Article Text
Abstract
Introduction Incidence of syphilis, chlamydia and gonorrhoea continue to rise among HIV-positive men who have sex with men (MSM) in Ontario. We previously observed an elevated risk of sexually transmitted infections (STI) among recreational drug users. Our aim was to determine the effect of a new STI diagnosis and recreational drug use on virologic failure (VF) among MSM successfully treated with antiretroviral therapy (ART). Our hypothesis is that any association between STIs and VF would be confounded by drug use.
Methods The OHTN Cohort Study follows people receiving HIV care in Ontario. STI results and viral load (VL) data were retrieved via linkage with the provincial laboratory. We restricted analyses to 2610 MSM who completed>=1 annual questionnaire in 2008–2014 and had two consecutive VL<50 within a six-month period on ART. VF was defined as a single VL>=1000 or two consecutive VLs>=200. Periods of STI exposure were set around the diagnosis dates for each STI. We modelled STI diagnosis exposures and drug use as time-varying covariates on risk of VF using Cox regression adjusting for age, region and income as confounders. Our model allowed for repeat STI exposures and repeated VF events using the marginal means/rates model.
Results There were 472 VFs with a 24 month cumulative incidence of 12.1% (95%CI 11.1, 13.1). VFs at time of a new chlamydia or gonorrhoea infection were close to nil. We did not observe an increased risk of VF at the time of a new syphilis infection (HR=1.2 95% CI 0.8, 2.0; aHR=1.1 95% CI 0.7, 1.7). Risk was higher among drug users (non-injection aHR=1.4 95% CI 1.1, 1.8; injection aHR=1.8 95% CI 1.1, 2.6). There was no significant interaction but some evidence of positive confounding between syphilis and VF by drug use.
Conclusion Regardless of drug use, we did not find an association between a new STI diagnosis and increased risk of VF among men on suppressive ART. Our data are limited by possible misclassification of STI exposures, because not all men were tested, and among those diagnosed, exact dates of acquisition were unknown.