Doxycycline versus azithromycin for the treatment of anorectal Chlamydia trachomatis infection in women concurrent with vaginal infection (CHLAZIDOXY study): a multicentre, open-label, randomised, controlled, superiority trial

Background: Anorectal infections with Chlamydia trachomatis are commonly found in women. Although the efficacy of doxycycline and azithromycin is comparable in the treatment of urogenital infection, their efficacies toward anorectal infection remain unclear. We therefore aimed to compare a single dose of azithromycin with a 7-day course of doxycycline for the treatment of anorectal C trachomatis infection in women with concurrent vaginal infection.

Methods: We did a multicentre, open-label, randomised, controlled, superiority trial involving four sexually transmitted infection screening centres and three pregnancy termination centres in France. We included sexually active adult women (≥18 years) with a positive C trachomatis vaginal swab who agreed to provide self-collected anorectal swabs for C trachomatis detection. Participants were randomly assigned (1:1), using block sizes of six and eight and stratification by each investigating centre, to orally receive either azithromycin (a single 1-g dose, with or without food) or doxycycline (100 mg in the morning and evening at mealtimes for 7 days [ie, 100 mg of doxycycline twice per day for 7 days]). All laboratory staff who did the bacteriological analyses, but not the participants and the investigators, were masked to the treatment groups. The primary outcome was the microbiological anorectal cure rate defined as a C trachomatis-negative nucleic acid amplification test (NAAT) result in anorectal specimens 6 weeks after treatment initiation among women who had a baseline C trachomatis-positive anorectal NAAT result. The primary analysis was done in the modified intention-to-treat population, with multiple imputation, which included all women who underwent randomisation and had a C trachomatis-positive vaginal and anorectal NAAT result at baseline. Adverse events were reported in all women who underwent randomisation. This study is registered with ClinicalTrials.gov, number NCT03532464.

Findings: Between Oct 19, 2018, and April 17, 2020, we randomly assigned a total of 460 participants to either the doxycycline group (n=230) or the azithromycin group (n=230). Four (1%) of 460 participants were excluded because they refused to take doxycycline or were found to be ineligible after randomisation. Among the 456 participants, 357 (78%) had a concurrent C trachomatis-positive anorectal NAAT result at baseline; 184 (52%) of 357 were in the doxycycline group and 173 (48%) were in the azithromycin group (ie, the modified intention-to-treat population). Microbiological anorectal cure occurred in 147 (94%) of 156 participants in the doxycycline group (28 missing values) versus 120 (85%) of 142 in the azithromycin group (31 missing values; adjusted odds ratio with imputation of missing values 0·43 [95% CI 0·21-0·91]; p=0·0274). Reported adverse events possibly related to treatment were notified in 53 (12%) of 456 women: 24 (11%) of 228 in the doxycycline group and 29 (13%) of 228 in the azithromycin group. Gastrointestinal disorders were the most frequently occurring, in 43 (9%) of 456 women: 17 (8%) of 228 in the doxycycline group and 26 (11%) of 228 in the azithromycin group.

Interpretation: The microbiological anorectal cure rate was significantly lower among women who received a single dose of azithromycin than among those who received a 1-week course of doxycycline. This finding suggests that doxycycline should be the first-line therapy for C trachomatis infection in women.

Funding: French Ministry of Health.

Translation: For the French translation of the abstract see Supplementary Materials section.