Endometriosis is a common disease affecting young women. Its clinical manifestations include pain and infertility, and it can dramatically affect quality of life. Treatments should be tailored to address the wishes of women according to the specific characteristics of the disease. Although many questions remain unanswered, strong evidence supports the use of laparoscopic surgery to improve pain and infertility. The systematisation of strategy is essential to make surgery more reproducible, safer and less time-consuming. Nevertheless, even in the most expert hands, complications may occur. Further investigations are needed to compare the different approaches. Outcomes must, however, include pain, fertility, organ dysfunction, and quality of life.
Introduction
Endometriosis is the presence of endometrial glands and stroma outside of the endometrial cavity. Its prevalence varies according to the population studied: 5–10% of women of reproductive age, 20–30% of subfertile women, and 40–60% of women with dysmenorrhoea.
Pelvic pain and infertility are two major manifestations that can dramatically affect general, physical, mental, and social wellbeing. Surgery is the cornerstone of diagnosing and treating the disease. The development of minimally invasive techniques has transformed laparoscopy into a gold-standard procedure for diagnosing and managing endometriosis.
Strong evidence indicates that surgery reduces pain associated with endometriosis in all stages of the disease. The role of surgery in treating infertility has been validated in a substantial subset of women. Laparoscopic surgery enables the most complex procedures to be carried out safely and effectively. Nevertheless, even in the most experienced hands, severe complications may arise.
Radical excision may ensure low recurrence rates, but has the risk of secondary functional disorders. For that reason, surgeons must achieve a reasonable balance between radical treatment and conserving functionality.
In addition, the management of endometriosis represents a challenge. Although it is a benign disease, it yields severe morbidity in active young women. For that reason, treatment must be tailored to each individual according to different clinical presentations, stages, personal expectations, and risk of recurrence.
In this chapter, we present state-of-the-art surgical treatments for endometriosis. Practical issues relating to surgical strategy will be highlighted in managing the different locations of the disease.
Endometriosis and pain
Pain is the most common complaint among women with endometriosis, and is a significant factor affecting quality of life. Endometriosis-related pain symptoms can be divided into non-specific and specific (organ-related) signs. Non-specific symptoms are the most frequent, and include dysmenorrhoea, deep infiltrating endometriosis (DIE), and non-menstrual pelvic chronic pain. In cases of DIE with bowel or bladder involvement, organ-specific symptoms can be found, such as dysuria or dyschezia. Importantly, more than 50% of women with DIE present with only non-specific symptoms.
Endometriotic implants and adhesions have been clearly associated with pain. Both are included in several scoring systems to describe the severity of the disease. No direct correlation, however, exists between symptoms and stages of the disease.
Two randomised-controlled trials (RCTs) comparing laparoscopic treatment of endometriosis with diagnosis alone have shown that either excision or ablation of the implants improves pain-related symptoms in all stages.
In the first RCT, Sutton et al. randomised 63 women with minimal-to-moderate endometriosis to adhesiolysis, excision of implants, and laparoscopic uterosacral nerve ablation or diagnostic laparoscopy. Pain was significantly reduced in 63% of women in the treated group. The effect was more favourable in the most advanced stages of the disease. In fact, when stage I cases were excluded, 74% of the women achieved pain relief, compared with 23% in the control group.
In the second RCT, Abbott et al. also found significant improvement in symptoms in 80% of women receiving treatment. Interestingly, 32% of the control group also experienced pain improvement, supporting a placebo effect of surgery similar to that reported in the study by Sutton et al.
A Cochrane review of five RCTs concluded that laparoscopic surgery results in improved pain outcomes compared with diagnostic laparoscopy alone. It was not possible, however, to determine which of the treatment modalities (excision or ablation) was superior. Although the group of severe endometriosis was potentially under-represented in the Cochrane analysis, other studies have shown the same benefits of surgery in all pain scores, including dysmenorrhoea, dyspareunia and non-menstrual chronic pelvic pain. In addition, significant improvement in both quality of life and sexual function was also reported in this group of women.
The effect of the different variations of surgical treatment is difficult to evaluate. At least, three RCTs found no any significant benefit of laparoscopic uterosacral nerve ablation in the management of endometriosis-related pain.
Evidence shows that a potential role exists for laparoscopic presacral neurectomy, particularly in the reduction of midline menstrual pain. Nevertheless, long-term complications, such as constipation and urinary urgency, are described with frequencies of 14% and 5% at 12 months. Finally, hysterectomy is a possibility that must be considered in severe pain recurrence after conservative treatment, as it provides longer pain relief. Removal of the ovaries also reduced the risk of future surgeries by 2.4-fold, but this benefit is marginal in women younger than 40 years.
Endometriosis and pain
Pain is the most common complaint among women with endometriosis, and is a significant factor affecting quality of life. Endometriosis-related pain symptoms can be divided into non-specific and specific (organ-related) signs. Non-specific symptoms are the most frequent, and include dysmenorrhoea, deep infiltrating endometriosis (DIE), and non-menstrual pelvic chronic pain. In cases of DIE with bowel or bladder involvement, organ-specific symptoms can be found, such as dysuria or dyschezia. Importantly, more than 50% of women with DIE present with only non-specific symptoms.
Endometriotic implants and adhesions have been clearly associated with pain. Both are included in several scoring systems to describe the severity of the disease. No direct correlation, however, exists between symptoms and stages of the disease.
Two randomised-controlled trials (RCTs) comparing laparoscopic treatment of endometriosis with diagnosis alone have shown that either excision or ablation of the implants improves pain-related symptoms in all stages.
In the first RCT, Sutton et al. randomised 63 women with minimal-to-moderate endometriosis to adhesiolysis, excision of implants, and laparoscopic uterosacral nerve ablation or diagnostic laparoscopy. Pain was significantly reduced in 63% of women in the treated group. The effect was more favourable in the most advanced stages of the disease. In fact, when stage I cases were excluded, 74% of the women achieved pain relief, compared with 23% in the control group.
In the second RCT, Abbott et al. also found significant improvement in symptoms in 80% of women receiving treatment. Interestingly, 32% of the control group also experienced pain improvement, supporting a placebo effect of surgery similar to that reported in the study by Sutton et al.
A Cochrane review of five RCTs concluded that laparoscopic surgery results in improved pain outcomes compared with diagnostic laparoscopy alone. It was not possible, however, to determine which of the treatment modalities (excision or ablation) was superior. Although the group of severe endometriosis was potentially under-represented in the Cochrane analysis, other studies have shown the same benefits of surgery in all pain scores, including dysmenorrhoea, dyspareunia and non-menstrual chronic pelvic pain. In addition, significant improvement in both quality of life and sexual function was also reported in this group of women.
The effect of the different variations of surgical treatment is difficult to evaluate. At least, three RCTs found no any significant benefit of laparoscopic uterosacral nerve ablation in the management of endometriosis-related pain.
Evidence shows that a potential role exists for laparoscopic presacral neurectomy, particularly in the reduction of midline menstrual pain. Nevertheless, long-term complications, such as constipation and urinary urgency, are described with frequencies of 14% and 5% at 12 months. Finally, hysterectomy is a possibility that must be considered in severe pain recurrence after conservative treatment, as it provides longer pain relief. Removal of the ovaries also reduced the risk of future surgeries by 2.4-fold, but this benefit is marginal in women younger than 40 years.
Endometriosis and infertility
Minimal and mild endometriosis
The relationship between minimal and mild endometriosis and infertility is still controversial. In addition, two RCT’s analysed the role of surgery to improve fertility in this group of women, and reached conflicting conclusions. The first study is a large multicentre, prospective evaluation of 341 women that were randomised to conservative laparoscopic treatment or diagnosis only. The results showed a benefit in fecundity in the treated group (4.7 v 2.4 per 100-persons months) and cumulative probability of pregnancy (30.7% v 17.7% at 9 months).
In the second study, a similar design was carried out involving 111 women, but no difference in pregnancy rate was found at 1-year follow up. Meta-analysis of these two studies, combining ongoing pregnancy and live birth rates, showed that surgery significantly improved fecundity in this subset of women.
Moderate to severe endometriosis
In advanced stages, endometriosis can exert a significant detrimental effect on fertility caused by anatomical distortion and adhesions. In a prospective evaluation, Busacca et al. found a beneficial effect of surgery, especially in moderated disease; however, fecundity rates remained low at 2.4% per month.
When considering surgery, the surgeon must consider symptoms and should exclude all other causes of infertility. Some evidence shows that, in the presence of tubal, ovarian, or male factor infertility, assisted reproductive techniques (ART) are the treatment of choice; however, in the absence of access to ART, surgery becomes a useful alternative.
General strategy
The aim of the strategy is to systematise all the steps to make this surgery more reproducible, safe, and less time-consuming. We divided the strategy into general and specific steps out of didactical purposes. The general strategy involves basic steps and ergonomic principles that were to all surgeries.
A low-residue diet is prescribed 5 days before surgery, which guarantees easy displacement of the bowel during surgery, allowing for adequate exposure. Rectal enema is used in cases of DIE the night before surgery.
The woman is placed in a lithotomy position with both arms located alongside the body and the coccyx at the edge of the table. The legs must be semi-flexed, and their position must give the third surgeon optimal access to use vaginal and rectal instruments. In some cases, endometriosis surgery can take longer than planned, and consequently strict caution must be paid to positioning, avoiding both vascular and nerve compression. A thorough rectovaginal examination under anesthesia must be systematically carried out to evaluate the endometriotic lesion before surgery begins.
A 10-mm, 0-degree laparoscope is introduced at umbilicus level, and three 5-mm accessory trocars are placed, respectively, in the iliac fossa and in the suprapubic midline. The midline trocar is located at the level of the lateral ones, or higher, to obtain a more ergonomic set-up. A systematic inspection of the abdominal and pelvic cavity must be carried out. This first assessment is fundamental to determine the severity of the disease and adapt the surgical strategy. Findings are registered according to the American Society of Reproductive Medicine score.
Adequate exposure of the operative field is mandatory in all cases, especially when dealing with DIE. Several progressive steps are used for this purpose. First, the woman is placed in a Trendelenburg position, and the uterine manipulator is used. Adhesiolysis is then carried out to re-establish normal anatomy. During this process, the physiological attachment of the sigmoid colon is freed to allow access to the left adnexa and ureter.
If endometriotic cysts are present, their adhesions are taken down, and usually, during this procedure, they are opened and drained. When dealing with more complex cases, different organs can be suspended. This manoeuver is helpful because it improves the exposure of the surgical field and it frees the assistant’s hand to help the surgeon. Suspension of both ovaries and sigmoid colon is usually conducted in DIE surgery using straight needles or special devices.
In obese women, in some cases, the small bowel can be also suspended using curved needles. Ureteral identification is recommended starting dissection. Of note, the left ureter crosses the common iliac artery and the right ureter crosses the external iliac artery. Surgery is then pursued according to each case, and its specific strategy is described further on.
Specific strategy
Ovarian endometriosis
Ovarian endometriosis is one of the most frequent locations, identified in about 22% of women with chronic pelvic pain or infertility. It is frequent to find the disease in other locations, and endometrioma must be considered as a marker of more extensive pelvic and intestinal disease. Caution must consequently be paid not to underdiagnose or undertreat these women.
The diagnosis of endometrioma is easily established through transvaginal ultrasonography, with high sensibility and specificity rates. The preoperative work-up must exclude the possibility of DIE, intestinal or genitourinary involvement, or both. An evaluation of the ovarian reserve is recommended in this group of women by means of the Anti-Müllerian hormone, and the ultrasonographic antral follicle count.
Laparoscopic management of endometrioma is the gold standard, the main indications being a symptomatic cyst or a cyst larger than 3 cm. Several techniques have been described to treat endometrioma. In most of these techniques, the procedure consists of opening and draining the cyst followed by either excision (stripping technique), fulguration, or vaporisation of the cystic wall (ablative technique). Drainage alone is not recommended because of the high recurrence rate.
A Cochrane review found that excisional surgery provides better outcomes than ablative treatment. Removal of the endometrioma was associated with a significant increase of spontaneous pregnancy in subfertile women (odds ratio: 5.21), with a reduction in the recurrence of endometrioma and pain symptoms.
In ovarian reserve studies, an inadvertent removal of ovarian parenchyma during cystectomy has been well documented. A study by Muzii et al. showed that the stripping technique removes ovarian tissue in about 54% of cases, although the specimen did not show the morphologic characteristics seen in normal ovarian tissue. It has been proposed that cystectomy might negatively affect the success of controlled ovarian hyperstimulation. Conversely, Canis et al., showed that the number of oocytes and embryos obtained from in-vitro fertilization embryo transfer cycles did not decrease significantly after laparoscopic cystectomy was carried out.
A significantly better response to gonadotropin stimulation with the excisional technique was found in the Cochrane revision, with no increase in the likelihood of pregnancy. No difference, however, was found in response when each woman acted as her own control by comparing the treated and untreated ovary.
Significant reduction in ovarian reserve assessed by Anti-Müllerian hormone levels has been documented in women treated for endometriotic cyst, and also with DIE. In addition, Busacca et al. reported 2.4% of ovarian failure immediately after bilateral endometrioma surgery. Consequently, particular attention must be paid to infertile women who will be treated for endometrioma, in women presenting with DIE, and when facing endometrioma recurrence, particularly in bilateral cases.
In our practice, we systematically carry out an ovarian reserve evaluation by means of AMH levels and ultrasonography immediately before and 3 months after surgery, and this valuable information was integrated into treatment planning. The woman is informed of the different alternatives of management, and the treatment is tailored to her wishes.
As previously outlined during the exposure step (see section on general strategy above), the ovary is freed from the broad ligament by traction. This procedure induces cyst rupture in almost all cases; if not, the cyst is intentionally emptied. The interior surface of the cyst is then washed out and inspected by cystoscopy. If dealing with DIE, we proceed with the ovarian suspension to the abdominal wall, leaving cystectomy to the end of the surgery. The crucial part of cystectomy is the adequate identification of the cleavage plane.
The first method is to enlarge the ovarian incision obtained by puncture or rupture, and to identify the correct cleavage plane. A second and efficient way to proceed is to evert the cystic fundus, incise it, and carry out cystectomy for both halves of the cyst. Two grasping forceps are used to exert traction and counter-traction in the incision margins. The cyst’s capsule is then detached from the ovarian parenchyma by stripping. Precise haemostasis of the ovarian bed is obtained with a bipolar coagulation electrode by means of short applications on the small vessels. Blind coagulation must be avoided. Generally, the ovary is left open without any need for suture. Anti-adhesion barriers are placed at the end of the procedure to decrease the risk of postoperative adhesions.
Deep infiltrating endometriosis of Douglas’s pouch
Deep infiltrating endometriosis is a specific disorder defined by the presence of endometriotic lesions penetrating more than 5 mm under the peritoneal surface. Its exact incidence is unknown as it is not always symptomatic, but it has been estimated to affect 20–35% of women with endometriosis.
Distribution of lesions is variable, but the disease has a typical multi-focal presentation. Lesions of DIE can be further classified either as anterior, when they invade the detrusor muscle or as posterior, when they are located at the pouch of Douglas. At this level, most frequent sites affected by the disease are the uterosacral ligaments, the posterior vaginal wall, and the anterior rectosigmoid serosa.
Characteristically, DIE is associated with severe symptoms, including dysmenorrhoea, deep dyspareunia, non-menstrual chronic pelvic pain, dyschezia, catamenial diarrhoea, and infertility. The severity of the symptomatology depends on the depth, location, and multifocal nature of DIE lesions.
Diagnosis is based on the presence of symptoms and evidence either of pain, palpable nodule, or both, or induration at the level of the posterior vaginal fornix, the uterosacral ligaments, or both. Lesions located high in the rectovaginal septum or the retrocervical area, however, might be missed out by bimanual examination. Consequently, further studies with images are fundamental for diagnosis.
Diagnostic accuracy of imaging studies has been extensively covered. Transvaginal ultrasound is the most accessible method, and has a good sensitivity for ovarian endometrioma diagnosis. In cases of pelvic DIE, its real capacity has been questioned. Nevertheless, recent publications advocate that, in expert hands, it may indeed be a powerful tool in the diagnosis of posterior DIE.
Some investigators have reported a sensitivity as high as 98% in case of lesions affecting the rectosigmoid and 95% for retrocervical disease, and a specificity of 98% and 100%, respectively.
Transrectal ultrasound has also been shown to be a valuable diagnostic tool. The method, however, has some drawbacks, as it requires special equipment and is known for discomfort inherent to the rectal examination.
In a study by Bazot et al., transvaginal ultrasound and rectal endoscopic sonography for the diagnosis of bowel endometriosis produced, respectively, a sensitivity of 92.6% and 88.9%; a specificity of 100% and 92.6%; a positive predictive value of 100% and 96%; and a negative predictive value of 87% and 80.6%. Colonoscopy or barium enemas are only of value in cases of bowel stenosis or to exclude malignancy. Magnetic resonance imaging has been advocated to be a powerful tool in the diagnosis of DIE, and the reported sensitivity is 83% for rectosigmoid and 76% for retrocervical disease, with a specificity of 98% and 68%, respectively. In experienced eyes, however, the diagnostic potential of magnetic resonance imaging lies under its capacity to explore completely the abdomen.
Posterior DIE is a benign condition, with a tendency to progress in less than 10% of women who are symptom free. For that reason, surgical treatment should be discussed with the patient, and the following considered: progressive painful clinical signs; unexplained infertility, ureteral obstruction, or both, with risk of silent renal loss.
Uterosacral ligaments involvement
The uterosacral ligaments are the most frequent location of posterior DIE involvement. The lesion is isolated in 83% of the cases.
Laparoscopic excision of DIE at the uterosacral ligaments has been shown to be efficient. The anatomical landmarks for uterosacral ligament excision include ureters, uterine arteries, hypogastric nerves, and the rectosigmoid. In isolated lesions, dissection should begin with a peritoneal window, medial to the ureter, to dissect the nodule until the ascending uterine artery and torus uterinus. Special attention should be paid not to get too close to the rectum. If there is no vaginal commitment, the surgeon should avoid opening the vagina.
On the contrary, in cases of extensive adhesive disease, the dissection should not extend below the deep uterine vein, avoiding harm splanchnic autonomic nerves responsible for urinary and bowel function.
In bilateral uterosacral involvement, the surgeon should decide whether a radical or conservative approach should be taken towards the function. Bilateral excision of nodules at the uterosacral ligaments has a high risk of hypogastric nerve damage, exposing the woman to post-surgical bladder voiding problems. Although transient in most women, in some cases, it might last for weeks requiring auto-catheterisation. Nerve-sparing surgery has been advocated to limit voiding dysfunctions. Nevertheless, extensive fibrosis and secondary inflammatory reaction make nerve dissection difficult, even in expert hands. This highlights the need for good anatomical knowledge, and the feasibility of nerve-sparing procedures in severe DIE needs to be questioned. In our opinion, it may be better to avoid surgery than leave the woman with permanent voiding dysfunction.
In small isolated lesions, the indwelling catheter might be retrieved within 24 h. In the case of extensive unilateral or bilateral uterosacral excision, however, we recommend carrying out a bladder sensitivity test to check the integrity of the urge sensation reflex with filled bladder. If the reflex is normal, the catheter is retrieved; if not, the woman should undergo bladder re-education with indwelling catheter being clamped every 3 h. The test is repeated after 24 h and, if positive, the catheter can be retrieved.
Endometriosis of the rectovaginal septum
Laparoscopic resection of DIE at the rectovaginal septum has also proven to be efficient and safe. Treatment of rectovaginal DIE involves multiple techniques, including exclusive laparoscopic procedure or laparoscopically assisted vaginal surgery. The surgical technique begins with the identification of both ureters, and sometimes ureterolysis is required. Pararectal fossas are avascular spaces that open easily with carbon dioxide dissection, and should be dissected medially to the course of the hypogastric nerves. Dissection should continue lateral to the rectum, and then to the rectovaginal nodule on both sides. Special attention should be paid not to cut the middle rectal artery. Once the surgeon has reached healthy connective tissue downwards in the posterior vaginal wall, it is possible to begin the dissection of the anterior rectal wall. It is essential to divide the nodule as close as possible to the rectal wall, leaving as much endometriosis tissue attached to the vaginal wall. Finally, once the rectum has been detached, the nodule can be dissected from the posterior vaginal wall. On the contrary, if the vagina is infiltrated, it can be opened and the nodule extracted through it. The vagina can then be closed using monofilament 2/0 with interrupted intra-corporeal knots.
One non-randomised trial, looking at reproductive outcome after surgical excision, showed that expectant management ( n = 60) compared with surgical excision of rectovaginal endometriosis ( n = 44) was not significantly different to the cumulative 24-month conception rate (46.8% v 44.9%). According to the investigators, burial of DIE lesions by adhesions at the pouch of Douglas may limit the potential effect of the inflammatory reaction over the fertilisation processes. This explains why the excision is unlikely to influence reproductive outcome.
In another study involving 500 women, of the 388 women wishing to conceive, 57% became pregnant naturally after a median 3-year follow up. In 167 women who were not pregnant, 64% conceived after in-vitro fertilisation owing to male-factor infertility or failure to conceive after 1 year.
A significant improvement in 24-month, pain-free survival after surgery compared with expectant management for deep dyspareunia (72.9% v 48.2%) and dyschezia (78.1% v 57.4%) has been reported. In addition, surgery has shown a significant improvement for dysmenorrhoea (38.9% v 24.5%) but not for non-menstrual chronic pain.
Bowel endometriosis
Endometriosis affects the bowel in 3–37% of cases. In more than 90% of cases, the rectum, the rectosigmoid junction or the sigmoid colon is involved. In more than 40% of cases, the disease is multifocal.
Digestive symptoms can be less frequent than general signs in women with colorectal endometriosis. In a study by Urbach, pelvic chronic pain was present in 69% of women with colorectal endometriosis, whereas constipation, diarrhoea and pain on defecation were present in 31.7 and 7% of cases, respectively. The main symptom of colorectal endometriosis was found to be pelvic pain irradiating to the rectum and anus.
The surgical treatment must be carried out by multidisciplinary surgical teams experienced with this disease. The patient should be informed of preoperative risks and bowel or bladder dysfunction before surgery. Two approaches can be proposed: a radical approach, with the aim of a complete resection of the lesions to prevent recurrence, and a more conservative approach that could decrease the rate of functional disorders. We believe bowel resection cannot be planned before surgery, because no complementary examination can specifically predict bowel involvement. The conservative approach must be attempted in almost every case, because vascularisation and nerves are better preserved, limiting functional disorders. We feel that bowel resection for endometriosis must be exceptional and adapted to this particular aspect.
Women with endometriosis are young, often without co-morbidity or vascular disorder, presenting with a benign pathology that does not require radical resection as in oncology. Consequently, resection must be as economically efficient and cosmetic as possible. An exception must be made for lesions of the sigmoid colon or of the small bowel that cannot tolerate conservative treatment.
Indications for a radical approach are multifocal lesions, sigmoid involvement, lesions larger than 3 cm, or involving more than 50% of the circumference of the bowel wall, or bowel stenosis. In almost all the cases of rectal involvement, a conservative approach can be attempted. To date, no consensus has been reached, but the conservative approach is favoured because segmental bowel resection can cause bowel dysfunction.
Bilateral opening of the pararectal fossas and the lateralisation of the ureters are mandatory before detaching the bowel from the rectovaginal septum. To do so, the assistant pulls the rectum cephalad by means of a flat forceps, and the posterior part of the uterus and the vagina are detached progressively turning the lesion around with scissors. Bipolar coagulation must be carefully performed, and the operator should leave as much of the disease as possible on the posterior vaginal wall. The nodule is then detached from the vagina, paying particular attention not to open the vagina if the disease does not infiltrate the mucosa. Once the bowel has been detached, a rectal probe is used to check if stenosis is present, and to control the circumference of bowel involvement.
Conservative treatment
Conservative treatment uses a stepwise strategy. Occasionally, adhesiolysis is sufficient and no more surgery on the bowel is required. Shaving techniques consist of resection of superficial lesions of the serosa or of the muscularis. Mucosal skinning consists of shaving until the mucosa is reached with the need for a suture to cover the bowel wall. This technique could have the advantage of preserving vascularisation and innervation of the bowel, and preventing an opening of the bowel wall, which can induce a potential risk of postoperative fistula. When the lesion is deeper than the mucosa and not larger than 3 cm, a discoid resection can be carried out using a circular stapler (discoid resection) or a linear stapler (wedge resection).
Radical treatment
For segmental bowel resection, the mesentery is dissected close to the digestive tract to preserve vascular lymphatic vessels as well as surrounding sympathetic and parasympathetic nerves. A linear endoscopic stapler is used to resect the bowel, leaving a margin of 1–2 cm from the nodule. It is proven that an intracorporeal anastomosis provides a faster recovery of bowel function, decreases postoperative narcotic use, length of stay, and morbidity. Conventionally, a 5-cm Pfannenstiel’s incision is made to resect the rectum and place the anvil, and an end-to-end or side-to-end anastomosis is carried out intracorporeally with a 28-mm or 32-mm PCEEA circular stapler transanally.
Less invasive approaches, such as transanal and transvaginal natural orifice specimen extraction, have been developed in colorectal surgery, and could easily be adapted to treat bowel endometriosis. This would avoid the use of mini-laparotomy, which can induce incisional hernia, pain or infection. The anastomosis is carried out laparoscopically. The anvil of the circular stapler is placed transanally or transvaginally. This technique has the advantage of requiring minimal mobilisation and less traction over the meso. The integrity of the anastomosis is systematically controlled with air and betadine tests. The two doughnut rings of tissue are also checked. The absence of tension or twists, and the vascularisation of the anastomosis, are systematically checked at the end of the procedure. When the sutures of the bowel and the vagina are at the same level, an omental flap can be placed to separate the anastomosis. A protective ileostomy is subject to debate, and we feel that it must be avoided, but when the resection is ultra-low (less than 6 cm from the anal verge) it might be considered according to the technical complexity and length of the operation, or when a defective anastomosis is suspected. Postoperative care must be cautious to detect early leaks of the anastomosis.
Bowel dysfunction after segmental bowel resection is an important issue. Colorectal resection may be followed by the occurrence of de-novo digestive symptoms or worsening of previous bowel disorders. In fact, no evidence supports the risk of recurrence as a valid argument in favour of colorectal resection. The aim of surgical treatment should be relief of pain rather than compulsive resection of all endometriotic lesions, with a symptom-guided approach.
Urinary tract endometriosis
Urinary tract endometriosis is a rare condition that represents 1.2% of endometriosis cases; however, in severe pelvic endometriosis, its prevalence may rise up to 20%.
Although sometimes underestimated, this condition can cause severe morbidity owing to ureteral extrinsic compression and secondary hydronephrosis, with progressive loss of renal function. Non-specific symptoms include dysmenorrhoea, deep dyspareunia, chronic pelvic pain, or both.
The most frequent locations are bladder (84%), ureter (10%), kidney (4%), and urethra (2%). Bladder and ureteral endometriosis can co-exist; in severe cases, multifocal disease is a possibility.
Bladder endometriosis
Bladder endometriosis is defined as the presence of endometrial glands and stroma in the detrusor muscle. Symptoms include dysuria (42%), haematuria (9–15%), and recurrent urinary tract infections (18%). The severity of dysuria might have a positive correlation with the diameter of the nodule and the presence of a lesion at the base of the bladder. Physical examination does not provide much information, except in the case of a palpable bladder nodule.
Preoperative work-up includes urine analysis and imaging. Endovaginal ultrasonography is the most accurate technique to diagnose this condition. Magnetic resonance imaging has the potential to provide a complete description of the pelvis, but it may not detect small lesions. Cystoscopy is useful but only in cases of mucosal involvement in advanced stages of the disease.
Cystoscopy should be carried out at the beginning of the procedure. Ureteral double-J stents should be placed when partial cystectomy is planned, when the lesion is located at the bladder trigone or when the lesion extends to the ureteral orifices.
Multiple surgical alternatives are available according to the degree of bladder-wall involvement. Shaving could be considered when superficial endometriosis is found on the bladder peritoneum. Mucosal skinning could be carried out in the absence of mucosal involvement, and the defect can be closed using monofilament 3/0 in a single-layer interrupted suture using intracorporeal knots. On the contrary, in cases of mucosal invasion, partial cystectomy should be carried out. After completing nodule excision, the defect can be closed in a single- or double-layer interrupted suture using intracorporeal knots. A methylene blue test must always be carried out to confirm bladder integrity, and ureteral stents should be left in place for 6–8 weeks. An indwelling catheter should be left in place for 10–15 days, and should be removed after confirmation of the absence of contrast medium leak by cystography.
Significant improvements in symptoms result after laparoscopic excision of bladder disease, with low recurrence rates, compared with resection by cystoscopy, which is associated with incomplete resection and high recurrence rates.
Ureteral endometriosis
Ureteral endometriosis is found in about 10% of women with urinary tract endometriosis, and the incidence might be higher in women with rectovaginal nodules. It is classified as extrinsic (85%), namely infiltration of surrounding connective tissue and adventitia, and intrinsic (15%), namely infiltration of the muscularis mucosa or uro-epithelium.
The most frequent configuration is unilateral, at the left ureteral channel where it crosses the uterine artery. Ipsilateral uterosacral ligament involvement is common.
Specific symptoms include cyclic colic flank pain, cyclic haematuria (reported in 15% of women), and in severe cases, unilateral ureteral obstruction, but non-specific gynecologic symptoms are more frequent.
Physical examination provides few data, but ureteral involvement should be always suspected in the presence of retrocervical nodules larger than 2 cm. The ideal imaging diagnostic modality is controversial. Ultrasound may depict hydronephrosis; however, unless the ureter is dilated, it is difficult to visualise the ureteral pathway. Nowadays, magnetic resonance imaging is the best imaging modality for ureteral evaluation, and allows for a more comprehensive evaluation of the complete pelvis. Renal scintigraphy is solely used to determine the degree of residual kidney function in the side where severe ureteral obstruction is found.
In women with important ureteral obstruction and severe hydroureteronephrosis, a double-J stent should be inserted. If that is not possible, the woman should be prepared for immediate ureteral surgical decompression.
Laparoscopic ureterolysis should always be carried out. Peri-ureteral fibrosis produces retraction and distortion, mainly in the lower third of the ureter, changing its lateral course to medial. Special attention should be paid not to harm the ureteral adventitia to avoid ureteral devascularisation. In most cases, ureterolysis is the only treatment required and, if severe devascularisation is observed, it is recommended that a double-J stent is inserted to decrease the risk of ureteral fistula.
If ureteral stenosis has been observed in the preoperative work-up, a double-J stent should be placed at the beginning of the surgery. When relieving extrinsic compression, stenosis might regress, but in case of severe persistent stenosis, or if intrinsic ureteral endometriosis is found after ureterolysis, double-J stent insertion and partial excision with end-to-end anastomosis must be carried out. The double-J stent should be left in place for 6–8 weeks.
When the lesion is located at the ureterovesical junction, ureteral reimplantation can be carried out. In cases where the ureter’s length is insufficient, a psoas hitch suspension can be carried out to ensure a tension-free anastomosis. A double-J stent and an indwelling catheter should be left in place 7–10 days. Before retrieving the bladder catheter, it is recommended that a cystography is carried out to test the integrity of the anastomosis.
In our study of 91 women with ureteral endometriosis operated on by laparoscopy, ureterolysis was the only surgical treatment required in 85.7% of women; in 10% of women, ureteral resection and end-to-end anastomosis were carried out. These results have been confirmed in another study of 30 women in whom ureterolysis was the only treatment in 73.3% of cases. Ureteral resection with end-to-end anastomosis was required in 16.7% and ureter reimplantation in 10% of cases.
Surgical outcomes after 3 years of follow up showed significant improvement in either painful urinary symptoms (i.e. suprapubic pelvic pain and dysuria) or non-specific symptoms (i.e. dysmenorrhoea, deep dyspareunia, and pelvic pain). This evidence complies with our experience in which a significant improvement in gynecologic symptoms was observed after a 24-month follow up.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
