West of Scotland Study Group: S Bjornsson (Urogynaecology Unit, Southern General Hospital, Greater Glasgow & Clyde Health Board, Glasgow, UK), M Allam (Department of Obstetrics & Gynaecology, Wishaw General Hospital, Lanrkshire NHS Trust, Wishaw, UK), A Paterson (Department of Obstetrics & Gynaecology, Royal Alexandra Hospital, Greater Glasgow & Clyde Health Board, Glasgow, UK), S Pringle (Urogynaecology Unit, Southern General Hospital, Greater Glasgow & Clyde Health Board, Glasgow, UK), C Hardwick (Urogynaecology Unit, Southern General Hospital, Greater Glasgow & Clyde Health Board, Glasgow, UK), J Tierney (Urogynaecology Unit, Southern General Hospital, Greater Glasgow & Clyde Health Board, Glasgow, UK), HM Ali (Urogynaecology Unit, Southern General Hospital, Greater Glasgow & Clyde Health Board, Glasgow, UK), R Hawthorn (Urogynaecology Unit, Southern General Hospital, Greater Glasgow & Clyde Health Board, Glasgow, UK), Z Parveen (Urogynaecology Unit, Southern General Hospital, Greater Glasgow & Clyde Health Board, Glasgow, UK), MU Lim (Department of Obstetrics & Gynaecology, Royal Alexandra Hospital, Greater Glasgow & Clyde Health Board, Glasgow, UK), N Aedla (Urogynaecology Unit, Southern General Hospital, Greater Glasgow & Clyde Health Board, Glasgow, UK), JS Sheridan (Interventional Radiology Department, Gartnavel General Hospital, Greater Glasgow & Clyde Health Board, Glasgow, UK).
Objectives To assess the complications and short-term outcomes of prolapse repair mesh devices used in the management of female pelvic organ prolapse (POP).
Design Retrospective cohort study.
Setting Multicentre study involving a tertiary referral urogynaecology unit and two district general hospitals.
Population 329 women who underwent surgical management of prolapse with various prolapse repair mesh devices in the period between January 2005 and December 2006.
Methods Women were identified from theatre records. An independent clinician performed a case notes review during the period March to May 2007.
Main Outcome Measures Complication rates and the short-term cure at 3-month follow-up (defined as ≤stage I prolapse on the POP-Q or Baden Walker scoring systems).
Results A total of 289 women were included: 219 (76%) used the Gynecare prolapse repair mesh devices, while 70 women (24%) used the American Medical Systems prolapse repair mesh devices. Operative complications included: bladder injury (1.6%), rectal injury (1.1%) and two women with serious vascular injuries. Postoperative complications included: buttock pain (5.2%), vaginal erosion (10%), one woman with bladder erosion and two women (0.7%) with serious infection, leading to necrotising fasciitis in one woman. Short-term cure rates in different groups varied from 94 to 100%, depending on vaginal compartment and device used. In total 15 women (5%) had persistent prolapse at 3-month follow up.
Conclusion The new prolapse repair mesh devices demonstrate excellent short-term cure and low morbidity rates. However, some complications are serious and require highly specialised management. Rigorous evaluation by means of independent, controlled studies is urgently required.
The lifetime risk of surgery for pelvic organ prolapse (POP) is estimated at 11%, with 29% receiving a second operation within 5 years.1 Pelvic floor repair involving anterior and/or posterior colporrhaphy has been the established treatment for POP over the last century, but carries a high rate of recurrence.1 Over the last decade, mesh augmented surgical repair is being increasingly used, mainly for secondary procedures. The use of synthetic mesh has become increasingly popular due to the fact that no autologous material is required and that the mesh materials are not biodegradable. However, reliable long-term data are lacking.
Recently a number of prolapse repair mesh devices have been designed by different companies and extensively marketed as a minimally invasive approach for pelvic floor repair. Manufacturers and devices include: Ethicon Women’s Health and Urology, Somerville, NJ, USA (Gynaecare Prolift System); American Medical Systems (AMS) Inc. Minnetonka, MN, USA (Apogee & Perigee) and CR Bard (Avaulta). Each of these prolapse repair mesh devices utilizes woven monofilament low-weight macroporous polypropylene meshes with extension arms to permanently reinforce the pubocervical and the rectovaginal fascia. The extension arms pass through the sacrospinous ligament, prespinous fascia, obturator fascia and muscles to restore the lateral supports of the vagina. Insertion techniques are very similar and involve the use of shaped insertion needles to anchor the mesh arms. The needle passage lies close to vessels, nerves and organs of the pelvis and pelvic side wall, with slight differences between devices as to which structures are closest to the needle tracks.
Unlike medications, the US Food and Drug Administration approval of prosthetic devices is less strictly regulated. The new devices do not require human testing prior to their release,2 and the main prerequisite is to demonstrate ‘equivalency’ to existing devices. Similarly, several mesh repair devices have been introduced into clinical practice without rigorous clinical and surgical evaluation of their safety, success rates or indeed any comparative data to prove their superiority over the standard pelvic floor repair procedures. Moreover, surgeons are under increasing marketing pressure from the industry towards early adoption of this new technology.
The aim of this study was to assess the complication rates and short-term outcome of the recently introduced mesh devices for surgical management of prolapse and to examine the recurrence and re-operation rates. The serious complications encountered are discussed in detail.
Patients and methods
This was a retrospective, multicentre study, which was approved by the local ethical committee. All women who underwent surgical management of POP with any of the mesh devices over the period January 2005 to December 2006 were included. Three centres were included: a tertiary referral urogynaecology centre and two district general hospitals in the West of Scotland. Women were identified from theatre records of the units, and 302 case notes were successfully traced during the study period of March to May 2007. An independent clinician in each unit performed a case notes review with the main outcome measures being the operative, postoperative and mesh-related complications. All relevant surgeons in the three units were contacted and asked to identify cases where complications may have occurred, which led to the procedure being abandoned or completed without the use of a device (i.e. those potentially not recorded as mesh repair procedure in the theatre records). Two further women with organ injuries were identified. The rate of organ injuries was calculated after excluding women with opposite compartment repair. For example, for bladder injuries, we excluded women who had posterior mesh insertion. The secondary outcomes were the short-term cure at 3-month follow up (defined as ≤stage I prolapse as classified on POP-Q3 or Baden Walker4 scoring systems), and re-operation rates for other compartment prolapse and/or urodynamic stress incontinence (USI) up to May 2007.
The two devices mostly used were those manufactured by Ethicon/Gynecare and AMS. Insertion techniques for both products are similar with only subtle differences in the pelvic structures involved. In the Gynecare procedure for posterior prolapse, the trocar is inserted 3 cm lateral and posterior to the anal orifice, crossing the ischiorectal fossa underneath the ileococcygeus muscle. At a level 2 cm posterior and medial to the ischial spine, the trocar should then pierce the sacrospinous ligament and pass through the predissected pararectal space and emerge in the vaginal incision.5 The AMS device has the same insertion technique with the main difference being a more helical design of the trocar, which pierces the ileococcygeus muscle rather than the sacrospinous ligament at the level of the ischial spine.
In the Gynecare device for anterior prolapse, the superior trocars are inserted through the upper medial angle of the obturator foramen at the level of urethral meatus, while the inferior trocars are inserted 1 cm lateral and 2 cm posterior to the upper arms. All trocars aim to pass through the arcus tendineus and emerge in the vaginal incision. While the superior trocars are passed directly behind the inferior pubic ramus and guided by the surgeon’s finger, the inferior trocars are directed blindly towards the ischial spine where they should pierce the arcus tendineus 1–2 cm above and medial to the ischial spine.5 The same technique is applied in the case of the AMS anterior device but the trocars are again more helical in shape with superior trocars being shorter than the inferior trocars.
The Gynecare mesh can be used for combined anterior and posterior repair, for which the insertion technique is a combination of the anterior and posterior device technique, with the middle part of the mesh lying behind the vaginal vault. There is no equivalent AMS product, although both AMS devices can be used together to achieve the same result.
Descriptive statistics were produced and between-group comparisons were performed using Fisher’s exact test. All analyses were performed using Minitab (version 14) with a significance level of 5%.
A total of 302 case notes were reviewed and 13 women (4%) who used the Bard device were excluded because the numbers were too small to draw any conclusions. In total, 289 women were included: 219 women (76%) used the Gynaecare device; 76 anterior, 70 posterior and 73 combined. Seventy women (24%) used the AMS devices: 32 posterior, 30 anterior and 8 combined. The choice of device used was primarily according to the surgeons’ preference for exclusively two indications: previous failed pelvic floor repair in the same compartment and/or severe prolapse (POP-Q stage III/IV3 or grade 3/4 Baden and Walker4). Table 1 shows the patient demographics, previous prolapse surgery and the associated procedures performed at the time of device insertion. The preoperative severity of POP is shown in Table 2.
Table 1. Patient demographics, previous prolapse surgery and associated procedures at time of mesh repair
BMI, body mass index; HRT, hormone replacement therapy; N/A, not applicable; TAH, total abdominal hysterectomy; TOT, transobturator tape; SSF, sacrospinous fixation.
Number of women
Mean age, years (range)
Mean parity (range)
Mean BMI (range)
28 (22– 38)
Smokers, n (%)
Postmenopause, n (%)
Previous surgery for prolapse/TAH, n (%)
Previous repeated surgery, n (%)
Associated procedures, n (%)
Table 2. Preoperative staging of prolapse and postoperative outcome
B/W grade, Baden Walker grade.
Number of women
Preoperative stage/grade of prolapse, n (%)
B/W grade 4/POP-Q IV
B/W grade 3/POP-Q III
B/W grade 2/POP-Q II
Postoperative stage/grade in same compartment, n (%)
B/W grade 4/POP-Q IV
B/W grade 3/POP-Q III
B/W grade 2/POP-Q II
B/W grade 1/POP-Q I or less
All women (n= 289) attended their first follow-up visit at 10–12 weeks. Twenty-four women (8.3%) initially declined pelvic examination due to various reasons (anxiety, n= 11; pain, n= 4 and undocumented reasons, n= 6) but were re-examined in a further follow up 2–4 weeks later. The short-term cure rate at 3 months was 95% with 274 women showing ≤stage I POP-Q or grade 1 Baden and Walker POP (Table 2). Fifteen women (5%) were considered failures, out of which 10 women (3.4%) had further surgery (Table 3).
Table 3. Management of persistent or other compartment prolapse
Fourteen women (4.8%) developed postoperative USI and all underwent surgical treatment (transobturator tape, n= 10 and colposuspension, n= 4). Twenty-one women (7.3%) re-presented with symptomatic prolapse in another compartment within a period of 4–22 months after the initial operation (mean 8.4 months), out of which 17 (6%) underwent further surgical treatment (Table 3).
Operative complications included three bladder injuries (3/189, 1.6%) and two rectal injuries (2/181, 1.1%) (Table 4). The three bladder injuries occurred during insertion of the superior trocars of an anterior Gynecare mesh device. Each was clinically suspected and confirmed on cystoscopy at the time. In two women, the trocars were re-inserted followed by recheck of cystoscopy and mesh insertion, while in one woman, the surgeon decided to abandon the procedure. In all three women, the bladder was left on free drainage for 7–14 days.
Table 4. Operative, postoperative, and mesh-related complications
The two rectal injuries occurred during the initial dissection of the rectum from the posterior vaginal wall due to presence of extensive scar tissue and therefore they were not related to mesh insertion by any particular device. In both women, the injury was in the lower rectum and was repaired before insertion of the mesh in one woman with no postoperative complications, and a traditional posterior colporrhaphy in the other women.
Six women (2%) had blood loss of over 400 ml, although three of these women were associated with vaginal hysterectomies, which may have contributed to the extra blood loss. In two of these women, it was caused by serious vascular injuries and one woman had a right internal pudendal artery injury following insertion of a posterior Gynecare mesh by a general gynaecologist working in the tertiary centre. On the first postoperative day, the woman’s haemoglobin had reduced from 12.6 g/dl to 7.8 g/dl and clinical examination revealed a large right-sided perineal haematoma extending through the ischiorectal fossa. Vaginal surgical exploration was performed with removal of the mesh and evacuation of a large infralevator haematoma. A puncture of the right internal pudendal artery, 1 cm medial to and at the level of the ischial spine was identified as the source of bleeding. This was secured using two deep stitches into the levator ani muscle and a further stitch into the lateral third of the sacrospinous ligament. The woman had a total blood loss of 3.6 l, received 5 units of blood and was admitted to the intensive care unit for 2 days. She was discharged home 9 days after surgery.
In the second case, the woman had left vaginal artery and right uterine artery injuries following an anterior Gynecare mesh insertion performed by a general gynaecologist in a district general hospital. The woman’s haemoglobin fell from 12.9 g/dl preoperatively to 7.1 g/dl on the first postoperative day although she was haemodynamically stable and an ultrasound scan showed a complex 9 cm diameter pelvic mass. She was transferred to the tertiary centre (14 miles away) for interventional radiology treatment. Bilateral selective angiography of the anterior division of the internal iliac artery demonstrated extravasation from the left vaginal artery (Figure 1a) and the right uterine artery. Both arteries were catheterised using a 2.5 F co-axial catheter and haemostasis was achieved by injection of gelatine sponge slurry (Spongostan film) followed by a 2 mm embolisation coil to occlude the proximal vessel (Figure 1b). The patient was transferred to a high dependency unit for 24 hours and received a total of 4 units of blood. She was transferred back to the referring unit on the third day and was discharged home 10 days after surgery.
Postoperative complications included fever in 24 women (8.3%), urinary tract infection in 7 (2.4%), voiding dysfunction (>48 hours) in 26 (9%), perineal haematoma in 3 (1%), vaginal adhesions in 3 (1%), buttock pain in 15 (5.2%), dyspareunia in 13 (4.5%) and vaginal erosion in 30 (10%) (Table 4). Fisher’s exact test showed no significant differences between both devices in all operative and postoperative complications (Table 5).
Table 5. Complications of Gynecare device versus AMS devices
Fisher’s exact test
NS, non significant.
Bleeding >400 ml
Same compartment prolapse
All vaginal erosions were detected on pelvic examination between 6 and 22 weeks postoperatively. All women except one were symptomatic with offensive vaginal discharge (29) and dyspareunia (8). In 28 women, erosions were less than 1 cm in size and were initially managed conservatively with local estrogen and antiseptic treatment, although they all subsequently required partial excision of the eroded mesh. Four women (1.3%) had persistent erosion requiring further excision. Two women (0.7%) had large erosions (4 and 8 cm) and presented with signs of local infection and pelvic pain. Both of them were managed with total removal of the mesh under anaesthesia 8 and 11 weeks postoperatively. The procedures were relatively easy as the infected meshes were not integrated into the surrounding tissues. One patient presented with severe bladder pain and recurrent urinary tract infections at 3 months postoperatively following combined anterior and posterior Gynecare mesh. Cystoscopy showed a few mesh fibres at the base of the bladder just distal to the trigone suggesting mesh erosion (Figure 2). The mesh was morbidly adherent to the bladder base and the anterior vaginal wall and could not be excised at cystoscopy. This patient required laparotomy and partial cystectomy.
Serious systemic infection occurred in two women (0.7%), both within 2 weeks of the operation. One case of systemic infection with Staphylococcus aureus occurred following an anterior Gynecare mesh insertion and was treated by intravenous antibiotics and total surgical removal of the mesh under general anaesthesia. The second woman developed necrotising fasciitis following vaginal hysterectomy and posterior Gynecare mesh insertion. She was readmitted 12 days after surgery with significant perineal pain, foul smelling vaginal discharge, mild lower abdominal pain, fever of 38°C and rigours. Two necrotic areas were noted in the perineal region with leakage of serous fluid. Pelvic computed tomography scan confirmed the diagnosis of necrotising fasciitis (Figure 3). Her general condition rapidly deteriorated and she was resuscitated and transferred to intensive care, from where she required extensive perineal debridement (Figure 4), laparotomy, end colostomy and extensive wound packing left in situ. She remained in intensive care for 2 weeks with considerable input from the tissue viability specialists. She was then transferred to the plastic surgery unit 32 days after the initial procedure where she subsequently underwent plastic perineal reconstruction. Detailed management of this patient will be a subject of further report.
The high success rates, lower morbidity rates and the wide popularity of the tension free vaginal tapes6 have revolutionised the concepts of urognaecological surgery. The simplicity and relative safety of the recently described transobturator tapes7,8 have undoubtedly led to a growing interest, both by the surgeons and the industry, for using the same approach for pelvic floor repair despite the blind insertion techniques. The blind insertion of the trocars into the obturator foramen, ischiorectal fossa, ileococcygeus muscle and the sacrospinous ligament exposes the patient to the potential injury to the rectum, bladder, inferior gluteal vessels, pudendal nerve and vessels, and the sciatic nerve. The anterior division of the internal iliac artery gives many branches as it enters the pelvis (Figure 5) and apart from the superior gluteal and the obturator, all the other branches pass medially in various courses and therefore are in theoretical risk of injury during insertion of the trocars. Detailed management of this patient will be a subject of further report.
In this study, the total short-term cure rate was excellent at 95% and was comparable to other studies by Fatton et al.9 and Flam10 who showed 95 and 100% cure rates, respectively, at 3-month follow up in 106 women and 55 women who underwent repair with Gynecare mesh.
In this study, the rate of women who presented with persistent/recurrent prolapse in the same compartment is low at 5% and only 3.5% required further surgery. The total reoperation rate was 13%, but this was mainly for prolapse of another compartment (6%) or unmasked stress incontinence (3.5%). It is unlikely that women involved in this study would have undergone further surgery elsewhere (as the regional tertiary referral centre was involved in the study) although these rates may be an underestimate as some women had only reached 3 months of follow up. The figures also exclude those with asymptomatic recurrence after 3 months.
The initial report of this new mesh repair technique by those who developed it, reported data from 277 women and showed low perioperative morbidity with one rectal injury (0.4%), four bladder injuries (1.5%) and one significant haematoma.11 The same group later reported a multicentre retrospective study9 of 110 women and showed 1% bladder injury rate and 2% significant haematoma rate. Flam10 reported bladder injuries in 3.6%, no rectal injuries and no cases of >150 ml blood loss. Our study has demonstrated a comparable rate of bladder injuries (1.6%), however the authors now recommend cystoscopy after insertion of all trocars during these anterior procedures. Some surgeons may argue that cystoscopy is not routinely required during transobturator tape procedures and so may not be required during anterior mesh insertion. However, we have shown previously that women with significant prolapse and those with previous pelvic floor surgery are at higher risk of bladder injury at the time of transobturator tape insertion and recommended cystoscopy to this subgroup.12 It is important to note that sites of the trocar insertion are slightly different in anterior mesh repair devices with the superior trocars being inserted at the level of the urethral meatus rather than the level of the clitoris and therefore are closer to the bladder during insertion.
In the case of internal pudendal artery injury, the authors believe that the trocar of the posterior Gynecare device has been passed through the sacrospinous ligament higher and lateral than recommended by the original authors.5 Cadaver studies have shown considerable anatomical variation in the relationship of the sacrospinous ligament to the adjacent nerves and vessels.13–15 However, most of these studies agreed that the average length of the sacrospinous ligament is 4–5 cm and that the inferior gluteal vessels emerge 1.5–2 cm above and medial to the ischial spine and courses close to its upper-lateral half while the pudendal nerve and vessels emerge 1 cm (range 0.5–3.3 cm) medial to the ischial spine and underneath the sacrospinous ligament. The inferior gluteal artery was found to be the most susceptible vessel during sacrospinous vaginal suspension procedures.16 In view of the above studies and our experience, we now emphasise the importance of inserting the posterior Gynecare trocars into the sacrospinous ligament 3 cm medial and 2 cm lower than the level of ischial spine to avoid the pudendal and inferior gluteal bundles. The posterior AMS trocars pass through the ileococcygeus muscle rather than the sacrospinous ligament and, therefore, have a lower risk of injuring the above structures; however, this may be at the expense of the strength of support provided.
In the case where the vaginal and uterine vessels were injured, it seems that the inferior trocars of the anterior Gynecare mesh were inserted quite deep and posterior in an attempt to pass through the arcus tendineus. However, in some women especially the elderly postmenopausal women, the arcus tendineus can be quite atrophic and difficult to identify. It is our experience that in such cases we pass the inferior trocars through the levator ani fibres.
Little information is available in the literature regarding the management of these women; however, it is known that the surgical ligation of the internal iliac artery is unlikely to control the bleeding due to the extensive anastomosis in this region.16 Multiple approaches might be needed in these women including packing, vascular clips and selective arterial embolisation. The later has been shown to be safe and effective in the management of severe perioperative bleeding following urogynaecology surgery.17–21 Angiography allows complete mapping of the pelvic vessels to identify the bleeders with the benefit of avoiding general anaesthesia, further major surgery and therefore shorter recovery time. However, it is important to emphasise that selective arterial embolisation will not always be successful and is not readily available in all units due to financial implications or geographical restrictions, therefore alternative strategies and surgical expertise must be in place to deal with such cases.
Infection can occur in any type of surgery, however the presence of mesh can impair the ability of the body defence mechanisms to eradicate the infection. The biomechanical properties of the mesh material and pore size have a major influence on the postoperative tissue healing and consequently mesh infection. However, both AMS and Gynecare devices use monofilament low-weight macroporous polypropylene meshes and therefore they fulfil the criteria of type I mesh according to Amid’s classification22 and are associated with good tissue incorporation and low rates of infection. In almost all the women we encountered or were referred to us, the infection was mild and localised. This type of localised infection is usually secondary to inadequate healing and respond to conservative measures such as systemic antibiotics, local antiseptic treatment with or without limited excision of the exposed portion of the mesh.11,23 Systemic signs of infection should direct us towards the possibility of a progressive infection of the mesh material, a rare yet a far more serious complication. It indicates a pelvic infection along the mesh and beyond, and warrants a complete and rapid removal of the mesh to avoid serious life-threatening complications.
The vaginal erosion rate (10%) in our study is comparable to the 12.7% reported by Collinet et al.11, yet significantly higher than the recently reported 4.7% by Fatton et al.9 The conservative management of these women was disappointing and all women proceeded for partial/total excision of the mesh. One further case of bladder erosion occurred and had the typical presentation of urinary tract infection (UTI) and bladder pain. The site of the polypropylene fibres at the base of the bladder rather than the lateral sides and the pattern of fibres (Figure 2) coincide with erosion rather than a missed bladder perforation. This would have been confirmed however, if a cystoscopy had been performed at the time of operation and excluded any operative injury.
The new mesh repair devices are associated with excellent short-term cure rates and low morbidity rates considering the surgically high-risk population involved. However, some of these uncommon complications are serious and can be life threatening and need highly specialised management. We, therefore, recommend that these procedures should only be performed by specialists with detailed knowledge and training in the anatomy of the pelvic sidewalls and in well-equipped units especially with facilities for interventional radiology. Cystoscopy should be considered in all cases involving anterior mesh repair kits. Randomised trials comparing these mesh repair kits to the established repair procedures are urgently needed before these procedures can be used in routine clinical practice.
The authors did not receive any external funding for this study.
Several members of the study group have received funding from pharmaceutical and medical devices companies to attend scientific meetings and medical conferences.
Contribution to authorship
M.A-f designed the study, obtained the ethics approval, liaised with all the centres, performed number of the procedures, supervised the data collection, performed the data analysed and wrote the paper. I.R. designed the study, supervised the data collection and edited the paper. The West of Scotland Study Group include those who have performed the procedures and those who have collected the data retrospectively.
The authors would like to thank Dr Rachel Connor (Consultant Radiologist) and Dr Jon Moss (Consultant Radiologist) for their support. Also, we thank the medical illustration departments for their efforts in supplying the figures in this article.