Discussion
We report a series of patients with deep endometriosis involving the sacral plexus or sciatic nerve, a complex disease which in a majority of cases involves other major pelvic organs before reaching pelvic nerves. The resulting diverse clinical presentations, due to multiple deep endometriosis localisations, required complex surgical procedures carried out by multidisciplinary surgical teams, with these procedures related to a significant risk of postoperative complications. A year after surgery, despite an overall improvement in pain and quality of life and a high pregnancy rate by natural conception in half of the cases, some patients experienced persistent bladder dysfunction and neurological symptoms directly related to the surgical manipulation of nerves. Due to the relative rarity of sacral and sciatic nerve endometriosis, we believe that our data may provide additional helpful information to existing knowledge available in the literature.
Our series has several weaknesses. Firstly it is a retrospective series of patients receiving surgical management, without a control group. In our practice we previously treated two patients for whom pelvic MRI revealed deep endometriosis infiltrating pelvic wall and entrapping the sacral plexus, and two others with deep endometriosis of the lumbosacral space involving the sciatic nerve. All four patients received medical treatment based on continuous contraceptive pill intake resulting in prolonged amenorrhea, and consider their residual symptoms and quality of life as satisfactory. These patients are however not included in the CIRENDO database as they have not had surgery in our department. Furthermore their endometriosis localisations have not intraoperatively been confirmed and the patient number is too small to allow any comparison with the cohort of women who benefited from surgery. Due to the lack of control group, we cannot definitively state that deep endometriosis removal leads to a significant improvement in pain and quality of life when compared to continuous medical treatment, though all 52 women underwent surgery due to the lack of efficacy of previous medical therapies.
The second limit is related to the heterogeneity of multiple deep endometriosis localisations recorded in our series and difficulties in demonstrating a precise relationship between specific localisations and corresponding symptoms. Accordingly baseline severe constipation may be attributed to mid/low rectum infiltration by deep endometriosis, or to splanchnic nerves or sacral root S2-S4 involvement; while postoperative constipation may result from postoperative low rectal resection syndrome (LARS)12 or rectal denervation following extensive removal of parametrium together with splanchnic nerves6. Consequently, assessment of the true efficiency of nerve surgical management on pelvic organ function may be challenging.
Our series presents however several strengths. This is a case-series including consecutive patients with strict inclusion criteria and prospective recording of preoperative, intraoperative and postoperative data, employing standardised questionnaires to assess pelvic pain, digestive and urinary function and quality of life. There is no loss to follow up. Inclusion criteria limited enrolment to only women for whom deep endometriosis nodules were in close contact with the sacral plexus or sciatic nerve, requiring specific nerve dissection or excision. Our results cannot therefore be overestimated by inclusion of women with neurological complaints associated with a less severe disease. All procedures were fully recorded and are available for review, and 5 procedures were broadcasted live during surgical meetings.
Our study provides detailed data on intraoperative findings and postoperative outcomes in patients managed for deep endometriosis involving large pelvic nerves. Due to the low prevalence of these localisations, clinical study available in the literature tends to rely on case-series, despite the afore-mentioned design weaknesses. The pooling of information reported by several surgical teams allows to further knowledge in this field and may be helpful for colleagues interested in the management of such patients.
Deep endometriosis infiltrating large pelvic nerves has been described as far back as 195513, but it was 50 years later that the first description of a laparoscopic approach to release the sacral plexus and sciatic nerve with deep endometriosis involvement was reported5. Successive case-series confirmed overall pain improvement in patients having benefited from surgery. The first large case series including 27 women with isolated endometriosis of the sciatic nerve and 148 women with deeply infiltrating parametric endometriosis with sacral plexus infiltration, was published by Possover et al2, and recorded a decrease in mean VAS pain score from 7.7 to 2.6. Later, the Negrar Hospital team (Verona, Italy) published several technical reports, focusing on the anatomy of pelvic nerves and the nerve-sparing technique3,14. Lemos et al reported on their experience in the management of 13 women with deep endometriosis responsible for the entrapment of lumbosacral plexus, and noted an overall improvement in 6 patients, while 6 other women were completely pain-free1. More recently, Possover et al reported a series of 46 women managed for intraneural endometriosis of the sciatic nerve which required resection of more than 30% of the nerve. Follow up continued over 5 years and revealed significant pain reduction and recovery of nerve function4.
Deep endometriosis nodules involving sacral roots differ from those of the sciatic nerve. Firstly, there is a higher prevalence of sacral roots involvement representing 94.2% of cases in our series and 85% (148 out of 175 cases) for Possover et al2. Secondly sacral roots entrapment appears to be much more frequent than actual nerve infiltration, due to the posterior development of large rectovaginal or parametrial nodules. Conversely, intraneural endometriosis was more frequently reported in sciatic nerves and may require partial resection of the nerve4.
Although pain reduction may occur early after surgery, various sensory or motor complaints may continue over months or years. When preoperatively affected, motor function recovery is long, and normal gait may take up to 5 years to restore4. Nerve entrapment or infiltration by fibrous endometriosis nodules require meticulous nerve dissection, excision of epineurium, fibre manipulation, as well as repeated haemostasis procedures responsible for thermal diffusion into the nerve, oedema and limited ischemia. These circumstances unavoidably lead to neuropraxia15followed by postoperative sensory, motor or vegetative temporary disorders. Neuropraxia refers to mild nerve injury involving impairment of both motor and sensory functions and is the first type of peripheral nerve injury according to the Seddon and Sunderland classification16,17 with no loss of axon but a temporary loss of myelin sheath. This demyelination leads to impairment of impulse conduction across the nerve segment15without Wallerian degeneration. The prognosis is favourable with complete recovery likely within weeks or months, due to axon remyelination18. Transitory bladder dysfunction requiring self-catheterisation for 4 to 6 weeks is a common clinical outcome of neuropraxia involving sacral or splanchnic nerves, and provides an explanation for bladder function recovery in three quarters of our patients with routine self-catheterisation during the early postoperative period. Administration of neuroleptic agents and postoperative physiotherapy play a key role in improving nerve recovery4.
However, neuropraxia is unlikely to be the most severe complication recorded in patients managed for deep endometriosis of the sacral plexus, particularly when the disease also involves low rectum, vagina or ureters. Concomitant rectal excision may be followed by specific complications such as rectovaginal fistula or low anterior rectal resection syndrome (LARS)12. Although our rectovaginal fistula rate may be considered high, it remains comparable to that reported in other series of patients undergoing low rectal resection for deep endometriosis associated with colpectomy19. The high rate of vaginal involvement requiring colpectomy also led to the relatively high prevalence of early postoperative pelvic abscess and frequent use of preventive stoma.
Nerve-sparing surgery, commonly applied in oncology for preservation of autonomic nerves, has been shown to improve quality of life by reducing the incidence of functional complications without compromising long-term survival20. The approach has to be tailored to suit surgical management of deep infiltrating endometriosis, which differs from cancer14,21. In cases where surgical indication exists for the treatment of early cancer, this approach works in disease-free tissue where nerves are not involved by the disease and may thus be easily separated and preserved. Conversely, preservation of the inferior hypogastric plexus is not possible and has to be sacrificed in patients who present with large infiltrations of the parametrium down to the levator ani muscles and sacral plexus. As performing small nerve sparing in large nodules of the parametrium is not feasible, bladder or rectal functional impairment may only be avoided by preserving the contralateral inferior hypogastric plexus, hypogastric and splanchnic nerves. In cases where endometriosis involves both parametria, surgeons tend to favour perfroming complete resection on the side with the more aggressive disease, and lesser treatment on the contralateral side, thereby preventing complete denervation of the bladder, rectum and vagina1,2,21. While unilateral sacrifice of small pelvic nerves may be considered, the conservation of large sacral and sciatic nerves remains a priority so as to avoid major sensory and motor somatic troubles.
It should be noted that women with deep endometriosis involving the sacral plexus and sciatic nerve in our series rarely had ovarian and fallopian tubes localisations, their disease being mainly sub-peritoneal. This resulted in an excellent pregnancy and delivery rate as early as 1 year after the procedure, with natural conception in half of the cases. This point is of particular importance, as a fear of fertility impairment following a complex surgical management may lead to unnecessarily privileging first line IVF in symptomatic patients with such severe disease and pregnancy intention.
All authors recommend referring patients with deep endometriosis of the sacral plexus and sciatic nerve to expert centres with high volume activity4,14,22. This recommendation highlights parameters which are important for ensuring the best outcomes: knowledge of anatomy and laparoscopic approaches to the sacral plexus, the availability of multidisciplinary teams for treatment of associated localisations in the digestive and urinary tract, experience in surgical procedures that ensure organ function preservation for each localisation, an ability to perform surgery in a reasonable operative time, and an ability to identify the early signs and symptoms of potential postoperative complications and to accurately manage them without delay. We believe this recommendation to be reasonable, despite the present lack of a large multicenter series or randomized trial to support it.
Study funding/competing interest(s): The North-West Inter Regional Female Cohort for Patients with Endometriosis (CIRENDO) is financed by the G4 Group (The University Hospitals of Rouen, Lille, Amiens and Caen) and ROUENDOMETRIOSE Association. No financial support was received for this study. The authors declare no competing interests related to this study.