Microbial etiology of endometritis: implications for screening, treatment and post-pelvic inflammatory disease sequelae

The polymicrobial etiology of pelvic inflammatory disease (PID, infection and inflammation of the female upper genital tract (UGT)) is incompletely understood, hindering optimal screening and treatment. Studies of T. vaginalis (TV) and PID are limited and none have prospectively investigated sequelae. TV prevalence was 12.8% among 647 women in the PID Evaluation and Clinical Health (PEACH) study, a multicenter randomized controlled trial of antimicrobial therapy for clinically suspected PID. Odds of histologically confirmed endometritis doubled among TV-infected vs. uninfected women (adjusted odds ratio (AOR):1.9, 95% confidence interval (CI):1.0-3.3). Prospectively, we observed borderline non-significant associations with persistent endometritis at 30 days (AOR:2.6, 95% CI:0.7-10.1), infertility (AOR:1.6, 95% CI:0.9-3.0), recurrent PID (AOR:1.2, 95% CI:0.7-2.1), pregnancy (AOR:0.7, 95% CI:0.4-1.1) and live birth (AOR:0.6, 95% CI:0.1-3.2) at seven years. Among 386 women with PID, UGT infection with C. Trachomatis (AOR:10.8, 95% CI:4.0-29.4), N. gonorrhoeae (AOR:21.6, 95% CI:8.8-53.1), M. genitalium (AOR:4.4, 95% CI:1.6-12.0), or two infections (AOR:21.9, 95% CI:2.8-172.9) was strongly associated with acute endometritis, and non-significantly associated with increased infertility and decreased pregnancy or live birth, irrespective of endometritis. Probabilistic bias analyses confirmed that lack of association between PID and outcomes was not due to exposure misclassification resulting from use of histologic endometritis as a proxy for laparoscopic salpingitis, the diagnostic gold standard. Despite a high rate of infertility among PEACH participants, not all women were affected post-PID. Using classification and regression tree analysis, we developed a decision rule with good discriminatory performance (AUC:0.72, 95% CI:0.66-0.77) for the identification of women at increased risk of infertility following treatment for clinically suspected PID. Among 643 women studied, readily available historical data including: age, smoking status, days of pain, history of infertility, and number of prior births were the strongest predictors of future infertility at the time of PID diagnosis. This work yields public health significance by identifying TV as a potential risk factor for PID and its sequelae, underscoring the importance of primary prevention of sexually transmitted infections (STIs). Studies of optimal STI screening and treatment strategies for secondary prevention of PID and post-PID sequelae should incorporate novel PID-associated microbes like M. genitalium and TV.