SEARCH

SEARCH BY CITATION

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Objective  To determine the long term response, quality of life and levels of pain following the radical excision of rectovaginal endometriosis.

Design  A cohort study.

Setting  A tertiary referral centre for the management of advanced endometriosis.

Sample  All patients who had undergone radical resection.

Methods  Case note review and patient questionnaire.

Main outcome measures  Surgical complications. Overall improvement. Dysmenorrhoea, dyspareunia, dyschezia and chronic pain were measured using a visual analogue scale. Quality of life was measured using the EQ-5D questionnaire.

Results  Twelve radical resections were performed by laparotomy, 48 by laparoscopy. Ten patients had a hysterectomy. Forty-eight patients underwent shaving of the pre-rectal fascia, two had a disc resection of the rectum, 10 had an anterior rectal resection. Two patients required a colostomy and two needed subsequent dilation of a stenosed anastomosis. Forty-four of the first 46 patients replied. The median follow up period was 12 months (range 2 to 22 months) and 86% (38/44) reported an improvement or whom 27 (61%) had a good response (pain completely gone or greatly improved). Patients having a hysterectomy or a disc or segmental resection of the rectum reported a good response and had a normal quality of life. Quality of life scores in the study group overall were lower than the background population.

Conclusions  Radical resection is an effective treatment for rectovaginal endometriosis. Hysterectomy and rectal resection were associated with a better response and quality of life.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Rectovaginal endometriosis causing obliteration of the cul-de-sac accounts for 5–10% of cases of endometriosis.1,2 Patients suffering this condition have severe pain and a dramatically impaired quality of life.3 Although medical treatments are effective for the management of peritoneal endometriosis,4 they seem to be less effective at treating advanced or rectovaginal disease, even when used in conjunction with surgery.5–7 There are significant differences both in the histological appearance and the expression of oestrogen receptors in rectovaginal endometriosis that may account for this.8,9

Because of the inadequacy of conservative measures, many authors advocate the complete dissection and surgical removal of rectovaginal disease, even if surgery involving the rectum is required.1,2,10,11 Our practice since September 2000 has been to offer patients the choice of full surgical excision of endometriosis in addition to medical options for hormonal manipulation and pain control. As aggressive surgery to rectovaginal disease is still not commonly practised in the UK, it is essential that close monitoring of the complications and the success of these procedures takes place. The purpose of this study is to assess the response that patients have to this surgery, to look at their current pain scores and quality of life and to report any peri-operative or longer term complications. We also wished to explore the response rate in relation to the extent of bowel or uterine surgery required, and whether patients having their surgery more than 12 months earlier had a similar response to those operated on more recently.

METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Patients referred with a history of chronic pelvic pain lasting more than six months were seen in the pelvic pain clinic. All patients were offered a laparoscopy. Patients with mild to moderate endometriosis were treated at the time using bipolar diathermy ablation or peritoneal excision where appropriate. Patients with severe disease, defined as disease causing partial or complete obliteration of the cul de sac, or disease involving any part of the bowel, underwent a staging laparoscopy and rectal examination. They were then extensively counselled about the risks of advanced surgery and the alternatives, and if they chose to proceed, were rebooked for a radical resection. A radical resection is defined as the complete excision of pelvic endometriosis that involves either the occluded cul de sac, the rectovaginal septum or the rectum or para-rectal fascia. Patients who have completed their families with severe menstrual dysfunction were offered a hysterectomy at the time. If the ovaries were obviously involved in the endometriosis, patients were advised to consider a bilateral oophorectomy with their hysterectomy.

The patients were placed in a low lithotomy position. Ureteric stents were inserted cystoscopically. A laparotomy was considered for patients choosing a hysterectomy where there was dense occlusion of the cul de sac, or where extensive bowel resection was anticipated, otherwise a laparoscopic approach was preferred. After insufflating carbon dioxide to a pressure of 18 mmHg, a 10 mm port was inserted through the umbilicus through which the laparoscope was passed. A 5 mm port was then inserted into the right and left lower quadrants, lateral to the obliterated umbilical vessels. A 10 mm port was inserted suprapubically to the right of the midline. A rectal probe was used to identify the margins of the rectum. For dissection, Metzenbaum scissors with a single patient use tip were used, and for haemostasis we used a combined suction, irrigation and bipolar diathermy probe (Surgiflex WAVE, ACMI, USA). Para-rectal dissection was performed using an Ultracision harmonic scalpel (Ethicon, USA) if a rectal resection was required. The surgical procedure used is similar to that described by Redwine.2 After dissecting down into the rectovaginal space, a decision was made about the extent of rectal surgery necessary. If endometriotic nodules in the pre-rectal fascia were mobile, the fascia was shaved off the front of the rectum. If nodules were fixed and appeared to involve the rectal muscle layer, then discrete nodules were removed by resecting a disc of rectal wall and larger areas were removed by anterior rectal resection.

Patients who underwent a radical resection were identified from a surgical database. The case notes were reviewed for post-operative complications, duration of stay and blood transfusion. Patients were contacted in the summer of 2002 and were sent a questionnaire by post. The first part of the questionnaire assessed pain and the response to surgery. A visual analogue scale was used to assess dysmenorrhoea, dyspareunia, dyschezia and chronic daily pain. Patients were asked if their pain had completely gone, was greatly improved, a little better, no better or worse. The second part of the survey assessed quality of life using the EQ-5D questionnaire and results were compared with a representative sample in the UK.12 This is an established tool for assessing quality of life after similar surgery.3,13 Data comparing pain and quality of life scores were analysed using an independent two-tailed t test.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Between September 2000 and March 2003, 60 radical resections were performed: 12 by laparotomy, 48 by laparoscopic techniques. Ten patients had a hysterectomy, 9 of these were total abdominal hysterectomies and 1 was a laparoscopically assisted vaginal hysterectomy. Six patients had a bilateral oophorectomy at the time of their hysterectomy. Forty-eight patients underwent shaving of the pre-rectal fascia, 2 had a disc resection and 10 had an anterior rectal resection. Details of the surgery are described in Table 1. Two patients required conversion to a laparotomy after an initial laparoscopic assessment suggested that more extensive rectal surgery would be required. Three patients had a hole in the vagina repaired intra-operatively. One patient required an intra-operative transfusion for an estimated blood loss of 1000 mL. Four patients required a two-unit blood transfusion post-operatively. Two patients required a temporary colostomy; one was performed electively after a very low rectal reanastomosis. The other was performed on a patient who underwent shaving of the pre-rectal fascia. After an initially good recovery, she became pyrexial after 48 hours and a laparoscopy revealed a small rectal perforation at the site of the surgery, which was oversewn laparoscopically. Two patients had short term ureteric stents inserted after extensive dissection of endometriosis from the ureters. These were removed under local anaesthetic using a flexible cystoscopy after six weeks. Two patients have subsequently required dilation of the rectal re-anastomosis site after difficulty passing a normally formed stool. Four patients have subsequently had a hysterectomy for continuing dysmenorrhoea, and four have had a laparoscopy for diathermy to minor recurrences of endometriosis.

Table 1.  Operative details.
 Mean (SD)Range
  • *

    The revised American Fertility Society Score.

Age (mean)31 (6.5)20–49
rAFS score*44 (34)10–162
Duration of surgery (minutes)146 (92)36–420
Estimated blood loss (mL)400 (319)100–1000
Length of stay (days)4.6 (2.6)1–10

Out of 58 cases where tissue was sent for histology, 55 demonstrated endometriosis and 3 showed scarring and fibrosis in the pre-rectal fascia. Of the 10 cases of anterior rectal resection, one showed endometriosis through to the submucosa, eight showed endometriosis in the muscularis propria and one showed endometriosis in the pre-rectal fascia. Of the 10 hysterectomy specimens, five had adenomyosis within the cervix or uterus and one showed benign leiomyomata.

The first 46 patients to undergo radical resection have been sent questionnaires. Non-responders were contacted by telephone. Two patients were untraceable. The median follow up period was 12 months (range 2 to 24 months). The response in terms of pain is reported in Table 2. Sixteen out of 23 (70%) women who had their surgery less than 12 months ago reported a good response compared with 13 out the 21 women (62%) who had their surgery more than 12 months ago. Quality of life comparisons with the UK normal values are reported in Table 3. Visual analogue pain scores are reported in Table 4.

Table 2.  Response to radical resection. Total numbers (%).
 OverallHysterectomySegmental or disc resection
Pain gone7 (16)4 (50)5 (63)
Greatly improved20 (45)3 (38)3 (38)
A little better11 (25)1 (13)0
No change5 (11)00
Pain worse1 (2.2)00
Total4488
Table 3.  EQ-5D quality of life scores. Data are mean, t values from independent two-tailed t test (P).
 Reference group*Study groupHysterectomySegmental or disc resection
  • *

    From Kind et al.12

Mean Self-rated Health Status8266, 4.3 (0.05)84, 0.29 (NS)89, −3.09 (0.05)
Mean Weighted Health State Index0.850.65, 3.99 (0.05)0.88, 0.60 (NS)0.92, 1.5 (NS)
Total in group19254488
Table 4.  Visual analogue pain scores by the time elapsed since resection. Mean scores out of 100 (SD).
 OverallUnder 12 monthsOver 12 monthst value, df (P)
Dysmenorrhoea60 (28)53 (30)68 (29)1.4, 34 (NS)
Dyspareunia26 (29)18 (29)34 (29)1.9, 42 (NS)
Dyschezia32 (33)26 (29)40 (36)1.4, 42 (NS)
Chronic pain29 (27)27 (25)31 (30)0.45, 42 (NS)

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Although not a randomised controlled trial, our series shows a generally favourable response to surgical excision of deep cul de sac and rectovaginal endometriosis and adds to the growing body of evidence to support excisional surgery for deeply infiltrating disease. The best reported response rates are those of Chapron's cohort of 29 women who reported 91% to 100% improvements in dysmenorrhoea, dyspareunia and chronic pelvic pain.14 Garry et al.3 and Wright and Shafik13 have found significant improvements in both pain and quality of life after similar surgery. This major surgery, however, does not seem to protect against a recurrence and our re-operation rate so far is 13%. After 35 months follow up, Varol et al.15 found a re-operation rate of 36% and recurrence of endometriosis in 15%.

Pre-operative MRI was not performed as scans could not be obtained within a reasonable time frame: In only one case was there mucosal involvement of the rectum by the endometriosis suggesting that pre-operative sigmoidoscopy and barium enema may be of limited value in assessment.

Other series reporting excision of colorectal endometriosis suggest favourable results. Jerby et al.11 reported 30 patients with colorectal disease, 7 of whom required rectal resection, with 28 showing improvement or great improvement after 10 months. Bailey et al.1 reported 130 patients with colorectal disease of whom 118 had disease in the rectum or cul de sac. One hundred and nine of these patients had an anterior rectal resection, with 93% reporting significant or complete relief of symptoms.

There is still continued discussion about whether a rectal resection is required for these patients.15 Our practice is to assess the rectum after dissection of the rectovaginal space and perform a resection if the muscular layer of the rectum appears to be involved. This is compatible with other reported techniques2 and failure to do this may theoretically reduce the efficacy of the surgery. It is interesting to note, although numbers are too small for meaningful conclusions, that in our cohort, patients having a rectal resection had a better response and quality of life than the overall group. In a study of patients undergoing rectovaginal dissection resection however, Hollet-Caines16 reported that 54 out of 61 had an improvement in pain. There is no doubt that rectal resection carries risks, principally breakdown of the anastomosis, abscess, rectal dysfunction and stricture at the anastomosis site,17–19 but it is feasible that a rectal resection may result in a more complete excision of disease. Clearly, a randomised trial would be necessary to demonstrate a difference between this and the more conservative rectal shaving.

Women having a hysterectomy at the same time as radical resection, also appear to have a very good response rate, although again our numbers are small, and this should be interpreted cautiously. Urbach et al.10 reported a cohort of 29 patients undergoing bowel resection. The 10 women having hysterectomy at the same time had a considerably higher cure rate. Possible explanations include differences in the ages, severity of disease and intention to conceive, as well as the effect of bilateral oophorectomy on the recurrence rates of endometriosis. We found adenomyosis in half of the hysterectomy specimens. It is proposed that the aetiology and structure of rectovaginal endometriosis is very similar to that of adenomyosis,8,9 and the presence of adenomyosis may therefore account for continuing pain and dysmenorrhoea in the patients where the uterus is conserved.

Radical resection of cul de sac and rectovaginal endometriosis is an option for women with advanced disease, but the decision to proceed to this surgery should be balanced against the risks. Further randomised, controlled trials to assess whether rectal resection or hysterectomy has a significant impact on the success of this surgery should be considered.

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. METHODS
  5. RESULTS
  6. DISCUSSION
  7. References

Accepted 7 January 2004