When attending endoscopic conferences you could be forgiven for thinking that little has changed in the management of endometriosis: ‘When you see it, you treat it – surgically.’

While that advice may be reasonable enough when dealing with severe disease causing debilitating pain, things are a little more complicated for endometriosis-related infertility.

The link between infertility and endometriosis

First of all, the link between infertility and endometriosis has not always been accepted as naturally evident. Nobody questions that connection in severe cases where the tubo-ovarian anatomy is critically altered through adhesion formation and fibrosis. But even in the recent past, some considered the presence of minimal and mild endometriosis inconsequential. The prevalence of endometriosis is increased in subfertile women (estimates vary between 30 to 50 per cent), but it was argued that this is mostly an epiphenomenon. Endometriosis happens to be more common in women with infertility, but one doesn’t cause the other.

However, those views are slowly changing. There is increasing evidence for a causal link. To begin with, genomic and proteomic studies have clearly shown that the uterine epithelium in women with endometriosis is functionally different from that in women without endometriosis.^1,2,3

From observational studies in women with minimal or mild endometriosis we know the monthly fecundity rate (MFR) is reduced following intrauterine insemination (IuI), both with autologous or donor sperm.⁴ A meta-analysis⁵ showed that pregnancy outcomes in IVF also worsen with the severity of the disease.

Perhaps the most convincing evidence supporting a causative link comes from a study in baboons confirming that the MFR in baboons with experimentally induced mild to severe endometriosis is reduced, compared with those with no or minimal endometriosis.⁶

Pain and infertility

The question then obviously arises how to treat the disease to improve a patient’s fertility. Ideally, such decisions should be informed by high quality evidence. Despite the difficulties undertaking such trials, it is pleasing to see there is a steadily growing body of such evidence, guiding us with the surgical treatment of endometriosis for both pelvic pain and infertility.

Patients with severe pain deserve the best surgery they have access to. Randomised controlled trials (RCTs) have clearly shown that endoscopic surgery for endometriosis is effective in reducing pain levels⁷, even though the recurrence rates are higher than we would like, estimated at 20 per cent after two years and 40 to 50 per cent after five years.⁸ In patients with co-existing infertility, such surgery obviously needs to be conservative.

Although surgery is justifiable for pain reduction, it is unclear to what extent the surgery also restores normal fertility. No RCTs have been done in infertile women with moderate to severe disease. The only evidence which is available is retrospective and thus potentially biased in favour of surgery, but it suggests that the cumulative pregnancy rate (CPR) in this patient group is increased by ten per cent to 25 per cent following surgery.⁹ Until better studies are performed, the case for surgery in patients with severe disease but no significant pain (they do exist!) is not overwhelmingly convincing. It is for the surgeon to carefully assess whether laparoscopic procedures in such patients will result in a positive benefit/risk ratio.

Treating women with minimal or mild disease

RCTs don’t always provide the full answer to our clinical questions. An example may illustrate the point. Does laparoscopic treatment of mild or minimal endometriosis improve a patient’s fertility? A meta-analysis is available to answer that question.10 It summarises the results of two RCTs, one showing no effect and the other larger one showing a favourable effect. When live birth rate and ongoing pregnancy after 20 weeks were combined, the odds ratio (OR) was 1.64 (95% confidence interval [CI] 1.05 to 2.57) in favour of laparoscopic surgery compared to a diagnostic laparoscopy.

Albeit statistically significant, this effect is relatively small. You need to dig a little deeper to find out that the number needed to treat is 12, which means that one extra live birth is achieved for every 12 women undergoing surgery for minimal or mild endometriosis. And it gets worse: if we assume a 25 per cent prevalence rate of minimal or mild endometriosis, almost 50 laparoscopies need to be performed to find those 12 women. Whether these statistics make it a worthwhile intervention depends on other factors. There may be additional reasons to perform the laparoscopy, such as the need for a tubal patency check or the co-existence of significant pain symptoms. The financial cost to the patient and the community needs to be taken into consideration too, as well as the risk of the intervention.

The search for a non-invasive diagnostic test

Part of the problem with the ongoing debate is that it is flawed with circular reasoning: if a woman suspected of having endometriosis doesn’t have a laparoscopy, we won’t know whether she needs one. This explains the feverish search for a non-invasive diagnostic test pursued by many teams, including some in Australia. Our own approach has been to look for a characteristic proteomic ‘fingerprint’ in the endometrium³, others are investigating siRNAs¹¹ and nerve fibers¹². Although it is unlikely that any such test will be perfect, it may improve patient selection and ultimately reduce the number of unnecessary laparoscopies. Conversely, Somigliana and colleagues¹³ have raised concerns that the number of unnecessary interventions might actually increase should the test be used as a screening tool. In their highly recommended paper, the authors warn of the real risk of ‘disease mongering’, a process that turns healthy people into patients, adds iatrogenic harm and wastes precious resources.

Used as a screening tool, the test would ‘diagnose’ endometriosis in a large number of asymptomatic women. It is uncertain to what extent the presence of non-symptomatic endometriosis presents a significant health concern. It is also important to draw attention here to the observation from two randomised placebo-controlled studies that endometrial deposits spontaneously disappear in up to a third of patients.^14,15

To laparoscope or not to laparoscope?

As yet, such a non-invasive diagnostic test does not exist and the question therefore remains: to laparoscope or not to laparoscope? That choice will eventually be influenced by the potential down-stream need for in vitro fertilisation (IVF) treatment. On the one hand, Barnhart’s meta-analysis of 22 observational studies⁵ demonstrated that the chance of achieving pregnancy with IVF is lower in endometriosis patients (OR 0.56; 95% CI, 0.44–0.70). Furthermore, pregnancy rates for women with severe disease are much lower than for women with mild endometriosis (OR 0.60, 95% CI 0.42–0.87).

So, at first glance this provides a strong argument in favour of surgery even when IVF is contemplated. unfortunately, we desperately need evidence to show that ablation/excision of endometriosis prior to IVF actually restores pregnancy rates to those of women without endometriosis. Without such evidence, it can be argued that a laparoscopy only delays and drains money away from IVF treatment without a proven benefit. This would be particularly true for example in older women and couples where there is a serious male factor problem. On the other hand, a laparoscopy is justified when pain symptoms are present. Another indication is when a young, subfertile woman presents with a normal transvaginal ultrasound and her partner’s semen analysis is normal. If minimal or mild endometriosis is diagnosed and adequately treated and both tubes are patent, level one evidence suggests her chances of conceiving naturally will be increased by 60 per cent.¹⁰

What about endometriomata?

This leaves the controversy around the treatment of endometriomata. The arguments in favour of early surgical treatment have been listed as the risk of spontaneous rupture, the impact on normal ovarian function, the high probability of other, often deeply infiltrating disease, the risk of abscess formation when inadvertently punctured during IVF, and the potential risk of malignant transformation. The latter is of particular concern, but the risk of ovarian cancer should be put in perspective. There is one extra ovarian cancer for every 10,000 women with endometriosis and this woman is likely to be close to menopause, with a large endometrioma in excess of 9 cm diameter.¹⁶ Vercellini puts it like this: ‘…in the worst scenario, the lifetime probability of developing ovarian cancer increases from 1/100 to 2/100. In other words, a woman with untreated endometriosis has a 98 per cent probability, instead of 99 per cent, of not developing an ovarian malignancy’.¹⁷

Excision of the cyst wall has proven to be superior to ablation of the endometriosis on the cyst wall. RCTs in women with an endometrioma larger than 3 cm have shown that laparoscopic excision was more successful in reducing all types of endometriosis-related pain.18 The recurrence rate was also lower (OR 0.41; 95% CI 0.18–0.93) and, more importantly, the subsequent spontaneous pregnancy rate was five-fold higher (OR 5.21; 95% CI 2.04–13.29). What remains unclear though is whether excision is actually better than no intervention at all. Indeed, a meta-analysis of five non-randomised studies concluded that surgical management of endometriomas has no significant effect on ovarian response to stimulation and IVF pregnancy rates compared with no surgery.¹⁹

Ovarian surgery also has an inherent risk. Benaglia and colleagues followed 93 women undergoing IVF following excision of a unilateral endometrioma.²⁰ Complete absence of follicular growth was observed in 12 operated ovaries while this event never occurred in the contralateral gonad (P<0.001). The frequency of severe ovarian damage following surgery was thus 13 per cent.

Conclusion

It is great to see that new, high quality evidence is emerging. Some questions can now be answered, but hopefully this short review has also planted some fertile seeds of doubt. Don’t get me wrong, I like my surgery, but it is vital to always re-examine why we are operating.

‘A good surgeon knows how to operate. A better surgeon knows when to operate. The best surgeon knows when not to operate.’