosporins no longer a recommended treatment for gonococcal infections. MMWR Morb Mortal Wkly Rep 2012; 61: 590–594.
8. Centers for Disease Control and Prevention (CDC). Cephalosporin susceptibility among Neisseria gonorrhoeae
isolates—United States, 2000–2010. MMWR Morb Mortal Wkly Rep 2011; 60: 873–877.
10. Mehta SD, Erbelding EJ, Zenilman JM, et al. Gonorrhoea reinfection in heterosexual STD clinic attendees: Longitudinal analysis of risks for first reinfection. Sex Transm Infect 2003; 79: 124–128.
11. Ota KV, Fisman DN, Tamari IE, et al. Incidence and treatment outcomes of pharyngeal Neisseria gonorrhoeae
and Chlamydia trachomatis
infections in men who have sex with men: A 13-year retrospective cohort study. Clin Infect Dis 2009; 48: 1237–1243.
12. De P, Singh AE, Wong T, et al. Predictors of gonorrhea reinfection in a cohort of sexually transmitted disease patients in Alberta, Canada, 1991–2003. Sex Transm Dis 2007; 34: 30–36.
13. Fung M, Scott KC, Kent CK, et al. Chlamydial and gonococcal reinfection among men: A systematic review of data to evaluate the need for retesting. Sex Transm Infect 2007; 83: 304–309.
14. Hosenfeld CB, Workowski KA, Berman S, et al. Repeat infection with chlamydia and gonorrhea among females: A systematic review of the literature. Sex Transm Dis 2009; 36: 478–489.
15. Baltimore City Health Department Sexually Transmitted Disease Clinics. Expedited partner therapy for partners of patients with gonorrhea or chlamydia: 2011 report for pilot program activities. Available at: dlslibrary.state.md.us/publications/Baltimore/HG18-214.1(f)_2011.pdf.
16. Shiely F, Hayes K, Thomas KK, et al. Expedited partner therapy: A robust intervention. Sex Transm Dis 2010; 37: 602–607.
17. Stephens SC, Bernstein KT, Katz MH, et al. The effectiveness of patient-delivered partner therapy and chlamydial and gonococcal reinfection in San Francisco. Sex Transm Dis 2010; 37: 525–529.
18. Centers for Disease Control and Prevention (CDC). GISP profile, 2010. 2012.
19. Centers for Disease Control and Prevention (CDC). Cephalosporin-Resistant Neisseria gonorrhoeae
Public health response plan. 2012.