Objective
To evaluate urinary symptoms before and after colorectal resection for endometriosis using validated questionnaires.
Study Design
We randomly assigned 52 patients with colorectal endometriosis to undergo laparoscopically assisted or open colorectal resection. The median follow-up was 19 months. Urinary symptoms were evaluated using the International Prostate Score Symptom and the Bristol Female Low Urinary Tract Symptoms questionnaires.
Results
Dysuria was observed in 29% of cases postoperatively. Using Bristol Female Low Urinary Tract Symptoms and International Prostate Score Symptom scores, an alteration was observed for voiding symptoms ( P = .01 and P = .006, respectively). No difference was observed between the laparoscopy and the open surgery group. An alteration of the International Prostate Score Symptom voiding symptoms was observed in the group that did not undergo nerve sparing surgery ( P = .048). An alteration of the International Prostate Score Symptom voiding symptoms was observed for patients who underwent vaginal resection ( P = .01) and parametrial resection ( P = .02).
Conclusion
Our findings confirm that colorectal resection for endometriosis is a source of urinary dysfunction whatever the surgical route.
Urinary dysfunction is of major concern after surgery for deep infiltrating endometriosis (DIE), especially when the colorectum is involved. Although previous studies have demonstrated that colorectal resection for endometriosis was associated with a significant improvement in gynecologic and digestive symptoms and quality of life, a high incidence of urinary side effects affecting quality of life has been reported to reach up to 17%.
Urinary dysfunction is linked to injury of the hypogastric plexus that contains the pelvic parasympathetic fibers responsible for the voiding function of the bladder detrusor. Previous retrospective studies have underlined that the risk of urinary dysfunction depends on extension of surgery. Dubernard et al reported incidences of urinary dysfunction after resection of the uterosacral ligaments, rectovaginal septum, and colorectum in 19.1%, 28.6% and 38.5% of cases, respectively. Nerve sparing surgery has been advocated to decrease the incidence of urinary side effects. Previous studies have confirmed that the use of this technique decreases bladder dysfunction after uterosacral ligament resection.
Few data from validated questionnaires are available about urinary dysfunction after colorectal resection for endometriosis. In 2005, Daraï et al reported a high incidence of urinary dysfunction after colorectal resection by laparoscopy. More recently, Dousset et al, by means of a qualitative questionnaire, have shown that 16% of patients had a transient peripheral neurogenic bladder in a series of patients undergoing open surgery for rectal endometriosis. However, there is a lack of prospective data to assess the exact incidence of urinary dysfunction after colorectal resection, especially taking into account the route and the use of nerve sparing surgery.
In this study, we report the results on the secondary endpoint of the first prospective randomized trial comparing laparoscopy with open surgery for colorectal resection for endometriosis, focusing on pre- and postoperative urinary symptoms evaluated by validated questionnaires.
Materials and Methods
Patients
This prospective randomized study was conducted between January 2006 and December 2008. Inclusion criteria were as follows: patients more than 18 years of age with suspected colorectal endometriosis based on symptoms, clinical examination and imaging techniques including transvaginal sonography, rectal endoscopic sonography, and magnetic resonance imaging (MRI) using published criteria. Exclusion criteria included patients with prior colorectal surgery for benign or malignant disease and those with a contraindication to laparoscopy because of severe medical illness. The study was approved by the institutional review board (CPP Ile de France I; IRB: 0511200) and was conducted according to the Institutional and Ethical rules ( NCT00939861 ). All patients provided written informed consent.
Of the 76 patients invited to participate in the study, 24 declined and underwent classic open surgery giving a study population of 52 patients. None of the remaining patients revoked consent or were lost to follow-up. Patients were randomly assigned at the department of gynecology by a minimization algorithm to undergo either laparoscopically assisted or open colorectal resection: 26 underwent laparoscopy and 26 open surgery ( Figure 1 ). Two patients in the laparoscopic group required conversion to open surgery: one for extensive adhesions and the other for perioperative hemorrhage. One patient in each group did not undergo colorectal resection because of superficial bowel infiltration. All results have been analyzed in intention to treat.
The primary endpoint of this study was improvement in dyschesia. Our results supported that laparoscopy is a safe option for women requiring colorectal resection for endometriosis because of the immediate benefits on pain and postoperative complications. Moreover, it offers a similar improvement in symptoms and in quality of life to open surgery. Secondary endpoints included improvements in other digestive and gynecologic symptoms, quality of life, assessment of urinary dysfunction using validated questionnaires, morbidity associated with surgery, and fertility outcomes.
Surgical procedure
All laparoscopically assisted and open colorectal resections were performed according to protocol guidelines with the same extent of resection, including ovarian cystectomy or salpingo-oophorectomy, uterosacral ligament resection, hysterectomy, ureterolysis, ureteral reimplantation, and multiple bowel resections when required for both groups. Conservative surgery was defined by removal of the colorectum with all endometriotic lesions without hysterectomy, and radical surgery was defined by removal of the colorectum with all endometriotic lesions, including hysterectomy. The endometriosis score of the American Society for Reproductive Medicine (ASRM score) was calculated for all patients. The nerve sparing surgery technique was introduced in our institution in the middle of the study. No randomization based on the nerve sparing technique criterion was done, hence, 26 patients (50%) underwent this technique: 13 in each group.
Follow-up
Patient complications were assessed at hospital discharge and at the 1-month postoperative visit. Patients were evaluated for symptoms using previously validated qualitative and quantitative questionnaires. Before and after surgery, the women completed a semiquantitative and quantitative (10-point analogue rating scale; 0 = absent, 10 = unbearable) interview-based questionnaire to rate the severity of the various symptoms believed to be associated with colorectal endometriosis. They completed the same questionnaire after the operation and were also asked about changes in bowel and urinary function at this time. To evaluate lower urinary symptoms, patients were asked to complete 2 validated symptom questionnaires: the Bristol Female Lower Urinary Tract Symptoms (BFLUTS) and the International Prostate Score Symptom (IPSS). The BFLUTS questionnaire is in 3 parts: 19 symptom questions, (most with a corresponding subquestion); 4 sexual function questions; and 11 quality-of-life questions. The IPSS consists of 7 questions in 3 parts: voiding symptoms (4 questions scored from 0 to 5 points), storage symptoms (3 questions scored from 0 to 5 points), and 1 quality-of-life question with a subquestion about the severity of symptoms. Each patient filled out BFLUTS and IPSS questionnaires preoperatively and at the first postoperative visit then each year. Results of the preoperative questionnaires were compared with those of the last postoperative visit.
Statistical analysis
The primary endpoint was improvement in dyschesia that occurs in at least 63% of patients with colorectal endometriosis. A 2-point improvement in dyschesia after surgery was taken to be relevant considering that the standard deviation in the population is 1.55. With a risk α = 2.5% and a risk β = 10%, at least 26 patients in each arm are required for this noninferiority trial. No power calculation was done on the secondary endpoints.
Univariate comparisons of surgical and postoperative data were conducted with the use of a 2-sample t test or Wilcoxon test for continuous data and a χ 2 test or Fisher’s exact test for categorical data. All reported P values were 2-sided, a P value of less than .05 denotes a significant difference.
Results
Baseline demographic and clinical characteristics of the population
In the whole population, median age and Body Mass Index (BMI) were 32 years (range, 25–44) and 22 (range, 16.3–34.5), respectively. Twenty-three patients (44%) had a prior history of infertility. Among these patients, median duration of infertility was 36 months (range, 12–168). Thirty-five patients (67%) had undergone prior surgery for endometriosis. The median number of DIE lesions at preoperative MRI was 5.5 (range, 2–9). Anatomic distribution of DIE lesions was: vagina in 15 cases (29%), uterosacral ligaments in 44 cases (84.5%), bladder in 3 cases (5.8%), parametrium in 22 cases (42.3%), and multiple intestinal locations in 10 cases (19.2%). No differences in these criteria were found between the groups.
For the 24 patients who declined the inclusion, median age and BMI were 32 years (range, 25–50) and 21.8 (range, 18–30), respectively. Ten patients (41.6%) had a prior history of infertility and 15 patients (62.5%) had undergone prior surgery for endometriosis. No difference was found with the study population.
Concurrent surgical procedures including ovarian cystectomy, salpingo-oophorectomy, hysterectomy, and resections of uterosacral ligament ( P = .23), vaginal ( P = .85), bladder ( P = .26), parametrial ( P > .99), and multiple bowel lesions ( P = .9) were similar in both groups. Total ASRM scores (91.5 in the laparoscopy group vs 94.4 in open surgery group) were similar in both groups ( P = .74).
Urinary symptoms before colorectal resection
In the whole population, 25 patients (48%) exhibited at least 1 urinary symptom preoperatively. Dysuria was observed in 23% of cases, urgency in 25% of cases, printing full bladder in 34% of cases, spontaneous urinary incontinence in 8% of cases, and stress urinary incontinence in 14% of cases.
For the whole population, preoperative IPSS and BFLUTS mean total scores were 8.6 (range, 0–30) and 25.6 (range, 18–56), respectively. According to the preoperative BFLUTS questionnaire on symptoms, hesitancy was observed in 41% of cases, strain to start in 36% of cases, stopping flow in 29% of cases, acute retention in 2% of cases, and incomplete emptying in 61% of cases. According to the preoperative IPSS questionnaire symptoms, stopping flow was observed in 38% of cases, decreased size of the flow in 23% of cases, and strain to start in 27% of cases.
Urinary symptoms after colorectal resection
The median follow-up was 19 months (range, 6–36). Fifteen patients (29%) required self-catheterization postoperatively (8 in the laparoscopy group vs 7 in the laparotomy group), with no difference between the groups ( P > .99). The mean duration of self-catheterization was 3 months (range, 1–12). Two patients did not totally recover satisfactory voiding function and still require self-catheterization.
Postoperatively, in the whole population, dysuria was observed in 29% of cases, urgency in 11% of cases, printing full bladder in 11% of cases, spontaneous urinary incontinence in 3% of cases, and stress urinary incontinence in 14% of cases. When comparing pre- and postoperative scores for urinary symptoms for each patient, an alteration was found for dysuria ( P = .004), urgency ( P = .04), printing full bladder (corresponding to the sensation by the patient of complete repletion) ( P = .04), although no change was found for spontaneous urinary incontinence and stress urinary incontinence.
The postoperative BFLUTS mean total score was 28.3 (range, 18–69). Hesitancy was observed in 41% of cases, strain to start in 36% of cases, stopping flow in 29% of cases, acute retention in 2% of cases, and incomplete emptying in 61% of cases. When comparing the pre- and postoperative BFLUTS scores, an alteration was observed for voiding symptoms ( P = .01). All the items of the symptom questionnaire were significantly altered (hesitancy [ P = .02], strain to start [ P = .01], stopping flow [ P = .01], and acute retention [ P = .02]), except incomplete emptying ( P = .8). No difference in storage symptoms and quality of life was observed.
The postoperative IPSS mean total score was 9.5 (range, 0–38). Stopping flow was observed in 50% of cases, decreased size of the flow in 41% of cases, and strain to start in 27% of cases. When comparing the pre- and postoperative IPSS scores, an alteration was observed for voiding symptoms ( P = .006), whereas storage symptoms and quality of life were not altered ( Table 1 ).
