Abstract Study question Is endometriosis related to worse outcome of assisted reproductive techniques (ART)? Summary answer Cumulative incidence of live birth in patients with and without endometriosis at laparoscopy was similar, although deep endometriosis and adenomyosis were negative prognostic factors. What is known already Whether endometriosis has a negative impact on the outcome of ART is still a matter of debate. Most published data report on one fresh cycle only, usually without taking frozen embryos into account. Further, a large heterogeneity in study population has been acknowledged by several meta-analyses, as in the control groups endometriosis was not always excluded by laparoscopy, and in case of endometriosis the prior treatment history was variable or unclear. Study design, size, duration Retrospective longitudinal cohort study of 1462 patients (779 with laparoscopically treated endometriosis of any rASRM stage, and 683 without endometriosis at laparoscopy) undergoing ART treatment between July 2003 and December 2014. Primary outcome studied was time to ART live birth. Secondary outcomes include -amongst others- number of cycles needed per ART live birth, time to ART or spontaneous live birth, cycle cancellation rate, and pregnancy outcomes like miscarriage and ectopic pregnancy, and per cycle analyses. Participants/materials, setting, methods All patients with a history of laparoscopy prior to the start of their first ART were included for analysis. The ART was performed in a tertiary referral center of a large University Hospital. Primary outcome studied was the time from initiation of the first ART cycle to delivery of the first live born. Survival analysis was conducted using cumulative incidence functions and cause-specific hazards regression. Main results and the role of chance The study included 1462 patients who initiated 4537 ART cycles, of which 3672 (81%) fresh and 857 (19%) frozen cycles. The unadjusted hazard ratio (HR) of live birth was 1.01 (95% CI 0.88–1.16). After adjustment for potential confounders (age, maternal BMI, maternal smoking, secondary infertility, duration of infertility, anovulation, reduced ovarian reserve, tubal factor, male factor and therapy started before 2007) the HR was 0.99 (95% CI 0.86–1.14). Within the endometriosis population (n = 779), covariate-adjusted analyses suggested that presence of adenomyosis (HR = 0.54; 95% CI 0.34–0.86) and a history of deep endometriosis (HR = 0.74 – 95% CI 0.58–0.94) were associated with a lower cumulative incidence of live birth. In contrast, there was little support of an association with diagnosis of stage III/IV (HR = 1.15; 95% CI 0.84–1.59) or a history of ovarian endometriosis (HR = 0.97; 95% CI 0.72–1.30). Beside the effect of the different variables directly linked to endometriosis, maternal BMI (HR = 0.80; 95%CI 0.71–0.91) and smoking (HR = 0.69; 95%CI 0.52–0.92) also negatively affected live birth delivery rate per patient. Limitations, reasons for caution Practices and success rates of ART may have changed during the 11-year recruitment period. Restricting to women who underwent laparoscopy, while providing evidence of the presence or absence of endometriosis, may have induced selection bias. However, the advantage of this time period, was the high rate of pre-ART laparoscopy (+/–50%). Wider implications of the findings: As deep endometriosis and adenomyosis represent negative prognostic factors within the endometriosis population, future studies should focus on optimalisation of ART in these subgroups. Trial registration number S57393
Read full abstract