Rectocele repair: A randomized trial of three surgical techniques including graft augmentation

American Journal of Obstetrics and Gynecology (2006) 195, 1762–71

www.ajog.org

Rectocele repair: A randomized trial of three surgical techniques including graft augmentation Marie Fidela R. Paraiso, MD, Matthew D. Barber, MD, MHS, Tristi W. Muir, MD, Mark D. Walters, MD Cleveland Clinic, Cleveland, OH Received for publication January 16, 2006; revised July 10, 2006; accepted July 20, 2006

KEY WORDS Rectocele Posterior colporrhaphy Randomized trial Biologic graft

Objective: This study was undertaken to compare outcomes of 3 different rectocele repair techniques. Study design: One hundred six women with stage II or greater posterior vaginal wall prolapse were randomly assigned to either posterior colporrhaphy (n = 37), site-specific rectocele repair (n = 37), or site-specific rectocele repair augmented with a porcine small intestinal submucosa graft (Fortagen, Organogenesis, Inc, Canton, MA; n = 32). Subjects underwent a physical examination and completed 3 validated pelvic floor instruments at baseline and 6 months, 1 year, and 2 years after surgery. Anatomic failure was defined as pelvic organ prolapse quantitation system (POPQ) point Bp R2 at 1 year. Results: Of 106 subjects who enrolled, 105 underwent surgery and of those 105, 98 subjects returned (93%) with a mean follow-up of 17.5 G 7 months. After 1 year, those subjects who received graft augmentation had a significantly greater anatomic failure rate (12/26; 46%) than those who received sitespecific repair alone (6/27; 22%) or posterior colporraphy (4/28; 14%), P = .02. There was a significant improvement in prolapse and colorectal scales and overall summary scores of the Pelvic Floor Distress Inventory short form 20 (PFDI-20), the Pelvic Floor Impact Questionnaire short form 7 (PFIQ-7) after surgery in all groups (P ! .001 for each) with no differences between groups. The proportion of subjects with functional failures was 15% overall, and not significantly different between groups. There was no significant change in the rate of dyspareunia 1 year after surgery and there were no differences between groups. Overall sexual function as measured by the Pelvic Organ Prolapse/Urinary Incontinence Sexual Questionnaire short form (PISQ-12) improved significantly in all groups postoperatively (P ! . 001), with no differences between groups. Conclusion: Posterior colporraphy and site-specific rectocele repair result in similar anatomic and functional outcomes. The addition of a porcine-derived graft does not improve anatomic outcomes. All 3 methods of rectocele repair result in significant improvements in symptoms, quality of life, and sexual function. Ó 2006 Mosby, Inc. All rights reserved.

Funded by an unrestricted research grant from Organogenesis, Inc, (Canton, MA). Organogenesis had no involvement in the design, implementation, analysis or writing of this manuscript. Presented at the Thirty-Second Annual Meeting of the Society of Gynecologic Surgeons, April 3 through 5, 2006, Tucson, AZ. Reprints not available from the authors. 0002-9378/$ - see front matter Ó 2006 Mosby, Inc. All rights reserved. doi:10.1016/j.ajog.2006.07.026

Approximately 200,000 women undergo prolapse surgery annually in the United States. An estimated three fourths of women with prolapse have a rectocele.1 Cundiff and Fenner2 reviewed and summarized outcomes after posterior colporrhaphy, site-specific repair, transanal repair, and rectocele repair with graft materials.

Paraiso et al Anatomic cure ranged from 76% to 96% for posterior colporrhaphy and from 56% to 100% for site-specific defect repair. Currently, the use of implants, both synthetic and biologic, in reconstructive pelvic surgery is expanding despite paucity of data. Proposed indications for the use of graft implantation include: suboptimal autologous tissue; need to augment weak or absent endopelvic tissue; unavoidable stress on the repair; need to bridge a space; insufficient vaginal length or caliber; denervated pelvic floor; and surgeon preference. Some investigators have recommended the use of graft material for recurrent prolapse.3 The objective of this investigation is to compare anatomic and functional outcomes of 3 different surgical techniques for treating rectoceles: posterior colporraphy, site-specific rectocele repair, and site-specific rectocele repair augmented with a porcine-derived, acellular collagen matrix graft (Fortagen, Organogenesis, Inc, Canton, MA).

Material and methods This investigation was approved by the Institutional Review Board at the Cleveland Clinic and all patients provided written informed consent for participation. Funding was provided through an unrestricted research grant from Organogenesis, Inc (Canton, MA), who played no role in the design, implementation, analysis, or writing of this manuscript. Patients undergoing surgery for stage II or greater posterior vaginal wall prolapse from June 2002 through December 2004 were invited to participate. Patients were included if they were 21 years or older and did not desire future vaginal delivery. Patients undergoing concurrent prolapse and/or incontinence surgery were also included. Patients were excluded if they underwent concomitant colorectal procedures; if they had an allergy to pork products; or if they were unwilling to accept porcine product implantation. At baseline, each subject underwent an evaluation that included a standardized history, gynecologic examination using the pelvic organ prolapse quantitation system (POPQ),4 rectovaginal examination and sacral neurologic examination. All examinations were performed in the supine lithotomy position. If the maximal extent of the prolapse could not be observed in the supine position, patients underwent a standing examination. Medical comorbidities were characterized by using the Charlson Comorbidity Index5 and functional capacity was estimated using the Duke Activity Status Index.6 Multichannel urodynamics were performed preoperatively for those patients with symptomatic urinary incontinence or pelvic organ prolapse that extended beyond the hymen. Each subject completed 2 condition-specific quality of life questionnaires (the Pelvic Floor Distress Inventory short form

1763 20 [PFDI-20], the Pelvic Floor Impact Questionnaire short form 7 [PFIQ-7]),7 and a condition-specific sexual function questionnaire, the Pelvic Organ Prolapse/Urinary Incontinence Sexual Questionnaire short form (PISQ-12).8 Patients were randomly assigned to 1 of 3 groups (posterior colporrhaphy, site-specific rectocele repair, and site-specific rectocele repair with graft augmentation) by a computer-generated randomization schedule. Group assignments were concealed in consecutively numbered, sealed, opaque envelopes. Postoperatively, it was not possible to blind the surgeons to the treatment allocation. Patients were blinded to treatment allocation in the immediate postoperative period. If they requested, they were informed of their treatment allocation at their 6-week postoperative visit. Institutional review board requirements did not allow us to blind the subjects beyond this point. However, all postoperative assessments and examinations were performed by a nurse who was blinded to treatment assignment. The patients were administered preoperative antibiotic prophylaxis with 1 g of cefazolin or 100 mg of vibramycin if penicillin-allergic. During the initial dissection for each of the procedures, the vaginal epithelium was opened transversely at the posterior fourchette. The posterior vaginal incision was made in the midline and extended 1 cm above the superior aspect of the posterior vaginal wall defect. Dissection of the vaginal epithelium away from the underlying fibromuscularis extended superiorly to identify the edge of the fibromuscularis, laterally to the medial aspect of the levator ani muscles, and inferiorly to the perineal body. Posterior colporrhaphy was performed using No. 2-0 braided polyester suture (Ethibond, Ethicon, Inc, Somerville, NJ) in interrupted mattress stitches to plicate the rectovaginal muscularis across the midline similar to the midline rectovaginal fascial plication described by Maher et al.9 Unlike traditional posterior colporrhaphy, we did not plicate the levator muscles in the midline. The site-specific posterior repair was performed using the technique described by Cundiff et al.10 Interrupted stitches of No. 2-0 braided polyester suture (Ethibond, Ethicon, Inc) were placed to reapproximate the broken edges of the fibromuscularis and correct all defects. The site-specific posterior repair with graft implant was identical to the procedure described above augmented by a 4 ! 8 cm Fortagen graft (Organogenesis, Inc). The graft was perforated with a scalpel 1 cm medial from its borders in 3 to 4 rows of 3-mm incision points as recommended by the manufacturer. The graft was secured superiorly to the posterior vaginal fibromuscularis and epithelium with No. 2-0 delayed absorbable polydioxanone suture (PDS, Ethicon, Inc). Laterally, the mesh was attached to the levator ani fascia with interrupted stitches of No. 2-0 braided polyester suture (Ethibond, Ethicon, Inc). In cases in which a

1764 concommitant uterosacral vaginal vault suspension or iliococcygeus fascia suspension was performed, the graft was secured superiorly with the respective suspension sutures. Inferiorly, the graft was secured to the perineal body with No. 2-0 polyglycolic acid suture (Vicryl, Ethicon, Inc). Concomitant perineorrhaphy was performed if a patient reported splinting her perineum to defecate and/or a perineal defect was noted at the time of surgery. No. 0 polyglycolic acid sutures (Vicryl, Ethicon, Inc) were used to reapproximate the deep and superficial transverse perineus muscles and bulbocavernosus muscles. The vaginal epithelium was trimmed and closed with No. 2-0 polyglycolic acid sutures in a running interlocking stitch continuing with a subcuticular stitch to close the perineum. Postoperatively, subjects were evaluated 6 weeks, 6 months, and 1 and 2 years after surgery. At the 6-month, 1-year, and 2-year visits, a nurse with extensive experience using the POPQ system, who was blinded to the patient’s treatment assignment, performed a standardized patient interview and a pelvic examination with POPQ in the supine position. Subjects also completed the PFDI-20, PFIQ-7, and PISQ-12 at each of these visits. In addition, subjects were asked to complete a global index of improvement. Subjects were instructed to complete the sentence ‘‘Compared to before surgery, I feel that I am ________’’ using a 7- point scale from ‘‘Very much worse’’ to ‘‘Very much better.’’ Subjects were also asked ‘‘If you had to do it all over again, would you choose the same treatment?’’ The primary outcome of this investigation was anatomic cure of the posterior vaginal wall 12 months after surgery. Subjects were defined as having an anatomic cure if POP-Q point Bp was less than or equal to –2 at the 12-month postoperative visit, consistent with the National Institutes of Health (NIH) Panel on Standardization of Terminology for Pelvic Floor Researchers definition of ‘‘satisfactory anatomic outcome’’ for posterior vaginal prolapse.11 The definitions of other anatomic and functional outcomes can be found in Table I. All methods and definitions conform to the standards proposed by the International Continence Society and the NIH unless otherwise stated.3,11 Assuming alpha of .05, a sample size of 32 subjects in each group provides an 80% power to detect a variance of proportions of 14% and an average failure rate of 17% by using a c2 test. Presuming a 10% loss of followup, our desired enrollment was 106 subjects. Sample size calculations were performed using nQuery Advisor 4.0 (Statistical Solutions Ltd, Boston, MA). Statistical analysis for the primary endpoint was performed by using the Pearson’s c2 test. Logistic regression was used to study the effect of known covariates on procedure success, including stage of prolapse preoperatively, primary versus recurrent repair, and

Paraiso et al preoperative defecatory dysfunction. Secondary outcomes were analyzed with Pearson’s c2 test or Fisher exact test for categorical data and analysis of variance (ANOVA) or Kruskal-Wallis test for continuous data as appropriate. Changes in the scales of the PFDI-20, PFIQ-7 and PISQ-12 were compared by using repeated measures ANOVA. Kaplan-Meier survival curves were generated for the development of anatomic failure and for the development of posterior vaginal wall prolapse to or beyond the hymen and comparisons were made with the log-rank test. The study was performed under the principle of intent-to-treat, so that each patient was analyzed in the group to which she was originally assigned, regardless of which procedure she underwent. Statistical analysis was performed using JMP 5.01 software package (SAS Institute, Cary, NC).

Results One hundred six patients were enrolled and randomly assigned in this trial. Allocation and follow-up are displayed in Figure 1. There were 3 protocol violations during the study period: 1 patient randomly assigned to the site-specific repair group underwent a posterior colporrhaphy; 1 patient in the graft augmentation group was implanted with SurgiSIS (Cook Urological, Inc, Bloomington, IN) because of insufficient supply of study material, and 1 patient in the traditional posterior colporrhaphy group underwent concomitant sigmoid resection and rectopexy. Each of these 3 subjects was included in the analysis and analyzed in the group to which they were originally assigned. Demographic and preoperative anatomic data were similar between groups (Table II). Overall 41% of subjects (n = 43) had stage II pelvic organ prolapse, whereas 53% (n = 56) had stage III and 6% (n = 6) had stage IV. All but 4 subjects (3.8%) underwent other surgical procedures at the time of rectocele repair. There was no difference between groups in type and number of concomitant procedures performed (Table III). Overall, 37% (n = 39) received a hysterectomy, 62% (n = 65) received an anterior colporrhaphy, and 70% (n = 74) underwent a vaginal vault suspension at the time of rectocele repair. Perineorrhaphy was performed in 78% (n = 82) with no difference between groups and was not independently associated with anatomic success, postoperative dyspareunia, postoperative defecatory dysfunction, or ‘‘any’’ postoperative splinting. However, subjects who underwent a perineorrhaphy were significantly less likely to have bothersome splinting 1 year after surgery after controlling for treatment group and baseline symptoms (OR .08, 95% CI 0.004-0.71; P = .038). No significant difference was noted in operating time, estimated blood loss, change in hematocrit, and hospital

Paraiso et al Table I

1765

Definitions of anatomic and functional outcomes

Outcome Anatomic outcomes Anatomic cure of posterior vaginal prolapse (primary outcome) Posterior vaginal support to or beyond the hymen Functional outcomes Functional failure

Dyspareunia

Defecatory dysfunction

stay between treatment groups (Table IV). In general, the intraoperative and postoperative complication rate was low and no differences were noted between groups. There were no graft exposures or graft complications noted at any point during the follow-up period. One year after surgery, 86% (24/28) subjects in the posterior colporrhaphy group and 78% (21/27) subjects in the site-specific group had an anatomic cure of their posterior vaginal wall prolapse, significantly greater than those subjects who received rectocele repair with graft augmentation (cure rate 54%; 14/26), P = .02. Time to the development of posterior vaginal wall recurrence is demonstrated in Figure 2. Time to development of posterior vaginal wall prolapse was significantly earlier in the graft augmentation group than in the posterior colporrhaphy group (P = .046). The site-specific group also had an earlier time to recurrence than the posterior colporrhaphy group, but this did not reach statistical significance (P = .18). Twenty percent (5/25) of subjects in the graft augmentation group had posterior vaginal wall prolapse develop to or beyond the hymen (Bp R 0) 1 year after surgery compared with 7.1% (2/28) in the posterior colporrhaphy group and 7.4% (2/27) in the site-specific repair group (P = .27). Subjects in the graft augmentation group had prolapse develop to or beyond the hymen earlier than the posterior colporrhaphy or site-specific repair group, but this did not reach statistical significance, P = .36. Baseline prolapse stage, baseline Bp measurement, history of previous prolapse surgery, preoperative defecatory dysfunction, and concurrent (or type of) vaginal vault suspension did not significantly influence the anatomic success rates of the treatment groups.

Definition POP-Q point Bp % to –2 at the 1 year after surgery. POPQ point Bp R 0 at 1 year after surgery Worsening of prolapse (POPDI-6) and/or colorectal (CRADI-8) scale scores of the PFDI-20 1 year after surgery A response of ‘‘sometimes,’’ ‘‘usually,’’ or ‘‘always’’ to the question ‘‘Do you feel pain during sexual intercourse?’’(PISQ-12, question 5) An affirmative answer to any of the following questions: ‘‘Do you feel you need to strain too hard to have a bowel movement?’’ ‘‘Do you feel you have not completely emptied your bowels at the end of a bowel movement?’’ or ‘‘Do you ever have to push on the vagina or around the rectum to have or complete a bowel movement?’’ (PFDI-20, questions 4, 7, and 8).

One year after surgery the median Bp measurement for subjects in the graft augmentation group was significantly greater than those in the posterior colporrhaphy or site-specific group (–2 [range –3 to C1], –3 [range –3 to 0], –3 [range –3 to 0], respectively, P = .02). The support of the vaginal cuff (POPQ point C) improved significantly from preoperative measurement (P ! .001) with a median measurement at one year after surgery of –6 (range –10 to C7) with no differences between the treatment groups. The median genital hiatus measurement decreased 1 cm (range –4 to C2) postoperatively (P ! .001), whereas the median perineal body measurement increased 1 cm (range –2 to C3.5) (P ! .001). The median total vaginal length decreased 0.5 cm (range –6 to C3) 1 year after surgery. The reoperation rate for pelvic organ prolapse (any segment) during the study period was 3% (n = 1/33) in the posterior colporrhaphy group, 5% (n = 2/37) in the site-specific repair group, and 10% (n = 3/29) in the graft augmentation group (P = .47). All groups had significant improvements in the prolapse, colorectal, and urinary scales of the PFDI-20 and the PFIQ-7 after surgery indicating decreased bother and improved health-related quality of life (P ! .001 for all) (Table V). There were no significant differences in any of the scales or subscales between groups. A worsening of prolapse or colorectal symptoms 1 year after surgery (ie, functional failure) was noted in 16% (5/31) of the posterior colporrhaphy group, 12% (4/29) of the site-specific repair group, and 21% (6/28) of the graft augmentation group (P = .61). Overall, defecatory dysfunction decreased significantly after surgery as did the individual symptoms of splinting to defecate, hard straining, and feeling of incomplete emptying (P ! .001

1766

Paraiso et al

Figure 1

Flow diagram of subject enrollment and follow-up.

for all), with no differences between treatment groups (Table VI). Fifty-two subjects (50%) considered themselves sexually active at baseline. Of these, 15% (8/52) were not sexually active postoperatively. Twenty-three percent of subjects who were not sexually active at baseline, became sexually active postoperatively (12/53). There was a significant improvement in PISQ-12 scores from baseline to 1 year after surgery in each treatment group (P ! .001 for each) with no differences between groups (Table VI). At baseline, 51% of sexually active subjects complained of dyspareunia (sometimes, usually or always). Postoperatively, 36% of subjects complained of dyspareunia (sometimes, usually or always) with no significant change in any of the groups. The percent of patients in each group who reported dyspareunia ‘‘usually or always’’ 1 year after surgery was 20% in the posterior repair group, 14% in the site-specific repair group, and 6% in the graft augmentation group with no significant difference between groups.

Postoperatively, there was no significant difference between groups in the global index of improvement. Seventy-four percent of subjects in the posterior colporrhaphy group indicated they were ‘‘much better’’ or ‘‘very much better’’ after surgery, whereas 19% indicated they were somewhat better. In the site-specific group, 88% indicated they were ‘‘much better’’ or ‘‘very much better’’ and 12% indicated they were ‘‘somewhat better.’’ In the graft augmentation group 90% indicated that they were ‘‘much better’’ or ‘‘very much better’’ and 7% were ‘‘somewhat better.’’ Overall, 93% of subjects indicated that they would choose the same treatment again, with no difference between treatment groups.

Comment In this investigation posterior colporrhaphy and sitespecific rectocele repair resulted in similar anatomic and functional outcomes. The addition of a porcine-derived

Paraiso et al Table II

1767

Demographic and preoperative anatomic data by treatment group Posterior colporrhaphy (n = 37)

Age (mean years G SD) Parity (median [range]) Menopausal status Premenopausal Postmenopausal with HRT Postmenopausal without HRT Race White Black Other Insurance status Private/HMO Medicaid/Medicare Charlson Comordibity Index* (median [range]) Functional capacityy (median METS [range]) Current smoker Prior hysterectomy Prior URPS Pelvic organ prolapse stage Stage II Stage III Stage IV POPQ measurements, cm (median [range]) Point Bp Point C Genital hiatus (GH) Perineal body (PB) Total vaginal length (TVL)

Site-specific rectocele repair (n = 37)

Site-specific rectocele repair with graft (n = 31)

61 G 12 3 (1-6)

62 G 9 3 (1-8)

60 G 11 3 (1-6)

5 (14) 13 (35) 19 (51)

2 (5) 15 (40) 20 (54)

5 (16) 12 (39) 14 (45)

35 (95) 2 (5) 0 (0)

34 (92) 1 (3) 2 (5)

30 (97) 1 (3) 0 (0)

24 13 0 5.5 3 22 10

23 14 0 5.5 3 20 5

20 11 0 6 1 15 9

P .66 .24 .49

.39

.97 (65) (35) (0-1) (3.5-8) (8) (59) (27)

15 (41) 20 (54) 2 (5) 0 2.5 4 3.5 8

(1-C8) (8-C8) (2-6) (3-7) (5-12)

(62) (38) (0-2) (2.8-9) (8) (54) (13)

(64) (36) (0-2) (5-8) (3) (48) (29)

12 (32) 24 (64) 1 (4)

16 (53) 12 (38) 3 (9)

0 2 4 3 8

0 4 2 3 8

(1-C8) (8-C8) (2-7) (2-5) (6-10)

(1-C10) (8-C10) (2-6) (2-6) (6-11)

.16 .47 .37 .66 .38 .27

.81 .79 .30 .14 .27

Data are presented as n (%) unless otherwise noted. HRT, hormone or estrogen replacement therapy; HMO, health maintanence organization; URPS, urogynecologic or pelvic reconstructive surgery. * Charlson Comorbidity Index6 is calculated by assigning certain comorbidities a weighted value. The assigned values are added to give a total score. y Functional capacity is expressed in metabolic units (METS) as determined by the Duke Activity Index.7

Table III

Concomitant procedures performed with rectocele repair

Hysterectomy Anterior colporrhaphy Paravaginal repair* TVT or TOT Burch colposuspension* Vaginal vault suspensiony Sacral colpopexy* Oophorectomy* Trachelectomy Inguinal hernia repair Sigmoid resection/rectopexy

Posterior colporrhaphy (n = 37)

Site-specific rectocele repair (n = 37)

Site-specific rectocele repair with graft (n = 31)

P

12 20 0 15 2 22 3 2 0 0 1

14 26 1 16 1 26 3 1 1 1 0

13 19 1 12 3 19 1 1 1 1 0

.71 .52 .67 .92 .46 .76 .66 .81 .65 .65 .39

(32) (54) (0) (41) (5) (59) (8) (5) (0) (0) (3)

(38) (70) (3) (46) (3) (70) (8) (3) (3) (3) (0)

(42) (61) (3) (38) (10) (61) (3) (3) (3) (3) (0)

Data presented as n (%). TVT, tension-free vaginal tape; TOT, transobturator tape. * Abdominal or laparoscopic. y Includes uterosacral vaginal vault suspensions, iliococcygeus suspensions and sacrospinous ligament fixation.

1768 Table IV

Paraiso et al Hospital data, intraoperative and postoperative complications

Mean operating time (mean G SD) Median estimated blood loss (mean [range]) Mean change in hematocrit (%) (mean G SD)* Median length of hospital stay (days) (mean [range])

Posterior colporrhaphy (n = 37)

Site-specific rectocele repair (n = 37)

Site-specific rectocele repair with graft (n = 31)

150 150 8 2

151 150 8 2

169 200 9 2

G 68 (50-950) G4 (1-19)

G 69 (50-600) G3 (1-7)

G 62 (50-3500) G3 (1-6)

P .61 .14 .60 .57

Intraoperative complications: Blood transfusion Bladder injury Ureteral injury Bowel injury

1 0 0 0

(3) (0) (0) (0)

0 2 0 0

(0) (5) (0) (0)

1 0 1 0

(3) (0) (3) (0)

.67 .15 .30 1.00

Postoperative complications:y Infection requiring antibiotics Cardiac/myocardial infarction Pulmonary Ileus Pelvic abscess Blood transfusion Urinary tract infection Rectovaginal hematoma Return to the operating roomz Readmissions Emergency room evaluationsx

4 1 0 1 1 2 4 0 0 2 6

(11) (3) (0) (3) (3) (5) (11) (0) (0) (5) (16)

0 0 1 0 0 0 7 0 0 1 5

(0) (0) (3) (0) (0) (0) (19) (0) (0) (3) (19)

3 0 1 0 0 0 6 1 1 0 4

(1) (0) (3) (0) (0) (0) (19) (3) (3) (0) (13)

.13 .39 .69 .39 .39 .15 .54 .30 .30 .41 .91

Data presented as n (%) unless otherwise stated. * Preoperative hematocrit minus postoperative hematocrit obtained on the morning of the first postoperative day. y Postoperative complications include all complications after surgical procedure but within the 6-week postoperative period. z Subject returned to the operating room on the day of surgery for postoperative bleeding and underwent an exploratory laparotomy, evacuation of a pelvic hematoma and ligation of an ovarian pedicle. x Reasons for emergency room evaluation includes: problems with bladder catheter (n = 4), constipation (n = 2), rectal bleeding (n = 1), pelvic abscess (n = 1), postoperative infection (n = 2), and other nonsurgery related conditions (n = 5).

Figure 2 Kaplan-Meier survival curve comparing the development of anatomic recurrence of posterior vaginal prolapse (POPQ point Bp O –2) after posterior colporrhaphy (----), site-specific rectocele repair (_._), and site-specific rectocele repair with graft augmentation (solid line). Time to development of posterior vaginal wall prolapse was significantly earlier in the graft augmentation group than in the traditional posterior colporrhaphy group (P = .046).

small intestinal submucosa graft resulted in worse anatomic outcomes than the repairs without augmentation. All 3 methods of rectocele repair resulted in significant improvement in prolapse and defecatory dysfunction symptoms and health-related quality of life. Contrary to our expectations, overall sexual function improved significantly in all 3 groups and there was no significant change in the dyspareunia rate after any of the 3 methods of rectocele repair. The available data on the use of biologic grafts or permanent synthetic mesh are sparse and largely consist of small retrospective series with short-term follow-up. Kohli et al12 reported a 93% cure rate at 13 months in 30 patients who underwent site-specific posterior repair augmented with cadaveric dermal allograft (Alloderm, Lifecell Corp, Branchburg, NJ). Altman et al13 recently published a prospective cohort of 32 patients who underwent rectocele repair with porcine dermis (Pelvicol, CR Bard, Murray Hill, NJ). At the 12-month followup, 24% had rectoceles (Rstage II) by clinical examination and 52% had rectocele during defacography. Three years after surgery, 41% had a rectocele develop of stage

Paraiso et al Table V

1769

Health-related quality of life outcomes Preoperative PC (n = 35)

SS (n = 36)

1 year after surgery GA (n = 31)

2 years after surgery

PC SS GA PC SS GA P (within (n = 28) (n = 29) (n = 24) (n = 15) (n = 17) (n = 13) group)

PFDI-20* 114 G 56 146 G 66 116 G 55 39 G 30 POPDI-6 43 G 24 55 G 27 41 G 24 11 G 16 CRADI-8 32 G 19 39 G 21 31 G 36 16 G 13 UDI-6 38 G 25 52 G 30 38 G 23 10 G 15 PFIQ-7y 65 G 69 87 G 66 63 G 64 10 G 18 POPIQ-7 20 G 29 26 G 27 19 G 22 1 G 3 CRAIQ-7 18 G 26 24 G 24 19 G 24 4 G 11 UIQ-7 26 G 27 37 G 25 26 G 27 6 G 10

46 12 17 16 22 3 6 14

G G G G G G G G

53 19 18 24 38 8 16 27

34 9 14 10 10 3 3 3

G G G G G G G G

37 15 14 14 23 12 8 7

44 10 16 17 16 3 6 7

G G G G G G G G

32 12 15 15 32 6 11 15

53 14 20 18 16 0 3 13

G G G G G G G G

46 15 18 25 31 0 8 25

32 7 11 14 5 1 1 3

G G G G G G G G

33 10 11 18 13 1 5 7

! ! ! ! ! ! ! !

.001 .001 .001 .001 .001 .001 .001 .001

for for for for for for for for

(between group) each each each each each each each each

.28 .36 .27 .33 .65 .69 .51 .81

Data presented as mean G SD. PC, Posterior colporrhaphy group; SS, site specific rectocele repair group; GA, site-specific rectocele repair with graft augmentation. * The Pelvic Floor Distress Inventory-20 (PFDI-20) has a range of 0 to 300 with higher scores indicating greater distress. It has 3 subscales: the Pelvic Organ Prolapse Distress Inventory-6 (POPDI-6), the Colorectal-anal Distress Inventory-8 (CRADI-8) and the Urinary Distress Inventory-6 (UDI-6), each of which has a range of 0 to 100. y The Pelvic Floor Impact Questionnaire-7 (PFIQ-7) has a range of 0 to 300 with higher scores indicating greater adverse impact on quality of life. It has 3 subscales: the Pelvic Organ Prolapse Impact Quesitonnaire-7 (POPIQ-7), the Colorectal-anal Impact Questionnaire-7 (CRAIQ-7) and Urinary Impact Questionnaire-7 (UIQ-7) each of which has a range of 0 to 100.

Table VI

Bowel and sexual function by treatment group Preoperative PC

Bowel function Defecatory Dysfunction* Splinting to defecate Hard straining Feeling of incomplete emptying Sexual function Sexually active PISQ-12y (mean G SD) Dyspareunia Never/Seldom Sometimes Usually/Always

28 15 24 22

Postoperative SS

(80) (43) (67) (62)

30 21 27 24

GA (85) (58) (77) (69)

29 15 21 27

PC (97) (51) (68) (93)

9 6 11 14

SS (32) (19) (35) (45)

10 7 14 17

P (between group)

GA (35) (21) (42) (51)

5 2 12 8

(21) (7) (41) (28)

.32 .26 .87 .17 .54 .24 .45

17 (46) 29 G 8

18 (46) 31 G 8

17 (53) 33 G 8

19 (61) 36 G 5

21 (63) 36 G 7

16 (57) 37 G 5

10 (45) 6 (26) 7 (30)

12 (52) 9 (39) 2 (9)

12 (50) 12 (50) 0 (0)

11 (55) 5 (25) 4 (20)

16 (72) 3 (14) 3 (14)

16 (84) 2 (11) 1 (6)

Data presented as n (%) unless otherwise noted. * See Table I for definition; A significant improvement was seen in all groups 1 year after surgery (P ! .001 for each). y PISQ-12 has a range from 0 to 48 with higher scores indicating better sexual function. All 3 groups had a significant increase in score from baseline after surgery (P ! .001 for each).

II or greater.14 In our study, subjects who received a sitespecific repair with graft augmentation were more likely to have an anatomic recurrence develop than those who received a site-specific rectocele repair alone or a posterior colporrhaphy. Our definition of anatomic cure was relatively strict (Bp ! –2), so it is difficult to know whether this statistically significant difference is clinically important, particularly given the similar functional outcomes of the treatment groups. Although not statistically significant, subjects who received graft augmentation were also more likely to have prolapse develop to or beyond the hymen than the other 2 treatment groups. On the basis of anatomic outcomes, we cannot recommend use of the Fortagen graft (Organogenesis, Inc) for rectocele repair augmentation.

There are a few explanations for why graft augmentation may have resulted in worse anatomic outcomes in our study. First, the graft material we used may have impaired the normal healing process and thereby weakened the posterior vaginal wall. Complications of wound healing have been reported with porcine dermis. Moore et al15 reported a graft-related complication rate of 20% in posterior repair interposition. A recent publication by Gandhi et al16 summarized histopathology of extirpated hexamethylene di-isocyanate (HMDI) cross-linked porcine dermis grafts (Pelvicol). Tissue reaction to the xenograft varied, including (1) limited remodeling, (2) foreign body type reaction, and (3) complete replacement of graft by fibroconnective tissue and moderate neovascularization. The authors questioned the overall

1770 tolerability and efficacy of HMDI–cross-linked porcine dermis grafts in pelvic reconstructive surgery. If a biologic xenograft is not acellular and contains intact DNA of its species of origin and/or it is cross-linked to increase graft permanence as is the Fortagen graft (Organogenesis, Inc), it may be perceived by the implant recipient as foreign. Fortagen (Organogenesis, Inc) is cross-linked with 1-ethyl-3(3-dimethylaminopropyl) carbodiimide hydrochloride (EDC). No histopathologic studies have been performed to date. A second explanation is that despite our efforts to standardized site-specific rectocele repairs across treatment groups, it is possible that we subconsciously lowered our efforts during the site-specific repair knowing that a graft was going to be placed as reinforcement in the graft augmentation group; for instance, fewer stitches, less attention to detail. However, we do not believe this occurred. The Cochrane Review17 noted that there are insufficient data on the effect of surgery on bowel symptoms and on whether one variation of vaginal repair (posterior colporrhaphy versus site-specific defect repair versus graft augmentation) is superior to others. In our investigation we found that splinting, hard straining to defecate, and incomplete emptying improved significantly at 1 year after surgery and did not differ among all 3 techniques. Postoperative sexual dysfunction has been of significant concern for a number of decades with the surgical management of rectoceles and perineal body defects. In a prospective observational study of sexual function after prolapse repairs, Weber et al18 found that performance of a posterior colporrhaphy, and especially a posterior colporrhaphy with Burch colposuspension, were the only variables that predicted postoperative dyspareunia. Vaginal dimensions in general did not predict dyspareunia and pain was often caused by ridging in the posterior vaginal wall. Maher et al9 reported on 38 women with significantly improved symptomatic rectoceles (Rstage II) and obstructed defecation 24 months after midline rectovaginal fascial plication. In this study the rate of dyspareunia actually improved, possibly because of minimal trimming of vaginal epithelium, and 97% of women were very satisfied. Many authors have abandoned traditional posterior colporrhaphy for sitespecific rectocele repair because several uncontrolled studies have noted that sexual function, particularly dyspareunia, improved or did not worsen after undergoing this technique.10,19-22 However, a recent retrospective review23 and our investigation show similar rates of dyspareunia with posterior colporrhaphy and sitespecific rectocele repair. Interestingly, we found a relatively high rate of preoperative dyspareunia compared with other studies. This may, in part, be explained by the fact that we used a validated sexual function questionnaire whereas others have not. Because of small

Paraiso et al numbers in general and even smaller numbers of sexually active women, our ability to compare dyspareunia rates between treatment groups is somewhat limited. However, surgeons can be reassured that performing a rectocele repair at the time of pelvic reconstructive surgery generally results in an improvement of sexual function. The principal strength of this investigation is that it is a randomized clinical trial that used validated outcome measures in multiple domains assessed by a blinded clinical examiner. The limitations of this trial are the small numbers in each group, the medium length followup and the use of a graft that is not, at present, commercially available. Another limitation is that all but a few subjects underwent concurrent pelvic reconstructive procedures. Thus, our study is primarily generalizable to patients receiving multiple prolapse procedures. For instance, our complication rates reflect those of women undergoing multiple reconstructive surgeries and are most likely an overestimate for women undergoing an isolated rectocele repair. Also, the improvements in anatomic, functional, and quality of life outcomes seen in this study are due to a comprehensive prolapse repair and not due to the rectocele repair alone. A larger study is needed to evaluate whether posterior colporrhaphy is superior to the site-specific repair without graft augmentation. There are few prospective studies comparing graft augmentation techniques versus traditional repairs for anterior and/or posterior vaginal wall prolapse. Our investigation demonstrated that traditional surgeries for rectoceles are superior to site-specific rectocele repair augmented with a bioengineered collagen matrix derived from porcine small intestinal submucosa. As properties of biologic grafts differ, our findings can only be applied to this specific biologic graft. However, surgeons should be discouraged from using biologic grafts routinely until clinical trials justify their use.

References 1. Department of Health & Human Services, Administration on Aging: Older population by age: 1900-2050. Available from: http://www.aoa.gov/prof/Statistics/online_stat_data/AgePop2050. asp. Accessed on May 9, 2004. 2. Cundiff GW, Fenner D. Evaluation and treatment of women with rectocele: focus on associated defecatory and sexual dysfunction. Obstet Gynecol 2004;104:1403-21. 3. Birch C, Fynes MM. The role of synthetic and biological prostheses in reconstructive pelvic floor surgery. Curr Opin Obstet Gynecol 2002;14:527-35. 4. Bump RC, Mattiasson A, Bo K, Brubaker LP, DeLancey JO, Klarskov P, et al. The standardization of terminology of female pelvic organ prolapse and pelvic floor dysfunction. Am J Obstet Gynecol 1996;175:10-7. 5. Charlson ME, Pompei P, Aless KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: development and validation. J Chron Dis 1987;40:373-83. 6. Hlatky MA, Boineau RE, Higginbotham MB, Lee KL, Mark DB, Califf RM, et al. A brief self-administered questionnaire to

Paraiso et al

7.

8.

9.

10.

11.

12. 13.

14.

determine functional capacity (Duke Activity Status Index). Am J Cardiol 1989;64:651-4. Barber MD, Walters MD, Bump RC. Short forms of two condition-specific quality-of-life questionnaires for women with pelvic floor disorders (PFDI-20 and PFIQ-7). Am J Obstet Gynecol 2005;193:103-13. Rogers RG, Coates KW, Kammerer-Doak D, Khalsa S, Qualls C. A short form of the Pelvic Organ Prolapse/Urinary Incontinence Sexual questionnaire (PISQ-12). Int Urogynecol J 2003;14:164-8. Maher CF, Qatawneh AM, Baessler K, Schluter PJ. Midline rectovaginal fascial plication for repair of rectocele and obstructed defecation. Obstet Gynecol 2004;104:685-9. Cundiff GW, Weidner AC, Visco AG, Addison WA, Bump RC. An anatomic and functional assessment of the discrete defect rectocele repair. Am J Obstet Gynecol 1998;179:1451-7. Weber AM, Abrams P, Brubaker L, Cundiff G, Davis G, Dmochowski RR, et al. The standardization of terminology for researchers in female pelvic floor disorders. Int Urogynecol J 2001;12:178-86. Kohli N, Miklos JR. Dermal graft-augmented rectocele repair. Int Urogynecol J 2003;14:146-9. Altman D, Zetterstrom J, Lopez A, Anzen B, Falconer C, Hjern F, et al. Functional and anatomic outcome after transvaginal rectocele repair using collagen mesh: a prospective study. Dis Colon Rectum 2005;48:1233-41. Altman D, Zetterstrom J, Mellgren A, Gustafsson C, Anzen B, Lopez A. A three-year prospective assessment of rectocele repair using porcine xenograft. Obstet Gynecol 2006;107:59-65.

1771 15. Moore RD, Miklos JR, Kohli N. Posterior repair with Pelvicol dermal graft augmentation. Int Urogynecol J 2003;14(Suppl 1):S2. 16. Ghandi S, Kubba LM, Abramov Y, Botros SM, Goldberg RP, Victor TA, et al. Histopathologic changes of porcine dermis xenografts for transvaginal suburethral slings. Am J Obstet Gynecol 2005;192:1643-8. 17. Maher C, Baessler K, Glazener CM, Adams EJ, Hagen S. Surgical management of pelvic organ prolapse in women. Cochrane Database Syst Rev 2004;4:CD004014. 18. Weber AM, Walters MD, Piedmonte MA. Sexual function and vaginal anatomy in women before and after surgery for pelvic organ prolapse and urinary incontinence. Am J Obstet Gynecol 2000;182:1610-5. 19. Kenton K, Shott S, Brubaker L. Outcome after rectovaginal fascia reattachment for rectocele repair. Am J Obstet Gynecol 1999;181: 1360-3. 20. Porter WE, Steele A, Walsh P, Kohli N, Karram M. The anatomic and functional outcomes of defect-specific rectocele repairs. Am J Obstet Gynecol 1999;181:1353-8. 21. Glavind K, Madsen H. A prospective study of the discrete fascial defect rectocele repair. Acta Obstet Gynecol Scand 2000;79:145-7. 22. Sing K, Cortes E, Reid WMN. Evaluation of the fascial technique for surgical repair of isolated posterior vaginal wall prolapse. Obstet Gynecol 2003;101:320-4. 23. Abramov Y, Gandhi S, Goldberg RP, Botros SM, Kwon C, Sand PK. Site-specific rectocele repair compared with standard posterior colporrhaphy. Obstet Gynecol 2005;105:314-8.