Surgical treatment is an excellent option for women with endometriosis and infertility

For a woman with suspected endometriosis and infertility there are only three options for becoming pregnant: It is estimated that 30–50% of women with endometriosis may have difficulty becoming pregnant,1 so how do women choose and which option(s) should they be directed to? Like so many areas of gynaecology, it is not a clear-cut answer and with no head-to-head study of surgery versus IVF for endometriosis-associated infertility, yet IVF is often recommended as the only pathway for women with endometriosis and infertility – and for a variety of factors. The first factor that should be examined is that expectant management for pregnancy is the optimal method since it is the safest, least invasive and cheapest of all of the options and detailed, direct and repeated counselling around timed intercourse is essential. However, time, vocational and social pressures often lead to this first simple step being overlooked as the primary method for dealing with infertility. Despite the long-term association of endometriosis and infertility, we still remain unclear as to exactly how the former leads to the latter, although there is no shortage of theories. Sadly perhaps, there are too few high-quality studies to direct debate, yet let us consider what those data are at the current time. The EndoCAN study from 19972 remains the largest and best of the randomised controlled trials (RCTs) on endometriosis and infertility. It reports a near-doubling in fecundity for women having excision of early-stage disease (only women with stages I–II disease were included in this cohort). Of note is that there are 12 laparoscopies required for each additional pregnancy gained in this manner and the question remains, is this acceptable? Of course the answer is, what are the alternatives and to that end, it is only, keep trying expectantly or have IVF. Surgical excision of endometriosis is a good choice for women with endometriosis-associated pain as this has been demonstrated in placebo-controlled trials.3, 4 It is also a reasonable choice for many women with infertility and no pain, since it avoids repeated hormonal exposures and may allow for more than one pregnancy at substantially lower cost. Since the majority of women with endometriosis will also have pain symptoms, the treatment by surgery may have dual effects of decreasing pain and enhancing fertility. While the Marcoux study examined only early-stage disease, there are a number of other studies that have assessed higher-grade disease and also identified very reasonable outcomes. From our group, only women with stages III–IV disease were evaluated after surgery for their disease with 72% of women wanting a pregnancy becoming pregnant and 56% of women having a live birth.5 The median time to pregnancy was 12 months (95% CI 7–17 months), with 63% achieved by expectant management post-operatively and 37% with assistance. In the most recent paper from our unit, we examined the use of the Endometriosis Fertility Index (EFI) only in stages III–IV disease,6 where there was a 63% live birth rate in an infertile population having surgery for endometriosis and 64% of these women spontaneously conceived. Perhaps the most important aspect of this study was that it tracked pregnancy to five years and showed that for women with an EFI score 9–10 that at one year, expectant management led to a live birth in 46% of women, 58% at two years and 91% at five years, with IVF only increasing this to 95% at five years. For EFI of 7–8, expectant management led to a live birth at one year of 40% (43% when IVF was added) and at five years this was 63% for expectant management and 81% when IVF was added. Finally for very poor scores (EFI = 1–2), no woman had a live birth with expectant management over the five-year follow-up and even with IVF, only 39% of women had a live birth at five years – in fact all occurred in the first two years only. This study serves a number of important functions. First, it shows that for severe disease, there is a good prediction for live birth based on the EFI. Second, it negates the long-held belief that pregnancy is only likely in the first 12 months after surgery, and in fact this study clearly shows that there is a benefit in ongoing expectant management with higher EFI scores (usually correlating to less severe tube and ovary damage). Of course with low EFI scores where there is substantial tubal and or ovarian damage, it is perhaps not surprising that IVF is superior and indeed the prediction of outcome is highly desirable for women.7 But so is the fact that in 2/3 cases, they may avoid IVF (if they want) to with surgery. Ours are certainly not the only data to reflect these outcomes. A large prospective study of over 200 women having surgery with or without bowel resection for severe endometriosis shows a cumulative live birth rate of 36%, 50% and 67% at 1, 2 and 3 years post-index surgery, with no difference whether bowel resection was undertaken or not, with half achieving pregnancy with expectant management and the remainder with assisted reproductive technology.8 Similar outcomes are reported from a French team performing open or laparoscopic bowel resection, where cumulative pregnancy rates following surgery were 60% with a median of 19 months follow-up9 although it was reported in an extended analysis from the same group that pregnancy after expectant management only occurred in the laparoscopy group.10 In the most recent series of 111 women having bowel resection for deeply invasive disease, 73% of women were pregnant at 12 months of follow-up with the majority after expectant management.11 Similar data are reported from other centres12-14 where deep disease involving bowel resection has resulted in expectant pregnancies following surgery, with substantial improvement in pain and bowel function in addition. As with any surgical procedures, these surgeries should only be undertaken by experts and the risk of complication is not insubstantial and must be a shared decision with the woman noting risks and benefits. Perhaps one of the most controversial areas in endometriosis-associated infertility is the presence of ovarian endometriomas. Women with this type of endoemtriosis have particularly severe disease and it is hypothesised (although unknown) that invasion of the ovary by disease may increase oxidative stress and lead to diminished follicular density.15 Surgery may lead to further reduction in ovarian reserves and consideration of a woman's ovarian reserve is entirely appropriate where there are endometriomas present. IVF may be warranted in this setting as a primary intervention (where fertility is the only issue), since the documented loss of ovarian cortex may further impair fertility. The problem arises when women do not want to have IVF, when they have had unsuccessful oocyte retrievals in the presence of endometriomas or have had successful retrievals, but unsuccessful implantations and ongoing pregnancy. In this situation, it is likely that surgery will only get one go at improving the situation and that first procedure must be the best. Contemporary guidelines16 as well as RCTs17, 18 demonstrate that microsurgical stripping of endometrioma walls and ablation of disease offer the optimal cytoreduction with lowest risk of recurrence. Surgery is often successful after previous failed IVF cycles and this has been previously reported, with 22/29 (75%) women with endometriosis having multiple failed IVF cycles treated surgically, with 15/22 (68%) pregnant with expectant management.19 This demonstrates that where IVF has been unsuccessful, the role of surgery should still be considered and in fact may be of substantial benefit. So how do we wrap all this up? It is clear that surgery has a number of advantages for the often symptomatic woman with endometriosis with regards to pain management and reduction in her pelvic symptoms of dyspareunia, dyschesia and non-menstrual pelvic pain is likely to improve pregnancy from expectant management as her quality of life is improved and she is more likely to have regular timed intercourse. It should also be apparent that the first attempt at fertility should be the best and the adage here is that it is better to peek and shriek than peek and poke. If you cannot finish the surgery with removal of disease, then you should not start and refer to a specialist centre which can undertake optimal disease removal with minimal risk and a multidisciplinary team where appropriate. Deeply invasive disease can and has been successfully removed with improvements in fertility as well as the benefits of symptom reduction. Many women, and their families, will want choices in how they may approach future parenthood and that is not always the one-size-fits-all approach of IVF. Certainly it is an option, but an invasive and expensive option, so at least there should be the choice of a second invasive and less expensive option in surgery.