An anatomic and functional assessment of the discrete defect rectocele repair Geoffrey W. Cundiff, MD, Alison C. Weidner, MD, Anthony G. Visco, MD, W. Allen Addison, MD, and Richard C. Bump, MD Durham, North Carolina OBJECTIVE: The aim of this study was to describe the anatomic and functional results of the discrete fascial defect rectocele repair. STUDY DESIGN: Sixty-nine women underwent rectocele repair at Duke University Medical Center during a 3-year period beginning January 1, 1994. Repair was limited to reapproximation of discrete defects in the rectovaginal fascia, without levator plication or perineorrhaphy. Outcome measures included Pelvic Organ Prolapse Quantitation measurements, prolapse stage, and a symptom questionnaire. Univariate and nonparametric tests were used as appropriate. RESULTS: Before the operation 46% patients (32/69) reported constipation, 39% (27/69) reported splinting, 32% (22/69) reported tenesmus, and 13% (9/69) reported fecal incontinence. The median preoperative posterior Pelvic Organ Prolapse Quantitation stage was 2 (1-4). Pelvic Organ Prolapse Quantitation stage had improved for all but 2 women at 6 weeks. Eighteen percent (8/43) had recurrent rectoceles at 12 months. Mean values for the points describing the posterior vaginal wall improved >2 cm (P < .0001). Although perineorrhaphy was not performed, the genital hiatus decreased by 2.3 cm (P < .0001), with no significant change in the length of the perineal body. Functional results mirrored anatomic results, with statistically significant improvements for all symptoms. CONCLUSIONS: The discrete defect rectocele repair provides anatomic correction of rectoceles with alleviation of associated symptoms for most women. (Am J Obstet Gynecol 1998;179:1451-7.)
Key words: Perineal descent, perineorrhaphy, posterior colporrhaphy, posterior repair, rectocele repair
Posterior colporrhaphy was born of the early 19th century surgeonsâ&#x20AC;&#x2122; efforts to cure complete tears of the perineum.1 Simple closure of the gaping perineum evolved into elytrorrhaphy, a surgical approach devised specifically for prolapse, in which denudation and subsequent closure of the posterior vaginal wall significantly narrowed the vaginal caliber.2 Simon,3 who coined the term posterior colporrhaphy in 1867, advocated a more aggressive plication in the inferior portion of the vagina than in the superior portion, creating a rigid inferior shelf and a superior pocket in which the cervix and uterus rested. In 1870 Hegar applied the same concept to the perineum in introducing colpoperineorrhaphy, with the classic triangular denudation of the perineal body that created a tight introital band within the vaginal introitus.1 During the late 19th century the supports of the geni-
From the Division of Gynecologic Specialties, Department of Obstetrics and Gynecology, Duke University Medical Center. Presented at the Twenty-fourth Scientific Meeting of the Society of Gynecologic Surgeons, Lake Buena Vista, Florida, March 2-4, 1998. Reprint requests: Geoffrey W. Cundiff, MD, Box 3192, Duke University Medical Center, Durham, NC 27710. Copyright Š 1998 by Mosby, Inc. 0002-9378/98 $5.00 + 0 6/6/93745
tal organs were largely a mystery, and there was little distinction between prolapses of the rectum, bladder, and uterus. Colpoperineorrhaphy was used to treat not only rectoceles but all forms of pelvic organ prolapse. Gradually surgeons developed anatomic concepts to suit their empirical surgical methods, asserting that the main support of the uterus was the vagina, which in turn was supported by the insertion of the levator ani muscles into the perineal body.1 This concept was the basis for the incorporation of plication of the levator ani muscles into colpoperineorrhaphy, which was believed to strengthen the normal uterine support. Thus the surgical goals of colpoperineorrhaphy were constriction of the vaginal tube, creation of a perineal shelf, and partial closure of the genital hiatus. Despite improved understanding of pelvic anatomy, many of the concepts from which traditional posterior colporrhaphy evolved have survived into the 20th century and are still incorporated into the surgical repair of rectoceles today. These anatomic misconceptions may help to explain the disappointing results of traditional posterior colporrhaphy in reliably alleviating symptoms associated with rectoceles. Although the elimination of the vaginal bulge after traditional posterior colporrhaphy ranges from 76% to 96%, the relief of associated 1451
1452 Cundiff et al
Fig 1. Potential locations for rents in the rectovaginal fascia, as seen through posterior colporrhaphy incision with cervix retracted anteriorly. (Adapted from Richardson AC. The rectovaginal septum revisited: its relationship to rectocele and its importance in rectocele repair. Clin Obstet Gynecol 1993;36:980.)
defecatory and sexual dysfunction symptoms seems to be considerably lower.4, 5 In 1961 Francis and Jeffcoate6 noted an association between traditional posterior colporrhaphy and sexual dysfunction, reporting apareunia or dyspareunia in 50% of a series of 243 women. Haase and Skibsted7 reported dyspareunia in 21% of women after posterior colporrhaphy. Other authors have actually reported an increase in dyspareunia after traditional posterior colporrhaphy. For example, Kahn and Stanton4 noted an increase in sexual dysfunction, from 18% before the operation to 27% after the operation. Worsening of symptoms was also noted with respect to defecatory function, with symptoms of constipation, incomplete emptying, and fecal incontinence all increased after the operation. A third of women continued to resort to splinting to complete defecation after the operation. In the colorectal surgery literature rectocele repair evolved as a transanal surgical procedure. Redding8 recognized the contribution of rectoceles to anorectal symptoms and recommended combining rectocele repair with other anorectal operations. Marks9 noted persistent anorectal symptoms after the vaginal repair of rectoceles as early as 1966 and recommended a concomitant transanal resection of redundant rectal mucosa. This combined approach was fraught with recurrence and rectovaginal fistulas. Sullivan et al10 favored a purely
December 1998 Am J Obstet Gynecol
transanal approach that included plication of the rectal muscularis and attachment of the plicated muscularis to the levator ani complex. Although the transanal approach is inadequate for high rectoceles and for rectoceles associated with enterocele, several authors have reported cure or improvement of constipation in as many as 85% of women.10-12 This is in contrast to a recent series in which nearly half of the patients required a second operation to reestablish the normal support of the perineum to relieve constipation.13 Moreover, in a large retrospective comparison of the transanal approach with traditional posterior colporrhaphy, Arnold et al14 found no significant difference between the 2 groups with respect to constipation, fecal incontinence, or dyspareunia but did note a higher incidence of rectal pain after the vaginal repair. Kahn et al15 reported similar findings in a prospective comparison between posterior colporrhaphy with levator plication and the transanal rectocele repair. In the last decade several authors have recommended a modified rectocele repair after careful study of normal pelvic anatomy. Although Denonvilliersâ&#x20AC;&#x2122; fascia, or rectovaginal fascia, was described in the 19th century anatomy literature, its clinical significance was not recognized until the 20th century.16 In 1948 Uhlenhuth et al17 attributed significant pelvic support to the rectovaginal fascia, yet the very existence of this fascial layer was challenged by histologic studies. Milley and Nichols18 contradicted these early studies by finding the rectovaginal fascia in 100% of surgical dissections and 90% of cadaveric dissections. They not only described detachment of the rectovaginal fascia as a cause of rectocele but also recognized the nonanatomic result from levator plication. They suggested that the rectovaginal fascia could be closed as a separate layer from the vaginal mucosa but still advocated transverse plication as the method of closure. Richardsonâ&#x20AC;&#x2122;s cadaveric dissections19 further supported the role of the rectovaginal fascia in the formation of rectoceles, which they attributed to discrete breaks in the rectovaginal fascia. They described 5 different locations of breaks in the rectovaginal fascia that could occur alone or in combination (Fig 1). Richardson19 advocated a rectocele repair limited to reapproximation of fascial breaks. The aim of this report is to describe the anatomic and functional results of the discrete defect rectocele repair. Methods The study population comprised all 69 women who underwent discrete defect rectocele repair at Duke University Medical Center during a 3-year period beginning January 1, 1994. Some women were treated with a traditional posterior colporrhaphy early in the series, and these patients were excluded. Women who underwent a rectocele repair in combination with an abdominal sacral colpoperineopexy were also excluded.
Cundiff et al 1453
Volume 179, Number 6, Part 1 Am J Obstet Gynecol
Table I. Surgical approach and concurrent operations Surgical approach
Operations
Posterior colporrhaphy alone Vaginal reconstruction Total vaginal hysterectomy Bilateral salpingo-oophorectomy Anterior colporrhaphy/paravaginal repair Culdoplasty Cuff suspension Suburethral sling Sphincteroplasty Abdominal/vaginal reconstruction Total abdominal hysterectomy Bilateral salpingo-oophorectomy Retropubic urethropexy Anterior colporrhaphy/paravaginal repair Culdoplasty Sacral colpopexy Sphincteroplasty
Surgical technique. Preoperative bowel sterilization was not routinely used. Patients were placed in a dorsal lithotomy position under general endotracheal anesthesia. The surgical approach to the discrete rectocele repair was similar to that described by Richardson.19 The vaginal epithelium was opened transversely at the posterior fourchette, longitudinally incised in the midline to a level above the bulge of the rectocele, and carefully dissected from the underlying rectovaginal fascia out to its lateral attachment at the arcus tendineus levator ani. Sharp dissection was used to stay just beneath the vaginal epithelium and was especially important in dissecting the hernia sac from the vaginal epithelium because there frequently was scarring in this portion of the dissection. Because there is an avascular plane between the vaginal epithelium and the rectovaginal fascia, bleeding generally occurred from the underlying rectal muscularis and therefore provided clues to the location of the break in the rectovaginal fascia. Irrigation, optimal hemostasis, and a rectally placed finger directed anteriorly also helped to define defects in the rectovaginal fascia and to locate the fascial margins. Once the rectovaginal fascial defects were located, they were reapproximated in a simple interrupted fashion with 0 or 2-0 permanent sutures. The levator ani muscles were not plicated. The vaginal epithelium was closed with interrupted simple sutures of 2-0 braided delayed absorbable suture material. Trimming of the vaginal epithelium was performed only to provide fresh edges for reapproximation, not to narrow vaginal caliber. Perineorrhaphy was not uniformly performed but was included only to reconstruct the perineal body in those women who had separation of the superficial transverse perineal muscles. The discrete defect rectocele repair was combined with other abdominal and vaginal reconstructive operations in 64 patients (93%, Table I).
No. 5 35 14 12 25 22 15 4 1 29 8 18 26 7 22 7 1
Follow-up and analysis. Patient follow-up included an initial postoperative evaluation at 6 weeks and long-term follow-up at between 1 and 3 years. Pelvic organ prolapse was quantified according to the International Continence Societyâ&#x20AC;&#x2122;s Pelvic Organ Prolapse Quantitation (POP-Q) system.20 In addition a questionnaire addressing symptoms and overall satisfaction was administered to 60 patients (87%) either in person or by telephone between 1 and 3 years after the operation. Outcome measures included the POP-Q measurements of the posterior vaginal wall and perineum and the prolapse stage and also the symptoms related to rectocele. Statistical analysis included descriptive statistics, the Wilcoxon signed rank analysis for POP-Q results, the Fisher exact test for correlation of patient satisfaction with anatomic and symptomatic cure, and the McNemar test for the comparison of symptom proportions. Results The patients were largely postmenopausal, with a mean age of 60 years (SD 11.8 years, range 30-83 years), although only 18% of the total population were both postmenopausal and not receiving estrogen replacement therapy. The subjects were largely parous, with a median parity of 3 (range 0-9). Because this is a referral service, there was a high prevalence of previous gynecologic operations, including hysterectomy 67% (46/69), previous reconstructive operations 44% (30/69), and previous anorectal surgery 6% (4/69). The prevalence of previous rectocele repair was 17% (12/69), including 1 woman who underwent 2 rectocele repairs as part of this series. All patients had symptomatic pelvic organ prolapse before the operation. Symptoms attributable to a rectocele included those related to protrusion, sexual dysfunction (36%, 25/69), and defecatory dysfunction (71%, 49/69). These symptoms are presented in more detail in Table II.
1454 Cundiff et al
December 1998 Am J Obstet Gynecol
Table II. Incidences of rectocele symptoms before and after operation Before operation (n = 68) Symptom
No.
Protrusion symptoms Pelvic pressure Protruding tissue Sexual dysfunction Apareunia Dyspareunia Defecatory symptoms Constipation Tenesmus Splinting Fecal incontinence
25 16 9 49 32 22 27 9
Long term (n = 61)
%
No.
%
37 24 29 71 46 32 39 13
17 11 9 22 16 6 24 8 9 15 5
28 16 15 36 24 19 39 13 15 25 8
Long-term follow-up data are based on postoperative questionnaire, median 24 months after the operation.
Table III. Location of rectovaginal fascial defects according to preoperative examination and intraoperative findings Preoperative (n = 61)
Intraoperative (n = 68)
Fascial defect
No.
%
No.
%
Superior Midline Right lateral Left lateral Inferior Combination
4 9 0 0 46 2
7 15 0 0 75 3
3 4 1 1 30 29
4 6 1.5 1.5 44 43
The median prolapse stage was 2 (range 1-4). Focusing on defects of the posterior wall alone revealed a median posterior wall prolapse stage of 2 (range 1-3). This was associated with perineal descent, according to a subjective assessment, in 18% of women. In addition to the POP-Q, the initial examination included vaginal and rectovaginal examinations to define the suspected location of the rectovaginal fascial defect, and these results are presented in Table III. Of the 69 women who underwent discrete defect rectocele repair, 94% (65/69) returned for their initial postoperative visit at a median of 6 weeks (range 4-192 weeks). Three percent of patients (2/69) had a rectocele at this initial examination. Seventy-three percent of women (50/69) had POP-Q parameters assessed at this initial visit. Of these women, 94% (47/50) had improvements in the posterior wall prolapse stage, with a median improvement of 2 (range 0-3, P < .0001). Seventy-one percent (49 of 69) of patients returned for
Table IV. POP-Q: Measurements of posterior vaginal wall before and after operation Posterior wall measurement
Before operation
Initial follow-up
–3.8 0.2 –0.4 4.8 3.5
–6.8 –2.7 –2.7 2.5 3.2
Point C Point Bp Point Ap Genital hiatus Perineal body
Long-term Statistical follow-up significance –7.0 –2.4 –2.4 2.9 3.3
P < .0001 P = .3
P values are calculated for change in parameters between preoperative and long-term follow-up values.
Table V. POP-Q: Stage according to posterior vaginal wall before and after operation Before operation (n = 68)
Initial follow-up (n = 50)
Long-term follow-up (n = 43)
Posterior wall stage
No.
%
No.
%
No.
%
Stage 0 Stage I Stage II Stage III Stage IV
0 14 42 12 0
0 20 62 18 0
42 6 2 0 0
84 12 4 0 0
27 10 5 1 0
63 23 12 2 0
a long-term evaluation at a median of 12 months (range 3-48 months). POP-Q measurements were assessed in 43 of these patients. Six percent (2/43) had worsening of the posterior wall prolapse stage, but overall there was improvement. The median improvement was 1 (range 03, P < .0001). The improvements in the posterior wall measurements are listed in Table IV. The mean decrease in genital hiatus was 1.9 cm (SD 1.6 cm, range 3.5-5 cm, P < .001), with no increase in the perineal body measurement. Recurrent rectoceles, defined as no change in or worsening of the preoperative posterior wall prolapse stage, were noted in 18% of these patients (8/43). The mean posterior wall prolapse stage of the patients with recurrent rectoceles was 2 (range 1-3) (Table V). No patients had persistent perineal descent. According to the questionnaire completed by 87% of subjects (60/69) at a median of 24 months (range 5-47 months) after the operation, there was a statistically significant improvement in rectocele symptoms (Table II). The prevalence of apareunia before the operation was 23% (16/69), and this did not change at long-term follow-up. Dyspareunia in those women who were sexually active improved from 29% (9/31) before the operation to 19% (6/31) at the long-term follow-up. Among the 6
Volume 179, Number 6, Part 1 Am J Obstet Gynecol
women who reported dyspareunia at long-term followup, this was a new postoperative symptom in 1 case, and that patient’s dyspareunia resolved with subsequent excision of a uterosacral suspension stitch. Bowel symptoms were also improved at long-term follow-up. The percentage of women with persistent constipation at long-term follow-up was 16%. Three subjects (4%) had new-onset constipation. Tenesmus improved in 59% of women with this symptom and was a new finding in 2 women (3%). Splinting was eliminated in 63% of women who had reported this symptom before the operation and was a new symptom in 5 (7%) patients. Fecal incontinence was also improved in 56% of women and developed after the operation in 2 subjects (3%). Finally, patients were asked to rate their overall satisfaction with the outcome of their operations. This was reported on a 10-point scale on which a score of 10 represented complete satisfaction and 1 represented no satisfaction. The mean score was 8.6 (SD 2.2, range 1-10). To correlate satisfaction with symptomatic improvement and anatomic correction, this parameter was converted to a dichotomous variable, with a score of ≥8 defined as satisfaction. Satisfaction did not correlate with anatomic cure (P = .19), but it did correlate with alleviation of defecatory symptoms (P = .04). Comment A rectocele is a herniation of the anterior rectal wall into the vaginal cavity and at times through the vaginal introitus. This anatomic distortion results from loss of the integrity of the intervening rectovaginal fascia. Complete separation from the perineal body results in concurrent perineal descent. In addition to the protruding vaginal mass, rectoceles can result in incessant pelvic pressure and can compromise normal defecatory and sexual function. Traditional surgical approaches to treating rectoceles have focused on eliminating the bulge of the vagina. When these procedures are performed through the transvaginal or the transanal approach, the success rates for elimination of vaginal bulge have ranged from 76% to 96%.4, 5 Unfortunately, these repairs have been less successful in relieving symptoms. The increase in sexual dysfunction after traditional posterior colporrhaphy presumably relates to failure to reestablish the normal vaginal anatomy.6, 7 Plication of the levator muscles creates a firm shelf between the rectum and vagina, which has been implicated as a cause of postoperative dyspareunia.18 Combining the levator plication with a perineorrhaphy that decreases the genital hiatus at the expense of creating an abnormally long perineal body adds to coital discomfort. Similarly, although this surgical technique creates a flat vaginal wall and narrow genital hiatus, it often hides the ballooning of the an-
Cundiff et al 1455
terior rectal wall rather than correcting it, resulting in a perineal rectocele. A repair that focuses on eliminating the vaginal bulge without normalizing rectal defects or addressing associated symptoms seems to be treating the gynecologist’s observations, rather than the patient’s concerns. In this series the discrete defect rectocele repair provided an anatomic cure of rectocele in 82% of patients. More important, it alleviated constipation in 84% of women with that symptom and dyspareunia in 66% of women with that symptom. The fact that symptom relief correlated with patient satisfaction whereas anatomic cure did not underlines the priority of symptom relief for the patient. The explanation for the ability of the discrete defect rectocele repair to relieve rectocele symptoms lies in the fact that it reestablishes the normal integrity of the rectovaginal fascia. We believe that this both alleviates the anatomic defect and facilitates the return of normal vaginal and rectal function. The discrete defect rectocele repair also contributes to the correction of perineal descent. This is reflected by the decrease in genital hiatus measurements without any increase in the perineal body measurements. Reestablishing the continuity of the rectovaginal fascia decreases the width of the genital hiatus by resuspending the perineum, provided that the support of the superior portion of the rectovaginal fascia is intact or reconstructed. Perineal descent has been shown to be an anatomic finding in a variety of anorectal disorders in addition to rectoceles and enteroceles, including constipation, fecal incontinence, rectal pain, and solitary rectal ulcer syndrome.21-24 Successful surgical repairs for rectoceles and enteroceles should correct associated perineal descent to avoid progressive pudendal neuropraxis neuropathy and myopathy of the pelvic floor and anal sphincter, which can result in fecal incontinence. We previously showed that abdominal repairs that resuspend the perineum dramatically improve defecatory dysfunction.25 The discrete defect rectocele repair appears to be a vaginal technique that can help to achieve support of the perineum when combined with adequate apical support. Definitive proof of this will require prospective preoperative and postoperative measurements of perineal descent with either defecography or a perineometer. Although this is a retrospective series with a relatively short follow-up, it is encouraging and suggests that the discrete defect rectocele repair may avoid the dysfunction associated with the traditional posterior colporrhaphy or the transanal rectocele repair. This improvement in symptom relief may be due to a more anatomic correction of the defect responsible for rectocele and to reestablishment of normal support of the perineal body. Long-term prospective studies are needed to confirm these suppositions.
1456 Cundiff et al
REFERENCES
1. Jeffcoate TN. Posterior colpoperineorrhaphy. Am J Obstet Gynecol 1959;77:490-502. 2. Fricke JC. Abtheil des allgemeinen KrankenHauses. Ann Chir 1833;2:142-5. 3. Simon G. Prag Viertelgahrsch 1867;3:112-4. 4. Kahn MA, Stanton SL. Posterior colporrhaphy: its effects on bowel and sexual function. Br J Obstet Gynaecol 1997;104: 882-6. 5. Mellgren A, Anzén B. Nilsson BY, Johansson C, Dolk A, Gillgren P, et al. Results of rectocele repair: a prospective study. Dis Colon Rectum 1995;38:7-13. 6. Francis WJ, Jeffcoate TN. Dyspareunia following vaginal operations. J Obstet Gynaecol Br Emp 1961;68:1-10. 7. Haase P, Skibsted L. Influence of operations for stress incontinence and/or genital descensus on sexual life. Acta Obstet Gynecol Scand 1988;67:659-61. 8. Redding MD. The relaxed perineum and anorectal disease. Dis Colon Rectum 1965;8:279-81. 9. Marks MM. The rectal side of the rectocele. Dis Colon Rectum 1967;10:387-8. 10. Sullivan ES, Leaverton GH, Hardwick CE. Transrectal perineal repair: an adjunct to improved function after anorectal surgery. Dis Colon Rectum 1968;11:196-14. 11. Sehapayak S. Transrectal repair of rectocele: an extended armamentarium of colorectal surgeons: a report of 355 cases. Dis Colon Rectum 1985;28:422-33. 12. Khubchandani AT, Clancy JP, Rosen L, Riether RD, Stasik JT. Endorectal repair of rectocele revisited. Br J Surg 1997;84:89-91. 13. Silvis R. Long term results of anterior rectal wall repair in symptomatic anterior rectocele and anterior rectal wall prolapse: a prospective study [thesis]. Utrecht (The Netherlands): Elinkwijk; 1997. p. 75-89. 14. Arnold MW, Stewart WR, Aguitar PS. Rectocele repair: four years’ experience. Dis Colon Rectum 1990;33:684-7. 15. Kahn MA, Stanton SL, Kumar DA. Randomized prospective trial of posterior colporrhaphy vs. transanal repair of rectocele: preliminary findings. In: Proceedings of the eighteenth annual meeting of the American Urogynecologic Society; 1997 September; New Orleans, Louisiana. New Orleans: The Society; 1997. 16. van Ophoven A, Roth S. The anatomy and embryological origins of the fascia of Denonvilliers: a medico-historical debate. J Urol 1997;157:3-9. 17. Ulenhuth E, Wolfe WM, Smith EM, Middleton EB. The rectovaginal septum. Surg Gynecol Obstet 1948;86:148-63. 18. Milley PS, Nichols DH. A correlative investigation of the human rectovaginal septum. Anat Rec 1969;163:443-52. 19. Richardson AC. The rectovaginal septum revisited: its relationship to rectocele and its importance in rectocele repair. Clin Obstet Gynecol 1993;36:976-83. 20. Bump RC, Bo K, Brubaker L, DeLancey J, Klarskov P, Shull B, et al. The standardization of terminology of female pelvic organ prolapse and pelvic floor dysfunction. Am J Obstet Gynecol 1996;175:10-7. 21. Henry MM, Parks AG, Swash M. The pelvic floor musculature in the descending perineum syndrome. Br J Surg 1982;69: 470-2. 22. Bartolo DC, Read NW, Jarett JA, Read MG, Donnelly TC. Johnson AG. Differences in anal sphincter function and clinical presentation in patients with pelvic floor descent. Gastroenterology 1982;85:68-75. 23. Snooks SJ, Nicholls RJ, Henry MM, Swash M. Electrophysiological and manometric assessment of the pelvic floor in solitary rectal ulcer syndrome. Br J Surg 1985;2:131-3. 24. Hudson CN. Female genital prolapse and pelvic floor deficiency. Int J Colorect Dis 1988;3:181-5. 25. Cundiff GW, Harris RL, Coates KW, Low VH, Bump RC, Addison WA. Abdominal sacral colpoperineopexy: a new approach for correction of posterior compartment defects and perineal descent associated with vaginal vault prolapse. Am J Obstet Gynecol 1997;177:345-55.
December 1998 Am J Obstet Gynecol
Discussion DR LESTER BALLARD, Cleveland, Ohio. Here is how I understand rectovaginal anatomy. From the vagina to the rectum, there are several layers: the vaginal epithelium, the vaginal muscularis, the rectovaginal fascia (Denonvilliers’ fascia), which is firmly adherent to the posterior surface of the vaginal muscularis, the rectovaginal adventitia (also known as the endopelvic fascia), the rectal muscularis (including the internal anal sphincter), and the rectal epithelium. Dr Cundiff’s use of the term rectovaginal fascia is not defined in terms of layers, but I think that he is referring to the vaginal muscularis and adherent Denonvilliers’ fascia. Dr Cundiff describes scarring encountered during dissection of the “hernia sac” from the vaginal epithelium. Usually we think of hernia sacs as involving the peritoneum with fascial defects, but there is no peritoneum in the posterior vagina. He describes an avascular plane between the vaginal epithelium and the rectovaginal fascia, but there are blood vessels throughout the layers of the vagina and rectum. If dissection is carried out in such a way as to create bleeding from the rectal muscularis, then the dissection is actually under Denonvilliers’ fascia. I think that most dissection for posterior colporrhaphy separates the vaginal epithelium from the vaginal muscularis and Denonvilliers’ fascia. The locations of defects found before and during operation were shown, but the technique used to distinguish between the different types of defects—both during the office examination and during the operation—was not described. Vaginal incisions and dissection may create the impression of defects, especially inferiorly and midline, because that is how the vagina is opened. In this study most defects were judged as inferior, 75% before the operation and 44% during the operation; 39% were a combination, although which defects were combined is not stated. In Fig 1, adapted from Richardson,1 the inferior defect looks like separation of the rectovaginal fascia (the vaginal muscularis and Denonvilliers’ fascia) from the perineal body. I would think that the clinical correlate of that separation would be perineal descent, not rectocele, yet perineal descent was present before the operation in only 18% of women. The technique used to assess perineal descent was not defined. There was no attempt to correlate patients’ symptoms with the clinical examination either before or after the operation. If the examinations and symptom assessments were done at different times in some or all cases, then the ability to draw conclusions from these data is impaired. The prevalence of postoperative symptoms is high. Symptoms of defecation were grouped together, and a statistically significant difference was reported between the percentages of women with preoperative symptoms (72%) and of women with postoperative symptoms (39%). However, no confidence interval or P values were provided to support this statement. The incidences of symptoms of sexual dysfunction were essentially the same
Cundiff et al 1457
Volume 179, Number 6, Part 1 Am J Obstet Gynecol
before and after the operation (37% and 36%), and this result is not addressed at all. The introduction describes at great length the symptoms associated with â&#x20AC;&#x153;traditionalâ&#x20AC;? posterior colporrhaphy, but the results of the proposed method show virtually the same numbers for some symptoms. For example, dyspareunia occurred in 19% of women after the operation in their series and in 21% in the series of Haase and Skibsted.2 In this series 25% of patients used splinting after the operation, compared with 30% of patients reported on by Kahn and Stanton3 (levator plication vs transanal repair), yet the authors propose that their site-specific defect repair rate is superior. The authors did not define or use standard terms to describe the symptoms of bowel dysfunction, such as splinting, constipation, tenesmus, or fecal incontinence, or of sexual dysfunction, such as apareunia and dyspareunia. In a retrospective study it is hard to compare preoperative and postoperative assessments unless symptoms are specifically assessed with the same standardized questionnaire. These definitions are necessary if apples are to be compared with apples and oranges compared with oranges. Without consistent terms, changes in defecation and sexual problems cannot be properly evaluated. I certainly concur with Dr Cundiff that we need prospective, comparative trials of posterior colporrhaphy techniques with a standard taxonomy of patient symptoms and staged physical preoperative and postoperative findings. There is enough talent in this room, as well as in this Society, to perform controlled studies that compare these different procedures to determine their effectiveness in both the short and long term. I have the following questions for Dr Cundiff: 1. What is your definition of rectovaginal fascia? 2. How do you distinguish the different types of defects during an office examination? During the operation? Is it possible that vaginal dissection, especially midline and inferior, actually creates defects rather than revealing preoperative defects? 3. How was perineal descent assessed? 4. Did you find any relationships between the type of defect found before the operation and preoperative symptoms, postoperative success or failure in curing rectocele, and postoperative symptoms? 5. Please comment on the possible explanations for the high prevalence of postoperative symptoms, especially in light of the fact that some symptoms (dyspareunia, splinting) were found at rates similar to those obtained with traditional posterior colporrhaphy. 6. Define the terms used to describe the symptoms of bowel and sexual dysfunction.
REFERENCES
1. Richardson AC. The rectovaginal septum revisited: its relationship to rectocele and its importance in rectocele repair. Clin Obstet Gynecol 1993;36:976-83. 2. Haase P, Skibsted L. Influence of operations for stress incontinence and/or genital descensus on sexual life. Acta Obstet Gynecol Scand 1988;67:659-61. 3. Kahn MA, Stanton SL. Posterior colporrhaphy: its effects on bowel and sexual function. Br J Obstet Gynaecol 1997;104: 882-6.
DR CUNDIFF (Closing). My definition of rectovaginal fascia is basically the same as that of Denonvilliersâ&#x20AC;&#x2122; fascia, a musculofascial layer that is found between the rectum and the vaginal mucosa or epithelium that is attached to the perineal body inferiorly, to the levator ani muscles at the arcus tendinea levator ani laterally, and to the uterosacral ligaments superiorly. I briefly described our examination in the office, which really was a brokenspeculum examination as well as a bimanual examination and, probably most importantly, a rectovaginal examination. The last seems to be the most useful for actually defining the defects in the rectovaginal fascia. I think that it is much more difficult to distinguish lateral defects on the clinical examination, but you can usually feel the inferior defects fairly easily. With respect to measuring perineal descent, we really based this on a subjective assessment of perineal descent and also on how much the genital hiatus increased with straining. We are actually looking right now to see how that correlates with perineometer measurements, but we consider it to be a good estimate of perineal descent. I think that among the most important points of this article is that rather than making a smaller genital hiatus by simply enlarging the perineal body, we are actually making a smaller genital hiatus by elevating the perineum. I think Dee Fenner mentioned that in their patients a decrease in genital hiatus has been correlated with symptom relief, and I believe that this is why we did show a better symptom relief than with the posterior colporrhaphy. I think that a lot of symptoms of rectocele are due primarily to the perineal descent. With respect to suture, we used several permanent sutures. Sometimes we used polyester and sometimes we used polypropylene (Prolene), and we did have some problems with erosion of the suture, which we switched around some. None of the patients underwent anorectal testing, manometry, or surface electromyography. Symptoms were all based on questioning the patient. Unfortunately, because this was a retrospective study, our questions are not as standardized as I would have liked them to be. All the postoperative questions were from a questionnaire, so they were standardized, but I think that this is one reason that we need to evaluate this prospectively.