Effects of endometriosis on in vitro fertilisation – Myth or reality?

Endometriosis affects about 10% of reproductively aged women and up to 40–50% of infertile women. That an association exists between endometriosis and fertility is undisputed. Endometriosis is thought to create an inflammatory pelvic milieu that is toxic to oocytes, sperm and embryos, thereby compromising natural (in vivo) conception.1 In vivo prospects may be further hampered by physical sequelae of inflammatory scarring such as tubal dysfunction/occlusion, distorted tubo-ovarian relationships and ovulatory dysfunction secondary to endometriomas. During in vitro fertilisation (IVF), oocytes and sperm are brought together outside the body (in vitro), thereby mitigating against toxic effects of the endometriotic pelvic environment. An ongoing debate has therefore centred around the extent to which endometriosis might compromise the performance of a treatment like IVF in which fertilisation and preimplantation embryo development occur outside the pelvis. A recent study based on Australian data addresses this debate by comparing outcomes of intrauterine insemination (IUI) – in which fertilisation and preimplantation embryo development occur in vivo – with IVF in women with endometriosis.2 This was a retrospective national cohort study involving 1322 participants from the Australian Longitudinal Study on Women's Health, 459 of whom had endometriosis and were born between 1973 and 1978. Participants were first surveyed in 1996, then every three years from 2000, and were identified as having had at least one cycle of either IUI or IVF through linked Medicare Benefits Scheme records, providing a unique longitudinal study design on an Australian community-based sample. It was found that women who used more IUI were more likely to do more treatment cycles and less likely to report a birth than women who predominantly did IVF. Conversely, outcomes for the predominantly IVF-treated women were comparable to those of women without endometriosis altogether, suggesting that IVF can effectively counter adverse fertility effects of endometriosis.2 The bedrock of successful IVF is the creation of top-quality embryos (typically at the blastocyst stage of development) and their transfer to a receptive endometrium. Because oocytes provide virtually all the cytoplasmic building blocks required by the embryo, oocyte quality is rate limiting for embryo quality.3 Consequently, and because oocyte quality declines with advancing female age, IVF success is inversely proportional to female age. The other key component, endometrial receptivity, is dependent upon appropriate hormonal exposure (oestrogen and progesterone) along with the regulated expression of a multitude of genes, many of which we are only just beginning to understand. Two critical issues at the centre of the endometriosis–IVF debate, therefore, pertain to whether endometriosis impacts oocyte quality and/or endometrial receptivity. Because embryo quality is a direct reflection of oocyte quality, studies have addressed this question by asking whether embryo quality is inferior in age-matched women with endometriosis. A recent meta-analysis of 22 studies compared embryo quality based on morphological grading in women with and without endometriosis and found similar rates of embryo formation, cleavage and production of high-quality embryos.4 Using different methodology, another study compared livebirth rates in women with endometriosis after the transfer of embryos derived from either autologous or donor oocytes, thereby enabling ‘endometriosis oocytes’ to be compared with a gold standard of high oocyte quality (donor oocytes).5 No significant difference in livebirth rates was observed between the two groups after either fresh or frozen embryo transfers (FETs).5 Foregoing studies relied on surrogate markers of oocyte quality, which was a disadvantage. In particular, morphological grading of embryos is highly subjective and therefore lacks predictive certainty, with 20–80% of good-graded embryos containing an abnormal chromosome complement depending on female age; despite having similar morphological grades, about one in four embryos is aneuploid in women aged below 32 years compared with a staggering three in four aneuploidy rate for those above 42 years.6 A significant progress has been the use of preimplantation genetic testing for aneuploidy (PGT-A), which, despite its shortcomings, arguably provides the most objective prospective readout of embryo quality available to date. PGT-A seeks to determine whether an embryo has a normal chromosome complement (euploid) or an abnormal one (aneuploid) by analysing a five- to ten-cell biopsy taken from the trophectoderm (placental precursor) of blastocyst-stage embryos.7 Using aneuploidy rates as a measure of embryo quality, one retrospective study analysed 1880 PGT-A-tested blastocysts from 305 women with endometriosis and compared them to 23 054 PGT-A-tested blastocysts from 3798 patients without endometriosis.8 If there is indeed an effect on oocyte quality, one would expect higher aneuploidy rates in endometriosis embryos. Highly significantly, however, no statistically significant difference in aneuploidy rates was identified between the two groups after stratifying by age.8 In keeping with this, another more recent study found that aneuploidy rates in 250 blastocysts from endometriosis patients were similar to those in 1332 blastocysts from patients with male factor infertility (45.2% vs 42.4%, P = 0.416).9 To rigorously address this question, one must first be confident that embryo quality is equivalent in both the test and control groups, thereby controlling for the variable with the overriding influence on pregnancy success. Based on the foregoing discussion, the most robust approach for achieving this involves PGT-A, which not only provides reassurances regarding ploidy but, with modern-day blastocyst-biopsy based regimes, also means that embryos of comparable developmental stages are being studied. Many earlier reports lacked rigour in controlling for the embryo variable by incorporating study groups in which embryo quality was defined by morphological grading, involved transfer of embryos of differing developmental stages or had varying numbers of embryos transferred. A landmark study recently investigated FET of PGT-A-tested euploid embryos in patients with endometriosis (54 cycles) and patients undergoing IVF for isolated male factor infertility (therefore not expected to have an endometrial defect, 355 cycles).9 It is notable as well that all patients in the endometriosis arm were diagnosed surgically, that patients in both groups underwent similar medicated FET cycles involving exogenous oestrogen and intramuscular progesterone and that mean female age and mean number of embryos transferred (predominantly single embryo transfers) were comparable between the two groups. Blastocyst development, numbers of good-quality blastocysts produced and euploidy rates were the same in both groups. Importantly, patients with endometriosis had similar rates of clinical pregnancy, miscarriage and livebirth to patients with male factor infertility.9 An important proviso here is that all FETs were undertaken using medicated cycles, which has been purported to suppress the detrimental effects of ectopic endometrium on uterine endometrium.1 It is possible that endometriosis might have a greater effect on endometrial receptivity with fresh embryo transfers or when FETs are undertaken using natural cycle regimes. Collectively, these results suggest that when known high-quality embryos are transferred, patients with endometriosis do not exhibit any marked defect in endometrial receptivity, at least during medicated FETs. The emerging picture, including from recent Australian data,2 increasingly indicates that endometriosis exerts its most detrimental impact when fertilisation occurs in vivo and that IVF can counter these negative effects. The utility of endometriosis-related surgery as an adjunct for improving IVF outcomes is therefore questionable but a critically important issue to address because about 35% of women undergoing assisted reproductive treatment in Australia have endometriosis.2 Endometriomas are perhaps the most significant form of endometriosis from the IVF perspective as they can reduce ovarian reserve and compromise access to normal ovarian tissue. Consistent with the former, a recent meta-analysis focusing on the effect of untreated endometriomas in IVF found that significantly fewer mature oocytes were obtained compared with unaffected controls.10 Notably, however, high-quality embryo formation and pregnancy rates were unaffected despite the continued presence of untreated endometriomas, suggesting that oocyte quality was not impaired and, importantly, that prospects remained very good in the absence of surgery.10 For women with normal ovarian tissue accessible for oocyte retrieval, surgical excision of endometriomas may not therefore improve either oocyte quality or oocyte numbers and could well have a negative impact on the latter given the likelihood of inadvertently removing normal ovarian tissue. H.A.H. is supported by the Christopher Chen Endowment Fund and NHMRC grants APP1078134, APP1103689 and APP1122484. The author declares no conflict of interest.