Vaginal paravaginal repair with an AlloDerm graft

发布时间：[2014-05-20]

 

 Am J ObstetGynecol 2003;V 189:1612-9.


Vaginal paravaginal repair with an AlloDerm graft
 
Jeffrey L. Clemons, Deborah L. Myers, Vivian C. Aguilar, andLily A. Arya
 
OBJECTIVE: This study was undertaken to describe outcomes of a technique of vaginal paravaginal repairthat used AlloDerm graft (LifeCell, Branchburg, NJ) in women with recurrent stage II or with primary or recurrentstage III/IV anterior vaginal wall prolapse.
STUDY DESIGN: This was an observational study. Thirty-three women underwent a vaginal paravaginal repairusing AlloDerm graft. Anterior vaginal wall prolapse was staged using the pelvic organ prolapse quantificationsystem preoperatively and every 6 months after surgery. Recurrence of prolapse, changes in functional status(urinary symptoms, prolapse symptoms, and sexual activity), and complications were recorded. Objectivefailure was defined as recurrent anterior vaginal wall prolapse, stage II or greater, and subjective failure assymptomatic recurrent anterior vaginal wall prolapse. Life-table analysis evaluated objective and subjectivefailure. Risk factors for recurrent anterior vaginal wall prolapse were evaluated.
RESULTS: The mean age was 65.2 years and 93% of the women were white. Preoperatively, 6 women hadrecurrent stage II, 24 women had stage Ill, and 3 women had stage IV anterior vaginal wall prolapse. Themedian length of follow-up was 18 months. Postoperatively, 12 women had asymptomatic stage II anteriorvaginal wall prolapse (not beyond the hymen) develop. and 1 woman had symptomatic stage II prolapsedevelop. Thus, there were 13 (41%) objective failures and 1 (3%) subjective failure. Life-table analysisdemonstrated the cumulative probability of an objective failure was 0.24 at 1 year and 0.50 at 2, 3, and 4 years.The cumulative probability of a subjective failure was 0.00 at 1 and 2 years and 0.11 at 3 and 4 years. No riskfactors for objective failure were identified. Voiding complaints resolved in 11 of 14 (79%) women (P=.004),incontinence symptoms resolved in 17 of 19 (89%) women (P<.001), and urgency symptoms resolved in 20of 23 (87%) women (P<.001) (all two-tailed Fisher exact test). Twenty-one women (64%) were sexuallyactive, and none complained of postoperative dyspareunia. Complications included 1 case of febrile morbidity,1 cystotomy, and 1 anterior wall breakdown secondary to hematoma formation caused by heparin therapy. Noother erosions or rejections were seen.
CONCLUSION: Vaginal paravaginal repair with AlloDerm graft in women with recurrent stage II or stage II I/IVanterior vaginal wall prolapse is safe and has good subjective but only fair objective success within the first 2years. (Am J ObstetGynecol 2003;189:1612-9.)
Key words: Graft, reconstructive pelvic surgery, anterior vaginal wall prolapse
 
Recurrent anterior vaginal wall prolapse remainsamajor problem in pelvic reconstructive surgery. Different failure rates have been reported in theliterature.[1-7] Failure rates may vary because of thediffereuces in the technique of repair used, the patientpopulation studied, the extent of preoperative prolapserepaired, and if the repair was performed for primary orrecurrent prolapse. Failure rates also depend on thedefinition of failure. For anterior colporrhaphy, failurerates range from 3% to 20%, when surgical failure wasdefined as recurrent symptomatic prolapse.[1,2] In a randomizedtrial of three techniques of anterior colporrhaphy,Weber et al[3] report 
 
failure rates from 54% to70%, but surgical failure was defined as recurrent stage IIor greater anterior vaginal wall prolapse, regardless ofsymptoms. For vaginal paravaginal defect repair, failurerates range from 1% to 4% for recurrent symptomaticprolapse and from 20% to 24% forrecurrent asymptomaticprolapse.[4-7] These studies included women withvarying degrees of anterior vaginal wall prolapse.
To improve success of the repair of the anteriorcompartment, various types of grafts, eg, Marlex mesh(CR Bard, Cranston, RI), Prolene mesh (Ethicon,Somerville, NJ), solvent-dried cadaveric fascia lata,and polyglactin mesh, have been used.[8-13]  AlloDerm(LifeCell Corp, Branchburg, NJ) is a cadaveric, acellular,immunologically, inert dermal matrix that has been usedas a graft in reconstructive facial surgery, dental surgery,coverage of skin graft donor sites, and intestinal fistularepair. AlloDerm has recently been used with good successin pelvic reconstructive surgery in site-specific rectocelerepair, [14]rectovaginal fistula repair,[15] a combined slingand anterior colporrhaphy procedure, 16 and anterior andposterior colporrhaphy.[17] The purpose of this study was todescribe the success of a technique of vaginal paravaginaldefect repair with the use of AlloDerm graft material inwomen with recurrent stage II or with primary or recurrentstage III/IV anterior vaginal wall prolapse. Asecondary purpose was to determine risk factors forrecurrent prolapse after surgery.
Methods
This was an Institutional Review Board-approvedobservational study performed at our institution.Between November 1998 and April 2002, 33 women witheither recurrent stage II or with primary or recurrent stageIII or IV anterior vaginal wall prolapse underwent a vaginalparavaginal defect repair using AlloDerm graft.Concurrent pelvic reconstructive surgical procedures,including anti-incontinence procedures, were performedwhen needed.
Preoperatively, demographic data, comorbidities, andfunctional status (urinary symptoms, prolapse symptoms,and sexual activity) were recorded. Pelvic prolapse wasstaged by using the pelvic organ prolapse quantificationsystem (POP-Q), and methods, definitions, anddescriptions conform to the standards recommended bythe International Continence Society.[18]Patientwereexamined on a standard pelvic examination table in thedorsal lithotomy position. All examinations were performedwith an empty bladder, after a catheterization fora postvoid urine residual. A split speculum examinationwas used to stage the prolapse. All women were examinedat maximal strain, and all patients confirmed that the sizeof the prolapse seen during the examination was asextensive as the most severe prolapse that she hadexperienced. The anterior and posterior vaginal wallsand the vaginal apex were each staged separately. Allwomen underwent a complete multichannel urodynamicevaluation (complex cystometry, leak point pressure,urethral pressure profile, and uroflowmetry with postvoidresidual). Prolapse reduction that used either a ringpessary or long cotton swabs was performed to diagnosepotential stress urine incontinence. Urinary retention wasdefined as a postvoid residual of more than 100 ml.
The surgical technique is described. All patientsreceived preoperative antibiotic prophylaxis. The anteriorvaginal wall was opened in the midline at the apex andthe 
incision was extended to 2 cm from the externalurethral meatus. Dissection was carried out paravaginallyalong the levator any muscles, with the levator musclespalpated laterally against the pubic bones. Specific defectsof pubocervical fascia were interrupted   with interruptedsutures of 2-0 polygylcolic acid. A 3×7cm strip ofAlloDerm was then cut into the shape of a trapezoid,sized, a nd tailored for each patient. The top of the graftwas approximately 4 em in width, the bottom wasapproximately 6 to 7 cm, and the height was 3 cm. Thegraft was positioned over the bladder base and secured tothe arcustendineus with four sutures of 2-0 braidedpermanent polyester sutures bilaterally. The top of thegraft was secured on either side of the urethra at the levelof the urethovesicaljunction with two 2-0 polygylcolic acidsutures and the base of the graft was attached to thevaginal apex with two 2-0 polygylcolic acid sutures (Fig 1) .Cystoscopy was performed to ensure integrity of thebladder and ureters. The anterior vaginal wall was closedin two layers: interrupted horizontal mattress sutures of2-0 polyglycolic acid were placed submucosally and intermpted sutures of 2-0 polyglycolic acid were used toclose the anterior vaginal wall. Redundan t vaginal wall wastrimmed. Other concomitant 1•econstructive and anti incontinence procedures were performed accordingly.Estimated blood Joss and surgical and postoperativecomplications were recorded.

Fig 1. Placement of the AlloDerm graft.
Patients underwent a voiding trial on postoperative clay2. Women with urinary rerention were discharged homewith a Foley catheter. All women used estrogen vaginalcream nightly. Women were seen at 2,5,and 8 weeks aftersurgery. A postvoid residual was obtained at the 2-weekvisit to rule out occult urinary retention. Patients werethen evaluated at 6-month intervals. At each postoperativevisit, functional status (utinary symptoms, prolapsesymptoms, and sexual activity) was assessed and patientswere examined by one of the authors.
Recurrence of pelvic prolapse was assessed with the POPQsystem, again at maximal strain. Anterior vaginal wallprolapse was staged by u sing points Aa and Ba. Outcomeswere assessed objectively and subjectively. Objective failurewas defined as recurrent anterior vaginal wall prolapse(either pointAa or Ba), stage II or greater, during maximalstrain. For objective success, the anterior vaginal wall (bothpointAa and Ba) had to remain at either stage 0 or 1 on allpostoperative examinations. Subjective failure was definedas recurrent anterior vaginal wall prolapse associated withsymptoms of prolapse, but not symptoms of prolapseassociated with recurrence of the posterior or apical compartments.Subjective success was defined as no symptomsof prolapse, even if there was recurrentanteriorvaginal wallprolapse. Symptoms of prolapse were vaginal bulge, pelvicpressure, and splin ting to empty the b owel or bladder.
Statistical analysis consisted of two-tailed t tests forcontinuous variables and two-tailed x2 tests (or Fisherexact test) for dichotomous variables to identify riskfactors for objective failure. Risk factors assessed weredemographics, comorbidities, preoperative prolapsestage, and prior reconstructive surgery. P values less than.05 were considered statistically 
significant. Kaplan-Meiersurvival curves for objective and subjective failure rateswere made by using the life-table analysis method. Allstatistical methods were performed using the SASsoftware package (SAS Institute Inc, Cary, NC).
Results
The mean age for the 33 women was 65.2 years (range42-83 years), mean parity was 2.8 (range 1-6), and meanweight was 149.5 pounds (range 104-197 pounds). Thirtyonewomen were white, 1 was Hispanic, and 1 was Asian.Thirty women (91%) were postmenopausal, of which 21(64%) were using estrogen re placement therapy. Twentyninewomen (88%) had a prior hysterectomy, and 19(58%) had prior surgery for pelvic organ prolapse.Comorbidities included the following: 7 women (21%)had chronic obstructive pulmonary disease, 4 women(12%) had diabetes mellitus, 6 women (18%) were obese(body mass index [BMI]≥30mg/ m2 ), 15 women (45%)reponed constipation , 15 women (45%) strained to detecate,and 1 woman (3%) smoked tobacco. Functionalstatus was notable for the following: 14 women (42%)reported voiding difficulty, 19 women (58%) reportedstress u rinary incontinence symptoms, 23 women (70%)reported urgency symptoms, and 21 women (64%) weresexually active. All women complained of a vaginal bulge.
The number of women with preoperative anteriorvaginal wall prolapse stages were as follows: 6 women hadstage II (all recurrent, and Aa at+ 1 cm), 24 women hadstage III (11 recurrent, 13 primary), and 3 women hadstage IV (2 recurrent, 1 primary). The number ofpreopetative anterior, posterior vaginal wall and vaginalapex prolapse stages are listed in Table 1. Preoperatively, 9women (28%) had urinary retention (range 120-300 ml).Ten women were diagnosed with genuine stress urinaryincontinence on urodynamic evaluation and underwenta concurrent anti-incontinence procedure.
Twenty-five (76%) women underwent 54 concurrentpelvic reconstructive procedures. The type and number ofconcurrent procedures performed were vaginal hysterectomywith McCall's culdoplasty(3), uterosacral ligamentvaginal vault suspension (10), sacrospinous ligament suspension(l), enterocele repair (11), posterior colporrhaphy(16), Tension-free vaginal tape sling (Gynecare,Somerville, Nj) (8), Mersilene mesh (Ethicon , Somerville,NJ) sling (2), and perineorrhaphy (3).
The mean estimated blood loss for all proceduresperformed was 378 mL (range 150-1000 ml). There wereno transfusions. One woman had febdle morbidity andone woman had a cystotomy during enterocele dissection.One woman had a postoperative myocardial infarctionand required anticoagulation . She subsequently had ananterior vaginal wall hematoma develop with breakdownof the vaginal mucosa but the graft remained intact. Hersurgery was uncomplicated. She died from congestiveheart failure 2 months later and has been lost to follow-up .No other patients have been lost to follow-up, and nograft-related complications such as rejection, erosion,infection , or fistula formation were seen during follow-up. Five women had erosion of one paravaginal permanentsuurre, and these sutures were removed in the office.
The median length of follow-up for was 18 months(mean 19.5 months, range 6-48 months). Postoperatively,12 women had recurrent asymptomatic stage II anteriorvaginal wall prolapse (not beyond the hymen) develop.These were objective failures. One woman had recurrentsymptomatic stage II prolapse with stress incontinencedevelop at 28 months after surgery; both points Aa and Bawere at+1 cm. She was an objective as well as a subjectivefailure. The remaining 19 women had stage 0 or 1 anteriorvaginal wall prolapse. Thus, there were 13 (41%) objectivefailures and 1 (3%) subjective failure .
Life-table analysis of the postoperative 6-month intervalexaminations demonstrated that the cumulativeprobability of an objective failure was 0.24 (95% CI 0.05-0.43) at 1 year, 0.50 (95% Cl 0.12-0.88) at 2 years, and 0.50at3 and 4 years (95% Cis both 0.00-1.0). The cumulativeprobability of a subjective failure was 0.00 at 1 and 2 years(95% Cis 0.00-0.06 and 0.00-0.08, respectively) and 0.11 at3 and 4 years (95% Cis 0.00-0.44 and 0.00-1.0, respectively).The 95% Cis were wide for years 3 and 4 becauseof the small numbers of patientat this time interval offollow-up. Kaplan-Meier survival curves for objectivefailure and subjective failure are illustrated in Fig 2.
Eighteen women voided normally by the second postoperativeday and 7 women voided by the seventhpostoperative day. One woman required a Foley catheterfor 14 days because of a cystotomy. Seven women (21%)had postoperative urinary retention more than 7 days(range 11-59 days) .Fou r of the 7 women who had urinaryretention were among the 8 women who underwenta concurrent tension-free vaginal tape sling. 
 
The placementof the tension-free vaginal tape sling along with theAlloDerm graft increased the risk of postoperative urinary retention (odds ratio 7.3, 95% Cl 1.2-46.1. Fisher exacttest, P=.04).
Changes in functional status were analyzed with theFisher exact test. Overall, urinary symptoms were improvedwith surgery. Preoperative voiding difficultvresolvedin 11 of 14 (79%) women (P=.004), incontinencesymptoms resolved in 17 of 19 (89%) women (P<.001),and urgency symptoms resolved in 20 of23 (87%) women(P<.001) . Twenty-one women were sexually active andnone reported postoperative dyspareunia. Six of the 9(67%) women with preoperative urinary retention hadnormal postvoidresiduals after surgery (P=.10); theother 3 women required prolonged self-catheterization,but eventually had normal residuals.
Prolapse symptoms recurred in six women (19%). Onewoman had recurrent anterior vaginal wall prolapse; theother five women had recurrent prolapse of the apical andposterior compartments. The woman with symptomaticrecurrent anterior vaginal wall prolapse has not yetundergone repeat repair. The other five women haveneeded additional surgery. In three of the five women, theanterior vaginal wall remained at stage 0-1; in two women,the anterior vaginal wall was asymptomatic stage II. Twowomen had recurrent stage liposterior vaginal wallprolapse develop and underwent repeat posterior colporrhaphy.One woman had recurrent stage III vaultprolapse (stage I anterior wall) developafter6monthsandwas treated with a sacrospinous ligament vault suspension.Two women had recurrent stage II vault prolapse developafter 6 and 12 months and were treated with abdominalsacrocolpopexy. One other woman (antelior wall at stage0-I) had genuine stress incontinence develop postoperatively and was treatecl with a urethral collagen injection.
There was only 1 objective failure among the 6 women(17%) with a preoperative diagnosis of recurrent stage IIanterior vaginal wall prolapse, compared with 12 objectivefailures among the 26 women (46%) with either stage IIIor IV anterior vaginal wall prolapse, but this was notstatistically different (Fisher exact test, P=.36). Noidentifiable risk factors for objective failure were foundamong demographics, comorbidities, and prior reconstructivesurgery (TableII). A concurrent tension-freevaginal tape sling did not prevent objective failure (Fisherexact test, P=.68).
Comment
In our patient population described, a vaginal paravaginalrepair with AlloDerm graft had an objective failurerate of 41% at a median follow-up of 18 months. There wasonly one subjective failure, occurring at 28 months.Therefore, the subjective failure rate was 3%. Functionally,these patients had significant improvement in symptomsof voiding difficulty, incontinence, urgency, and bulge.There were no complaints of dyspareunia with this technique.Our subjective failurerates are similar to that reponedin the literature, but our objective failure rates arehigher. These differences may be due to definition of failure,the population studied, or length of follow-up. Objectivefailure rates may have been unclerreported in otherstudies 
 
because of nonuse of the POP-Q staging system.Shull et al[4] evaluated 56 women (grade 1-4) for a mean of1.6 years and reported 2 (4%) with symptomatic recurrentgrade 3 anterior wall prolapse and 13 (24%) withasymptomatic grade 1-2 prolapse. Young et al[5] evaluated100 women (grade 2-4) for a mean of 10.6 months andreported one (1%) symptomatic recurrent grade 3 anterior vaginalwall prolapse and 23 (23%) with asymptomaticprolapse. Mallipeclcli et al[6] reported only one (3%) recurrentsymptomatic grade 2 cystocele in 35 women (grade 2-4) who were evaluated for a mean of 1.6 years. Farrell[7]reported a failure rate of20% (not defined) in 27women(grade 1-3) who were evaluated for a mean of 8 months.
All our objective failures p roved to be fai lures of pointAa; that is, the anterior vaginal wall was well supportedproximally but not distally. This may be related to thesurgical technique or to the graft material. The techniquewe described places the top edge of the graft at the level ofthe urethrovesicaljunction (app roximately at poimAa).Perhaps there would be better success of repair if the graftwere placed more distally under the urethra or used asa sling. Goldberg et al[19] demonstrated that a concurrent suburethral sling was associated with a reduction in therate of recurrent cystocele after 1 year offollow-up ( 42% vs19%). Groutz et al[10] performed a "cystocele repair" witha 5×7cm patch of Tutoplast (solvent-dried ) cadavericfascia lata (Mentor, Santa Barbara, Calif) in 21 womenwith a severe cystocele, and 19 (90%) women hada concurrent sling with the same material. The description of the procedure is similar to our paravaginalrepair because the patch was anchored laterally to thearcustendineus and to the cardinal ligaments proximally(and posteriorly to the uterosacral ligaments if a hysterectomywas performed) with absorbable sutures. At a meanfollow-up of 20 months, therewere no recurrent cystoceles.The larger patch size may have offered berterproximal support and the use of a concurrent sling mayhave offered better distal support.
Chung et al[16] used an AlloDerm graft as a concominantsuburethral sling and patch for anterior colporrhaphy totreat 19 women with stressurinary inconrinence andgrade 3 cystocele . They, too, used a 3×7cm graft, butplaced it in a longitudinal direction, supporting theurethra, and the remainder was sutured to the pubocervicalfascia. With a mean follow-up of 28 months,there were three (16%) objective failures. Flood et al[12]andNicita[13] both describe the use of a Marlex patch torepair cystoceles, but review of each technique shows thatthe graft is anchored distally beneath the urethra, likea sling. Flood et al[12] anchored a 1×4cm mesh grafttransversely at the bladder neck only, and reponed norecurrent anterior vaginal wall prolapse in 140 womenafter a mean follow-up of 3.2 years, with three vaginalmesh erosions. Nicita[13] used a large hammock-shapedgraft ( 10cm transverse, and 3-5cm anteroposteriorly,anchored at the levators, the cervix, and around thebladder neck). andreponed no recurrent cystocele in 44women after a mean follow-up of 13.9 months, with onevaginal mesh erosion. Further investigation into theprevention of recurrent anterior vaginal wall prolapsewith a concurrent sling is warranted.
Failures may also be due to the elasticity of theAlloDermgraft[20] or due to remodeling by host connectivetissue. AlloDerm is a cadaveric dermal graft that isdecellularized to remove the risk of rejection or 
inflammationand then freeze dried through a process thatmaintains collagen, elastin, and proteoglycans (http://www.lifecell.com). It is designed to serve as a biologicscaffold for normal tissue remodeling. By clay 7 to day 10,revascularization and normal tissue remodeling begins,and host collagen deposition occurs over the next severalmonths. By 6 to 8 months, AlloDerm is naturally remodeledinto the patient's own tissue, without fibrosis.These properties may account for the lack of dyspareuniaand minimal complications in wound healing but mayalso con tribute to failure. Seven of the 12 women withrecurrent asymptomatic stage II anterior vaginal wallprolapse had a "hypermobile" anterior vaginal wall: atrest, the anterior vaginal wall is at stage 0, but with Valsalvamaneuver it temporarily stretches down toward thehymen and then returns to it normal position when theValsalva effort ends. Because AlloDerm retains moreelasticity, it may allow the anterior vaginal wall to stretchwith strain, and then return to its original position oncethe strain ends. Alternatively, the hypermobile anteriorvaginal wall may represent early failure of the AlloDermgraft because it may be stretching out. Longer follow-up isneeded to determine whether the AlloDerm remainselastic and successful or is stretching out and failing.Jelovsek et al[17] used an AlloDerm graft with anterior (9) orposterior (15) colporrhaphy in 24 women with stage II-III anterior or posterior vaginal wall prolapse and reportedan 87% success rate (defined as stage 0-1, same as ourstudy) in the posterior wall but only a 44% success rate inthe anterior wall , after a median follow-up of 18 months.The success in the posterior vaginal wall is similar to thatreponed by Miklos et al,[14 ]where they reported a 95%success rate (using a pointAp measurement of -0.5 cm orgreater to define failure) in 57 women who underwenta site-specific rectocele repair augmented with an AlloDerm graft. Perhaps AlloDerm is better suited forrepair of posterior vaginal wall prolapse than anteriorvaginal wall prolapse.
Several studies have used various types of syntheticmesh with anterior colporrhaphy[8,9,12,13]. Julian[8] performeda vaginal paravaginal repair and anterior colporrhaphyin 24 women with recurrent grade 3-4 anteriorvaginal wall prolapse, with (12) or without (12) a Marlexpolypropylene mesh patch. The mesh was used to preventmidline anterior recurrence, and was anchored withnonahsorbable suture in a trapezoidal area under theanterior vaginal wall, created after performing anteiorcolporrhaphy, paravaginal repair, and needleurethropexy. After 2 years follow-up. there were norecurrences in the mesh group compared with 4 cases(33%) ofrecurrenc.e in the control group, but there werethree mesh erosions (25%) , all managed conservatively.Migliari et al[9] also performed an anterior colporrhaphywithMarlex mesh in 12 women with grade 3 anteriorvaginal wall prolapse. The mesh was attached to thepubourethral ligaments and pubocervical fascia anteriorlyand to the 
 
 
cardinal ligaments and pubocervical fasciaposteriorly with absorbable sutures. Three women (25%)had recurrent asymptomatic grade 1 prolapse developafter a mean of 20 months follow-up (no symptomaticprolapse), and there were no erosions. The use ofa permanent mesh, however, runs the long-term risk oferosion and fistula formation.
Interestingly, Sand et al[11]reported on the use ofpolyglactin mesh, a synthetic absorbable mesh in a randomizedtrial of 160 women who underwent anterior andposterior colporrhaphy. The mesh was simply placed to tryto enhance scarring and not sutured to provide support.Of 143 women with 1 year of follow-up, 8 women ( 11%)without mesh and 2 women (3%) with mesh had recurrentgrade 2 prolapse, and an addition al 31% and 22%,respectively, had recurrent grade 1 anterior vaginal wallprolapse.
Our technique had few surgical complications. Therewere no blood transfusions, no wound infections, and nograft erosion or rejection. Our complication rate is similarto complication rates reported in other studies on vaginalparavaginal repair.[4,7]We did have seven women (21%)who had urinary retention more than 7days. The placementof a tension-free vaginal tape sling along with theAlloDerm graft significantly increased the risk of postoperative urinary retention, as 50% of women who underwenta concurrent tension-free vaginal tape sling hadretention. This rate of urinary retention is similar to that reported in a study by Partoll,[21] in which urinary retention(defined as a need for catheterization for more than 3days) occurred in 43% of37 women undergoing a tension-freevaginal tape sling with concurrent pelvic reconstructivesurgery.
There are several limitations of our study. The follow-uptime is short (median 18 months, mean 19.5 months),therefore long-term safety and success are unable to beassessed. Another limitation is that the examiners atpostoperative visits were not blinded. We also did not havea control group. Hence, our study does not indicate ifthe use of a graft improves outcomes compared witha traditional vaginal paravagin al repair without a graft.Finally, the small mumber of women in the study may have limited our ability to identify for recurrent prolapse are not yetidentified in the literature.
Assuming that grafts would be useful, which grafts arebest suited for pelvic reconstructive surgery? Ideally, thegraft would be strong and durable, yet remain pliable inthe vagina, so as not to interfere with sexual intercourse.Various studies have compared the biomechanical propertiesof autologous and homologous graft materials.Lemer et al[22] examined maximum load to failure andstiffness in 70 1 X 1cm specimens of four types of graftswith a tensiometer: autologous rectus fascia, solvent-driedcadaver fascia lata, freeze-dried cadaveric fascialata,and cadaveric dermis. He found no difference amongautologous rectus fascia, solvent-dried cadaver fascia lata, and cadaveric dermis, but the maximum load to failureand 
stiffness were significantly lower with freeze-driedcadaveric fascia lata. Choe et al[23]  
 
examined the tensilestrength of various pubovaginal sling graft materials.Synthetics and dermis had the highest tensile strength,followed by cadaveric fascia lata, autologous rectus fascia,and vaginal mucosa. Scalafini et al[20] compared 
themicroscopic structure and physical properties of solvent-driedcadaver fascia lata, freeze-dried cadaveric fascia lata,and two cadaveric dermis grafts, AlloDerm and DuraDerm(CR Bard, Murray Hill, NJ ). He found that AlloDerm andsolvent-dried cadaver fascia lata were stronger and retainedmicroscopic architecture closer to that of untreatedtissue than DuraDerm or freeze-dried cadavericfascia lata, and AlloDerm retained significantly moreelasticity• The above differences suggest that the specific processing method can preserve or alter the physicalproperties of graft materials. andthat AlloDerm andsolvent-dried cadaveric fascia lata likely represent the bestallograft materials.
With our technique, we had good subjective outcomes,but only fair objective outcomes. Most providers wouldagree that resolution of the symptom(s) for which thepatient presented is important. Whether Allodermremains a durable graft remains to be seen. Longerfollow-up is needed to determine du rability of therepair. Ideally, randomized controlled trials will direct us toward the best approach to repair anterior vaginal wallprolapse.
 
REFERENCES
1. Porges RF, SmilenSW. Long-term analysis of the surgicalmanagement of pelvic support detects. Am J ObstetGynecol 1994;171:1518-26.
2. Macer GA.Transabdominal repair of cystocele, a 20 year experience, compared with the traditional vaginal approach. Am J ObstetGynecol 1978; 131:203-7.
3. Weber AM, Walters MD,PiedmonteMR, Ballard LA. Anterior colporrhaphy: a randomized trial of three surgical techniques. Am J ObstetGynecol 2001;185:1299-304.
4. Shull BL, Benn SJ, Kuchl TJ. Surgical management ofprolapse of the anterior vaginal segment: an analysis of support defects, operative morbidity, and anatomic outcome. Am J ObstetGynecol 1994; 171:1429-36.
5. Young SB, Daman JJ, Bony LG. Vaginal paravaginal repair. One-year outcomes. Am J ObstetGynecol 2001;185:1360-6
6. Mallipeddi PK, Steele AC, Kohli N, Karram MM. Anatomic and functional 
 
outcome of vaginal paravaginal repair in the correction of anterior vaginal wall prolapsel. IntUrogynecol J Pelvic Floor Dysfunct 2001;12；83-8
7. Farrell SA,Ling C. Currycombs for the vaginal paravaginal defect repair. ObstetGynecol 1997;90:845-7
8. Julian TM. The efficacy of Marlex mesh in the repair of severe,recurrent vaginal prolapse of the anterior midvaginal wall. Am J ObstetGynecol 1996;175:1472-5
9. Migliari R, De Angelis M, Madeddu G, Verdacchi T. Tension-free vagina mesh repair for anterior vaginal wall prolapse. EurUrol 2000;38:151-5
10.Groutz A, Chaikin DC, Theusen E, Blaivas JG. Use of cadaveric solvent-dehydrated fascia lata for cystocele repair-preliminary results. Urology 2001;58:179-83
11. Sand PK, Koduri S, Lobel RW, Winkler HA, Tomezsko J, CulliganPJ,et al. Prospective randomized trial of polyglactin 910 mesh to prevent recurrence 
ofcystoceles an rectoceles. Am J ObstetGynecol 2001;184:1357-64
12. Flood CG, Drutz HP.  Waja L. Anterior colporrhaphy reinforced with Marlex mesh for the treatment of cystoceles. IntUrogynecol J Pclvic Floor Dysfunct 1998;9:200-4
13. Nicita G. A new opreatione for genitourinary prolapse. J Urol 1998;160:741-5
14. Miklos JR, Kohli N, Moore R. Levarorplasty release and reconstruction of rectovaginal septum using allogenic dermal graft. IntUrogynecol J Pelvic Floor Dysfunct 2002;13:44-6
15.Miklos JR, Kohli N. Rectovaginalforsula repair utilizing a cadaveric dermal allograft. IntUrogynecol J Pelvic Floor Dysfunct 1990;10:405-6
16. Chung SY, Franks M, Smith CP, Lee JY, Lu SH, Chancellor M. Technique of combined pubovaginal sling and cystocele repair using a single piece of cadaveric dermal graft. Urology 2002;59:538-41
17. Jelovsek JE, Amundsen C, Addison WA, Weidner AC. Outcomes of pelvic organ prolapse surgery with alloderm graft. Proceedings of the 23rd annual meeting of the American Urogynecologic Society;2002 Oct 17-19; San Francisco: The society; 2002
18. 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 ObstetGynecol 1996; 175:10-7
19. Goldberg RP, Koduri S, Lobel RW, Culligan PJ, Tomezsko JE, Winkler HA, et al. 
 
Protective effect of suburethralslings onpostoperative cystocele recurrence after reconstructive pelvic operation. Am J ObstetGynecol 2001;185:1307-13
20. Sclafani AP, McCormick SA, Cocker R. Biophysical and microscopic analysis of homologous dermal and fascial materials for facial nestheticandreconstuctive uses. Arch Facial PlastSurg 2002;4:164-71
2 1. Panoll LM. Efficacy of tension-free vaginal tape with otherpelvic reconstructive surgery. Am J ObstctGynccol 2002;186:1292-8.
22. Lemer ML, Chaikin DC, Blaivas JG. Tissue strength analysis ofautologous and cadaveric allografts for the pubovaginalsling.NcurourolUrodyn 1999; 18:497-503.
23. Choc JM, Kothandapani R, James L, Bowling D. Autologous, cadaveric, and synthetic materials used in sling surgery: comparativebiomechanical analysis. Urology 2001; 58:482-6.
 
Discussion
DR PAUL FINE. This is a prospective observational studyof a surgical :.echnique for correcting vaginal wallprolapse. It inc01-porates bilateral pantvaginal attachmentof an AlloDenn graft by a vaginal approach. Dr Clemonsand his coauthors point out the difficulty in comparingcurrently published studies of anterior wall reconstructionbecause of variable subjective and objective outcomes.Recurrence of anterior vaginal wall prolapseoccurs too frequently, even in the hands of experiencedpelvic reconstructive surgeons. All 33 enrolled womenhad either recurrent POP-Qstage II or greater or primarystage III or greater anterior wall prolapse. Twenty-fivewomen (76%) had concurrent prolapse repairs of othervaginal compartments or anti-incontinence procedures.Both subjective and objective POP-Q outcomes wereassessed. Results were a disappointing 41% objectivefailure (POP-Q stage II or greater), and a 3% subjectivefailure. I have the following questions for the authors:
1. POP-Q measurements do not assess paravaginaldefects. How was paravaginal support assessed preoperativelyand postoperatively? Were objective failures central or lateral?
2. Technical details are unclear. Was the fibromuscularis("fascia") first split from the vaginal mucosa? Weresite-specific defects or paravaginal fascial defects firstrepaired before placement of the graft? Were paravaginaldefects created to attach the graft? 
 
How wasoptimal graft tensioning accomplished? Four sutureswere placed at the 4-cm superior aspect of the graftbut only two at the 7-cm inferior portion of the graft should more sutures have been placed?
3. Subjective outcome questionnaires are not specified.Are they validated?
4. How many of the anterior wall failures were afterprimary versus repeat repairs?
5. How many of the five apical and posterior wall failureswere primary (de novo) or secondary failures (afterconcurrent repair during the AlloDerm paravaginalrepair)?
6. The postoperative application of estrogen cream isintuitively attractive, but not evidence based. Howsoon postoperatively did you begin using it?
7. Why did you use weight and not BMI?
8. The authors acknowledge the major study limitations,including a small number of women, lack of a control group, and short follow-up time (median 18 months).Are the authors still studying this technique ora modification of it?