INTRODUCTION

Endometriosis afflicts 1 in 9 (11%) people assigned female at birth,¹ taking an average of 6.4 years from first symptoms to diagnosis. It is associated with period pain and on average eight additional comorbid symptoms.² The incidence of infertility is 30–50% higher for people with endometriosis, although many people with endometriosis can conceive without fertility treatment. There is also a 25–50% higher incidence of endometriosis in infertility populations.^3,4

The association between endometriosis and reduced fertility is well documented. Without treatment, the monthly fecundity rate (MFR; monthly chance of conception) in endometriosis cohorts ranges from 2–10%, compared to 15–20% in healthy women.^5,6 This is reflected in both lower livebirth rates (LBR; the number of livebirths per year per 1000 couples, 5.6 versus 12), and cumulative livebirth rates (CLR; the number of livebirths in 36 months per 1000 couples, 16 versus 25), when compared to a general infertility cohort. Women classified with severe (rASRM stage III and IV) endometriosis were particularly impacted, with a LBR of 1 and CLR of 5.1. While older studies did not demonstrate a significant difference in MFR between women with and without mild/moderate endometriosis,⁷ a prospective study in 2004 reported lower pregnancy rates over 3 years in an untreated cohort of 75 women with mild/moderate endometriosis compared to a cohort of 117 women without endometriosis (36% versus 55% respectively; P <0.05).⁸ Similarly, when women undertook donor sperm insemination in natural cycles (controlling for confounding by suboptimal sperm quality)⁹ the MFR was 3.6% in women with untreated endometriosis versus 12% in those with a normal pelvis. Finally, women with endometriosis demonstrate lower MFRs in ovulation induction (OI) and intrauterine insemination (IUI) cycles.¹⁰

In our recent online survey, insensitive fertility advice was commonly reported. This included advocating pregnancy ‘as soon as possible’, often at an inappropriate time (during teenage years) or circumstance (no current partner).¹¹ With improved understanding of the pathophysiology of endometriosis-associated infertility, better non-invasive diagnostics for detecting endometriosis, increasing access to fertility preservation through egg and embryo freezing, and evidence of benefit from tailored, endometriosis-specific fertility treatments, medical professionals are compelled to improve fertility dialogs with endometriosis patients.¹²

DOES DIAGNOSIS MAKE A DIFFERENCE FOR FERTILITY OUTCOMES?

Between 43 and 63% of women with a diagnosis of unexplained infertility are subsequently found to have endometriosis when a laparoscopy is undertaken.^{13,14,15,16,17} In an Australian Women’s Health cohort of 1322 couples, 34.7% of women had an endometriosis diagnosis, with 65.6% diagnosed before and 34.4% after their first IVF treatment. If endometriosis was diagnosed after their first IVF cycle, women were twice as likely to have used IUI, four times more likely to have had 11–36 IVF cycles and less likely to report a birth.¹⁸ Investigating endometriosis prior to using assisted reproductive technologies (ART) optimizes treatment outcomes.

Laparoscopy is costly, difficult to access and carries surgical risk.¹⁹ Prior to 2022, keyhole surgery was considered the ‘gold standard’ and only diagnostic test for endometriosis. The diagnostic test accuracy of transvaginal ultrasound scans (TVUS) and magnetic resonance imaging (MRI) for endometriosis has improved with advances in scan technology and improved skills in gynecological imaging.²⁰ In February 2022, the European Society for Human Reproduction and Embryology (ESHRE) strongly recommended the use of specialist endometriosis TVUS and MRIs in the diagnostic workup for endometriosis.²¹ Subsequently the National Institute for Health and Care Excellence (NICE)²² and Royal Australian and New Zealand College of Obstetricians and Gynaecologists (RANZCOG) guidelines²³ were modified to be concordant, albeit without MRI imaging being recommended as a first line diagnostic test. A caveat of all guidelines is that when a positive finding is seen, the diagnosis is highly accurate (specificity 97% and 95% with TVUS and MRI, respectively), but if endometriosis is not detected, then sensitivity is only moderate (63% and 73%, respectively),¹² and endometriosis cannot be excluded. When a patient presents for fertility investigation, a non-invasive diagnostic ultrasound scan may detect endometriosis but also permits assessment of both ovarian reserve (antral follicle count) and ovarian accessibility for oocyte retrieval.

LOW OVARIAN RESERVE AND FERTILITY PRESERVATION

Early natural menopause was associated with laparoscopically diagnosed endometriosis in two large prospective cohorts. Biopsies from the cortex of ovaries containing endometrioma contain reduced numbers of oocytes compared to contralateral endometriosis-free ovaries from the same person.^24,25 The follicles were smaller and markers of activated follicular recruitment and increased apoptosis were evident, suggesting follicular ‘burnout’.²⁶ Consistent with these biological findings, women with endometriosis have lower anti-mullerian hormone (AMH) levels,²⁷ lower antral follicle counts²⁸ and higher day 3 follicle stimulating hormone (FSH) levels²⁹ which are biomarkers of ovarian reserve.^30,31

Endometriosis alone confers risk for low ovarian reserve, which is likely due to oxidative stress and increased apoptosis of oocytes.³² Analysis of follicular fluid composition shows high levels of reactive oxygen species (ROS), inflammatory cytokines^33,34 and low levels of antioxidant enzyme activity when endometriosis is present.^35,36,37 Further, granulosa cells from endometriosis patients display markers indicating high oxidative DNA damage and apoptosis. Interestingly, bovine oocytes incubated with endometriotic follicular fluid have greater disruption of meiotic spindles compared to oocytes incubated with follicular fluid from people without endometriosis or media-only control fluid.³⁸ An inflammatory intrafollicular environment appears to contribute to oocyte damage and degeneration.

Assessment of ovarian reserve is relevant for anyone with endometriosis. It provides both information about the length of their fertility window and opportunities to preserve fertility by egg and embryo freezing. In 2013, the American Society for Reproductive Medicine (ASRM) endorsed the clinical use of egg freezing for fertility preservation.^39,40 Advances in vitrification improved oocyte survival and fertilization rates when compared to slow freezing^39,41,42 and led to its rapid uptake. Although egg utilization is impacted by loss at thawing and lower rates of fertilization when compared to fresh eggs, embryos created from fresh and frozen eggs have similar pregnancy rates. There is no maximal number of frozen eggs that will guarantee pregnancy and most studies report a freeze-thaw oocyte to livebirth efficiency of approximately 6%.⁴³ In those under 36 years, the cumulative livebirth increased with every additional oocyte until a plateau at 14 mature oocytes, with a cumulative livebirth rate of 85%.⁴⁴ When fertility preservation was undertaken in the presence of endometrioma, less oocytes were retrieved (5.4 ± 3.8 versus 8.1 ± 4.8).⁴⁵ Thus, more than one stimulated cycle may be required to store reasonable numbers of frozen oocytes. It may not be possible financially or emotionally, for people with endometriosis and a low ovarian reserve to freeze large numbers of eggs or it may entail multiple IVF cycles. It is reasonable to discuss fertility preservation with young people after a diagnosis of endometriosis as success rates from egg freezing are highest before 36 years of age.^46,47

Cobo et al. assessed the outcomes of oocyte freezing for women with endometriosis in a large retrospective cohort of 485 patients.⁴⁸ Inclusion criteria included having an endometrioma of at least 1 cm and an antral follicle count of at least 3. Oocyte survival rates after thawing were high at 83.2% and there was a cumulative LBR of 46.4%. People with endometriosis had a high oocyte utilization rate of 43% compared to other cohorts and a short interval between freezing and thawing oocytes (mean 1.5 years).⁴³ This evidence of clinical benefit supports oocyte preservation as a suitable treatment option for endometriosis.

Endometriosis confers an increased risk of infection after transvaginal oocyte retrieval,^49,50,51 although this complication is rare and difficult to quantify statistically. If the ovaries are adherent to the bowel or pelvic side wall, organs and blood vessels can be damaged, whereas accessing the aberrantly located ovaries through the myometrium is unlikely to cause complications. In these situations, prophylactic antibiotic cover in oocyte preservation cycles is a sensible precaution.

Endometriosis surgery is associated with a reduction in markers of ovarian reserve,^52,53 especially when excising endometriomas, high stage disease, or when oophorectomy is contemplated. Surgery for bilateral endometrioma is associated with a lower age at onset of menopause⁵⁴ and a small risk (2.4%) of immediate post-surgical premature ovarian insufficiency.⁵⁵ Attempts at conception can also be delayed by surgery (myomectomy, removal of hydrosalpinges, colostomy and its revision, long waiting lists). Pre-operative embryo freezing permits a planned approach where an embryo can be transferred at an appropriate later date. This approach preempts a post-surgical reduction in oocytes and maintains the high pregnancy and low miscarriage rates of the age of the patient when the embryo was created and frozen, rather than of the more advanced age when the embryo is transferred. If an embryo is not transferred in a fertility preservation cycle, a progestogen containing intrauterine device or subcutaneous, slow-release implant can be left in situ, and progesterone primed ovarian stimulation (PPOS) regimens can be utilized, thus reducing proliferation of the endometrium and endometriotic lesions.

INFLAMMATORY MEDIATORS IN PERITONEAL FLUID ALTER GAMETE FUNCTION AND EMBRYO DEVELOPMENT IN ENDOMETRIOSIS

In women, endometriotic lesions display estrogen and progesterone receptors,^56,57,58 but it is not ethically possible to observe cyclical or longitudinal, visual, or histological changes with repetitive laparoscopy. Our understanding of ectopic endometrial tissue responses to reproductive hormones have thus been delineated in animal models. Baboons with spontaneous and iatrogenic endometriosis have increased levels of inflammatory leucocytes, chemokines and cytokines in their peritoneal fluid and blood.⁵⁹ A cycle of tissue breakdown and repair is seen in engrafted ectopic lesions in endometriosis mouse models of endometriosis^60,61 (see Figure 1 below). The main difference in steroid response between eutopic and ectopic endometrium is that in endometriosis, dying tissues and cells in ectopic lesions are unable to leave the body. The immune system is thus required to launch an inflammatory, innate immune response to phagocytose dying endometrial cells in the pelvis. When the tissue is not completely cleared chronic inflammation, fibrosis, neovascularization and neural sensitization results.⁶² Endometriosis patients experience these events as period pain, bloating and heavy menstrual loss.

Endometriosis-exposed peritoneal fluid has high numbers of peritoneal macrophages associated with an increased inflammatory response and activation of the NFκB pathway.⁶³ They secrete proinflammatory chemokines and cytokines such as monocyte chemotactic protein 1 (MCP-1),⁶⁴ prostaglandin E2 (PGE2),⁶⁵ interleukin 1 (IL-1),⁶⁶ IL-6,⁶⁷ IL-10,⁶⁷ tumor necrosis factor (TNF) and regulated on activation normal T cell expressed and secreted (RANTES).^68,69,70,71 Markers of oxidative stress are dysregulated in the peritoneal fluid from people with endometriosis⁷² including glutathione peroxidase, superoxide dismutase,³² lipid peroxidase⁷³, xanthine oxidase and catalase.⁷⁴ Inflammation and oxidative stress associated with endometriotic peritoneal fluid, negatively impacts the function of gametes and embryos.

Reduced sperm motility⁷⁵ and binding to fallopian tube epithelium⁷⁶ were seen after exposure to peritoneal fluid from endometriotic women. In combination but not in isolation, IL-6 and IL-6R suppressed sperm motility and the number of rapidly moving sperm. Increased sperm DNA damage was also seen when sperm were incubated with endometriosis-associated peritoneal fluid for up to 3 hours.⁷⁷

In a mouse model of endometriosis, the presence of ectopic endometrial lesions did not impact the number of mature MII oocytes per mouse but was associated with a lower proportion of normal oocytes (61 versus 83%) and incomplete extrusion or division of the first polar body and spindle abnormalities.⁷⁸ When metaphase II mouse oocytes were incubated with peritoneal fluid from women with endometriosis, the proportion of oocytes with microtubule damage or chromosomal alterations was higher (68% and 64%, respectively) compared to control peritoneal fluid (24% and 15%) and tubal fluid groups (13% and 13%).⁷⁹

Exposure to endometriosis tissue in vitro and endometriosis-like lesions in vivo also negatively impacts fertilization and embryo development. Fertilization was reduced in an in vitro co-culture system where mouse oocytes were exposed to human endometriotic stromal cells. A mouse model of endometriosis also demonstrated lower numbers of zygotes compared to control mice that were not exposed to endometriosis-like lesions.^80,81

ENDOMETRIOSIS PRIMARILY INTERFERES WITH OOCYTE QUALITY

The comparative influence of endometriosis on oocyte quality and endometrial receptivity can be evaluated in human donor oocyte programs. In a case-control trial, when egg donors had rASRM III/IV endometriosis, lower pregnancy rates were observed in recipients independent of their endometriosis status. When oocytes from the same healthy egg donor were used to create embryos which were then transferred to recipients with or without endometriosis, similar implantation rates were seen.^82,83,84 There is evidence that there is a reduced effect of progesterone on decidualization and endometrial receptivity⁸⁵ in endometriosis. Any impact of this effect, however, appears to be secondary to the impact of endometriosis on oocyte quality.

ENDOMETRIOSIS AND INFERTILITY TREATMENTS

Apart from surgery, there are few fertility therapies that are specifically tailored for endometriosis patients. Treatment strategies have often overlapped with approaches to ‘unexplained infertility’, as people with endometriosis commonly commence fertility treatment without a diagnosis. From a biological perspective, strategies that enhance oocyte utilization and reduce an inflammatory pelvic environment are likely to be most effective.

Improving natural conception

Pursuing natural conception is an option for those with endometriosis. The inflammatory pelvic environment delays rather than prevents conception and many people with endometriosis conceive without using artificial reproductive technologies (ART). Optimizing lifestyle factors, such as reducing smoking, and marijuana use, making healthy dietary changes, increasing exercise, and limiting coffee and alcohol intake, decreases oxidative stress in cells and improves fertility outcomes. This is particularly relevant when endometriosis-derived inflammatory changes are difficult to modify. Suboptimal sperm concentration or quality will contribute to lower fertilization rates and poor embryo development and should be assessed and treated. When sperm counts are significantly low, intracytoplasmic sperm injection (ICSI) and assisted reproductive technologies (ART) should be considered as a first treatment option.

Recurrence of pain often occurs when people with endometriosis stop period-suppressing contraception and try to conceive naturally. Dysmenorrhea is the hallmark of endometriosis; however, if a complex pain syndrome develops, ovulation pain is a common sequela. Subsequent sexual dysfunction can be a barrier to successful natural conception. Women with endometriosis have a nine-fold increased risk of dyspareunia compared to the general female population.⁸⁶ Pelvic floor physiotherapy (stretching of the pelvic floor muscles) or low doses of neural modulators such as amitriptyline can improve pain symptoms. Higher levels of estradiol produced by growing follicles in ovulatory or ART cycles promotes endometriosis tissue growth and can negatively impact pain symptoms over time. It is important to reassess pain throughout fertility therapy and continually reconsider surgically removing lesions or reducing time to pregnancy with in vitro fertilization (IVF) thus limiting exposure to repeated menstrual pain triggers.

Ovulation suppression is not recommended during fertility treatment

Gonadotropin-releasing hormone (GnRH) agonists, antagonists, danazol and the combined oral contraceptive (COC) pill suppress ovulation and inhibit estradiol-dependent proliferation of endometriosis. This explains why pain and disease recurrence are reduced when these medications are prescribed after excisional endometriosis surgery.⁸⁷ A Cochrane library review evaluated 23 randomized controlled trials to determine whether ovulation suppression improved subsequent pregnancy rates upon cessation of these medications.⁸⁸ No benefit for pregnancy was seen when ovulation suppression was used compared to placebo or no treatment (OR 0.97, 95% CI: 0.68 to 1.34, P = 0.8). Supressing ovulation did not improve natural fertility in endometriosis patients and may delay conception by reducing ovulatory cycles. Postoperative ovulation suppression is only recommended by ESHRE to improve pain management for people not desiring immediate pregnancy.²¹

Antioxidants to improve fertility

In endometriosis pain management, the use of antioxidants including vitamin C, vitamin E, melatonin,⁸⁹ resveratrol,⁹⁰ curcumin⁹¹ and co-enzyme Q10 (CoQ10) are associated with lower inflammatory markers in peritoneal fluid and animal models, and pain reduction in clinical endometriosis trials.^92,93 There is evidence of improved fertility outcomes when antioxidants are used. For example, CoQ10 reduced endometriotic tissue volume and inflammatory and angiogenic markers in a rat model of endometriosis.⁹⁴ Additionally, mouse oocytes that were exposed to follicular fluid from people with endometriosis in vivo, had a low maturation rate that trended to recovery when CoQ10 was used as a media supplement.⁹⁵ However, there have been no randomized clinical trials assessing CoQ10’s benefit on fertility outcomes in women. Sixty-three randomized controlled trials were reviewed in a Cochrane study of antioxidant use in women attending fertility clinics, including people with endometriosis.⁹⁶ The evidence was low quality and there were no studies that evaluated its effect on people with endometriosis as an independent group. Evidence that antioxidants may improve clinical pregnancy rates [odds ratio (OR) 1.65, 95% confidence interval (CI) 1.43–1.89; P <0.001, I² = 63%; 35 RCTs, 5165 women] and with less certainty livebirth rates (OR 1.81, 95% CI: 1.36–2.43; P <0.001, I² = 29%; 13 RCTs, 1227 women) was presented. Based on this evidence and biological plausibility, it seems reasonable to support the use of CoQ10 or other antioxidants to women with endometriosis during fertility treatment.

Tubal flushing with oil-based contrast medium (Lipiodol^®)

If couples affected by endometriosis wish to conceive without IVF, fallopian tube patency should be assessed, as endometriosis-associated adhesions and anatomical distortion can disrupt tubal function and hydrosalpinges are more common in endometriosis. The brackish fluid in hydrosalpinges trickles back into the uterine cavity creating inflammation and compromising endometrial receptivity for embryos, which negatively impacts IVF conception rates.^97,98 Tubal patency is assessed by flushing saline, contrast media or dye through the fallopian tubes and assessing its flow into the peritoneal cavity during a hysterosalpingogram (HSG, X-ray), hystero-salpingo contrast sonography (HyCosy, ultrasound) or laparoscopy and dye surgery. Lipiodol is an oil based, iodinated, contrast medium used for fallopian tube patency testing, which is safe in natural conception.⁹⁹

Lipiodol (ethiodized oil) has a long half-life (~50 days)¹⁰⁰ and after traveling through the fallopian tubes, may alter the pelvic environment in endometriosis. Other postulated mechanisms of action include flushing of debris from the fallopian tube, modification of peritoneal macrophage function and altered regulation of implantation-related factors in the endometrium.¹⁰¹ Further studies are warranted to assess the biological effect of lipiodol on the pelvic environment in people with and without endometriosis.

Oil-based contrast media (Lipiodol) was compared to water soluble media at hysterosalpingography in a large multi-centered randomized controlled trial of 2238 couples with unexplained fertility.¹⁰¹ In the oil-based contrast group, 220 of 554 (39.7%) women had an ongoing pregnancy compared to 161 of 554 (29.1%) women in the water group (rate ratio, 1.37; 95% CI: 1.16–1.61). Endometriosis was not evaluated in the enrolled couples, although the proportion with this condition in an unexplained fertility cohort is likely to be 40–60%.

The FLUSH [Flushing with Lipiodol for Unexplained (and endometriosis-related) Subfertility by Hysterosalpingography] study randomized 158 women with normal fallopian tubes to an HSG using lipiodol or no treatment. A subgroup analysis compared those with proven endometriosis (n = 62) to those without proven endometriosis (n = 96). In the endometriosis group, lipiodol flushing was associated with a higher 6-month pregnancy rate (48.0% vs. 10.8%, RR 4.44, 95% CI: 1.61–12.21) and live birth rate (40.0 vs. 10.8%, RR 3.70, 95% CI: 1.30–10.50) which was not seen in the group with unknown endometriosis status. A network meta-analysis that compared interventions for endometriosis related infertility showed that lipiodol led to higher odds of clinical pregnancy when compared with placebo/no intervention (OR = 7.56; 95% CI: 2.02–29.37).¹⁰² A Cochrane review concluded that flushing fallopian tubes with oil contrast media may improve the clinical pregnancy rate (OR 1.42, 95% CI: 1.10–1.85, 6 RCTs, 2598 women, I² = 41%, low‐quality evidence), but with only three studies, there was not enough evidence to determine if the livebirth rate was affected.¹⁰³

Lipiodol is generally well tolerated, although can cause cramping pain and bleeding. Extremely rare side-effects include allergy and embolism. Lipiodol contains a high dose of iodine which can lead to temporary subclinical hypothyroidism in some women. Assessment of thyroid function and appropriate thyroxine supplementation should be instituted for 6 months after use, particularly in early pregnancy.¹⁰⁴ The evidence supports the use of lipiodol oil-based contrast in unexplained infertility and there is limited evidence of its benefit for people with endometriosis desiring fertility. It is thus reasonable to consider its use, particularly if tubal patency testing is required, for those with endometriosis who wish to conceive without using ART.

Intrauterine insemination and endometriosis

Intrauterine insemination (IUI) improves pregnancy rates by ensuring sperm are present in sufficient numbers around a recently ovulated oocyte when it is receptive to fertilization. When gonadotropins or ovulation induction agents are used, more than one follicle and higher estradiol levels may be seen. IUI can benefit couples with mild male factor defects, who struggle to have intercourse close to ovulation, who find it difficult to track ovulation or who live in different geographical locations. However, IUI does not alter the peritoneal environment in endometriosis and tubal patency is a requirement. The number of chances to fertilize an egg per month is capped at 1 or 2, due to risks of multiple pregnancy.

Randomized controlled trials that assessed IUI in endometriosis patients were conducted in the 1990s in small cohorts. In 1997, a 5.6-times higher livebirth rate was observed when 53 people with minimal to mild endometriosis were randomized to ovarian stimulation with follicle stimulating hormone (FSH) and IUI compared to 50 who received expectant management.¹⁰⁵ FSH use significantly impacted the rate of human chorionic gonadotropin (hCG) positive pregnancies in a partially randomized study in 57 couples with minimal to mild endometriosis.¹⁰⁶ Thirty-one couples were randomized to FSH and IUI and 15 of 127 cycles led to a positive pregnancy test (11.8%). This was 5.1 times higher than 26 couples not using FSH, where only 2 of 96 cycles resulted in pregnancy. Compared to expectant management and IUI alone, IUI with ovarian stimulation appears to improve pregnancy rate in minimal to mild endometriosis and underpins ESHRE’s recommendations that IUI may be offered to couples affected by rASRM stage I/II endometriosis.

When compared to couples with unexplained infertility, IUI and ovarian stimulation appeared less effective for people with endometriosis. The relative risk (RR) for pregnancy in unexplained fertility was 8.8 (95% CI: 1.1–71.3) compared to 5.1 (95% CI: 1.1–22.5) in the endometriosis group.¹⁰⁶ Similarly, higher pregnancy rates after IUI and ovarian stimulation when 119 couples with unexplained infertility (33.6%) were compared to those who had endometriosis at laparoscopy (16.3%).¹⁰⁷ The presence of lesions may influence IUI outcomes as there was no difference in pregnancy rates after FSH-IUI, when women who had surgical excision of rASRM stage I/II in the previous 6 months, were compared to those without endometriosis.

In a retrospective review of 65 patients having IUI with rASRM stage III/IV endometriosis, ovarian stimulation was associated with higher cumulative ongoing pregnancy rates (40% vs. 15.6%).¹⁰⁸ Kim et al. compared ultralong and long GnRH agonist use before IUI in 80 women with endometriosis.¹⁰⁹ The pregnancy rates were not influenced by GnRH agonist use with minimal/mild endometriosis, but they were increased in patients using an ultra-long downregulation GnRH protocol when higher stage endometriosis was present. The utility of IUI in rASRM III/IV endometriosis is less evident, however ESHRE supports its use, even though IVF provides a better chance of conception.¹⁰⁹

Surgery

Excision of endometriosis removes tissue prone to breakdown and repair in response to cyclic reproductive hormones, limiting an inflammatory immune response to dying endometrial cells in the pelvic cavity. Surgical removal of endometriosis is associated with reduced concentrations of inflammatory cytokines in plasma from women with endometriosis,¹¹⁰ although its impact on the peritoneal environment requires exploration. In a cohort of women with endometriosis who were trying to conceive, 50% conceived naturally within 6 months after an operative laparoscopy.¹¹¹

A 2020 Cochrane review assessed whether surgical treatment increased the chance of spontaneous pregnancy if endometriosis was present.¹¹² Moderate quality evidence from three RCTs^113,114,115 that included women with in rASRM stage I/II disease, showed surgical excision of endometriosis increased viable clinical pregnancy rates (OR 1.89; 95% CI: 1.25–2.86) when compared to diagnostic laparoscopy only. A similar conclusion was reached in a network meta-analysis which demonstrated an increased odds ratio for clinical pregnancy following surgical laparoscopy compared with placebo (OR 1.63; 95% CI: 1.13–2.35).¹⁰² Jin et al. 2014 reported that live birth rate was significantly increased after laparoscopic surgery for endometriosis (relative risk [RR] 1.52; 95% CI: 1.26–1.84, 4 studies; 741 patients).¹¹⁶ Based on this evidence the ESHRE guidelines recommend operative laparoscopy as a treatment option for infertility in rASRM stage I/II endometriosis as it improves the rate of ongoing pregnancy.

It is less clear whether surgery benefits fertility outcomes in the presence of endometriomas and deep infiltrating endometriosis, as no studies have compared conception after surgery for moderate to severe endometriosis with diagnostic laparoscopy or no treatment. There are also no comparisons between operative endometriosis surgery and IVF.¹¹² However, expectant management of moderate to severe endometriosis was associated with very low pregnancy rates (MFR 3.2%) in a prospective cohort of 123 infertility patients with endometriosis. Conversely, pregnancy rates after surgical treatment of endometrioma were 43.8% (95% CI: 22.5–66.4) in a review of eight studies,¹¹⁷ whereas two reviews indicated postsurgical pregnancy rates of 37–50.5% when deep endometriosis was excised,^118,119 although pregnancy rates were lower (21–28%) if the bowel was involved.^118,120 With less compelling evidence, ESHRE guidelines have formulated weak recommendations that clinicians may consider operative laparoscopy for the treatment of endometrioma, and deep endometriosis associated infertility, as it may increase the chance of natural pregnancy and represents a treatment option.^21,121 The ESHRE guideline development group recommended that the decision to perform surgery should be guided by the presence or absence of pain symptoms, patient age and preferences, history of previous surgery, presence of other infertility factors, ovarian reserve, and estimated endometriosis fertility index.

In vitro fertilization (IVF)

Stimulating multifollicular development with gonadotropins and retrieving oocytes preovulation for fertilization in the lab (IVF), moderates the impact of both lower oocyte numbers and the inflammatory pelvic environment for people with endometriosis. IVF also addresses fallopian tube dysfunction, sexual dysfunction and reduces time to conception compared to treatments where oocytes are left in situ. The disparity in pregnancy rates and birth outcomes seen in endometriosis when fertilization and embryo development occur in the reproductive tract, is ameliorated when ART is used.

Fertility outcomes for women with endometriosis have been compared to other infertility causes in cohorts and in large databases of reported ART outcomes. In 2016, 347,185 autologous fresh and frozen ART cycles recorded in the USA’s Society for Assisted Reproductive Technology’s (SART) national database from 2008 to 2010.¹²² Couples with an endometriosis diagnosis were more likely to have a canceled cycle (11.3%) than those with isolated endometriosis, tubal factor, or unexplained infertility (8.5%, 8.3%, 8.1% respectively, P <0.0001). When fresh cycles were analyzed, endometriosis was associated with a reduction in oocyte yield (RR 0.91 [0.91–0.92]), and the proportion of blastocysts transferred (RR 0.96 [0.93–0.99]).¹²³ IVF cycles are more likely to be canceled, demonstrate a reduced response to FSH¹²⁴ and have a lower number of oocytes retrieved and embryos created in endometriosis cohorts.¹²⁵

Interestingly, most studies report that IVF livebirth rates for couples with rASRM stage I/II endometriosis are comparable to couples with non-endometriosis associated infertility, despite a lower number of oocytes retrieved. A large systematic review did not find a significant reduction in implantation rates (RR = 0.83, 95% CI: 0.68–1.01, P = 0.07), clinical pregnancy rates (RR = 0.94, 95% CI: 0.83–1.07, P = 0.35) or livebirth rates (RR = 0.92, 95% CI: 0.83–1.02, P = 0.10) in mild/moderate endometriosis, although fertilization rates were significantly lower (RR 0.93; 95% CI: 0.87–0.99; 7 studies; 2044 patients). A similar conclusion was reached when patients with endometriosis only were compared to other infertile groups in the SART analysis¹²² and in an analysis of 1749 patients from the Latin American Registry maintained by the Latin America Network of Assisted Reproduction (REDLARA).¹²⁶ An explanation for this consistent finding is that, in mild/moderate endometriosis, even a smaller cohort of oocytes likely contains at least one of sufficient quality to support a term pregnancy.

In contrast, the literature reviews that compare rASRM stage III/IV endometriosis to other causes of infertility, indicate that pregnancy rates are significantly lower. Harb et al. 2016 found a lower implantation rate (RR = 0.79, 95% CI: 0.67–0.93, P = 0.006) and clinical pregnancy rate (RR = 0.79, 95% CI: 0.69–0.91, P = 0.0008) in rASRM stage III/IV disease, although the lower livebirth rate did not reach statistical significance (RR = 0.86, 95% CI: 0.68–1.08, P =0.19, 9 studies; 312 patients).¹²⁷ Similarly, Hamdan and colleagues described a significantly lower clinical pregnancy rate (OR 0.60; 95% CI: 0.44–0.81; 15 studies; 9,471 patients) and live birth rate (OR 0.77; 95% CI: 0.64–0.92; 8 studies) when those with stage III/IV endometriosis were compared to those with no endometriosis.¹²³ In a retrospective cohort analysis of 164 IVF-ET cycles in 148 women, a significantly lower pregnancy rate was seen when IVF cycles impacted by stage III/IV endometriosis were compared to stage I/II disease or tubal factor infertility (PR 9.7, 25 and 26.1%, respectively).¹²⁸ There is good evidence that the stage of endometriosis is a factor to take into consideration when planning IVF with rASRM stage III/IV disease being associated with poorer IVF outcomes.

Concomitant fertility risk factors such as male factor infertility, tubal factor or diminished ovarian reserve also significantly increase the likelihood of poor IVF outcomes when endometriosis is present. In the SART database review, endometriosis in association with other reproductive pathology was associated with the lowest chance of livebirth. Live birth rates were reduced by 19–26% in fresh cycles and 12–18% reduction in frozen cycles in comparison to unexplained, tubal factor and all diagnostic groups combined. In contrast endometriosis in isolation was associated with a similar or slightly higher live birth rate compared to other infertility diagnoses.¹²²

Surgical removal of endometriomas prior to IVF did not appear to impact pregnancy outcomes in two systematic reviews studies.^123,129 Although cycle cancelation rates were higher and oocyte yield was lower in the group with endometrioma, comparable clinical pregnancy and livebirth rates were observed when compared to those without endometrioma. Surgical management of endometrioma prior to IVF did not alter pregnancy outcomes and the authors recommended individualizing care. ESHRE guidelines recommend antibiotic prophylaxis as a good practice point if endometriomas are present, although the risk of pelvic infection is low.⁴⁹

Stimulation regimens

When embryos are transferred in fresh IVF cycles the luteinizing hormone (LH) surge is suppressed to prevent early ovulation, which occurs when multifollicular development induces supraphysiological estradiol levels. GnRH agonists can be used from the midluteal phase of the previous cycle to drain LH from the pituitary and prevent the LH surge. If GnRH agonist are used for up to 3 months, endometrial proliferation is suppressed, and endometriosis lesions become quiescent reducing pain symptoms when periods cease. In one small study, women with endometriosis were randomized to ultra-long 3-month downregulation with a GnRH agonist (n = 25) or a standard GnRH agonist IVF cycle (n = 26). The ultra-long downregulation demonstrated significantly higher ongoing pregnancy rates (80% vs. 53.85%) and a trend toward higher implantation rates (42.68% vs. 30.38%).¹³⁰ However ultra-long downregulation with GnRH agonists may down regulate FSH receptors on granulosa cells and may suppress follicular recruitment and oocyte numbers.

GnRH antagonists block GnRH signaling to the pituitary preventing the release of LH and be used in shorter, patient friendly IVF cycles with less side effects. Benschop et al. reviewed four trials that assessed stimulation regimes on IVF outcomes in endometriosis.¹³¹ Standard GnRH agonist and GnRH antagonist treatment in women with endometriomas resulted in similar clinical pregnancy rates, but women using agonist cycles had slightly more mature oocytes retrieved (MD 1.6; 95% CI: 2.4–0.76).

A recent Cochrane review (8 studies, 640 participants) was not able to determine if ultra-long GnRH agonist cycles impacted livebirth (RR 0.48, 95% CI: 0.26–0.87) and clinical pregnancy rates (RR 0.48, 95% CI: 0.26–0.87) or mean number of oocytes (MD 0.72, 95% CI: 0.06–1.38) or embryos (MD -0.76, 95% CI: -1.33 to -0.19) as the quality of evidence was very low.¹³² It is important to weigh up each individual’s ovarian reserve, degree of endometriosis/adenomyosis and the possible side effects of treatment when choosing an IVF stimulation regimen.

Managing miscarriage

Several large population cohorts and linkage studies have reported an association between endometriosis and pregnancy loss;^133,134 however, smaller prospective and retrospective studies have reported conflicting and uncertain results.¹³⁵ A 2020 systematic review (39 studies, 697,984 women) identified a higher risk of miscarriage in women with endometriosis with a spontaneous conception (OR: 1.81, 95% CI: 1.44–2.28, I² = 96%).¹³⁶ Concordantly, a 2019 systematic review (104 studies) detected an increased risk of miscarriage (OR 1.30, 95% CI: 1.25–1.35) and other obstetric complications in endometriosis patients.¹²⁵

There is some evidence that miscarriage risk is impacted by exposure of eggs and embryos to factors in the peritoneal environment when endometriosis is present. For women with endometriosis who undertook IVF (and the oocytes were taken out of the pelvic environment), the risk of miscarriage was comparable to those with tubal infertility (OR: 1.03, 95% CI: 0.92–1.14, I² = 0%). Additionally, endometriosis, but not adenomyosis, conferred a risk of miscarriage when a small cohort of 214 of women with MRI identified endometriosis and adenomyosis were compared to those with adenomyosis alone (OR 3.2, 95% CI: 1.1–9.65). This risk was significantly higher with deep infiltrating endometriosis (OR 4.37, 95% CI: 1.32–14.53).¹²⁵ It is important to continue to support women with endometriosis in the first trimester of pregnancy, and ensure their obstetrician is aware the increased risk endometriosis attributes to pregnancy.

Laboratory incubation in IVF may improve the conditions for fertilization and early embryo development. Pre-implantation screening can also be undertaken, which reduces miscarriage risk via the transfer of euploid embryos. Progesterone supplementation has been shown to reduce the risk of miscarriage risk in those with bleeding in the first trimester and a history of three previous miscarriages.¹³⁷ Progesterone supplementation is commonly used in ART cycles and may especially benefit endometriosis patients, as it is theorized that the endometrium is ‘resistant’ to progesterone.⁸⁵ IVF may prove to be beneficial in reducing miscarriage risk for women with endometriosis, but more research is required to establish this premise in prospective cohorts and trials.

SUMMARY (ALSO SEE FIGURE 2)

Endometriosis is associated with low ovarian reserve and reduced fertility. Specialist transvaginal ultrasound scans and MRIs improve detection of endometriosis before fertility treatment, which enhanced fertility outcomes in a large cohort of women. This information and accurate ovarian reserve testing, empowers young people with endometriosis with choices that include preserving fertility by freezing eggs.

Effective treatments for endometriosis-associated infertility optimize utilization of oocytes and reduce exposure to an inflammatory peritoneal environment. Minimizing pelvic pain and dyspareunia and reducing other causes of oxidative stress optimizes natural conception when surgery or ART is not desired. Limited research suggests flushing the fallopian tubes with oil-based contrast media (Lipiodol) may specifically benefit women with endometriosis, but the mechanism of effect needs to be elucidated.

Surgery improves spontaneous pregnancy rates in mild/moderate endometriosis, but has less impact in more severe disease. When lowering ovarian reserve is a risk, oocyte or embryo freezing can be undertaken beforehand and the embryos transferred after surgical recovery. Age, ovarian reserve, pain symptoms and patients’ preference are factors to take into consideration when contemplating surgery as a treatment option.

IVF extracts oocytes and embryos from an inflammatory peritoneal environment and utilizes oocytes efficiently. Pregnancy rates are comparable to other causes of infertility in mild and moderate stage endometriosis but decreased in late-stage disease. This is despite lower ovarian reserve, less oocytes retrieved, less mature oocytes and less embryos created in IVF cycles undertaken by women with endometriosis.

There is a paucity of evidence to support a particular preference between GnRH agonist and antagonist cycles. Progesterone primed ovarian stimulation (PPOS) has several theoretical advantages in endometriosis treatment when an embryo transfer is not contemplated, including suppressing lesion growth and minimizing menstrual loss and pain at the end of the cycle.

Finally, there is an increased risk of miscarriage in those with endometriosis and support in early pregnancy is recommended. Further research is required to determine if IVF or progesterone supplementation will reduce the risk of miscarriage.

PRACTICE RECOMMENDATIONS