By Jeffrey T. Jensen, MD, MPH
Synopsis: In a large, multicenter, randomized study of women in Europe who presented with ectopic pregnancy, there was no significant difference in the rate of subsequent intrauterine pregnancy observed following early ectopic pregnancy managed with medical therapy or conservative (salpingostomy) surgery, or following acute ectopic pregnancy managed with radical (salpingectomy) or conservative surgery.
Source: Fernandez H, et al. Fertility after ectopic pregnancy: The DEMETER randomized trial. Hum Reprod 2013; Mar 12 [Epub ahead of print].
In the DEMETER trial, the authors designed a study to address whether treatment for ectopic pregnancy (EP) affects subsequent spontaneous fertility. This multicenter trial was conducted at 17 participating centers in France from 2005 to 2009. Eligible subjects were women with an ultrasound-confirmed EP. Key exclusions included EP as a result of contraceptive failure or in vitro fertilization (since these might confound the time to subsequent pregnancy) or a single fallopian tube (randomization to salpingectomy considered unethical). Women presenting with EP were divided into two arms depending on the activity of the EP. In Arm 1 (less active ectopic pregnancies considered hemodynamically stable), medical treatment was considered practical, and women were randomly allocated to conservative surgery (e.g., salpingostomy with postoperative methotrexate) or methotrexate injection alone. In Arm 2 (active ectopic pregnancies), medical treatment was not considered practical, and subjects were randomly allocated to undergo either a radical (e.g., salpingectomy) or conservative (plus postoperative methotrexate) tubal procedure. Sample sizes were computed to provide a statistical power of 80% to detect a 20% difference in subsequent cumulative fertility rates between treatments in each arm. Cumulative fertility curves were drawn with the Kaplan–Meier method and compared with the log-rank test. Hazard ratios (HRs) were computed with the Cox model. Analysis was performed according to the intention-to-treat principle.
Over 5 years, 207 women were randomized in Arm 1 and 199 in Arm 2. For early EP (Arm 1), the cumulative fertility curves were not significantly different between medical treatment and conservative surgery. The 2-year rates of intrauterine pregnancy were 67% after medical treatment and 71% after conservative surgery (HR, 0.85; 95% confidence interval [CI], 0.59-1.22) Findings were similar among those women presenting with advanced EP (Arm 2), with cumulative 2-year fertility curves not significantly different between conservative (70%) and radical (64%) surgery (HR, 1.06; CI, 0.69-1.63).
DEMETER was the Greek goddess of fertility. EP remains one of the most important surgical problems faced by gynecologists. In the United States, EP occurs in about 1 in 80 pregnancies.1 Over the last 20 years, the surgical treatment of EP has become less needed as early diagnosis and highly effective medical management protocols using methotrexate have developed. Still, the question that occupies our mind as we counsel women with a first EP is “how will this treatment affect her subsequent fertility?” Our job requires us to balance all of the competing goods — lowest overall cost, need for follow-up, invasive nature of surgical procedures vs uncertainty of medical outcomes, risk of complications — and package this into something our patient (and her family) can digest during an acute health care event. The information in this study helps provide more clarity regarding how the treatment decision may impact subsequent fertility.
To study this question, the authors created a protocol designed to look at the real-life treatment decisions involved in early (medical management or conservative surgery) and advanced (conservative or radical surgery) EP. They studied only those women with EP as a result of a spontaneous conception (not contraception failure or IVF), as the primary outcome was subsequent spontaneous intrauterine pregnancy and this group was felt most likely to attempt repeat pregnancy in the next 2 years. However, interest in subsequent fertility was not a requirement for participation. In fact, enrollment was slow in Arm 2, as many women refused participation, apparently out of concern for randomization to the salpingectomy arm. It may be that those women most concerned and motivated about subsequent pregnancy did not participate. Given that, the finding that the 2-year rate of pregnancy was only 64-71% is not bad and may not reflect the overall chance of subsequent fertility.
The noteworthy result is that outcomes are generally good and similar between all of the treatments. Although the two treatment arms represent different severity of disease, the 2-year rate of subsequent pregnancy was similar (71% Arm 1, 70% Arm 2) among women treated with conservative surgery. In Arm 1, the pregnancy rates with medical therapy (67%) were not statistically different from conservative surgery. Although pregnancy rates were slightly lower (64%) among women randomized to salpingectomy in Arm 2, this was not statistically significant.
All studies have limitations and caveats. One important consideration in this study is that all women managed with conservative surgery received routine postoperative methotrexate. In the United States, this practice is less common, with careful follow-up of postop hCG levels used to guide the need for methotrexate treatment at many centers. The DEMETER protocol was guided by evidence reported by Graczykowski and Mishell that the relative risk of developing persistent EP after prophylactic methotrexate was 0.13 (CI, 0.02-0.97) in a study of 129 women with EP randomized to linear salpingostomy with or without postoperative methotrexate (1 mg/kg intramuscularly, within 24 hours postoperatively).2 The DEMETER groups also were unbalanced at baseline, suggesting some irregularity in randomization. However, appropriate statistical adjustments confirmed the main findings of no difference between treatments.
Another factor to consider is that this clinical trial was not designed as a non-inferiority study. The ad hoc statistical margin was a robust 20% difference in fertility rates. Therefore, while the observed effect size of 4-6% noted between treatment groups was not statistically significant, it might be worth considering if this is real. Consider that a 4-6% difference in survival would generally be considered highly relevant in a cancer study. Thus, it is appropriate to reflect on whether the observed difference would affect clinical judgment if a larger study confirmed these results with statistical significance. For early EP (Arm 1), medical therapy was slightly inferior to conservative surgery. This margin is small (4%) and would not motivate me to suggest surgery as an option to women who were good candidates for methotrexate. However, subsequent fertility should not be a factor in recommending against surgery in women interested in that approach. For women with advanced EP, the larger (6%) difference in subsequent pregnancy still argues for a conservative approach with linear salpingostomy.
1. Creanga AA, et al. Trends in ectopic pregnancy mortality in the United States: 1980-2007. Obstet Gynecol 2011;117:837-843.
2. Graczykowski JW, Mishell DR, Jr. Methotrexate prophylaxis for persistent ectopic pregnancy after conservative treatment by salpingostomy. Obstet Gynecol 1997;89:118-122.