Journal of Colon and Rectal Cancer
ISSN: 2471-7061
Current Issue
Volume No: 1 Issue No: 1
share this page

Research Article | Open Access
  • Available online freely | Peer Reviewed
  • Primary Abdominal Wall Reinforcement with Synthetic Mesh Following Harvesting of Vertical Rectus Abdominis Myocutaneous Flaps in Multivisceral Pelvic Resections

    Evita Zoucas 1       Caroline Bjarnevik 2     Kristina Nätterlund 2    

    1Colorectal Section, Division of Surgery and Urology, Skane University Healthcare, SE-205 02 Malmö, Sweden.

    2Department of Clinical Sciences, Lund University, SE-221 85 Lund, Sweden.

    Abstract

    Following multivisceral pelvic resections, the pelvis and perineum are often reconstructed using myocutaneous flaps. Abdominal wall defects after harvesting rectus abdominis flaps can be reinforced with mesh. Primary reconstruction using synthetic mesh was presently evaluated. Fifty-eight patients who underwent multivisceral pelvic resection and perineal reconstruction with a vertical rectus abdominis myocutaneous (VRAM) flap, January 2004 to February 2014, were retrospectively reviewed. The abdominal wall was reinforced in 26. Demographics, treatment procedures, surgical procedures, length of hospital and ICU stay, early and late morbidity at the recipient and donor sites were recorded. Patients with mesh reinforcement were significantly younger than those without. There were no further significant differences in patient demographics or treatment procedures between the two groups. In 31% of the patients with mesh, surgery was performed on two consecutive days, although total operating time did not differ significantly. Patients without mesh bled more. Surgery was associated with considerable morbidity, without significant differences in overall complication rate between the two groups. At the recipient site, wound infection/dehiscence was the most common early complication. The group with mesh had higher rate of total flap necrosis necessitating re-operation. At the donor site, wound infection /dehiscence, hernia, or bulge were recorded. Patients with mesh had lower rates of donor site morbidity. Perineal reconstruction with VRAM flap and primary abdominal wall reinforcement with mesh is feasible after multivisceral resection. Our study indicates that primary use of mesh can be applied in potentially contaminated surgical fields in oncologic patients without increasing morbidity and with improved long-term cosmetic results.

    Received 01 Apr 2015; Accepted 01 Dec 2015; Published 05 Dec 2015;

    Academic Editor:Pakula A, Department of Surgery and Surgical Critical Care. Kern Medical Center, Bakersfield, CA, USA.

    Checked for plagiarism: Yes

    Review by: Single-blind

    Copyright©  2015 Evita Zoucas, et al.

    License
    Creative Commons License    This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

    Competing interests

    The authors have declared that no competing interests exist.

    Citation:

    Evita Zoucas, Caroline Bjarnevik, Kristina Nätterlund (2015) Primary Abdominal Wall Reinforcement with Synthetic Mesh Following Harvesting of Vertical Rectus Abdominis Myocutaneous Flaps in Multivisceral Pelvic Resections. Journal Of Colon And Rectal Cancer - 1(1):35-44.
    Download as RIS, BibTeX, Text (Include abstract )
    DOI10.14302/issn.2471-7061.jcrc-15-661

    Introduction

    Multivisceral pelvic exenteration often represents the only option for sustainable treatment of patients with primary or recurrent advanced pelvic malignancy 1, 2. Musculocutaneous flaps are commonly used to fill the resulting intra-abdominal void created by the exenteration and to reconstruct the pelvic floor and perineum. VRAM flaps weaken the abdominal wall, which can be reinforced using mesh 3, 4. Due to potential contamination of the surgical field, there has been considerable controversy regarding implantation of foreign material, primarily, in the abdominal wall 5.

    The aim of the present study was to ascertain whether primary reconstruction of the abdominal wall with mesh after pelvic exenteration and harvesting of VRAM flaps is associated with increased risk of complications. An additional objective was to determine whether patient demographics or treatment procedures could identify specific indications for use of abdominal wall reinforcement in this setting.

    Patients and Methods

    We conducted an observational retrospective cohort study of all patients at our hospital who underwent extended perineal resection for pelvic malignancy and received a VRAM flap for perineal reconstruction. Prior to the surgery, the patients were given radiotherapy (50–60 Gy) in the lower pelvis and perineum and concomitant chemotherapy. Surgery was planned five weeks later and was performed on average 41 days ( range 38-46 d ) after neoadjuvant treatment. A total of 58 patients were operated between January 2004 and January 2014. No reinforcement of the flap donor site was done during the period 2004–2009 (32 patients), whereas from 2009 to 2014 the abdominal wall was reinforced using synthetic mesh (26 patients).

    All in- and out-patient medical records were systematically reviewed, and data on the two

    groups (i.e., those with and those without mesh reinforcement) were analyzed. Clinical data on patient demographic characteristics, comorbidity status, and tumor histology are shown in Table 1. There were fewer female patients in the group without mesh reinforcement. The patients with mesh were significantly younger, and one fourth of them were receiving immunosuppressive therapy other than the oncological chemotherapy. There were no differences in BMI or comorbidity between the two groups. Tumors were either adenocarcinomas of the rectum or anal squamous cell carcinomas.

    Table 1. Demographics of the 58 patients who received VRAM flaps for abdominal wall reconstruction with and without mesh reinforcement
        No mesh n = 32 Mesh n = 26  
    Gender Female n (%) 18 (56) 20 (77)  
    Age (years) Median (range) 67 (50–83) 63 (32–79) p < 0.05
    BMI Median(range) > 30 (n) 23.7 (15.6–36.0)3 24.0 (19.4–36.6)4  
    Comorbidity (n)        
      Lung disease 4 2  
      Heart disease 5 9  
      Kidney disease 1 1  
      CVI 5 2  
      Diabetes 2 1  
    Smoking (n)   18 11  
    Immunosuppression (n) SteroidsOther 20 52  
    Histological tumor type (n) Adenocarcinoma 25 16  
    Squamous cell carcinoma 7 10  

    Fisher’s Exact Probability Test

    Surgical Techniques

    Operations were conducted as one- or two-session procedures depending on the duration of the oncological part of the surgery and the expected duration of the reconstruction. In two- session procedures, the first stage (day 1) included removal of the tumor, and the second stage (day 2) comprised reconstruction of the perineum and reinforcement of the abdominal wall as appropriate. Perineal resection was performed with the patient in prone jackknife position. Patients were kept on a ventilator in the intensive care unit (ICU) during the non- procedural overnight interval.

    In all procedures, general anesthesia was combined with epidural analgesia. When possible, VRAM flaps were rised from the contralateral to the stoma site as described by Tei et al. 6. The flaps consisted of muscle and the overlying adipose tissue and skin, and were raised based on the inferior epigastric artery, including the medial but sparing the lateral perforating vessels. The flap was rotated intra-abdominally to fill the pelvic defect, with the skin paddle oriented vertically, to provide ample tissue for reconstruction of the pelvic floor and the vagina as necessary.

    The abdominal wall at the donor site was reinforced with polypropylene/PVDF mesh (Dynamesh, Fek Textiltechnik mbH, Aachen, Germany) using a variation of the Rives-Stoppa sublay technique 7. The fascial plane between the rectus muscle and the posterior rectus sheath was dissected on either side of the midline incision as far laterally as possible, including the stomas and sparing the lateral perforating vessels and nerves. The posterior rectus sheath and the peritoneum below the arcuate line were re-approximated with running monofilament suture. The mesh was cut to size and anchored in the posterior sheath by interrupted non-absorbable sutures in the periphery and running suture in the midline. The anterior rectus sheath was closed either in the midline or by further running sutures on the mesh to ensure minimal tension. The fatty tissue and skin were approximated with running sutures and clips, respectively. Suction drains were placed subcutaneously and kept in place a median of 60h (range 36- 108h) post operatively.

    All patients were treated with low-molecular-weight heparin and a prophylactic antibiotic regimen on the day of the procedure and until drain removal in cases of donor site reconstruction. Patients operated between 2004 and 2009 received intravenous dextran (Macrodex, Pharmacia, Uppsala, Sweden) 1, 3, and 5 days after surgery.

    Surgical data included the type of donor site reconstruction, as well as the type of extended resection, duration of operation, and blood loss. Length of hospital stay, number and type of complications, and unplanned return to surgery were recorded, as were complications at the donor and the recipient site. The complications were categorized as early or late (i.e., occurring within 30 days post-operatively or during follow-up, respectively), and were identified by physical examination and CT scans.

    Post-operative Regimen

    After surgery, the patients were instructed to lie in supine or lateral position on an air-fluidized therapy mattress. Torso elevation of maximum 45 degrees was accepted the first 15 days, and standing and walking, but not sitting, were permitted during the same period. Thereafter, sitting was allowed for gradually increasing lengths of time several times a day. Patients with no reinforcement of the abdominal wall were given an abdominal brace for optional use after discharge from the hospital.

    Follow-up assessments were conducted every 6 months. CT scanning of the chest, abdomen, and pelvis or a CT-PET was performed twice a year for up to 2 years and thereafter once a year for another 5 years. Follow-up data were available for 55 patients. The median follow-up time for late complications at the donor and recipient sites was 12 months (range 7–48 months).

    Statistical Analyses

    Continuous data are presented as median (range). Intergroup comparison was carried out using Fisher’s exact probability test and Levene’s test for equality of variances. Statistical significance was set at a level of 5%. Statistical analyses were conducted using SPSS software (SPSS Inc., Chicago, IL, USA).

    Results

    Table 2 lists the types of surgical procedure used for tumor removal and the preoperative treatment given. The majority of patients had received radiation at the recipient site. There were no significant differences between the patients with and those without abdominal wall reinforcement in terms of the defects created after the various oncological procedures. There were more patients with sacral resection in the groupwithout mesh reinforcement, whereas there were more vaginal resections in the group with reinforcement.

    Table 2. Surgery and preoperative treatment in the 58 patients
        No mesh n = 32 Mesh n = 26
        n n
    Surgery TPE 5 8
      PPE 9 4
      Resection of the bony pelvis 7 4
      APR 9 3
      APR + resection of vagina 2 8
    Preoperative treatment      
      Radiation 1 2
      Chemoradiation 29 23

    Abbreviations:
    APR = abdominoperineal resection;
    TPE = total pelvic exenteration (en bloc resection of the pelvic viscera in the anterior and posterior urogenital compartments); PPE = posterior pelvic exenteration (en bloc resection of the genital viscera and rectum);
    VRAM = vertical rectus abdominis myocutaneous flap.

    Procedures performed without mesh reinforcement led to significantly greater blood loss. However, there was no difference in the duration of the surgical procedures between the two groups, but significantly more patients underwent a two-session procedure in the mesh group (31 versus 6%). Differences between the groups with regard to the number of days of hospital stay were not statistically significant (Table 3).

    Table 3. VRAM flap and abdominal wall reconstruction: details of surgical procedures
      No mesh n = 32 Mesh n = 26  
    Blood loss (mL), median (range) 1500 (350–19000) 1350(200–3000) P < 0.05
    Operation time (min), median range) 725 (365–1800) 690 (445–1100)  
    Two-session procedure, n (%) 2 (6) 8  (31) P < 0.05
    Hospital stay (days), median (range) 28 (14–77) 33 (16–110)  

    Fisher’s exact probability test

    Table 4 shows early and late morbidity after VRAM flap and abdominal wall reconstruction. The procedures were associated with considerable morbidity, although there was no statistically significant difference in overall complication rate between the two groups.

    Table 4. Morbidity after VRAM flap and abdominal wall reconstruction in the 58 patients with pelvic tumors
      No mesh n = 32   Mesh n = 26    
        n (incidence %)     n (incidence %)  
    Early complications (< 30 days) 23 (71.8) 15 (57.6)  
    Late complications (> 30 days) 18 (56.2) 8 (31.0)  
    ICU 3 (9.3) 11 ( 42.3) P < 0.05

    Fisher’s exact probability test

    Patient demographics were not associated with the frequency of complications neither at the donor nor the recipient site. (Levene´s test). Larger numbers of patients with mesh reinforcement required ICU care.

    Morbidity at the recipient site is outlined in Table 5. Wound infection/dehiscence at this site was the most common early complication. Total flap necrosis necessitating re-operation occurred in two patients in the group without mesh and in five patients in the group with mesh. Late complications observed at follow-up included delayed wound healing and discrepancy in flap size. There was no statistically significant difference in recipient-site morbidity between the two groups.

    Table 5. Recipient-site morbidity after VRAM flap and abdominal wall reconstruction in the 58 patients with pelvic tumors.
        No mesh n = 32   Mesh n = 26  
        n (incidence %)   n (incidence %)
    Early complications (< 30 days) Partial flap ischemia 5 (15.6) 4 (15.4)
      Flap necrosis 2 (6.3) 5 (18.0)
      Wound infection/ dehiscence 9 (28.0) 5 (18.0)
      Reoperation 3 (9.3) 5 (18.0)
    Late complications (> 30 days) Delayed healing 6 (18.7) 5 (18.0)
      Size incongruence 1 (3.1) 3 (11.1)

    Donor-site complications were wound infection with or without dehiscence (early), and hernia or bulge (late), as shown in Table 6. One patient without mesh reinforcement was returned to the operating room due to wound infection, and another was re-operated for an incarcerated hernia at the stoma. One patient with mesh required secondary suture of the abdominal incision. The bulge rate was significantly higher in patients without mesh. No mesh had to be removed due to any complication.

    Table 6. Donor-site morbidity after VRAM flap and abdominal wall reconstruction in the 58 patients with pelvic tumors
        No mesh n=32    Mesh n=26   
        n ( incidence %)   n ( incidence %)    
    Early complications(< 30 days) Wound infection/dehiscence 7 (21.8) 1 (3.8)  
      Re-operation 2 (6.3) 1 (3.8)  
    Late complications(> 30 days) Hernia 8 (25.0) 2 (7.7)  
      Bulge 7 (21.8) 0 P < 0.05
      Delayed healing 1 (3.1) 3 (11.5)  

    Fisher’s exact probability test

    Discussion

    We studied patients with tumors of the rectum and anus to investigate the use of pedicled VRAM flaps for reconstruction of pelvic floor defects and implantation of non-absorbable mesh for reinforcement of the abdominal wall. The data we obtained show that harvest of rectus abdominis muscle was not associated with undue major morbidity at either the donor or the recipient site. The use of synthetic mesh in the primary closure of the abdominal wall after colorectal procedures did not increase the incidence of wound infection and in

    stead led to fewer hernias and better body contour in the long run. However, our results did not identify a specific group of patients that would probably benefit from simultaneous perineal and abdominal wall reconstruction.

    VRAM flaps have been used in association with removal of advanced pelvic tumors involving wide perineal resection 8, 9. The resulting large pelvic and perineal defects create wounds in irradiated, poorly vascularized tissue that are prone to infection and delayed healing 3, 10, 11, 12. As shown in earlier investigations 9, 13, 14, a VRAM flap introduces well-vascularized tissue that fills the pelvic defect, providing structural support for internal organs and interposing healthy skin in the perineal wound. Maricevich et al. 15 demonstrated that VRAM flaps represent a reliable option for reconstruction of the large-volume defects that arise after sacrectomies. In the series evaluated by those authors, skin dehiscence was 25% at the recipient site and 5.4% at the donor site, whereas wound infection was 9.6% and 8.1%, respectively. Preoperative chemotherapy and radiotherapy did not affect early morbidity. Sunesen et al. 13 observed no major perineal wound complications and no major wound infections when using VRAM flaps after pelvic salvage surgery. Nevertheless, in contemporary studies of large cohorts 10, 16, VRAM flaps were found to be associated with 11% major perineal wound dehiscence requiring secondary sutures or vacuum dressings. Also, in yet another study 17, it was demonstrated that wound healing was adversely affected by overweight, diabetes, and malignancy, but not by previous irradiation of the recipient site.

    The above-mentioned findings are corroborated by our results showing a high frequency of wound infection at the recipient site in the early post-operative period and also delayed perineal healing. We observed major morbidity requiring re-operation in 12% of all cases. Early complications were apparently associated with flap ischemia or necrosis, indicating that irradiation was not a significant contributing factor. We did not find any correlations between complications and BMI, comorbidity, or smoking, possibly due to the small numbers of patients in the studied cohort.

    The risk of morbidity at the donor site has been a drawback of the use of VRAM flaps. Improper healing and loss of muscle tissue can potentially cause abdominal wall incompetence or hernia, and several studies have noted fascia wall dehiscence and herniation ranging from 9% to 17% 6, 8, 13, 18. However, another investigation revealed no increase in complications related to the VRAM flap donor site 3.

    In the present study, we observed a considerable number of wound infections at the VRAM flap donor site in patients without abdominal wall reconstruction, although the majority of those infections were minor and did not require re-intervention. This group also had longer operation time and increased peri-operative bleeding, which might account for a higher risk of contamination. Patients who underwent reconstruction with mesh received an extra dose of antibiotics prior to mesh implantation, and treatment was continued with a broader spectrum prophylactic regimen several days peri-operatively, which might have affected the bacterial flora at the site of the wound. A significantly higher number of patients with mesh reinforcement had undergone two-session procedures, but no adverse effects on wound healing were apparent in this group compared to the early morbidity noted in the group without reconstruction. Herniation and bulging accounted for almost half of the late complications in the patients without abdominal wall reconstruction, which is higher than the level reported by other investigators 10, 17. No late morbidity due to incarceration was observed, and the abdominal brace that was provided was not used continuously by many of the patients.

    Primary reinforcement of the abdominal wall with prosthetic mesh was introduced to circumvent abdominal wall deficiency during the later part of the present study period. In colorectal surgery, primary closure of the abdominal incision using non-absorbable mesh is still controversial due to potential contamination of the surgical field and an associated increase in the risk of infection 3, 19. To our knowledge this is the first report comparing specifically the use of synthetic mesh for primary abdominal wall reinforcement after harvesting of the rectus muscle in pelvic cancer procedures. Although mesh reinforcement did not prolong operating time, significantly more patients with implanted mesh had two-session procedures and had to remain intubated in the ICU overnight. Furthermore, three patients with mesh were kept in the ICU for hemodynamic stabilization. Also, hospital stay was longer in patients with mesh than in those without such reinforcement, although this difference was not statistically significant.The patients were fully mobilized and could sit uninhibited before discharge which resulted in rather long hospital stay periods. Our data suggest that non-absorbable mesh can be applied without increasing wound infection or dehiscence. Even though the patients with mesh had undergone neoadjuvant chemoradiotherapy, and 27% of them were receiving systemic immunosuppressive therapy, there was no apparent abscess formation, and it was not necessary to explant any of the meshes. Long-term outcomes included significantly better body contour, as indicated by the absence of wall deficiency.

    In a recent review of materials used for abdominal wall reinforcement, Lee et al. 20 noted that a wound infection rate of 6.4% and a hernia rate of 3.2% were associated with synthetic non-absorbable prosthetics in clean-contaminated cases, which confirms the present results.

    We did not find any correlation between the rate of complications and comorbid conditions in our patients.

    The present study had several limitations, one of which was the retrospective design with a relatively small number of patients. Also, there is an inherent bias in the outcomes, because the two groups of patients underwent surgery at different times with variable oncological procedures. Furthermore, the patients were not randomly assigned to receiving a VRAM flap or having abdominal wall reinforcement, and we did not evaluate outcomes with regard to restriction of abdominal wall strength.

    Conclusion

    In conclusion, primary implantation of synthetic mesh appeared to be safe even in the current potentially contaminated field. Nevertheless, it should be noted that implantation of mesh led to operative procedures extending over two days for one third of the current patients, and the patients without abdominal wall reinforcement did not use an abdominal brace. Accordingly, further studies should be performed that include evaluation of abdominal wall functionality and measurement of quality of life in order to identify patients that can clearly benefit from abdominal wall reinforcement after harvesting of VRAM flaps.

    Acknowledgments

    The authors wish to thank Mr Lars Wallgren, MSc, Lecturer, Dept of Statistics, Lund University, Lund, Sweden, for his contribution in the statistical analysis and Ms Monica Keidser ,Dept of Clinical Science, Lund university, Lund, Sweden for excellent secretarial assistance.

    References

    1.Zoucas E, Frederiksen S, Lydrup M-L, Månsson V, Gustafson P. (2010) Pelvic exenteration for advanced and recurrent malignancy| World. , J. Surg 34, 2177-2184.
    2.Ferenschild F T, Vermaas M, Verhoef C, Ansink A C, Kirkels W J. (2009) Total pelvic exenteration for primary and recurrent malignancies World. , J. Surg 33, 1502-1508.
    3.Chessin D B, Hartley J, A M Cohen, Mazumdar M, Cordeiro P. (2005) Rectus flap reconstruction decreases perineal wound complications after pelvic chemoradiation and surgery : a cohort study. , Ann. Surg. Oncol 12, 104-110.
    4.Patel K, Shuck J, Hung R, Hannan L, Nahabedian M. (2014) Reinforcement of the abdominal wall following breast reconstruction with abdominal flaps: a comparison of synthetic and biological mesh. , Plast. Reconstr. Surg 133(3), 704-707.
    5.Kurmann A, Barnetta C, Candinas D, Beldi G. (2013) Implantation of prophylactic nonabsorbable intraperitoneal mesh in patients with peritonitis is safe and feasible. , World. J. Surg 7, 1656-1660.
    6.Tei T M, Stolzenburg T, Buntzen S, Lauberg S, Kjeldsen H. (2003) Use of transpelvic rectus abdominis musculocutaneous flap for anal cancer salvage surgery. , Br. J. Surg 90, 575-580.
    7.Mc Lanahan D, King L T, Weems C, Novotney M, Gibson K. (1997) Retrorectus prosthetic mesh repair of midline abdominal hernia. , Am. J. Surg 173, 445-449.
    8.C E Butler, Gundeslioglu Ö, Rodriguez-Bigas M. (2008) Outcomes of immediate vertical rectus abdominis myocutaneous flap reconstruction for irradiated abdominoperineal resection defects. , J. Am. Coll. Surg 206, 694-703.
    9.F W Buchel, Finical S, Johnson C. (2004) Pelvic reconstruction using vertical rectus abdominis musculocutaneous flaps. , Ann. Plast. Surg 52, 22-26.
    10.R E Horch, Hohenberger W, Eweida A, Kneser U, Weber K. (2014) A hundred patients with vertical rectus abdominis myocutaneous (VRAM) flap for pelvic reconstruction after total pelvic exenteration. , Int. J. Colorectal Dis 29, 813-823.
    11.B C Paun, Cassie S, A R MacLean, Dixon E, Buie W D. (2010) Postoperative complications following surgery for rectal cancer. , Ann. Surg 251, 807-818.
    12.Nilsson P, Svensson C, Goldman S, Glimelius B. (2002) Salvage abdominoperineal resection in anal epidermoid cancer. , Br. J. Surg 89, 1425-1429.
    13.K G Sunesen, Buntzen S, Tei T. (2009) Perineal healing and survival after anal cancer salvage surgery: 10-year experience with primary perineal reconstruction using the vertical rectus abdominis myocutaneous (VRAM) flap. , Ann. Surg. Oncol 16, 68-77.
    14.S W Bell, Dehni N, Chaouat M, J C Lifante, Parc R. (2005) Primary rectus abdominis myocutaneous flap for repair of perineal and vaginal defects after extended abdominoperineal resection. , Br. J. Surg 92, 482-486.
    15.Maricevich M, Marcevich R, Chim H, S L Moran, P S Rose. (2014) Reconstruction following partial and total sacrectomy defects: an analysis of outcomes and complications. , J. Plast. Reconstr. Aesthet. Surg 67, 1257-1266.
    16.T A Creagh, Dixon L, Frizelle F A. (2012) Reconstruction with Vertical Rectus Abdominus Myocutaneous flap in advanced pelvic malignancy. , J. Plast. Reconstr. Aesthet. Surg 65, 791-797.
    17.Daigeler A, S-Belyaeva M, Drucke D, Goertz O, Hirsch T. (2011) The versatility of the pedicled vertical rectus abdominis myocutaneous flap in oncologic patients. , Langenbecks Arch. Surg 396, 1271-1279.
    18.J H Lefevre, Parc Y, Kernéis S, Shields C, Touboul E. (2009) Abdomino-perineal resection for anal cancer: impact of a vertical rectus abdominis myocutaneus flap on survival, recurrence, morbidity, and wound healing. , Ann. Surg 250, 707-711.
    19.Breuing K, Butler C E, Ferzoco S, Franz M, Hultmann C S. (2010) Incisional ventral hernias: review of the literature and recommendations regarding the grading and technique of repair. , Surgery 148, 544-558.
    20.Lee L, Mata J, Landry T, Khwaja K, Vassiliou M. (2014) A systematic review of synthetic and biologic materials for abdominal wall reinforcement in contaminated fields. , Surg. Endosc 28, 2531-254.