Endometriosis and Assisted Reproductive Technologies: Maximizing Outcomes

Eric S. Surrey, MD


Semin Reprod Med. 2013;31(2):154-163. 

In This Article

Does Pre-IVF Cycle Medical Suppression Improve Outcomes?

Traditional medical therapy for symptomatic endometriosis such as progestins, danazol, and gonadotropin-stimulating hormone (GnRH) agonists has been shown to have little impact on enhancing spontaneous pregnancy rates in infertile endometriosis patients.[42] However, if the negative effect of this disease process on fertility returns rapidly after discontinuation of medication, then one could hypothesize that any benefits of medical suppression on enhancing fertility would be most evident if pregnancy could be achieved during a time of maximal suppression. This could only occur with the use of the assisted reproductive technologies.

Most of the investigations in this regard have examined prolonged use of GnRH agonists prior to IVF. In a prospective randomized multicenter trial, Surrey et al evaluated the effect of a 3-month course of a GnRH agonist administered immediately prior to initiating controlled ovarian hyperstimulation (COH) in preparation for IVF in 25 patients with surgically confirmed endometriosis.[43] Significantly higher ongoing pregnancy rates with a trend toward higher implantation rates were appreciated in comparison to controls with endometriosis who underwent standard COH protocols and IVF without prolonged GnRH agonist therapy. Of note is the fact that a higher percentage of patients who received prolonged agonist therapy had more advanced disease, a group that one would expect to have inherently poorer outcomes (Fig. 2).

Figure 2.

Ongoing pregnancy and implantation rates in endometriosis patients after a 3-month course of gonadotropin-releasing hormone agonist (GnRHa) versus control prior to in vitro fertilization. Modified from Surrey et al.[43] (Fig. 1). *p < 0.05.

These findings have been demonstrated by others. Seven previous studies of varying design have assessed the effect of suppression with a GnRH agonist (GnRHa) before IVF or GIFT.[44–50] The length of suppression varied from 6 weeks to 7 months. Some studies lacked control groups, but a beneficial effect of pretreatment was suggested by all.

Rickes and colleagues evaluated the effect of pre-cycle surgical treatment of endometriosis alone or in combination with a 6-month postoperative treatment course of a GnRHa on IVF or COH-intrauterine insemination (IUI) outcome in a prospective randomized trial of 110 patients.[51] The pregnancy rates were significantly higher for both forms of fertility therapy in those patients treated with a prolonged postoperative GnRHa course. However, when patients were stratified based on disease stage, a statistically significant difference was only appreciated among patients with stages III/IV endometriosis who underwent IVF. A summary of the results of the randomized trials is displayed in Table 4 .

Sallam et al more recently performed a Cochrane Database analysis of three of these prospective randomized trials including 163 endometriosis patients undergoing 3 to 6 months of pre-cycle GnRHa treatment.[52] This intervention resulted in significantly improved rates of both live birth (OR: 9.1%; 95% CI, 1.08 to 78.22) and clinical pregnancy (OR: 4.28; 95% CI, 2.0 to 9.15).

There are no studies that compare varying lengths of suppressive therapy or whether this approach should be offered to specific subgroups of endometriosis patients given the associated increased expense and time delay before pregnancy can occur.

The mechanism of action of this effect has not been clearly established. It has been suggested that, aside from their primary mechanism, GnRHa may act to diminish concentrations of peritoneal fluid metalloproteinase tissue inhibitors, downregulate peritoneal fluid inflammatory proteins, and increase apoptosis and expression of pro-apoptotic proteins.[53,54,55] Others have shown that GnRHa may significantly decrease endometrial nitric oxide synthase expression.[56]

A great deal of interest has surrounded the role of endometrial β3 integrin expression in this setting. Lessey et al previously demonstrated that administration of a GnRHa for 3 months to women with stage I/II endometriosis and aberrant endometrial β3 integrin expression resulted in a 64% rate of return of expression.[57] Ruan and coworkers reported that, in a murine model, impaired endometrial β3 integrin and leukemia-inhibitory factor expression as well as uterine receptivity resulting from ovarian stimulation were partially restored after GnRHa administration.[58]

Given that the pregnancy rate in control patients who were not administered prolonged agonist therapy still remained relatively high in the trial by Surrey et al, it is clear that not all endometriosis patients require this intervention.[43] These data beg the question of whether endometrial β3 integrin expression can be used as a marker to determine which patients might be candidates for prolonged pre-cycle GnRHa therapy. In a case-control study of 74 consecutive IVF candidates believed to be at high risk for implantation defects due to prior IVF failure despite adequate embryo quality and/or endometriosis, we reported a 48.6% prevalence of absent endometrial integrin expression.[59] Of those who had undergone laparoscopy, 52.8% had a diagnosis of endometriosis of whom 57.1% had stage III/IV disease. Miller and coworkers reported that live-birth rates from IVF were significantly increased in patients with positive versus negative integrin expression (38% versus 7%; p < 0.05).[60] Farrell and colleagues reported a small series of 11 patients with aberrant integrin expression and in phase endometrial biopsies who were administered an 8-week course of a GnRH agonist in conjunction with norethisterone acetate prior to IVF. Nine experienced ongoing pregnancies.[61]

In an effort to resolve this issue, Surrey and colleagues recently reported the results of a prospective randomized pilot trial of 36 endometriosis patients undergoing IVF.[62] The patients were randomized after assessing β3 integrin expression to receive either 3 months of GnRHa prior to initiating ovarian stimulation or to proceed directly to ovarian stimulation. Interestingly, a trend toward higher pregnancy rates that did not achieve statistical significance was noted in integrin-positive patients administered prolonged GnRHa. This is the opposite of what one would have predicted. In this study, the value of a negative biopsy in predicting ongoing IVF pregnancy after integrin expression was only 44.4% (Fig. 3). These results would either suggest that evaluating integrin expression is of little value in determining which patients would benefit from prolonged GnRHa, or they may have been confounded by a limited sample size or by the fact that patients in the control groups moved directly to IVF after biopsy, whereas study group patients did not undergo stimulation for 3 months. There are data suggesting that the performance of an endometrial biopsy alone may enhance implantation rates as a result of the localized injury, particularly in patients with a history of implantation failure.[63,64]

Figure 3.

In vitro fertilization cycle outcomes after randomization of endometriosis patients with negative pre-cycle endometrial β3 integrin expression to 3 months of gonadotropin-releasing hormone agonist (GnRHa) or no therapy prior to controlled ovarian hyperstimulation and oocyte aspiration. No differences were statistically significant. Modified from Surrey et al62 (Fig. 3).

There are a host of other unresolved issues. We do not know the ideal duration of therapy. Is there a need for repeated courses of therapy if an initial cycle is unsuccessful? A significant concern with the administration of GnRHa to patients with compromised or diminished ovarian reserve is the deleterious effect on ovarian response to subsequent gonadotropin therapy. In this circumstance, it may be wise to vitrify all embryos after appropriate stimulation and then administer the GnRHa prior to endometrial preparation for frozen embryo transfer.

There is even more limited data regarding the use of other suppressive medical therapies. Tei et al reported nine patients with reduced endometrial integrin expression. Repeated IVF failures who were treated with a 12-week course of danazol 400 mg showed a significant increase in expression in the first ovulatory cycle after completion of therapy.[65] No clinical outcomes were reported, however.

More recently, Miller and colleagues reported that the administration of the aromatase inhibitor letrozole 5 mg daily for days 2 to 6 of gonadotropin stimulation resulted in cycle outcomes that were similar to patients who were integrin receptor positive and not treated with this agent.[60] The weakness of this retrospective trial is the fact that the authors did not randomize integrin-negative patients to similar protocols with or without the use of letrozole. Although the results are encouraging, it is difficult to draw definitive conclusions.

De Ziegler and coinvestigators recently evaluated the role of a 6- to 8-week course of oral contraceptives in patients planning IVF with either surgically diagnosed endometriosis or those with sonographic suspicion of the presence of endometriosis.[66] The use of oral contraceptives resulted in higher pregnancy rates per retrieval than in controls (35% versus 12.9%, p = 0.01). This impact was even greater in those with presumed endometriomas. It is important to note that this was a retrospective trial and that control patients were both significantly older and had higher baseline follicle-stimulating hormone levels. An additional concern is the lack of documentation of endometriosis in all patients.

The dearth of appropriately designed randomized trials makes it difficult to determine which subset of endometriosis patients would benefit from pre-IVF cycle GnRHa (or potentially other suppressive therapy). I would propose that suppressive intervention be considered in the subsets of infertile endometriosis patients with significantly advanced disease, those with severe pain, and/or those with a history of prior IVF cycle failure particularly after transfer of good quality embryos.