- Email: [email protected]
Use of human acellular dermal matrix for hernia repair: Friend or foe?
Use of human acellular dermal matrix for hernia repair: Friend or foe? Raymond Candage, MD, Keith Jones, MD, Fred A. Luchette, MD, James M. Sinacore, PhD, Darl Vandevender, MD, and R. Lawrence Reed II, MD, Maywood, Ill
Background. Surgeons continue to search for the ideal prosthetic material to repair complex abdominal wall hernias. Recently, a new biologic material was introduced into the surgeon’s arsenal. The purpose of this study is to review a single institution’s experience with the use of human acellular dermal matrix (HADM [AlloDerm]) for repair of hernias. Methods. This was a retrospective review of all patients who received HADM for repair of an abdominal wall hernia. Patient demographics, comorbidities, wound contamination, operative technique, complications, and hernia recurrence were analyzed. Results. Between May 2004 and October 2007, HADM was implanted in a total of 46 patients undergoing repair of a ventral hernia. The average age was 54 years (range, 26--77), with an average American Society of Anesthesiologists classification of 2.5 (range, 1--4). Indications for use of HADM included complex ventral hernia repair (n = 34), mesh infection/enterocutaneous fistula (n = 10), and peritonitis (n = 2). The incidences of comorbidities were hypertension in 47%, diabetes mellitus in 16%, and coronary artery disease in 11%. The majority (87%; n = 40) of the procedures were performed on an elective basis. Seventeen procedures were performed in contaminated wounds. The HADM was placed as reinforcement to the hernia repair in 26 patients and as a ‘‘bridge’’ between the fascial edges in 20 patients. The average follow-up was 12.1 months. Wound complications were frequent at 54%. There were 6 recurrent hernias and 8 patients with eventration of the bioprosthesis so that the recurrent hernia rate was 30%. None of the recurrences were associated with a postoperative wound infection. The majority (88%) of patients who developed eventration of the HADM had a repair using the bioprosthesis to ‘‘bridge’’ an abdominal wall defect. Hernia recurrence and eventration were not associated with use of HADM in a contaminated/infected wound. Conclusions. HADM is a suitable prosthesis for repair of complex and routine abdominal wall defects. This bioprosthesis can incorporate into contaminated tissue without becoming infected. Eventration occurs when HADM is utilized as a fascial replacement rather than as a reinforcement. (Surgery 2008;144:703-11.) From the Divisions of General Surgery and Plastic Surgery, Department of Surgery, Stritch School of Medicine, Loyola University Medical Center, Maywood, Ill
APPROXIMATELY 2 MILLION ABDOMINAL OPERATIONS are performed in the United States each year. The incidence of postoperative incisional hernia ranges from 3% to 20% and represents a tremendous financial burden to the health care system. As compared with primary repair, the introduction of prosthetic mesh appliances for the repair of elective ventral hernias has had a significant impact on reducing the rate of hernia recurrence. However, the use of synthetic mesh has its own set of
Accepted for publication June 19, 2008. Reprint requests: Fred A. Luchette, MD, Department of Surgery, 2160 South First Avenue, Maywood, IL 60153. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2008 Mosby, Inc. All rights reserved. doi:10.1016/j.surg.2008.06.018
complications, including bowel obstruction, seroma formation, fistula formation, and infection. There is near universal agreement that the use of permanent synthetic prosthesis is contraindicated in the setting of contaminated wounds owing to the high rate of infection of the mesh and recurrence of the hernia. Before the development of biologic substitutes, surgeons either implanted an absorbable synthetic mesh or allowed the wound to granulate in and heal by secondary intention with or without an overlying skin graft. Each of these situations predisposes the patient to a staged repair and adds considerable morbidity to what is already a daunting surgical procedure. The development of human acellular dermis (Alloderm; LifeCell Corporation, Branchburg, NJ) as well as other biologic substitutes, allows surgeons to achieve definitive abdominal wall reconstruction in a single staged procedure in the SURGERY 703
704 Candage et al
setting of contaminated abdominal wall defects. Human acellular dermal matrix (HADM) is a cadaveric dermal graft (bioprosthesis) in which the epidermis has been separated and all of the cellular elements in the dermis that can lead to tissue rejection and graft failure are removed. This biologic material is thought to act as a scaffold that permits tissue regeneration with revascularization and repopulation of cellular elements. The material is entirely degraded and replaced with host collagen over time after implantation. Although numerous reports in the literature demonstrate the safety and efficacy of HADM for use in the setting of contaminated ventral hernias, there are not enough data or long-term reports to support a universal consensus for its use. A few reports in the literature comment on the rate of recurrent hernia and an undesirable property of HADM described as laxity of the material or its ability to stretch over time. This unique characteristic is also described as eventration of the bioprosthesis and is considered to be a recurrent hernia.1-3 The purpose of this study was to identify factors that contribute to the rate of recurrence for hernias repaired with HADM specifically when used in a nonsterile surgical wound. METHODS The Loyola University Medical Center Institutional Review Board approved this study. All patients undergoing hernia repair with HADM from May 2004 through October of 2007 were reviewed. Patient demographics including age, gender, comorbidities, American Society of Anesthesiologists (ASA) classification, preoperative plasma concentration of albumin, indication for repair, and number of previous abdominal operations were recorded. Perioperative data included wound classification (clean/contaminated), operative technique, skin closure status, elective versus emergent nature of the procedure, and placement of the mesh in an open versus closed abdomen at the time of operation. For study purposes, surgical wounds were considered to be contaminated actively in the following clinical settings: in the presence of an infected synthetic mesh, enterocutaneous fistula with hernia, feculent peritonitis (perforation or anastomotic leak), necrotizing fasciitis, and intraabdominal abscess. Operative wounds were considered to be potentially contaminated at the time of HADM implantation with any of these scenarios: prior history of feculent peritonitis, iatrogenic enterotomy with contamination during an elective hernia repair, peristomal hernia
Surgery October 2008
repair, incarcerated hernia without bowel necrosis or perforation, bowel resection performed at the time of the hernia repair, and fascial dehiscence with evisceration. Complex ventral hernias were defined as recurrent incisional/ventral hernias with or without removal of noninfected mesh, or repair of an initial incisional/ventral hernia with performance of another clean procedure at the time of operation (e.g., nephrectomy). Operative techniques. Various sizes of thick (2 mm) sheets of acellular cadaveric dermis were reconstituted at room temperature in normal saline solution for 20 to 30 minutes before implantation. The material was prestretched before its use and placed in the wound with an appropriate amount of tension. Thirteen surgeons utilized HADM to repair hernias. Details of the operative procedure were determined by a review of the operative report and by direct conversation with the attending surgeon of record where appropriate. Before implantation of the mesh, all wounds were first debrided of foreign body and any nonviable tissue and then irrigated copiously. The HADM was secured to the fascia with one of two techniques---continuous running suture or interrupted mattress sutures using permanent suture material. The running technique attached the bioprosthesis to the posterior rectus sheath fascia around the perimeter of the HADM. The interrupted method used horizontal mattress sutures to secure the bioprosthesis to the fascia using horizontal mattress sutures with the knots tied in the subcutaneous tissue. Placement of the HADM was classified as either bridging a fascial defect or reinforcing the fascial repair. In the bridged group, the bioprosthesis was placed in an intraperitoneal position with a variable degree of fascial underlay (1--4 cm) and then ‘‘bridged’’ a defect in the abdominal wall fascia. For a reinforced repair, the HADM was placed as an underlay, onlay, or a sandwich technique before and/or after primary reapproximation of the fascia. The use of a component separation to facilitate primary closure of the fascia was also recorded. If the hernia defect was larger than the size of a single piece of HADM, individual sheets were sewn together with the use of permanent monofilament polypropylene sutures in an interrupted or running fashion according to surgeon preference. If the incision was left open to avoid a wound infection, the wounds were managed with either vacuum-assisted closure therapy or with 0.9% NaCl-moistened dressings. After discharge, clinic records were reviewed for the presence of wound complications, hernia recurrence, or the presence of eventration. Patients
Candage et al 705
Surgery Volume 144, Number 4
who had not had a clinic visit since July 2007 were contacted by telephone. Although not ideal, as a substitute for a focused clinical examination, the patient was asked if there were any changes in their medical condition with regard to the most recent hernia repair. Patients who reported a subjective increase in their abdominal girth and/or change in waist size were defined as having eventration of the HADM. Patients who noted a significant bulge with pain or who underwent a subsequent procedure to repair a hernia after their initial operation at our institution were defined as having a hernia recurrence. Statistical analysis was performed on the nominal data using chi square analysis and MannWhitney test. Inquiries included the correlation between hernia recurrence and each of the following: wound infection, ASA classification, albumin level, type of surgical repair, patient comorbidities, primary wound closure, and status of the abdomen (open vs closed) at the time of operation. v2 analysis was also used to determine whether there was a correlation between the above variables and eventration of the mesh. Statistical significance was defined as P < .05. RESULTS A total of 53 patients underwent a hernia repair with HADM during the study period. Indications for the use of HADM included complex ventral hernia repair, mesh infection/enterocutaneous fistula with recurrent hernia, umbilical hernia, peritonitis, peristomal hernia, incarcerated femoral hernia, and inguinal hernia (Table I). Patients who underwent umbilical, peristomal, femoral, or inguinal hernia repairs were excluded from the final data analysis. There were 2 deaths during the immediate postoperative period. One patient death was the result of a cerebrovascular event 48 hours after hernia repair. Ten days later, the family decided to withdraw care. The second death was the result of a myocardial infarction with cardiac arrest on postoperative day 1. Forty-eight hours later, the family decided to withdraw care. Given the short time interval from hernia repair to patient death, we were unable to infer any meaningful outcome data for the study and these patients were excluded from the data analysis. Accordingly, there were 46 patients for final data analysis. A total of 15 (32%) patients were contacted by telephone for additional clinical information. Patient demographics are listed in Table II. HADM was utilized in 40 elective operations and in an emergent setting in 6 patients. Seventeen
Table I. Indications for use of human acellular cadaveric dermis Indications
Number of patients
Complex ventral hernia repair Mesh infection/enterocutaneous fistula Umbilical hernia Peritonitis Peristomal hernia Incarcerated femoral hernia Incarcerated inguinal hernia
34 10 3 2 2 1 1
Table II. Patient characteristics ASA classification No. of previous abdominal operations No. of sheets of HADM Albumin Age (yrs)
Mean
Median
2.8 3.0
3.0 3.0
1.7 3.0 56.5
1.0
patients underwent abdominal wall reconstruction in a contaminated field. An additional 13 patients undergoing an elective hernia repair were repaired with HADM owing to a history of significant intraabdominal contamination, open abdomen for damage control in trauma, or enteric contamination. Thirty-eight patients had HADM placed in the setting of a closed abdomen and the remaining 8 patients had HADM placed in an open abdominal wound. A total of 20 patients underwent a bridged repair; 26 patients were repaired using the HADM to reinforce the fascial closure. Patient demographics in both the bridged and reinforced group were similar (Table III). On review of the operative record, the operating surgeon commented on whether the HADM was sutured in running or interrupted fashion in 42 of the 46 patients. The majority of patients (n = 34; 81%) had the HADM sutured using an interrupted technique. Component separation was also performed in 7 patients (27%) in the bridged group and 6 patients (30%) in the reinforced group to facilitate approximation of the fascia. It is reasonable to infer that those defects requiring component separation and a bridged repair were large defects with considerable loss of domain. In contrast, if a reinforced repair was performed with primary fascial reapproximation, the original hernia was likely a smaller defect.
706 Candage et al
Surgery October 2008
Table III. Patient characteristics stratified by placement of HADM
ASA classification No. of previous abdominal operations No. of sheets of HADM Albumin Age
Bridged (n = 20)
Reinforced (n = 26)
2.9 ± 0.5 3.0 ± 2.1
2.6 ± 0.7 2.9 ± 1.7
2.0 ± 1.3 2.9 ± 0.9 54.7 ± 11.2
1.2 ± 0.5 3.1 ± 0.8 58.7 ± 15.4
Values are presented as means ± SD.
Of the 39 patients with a primary closure of the initial surgical wound, 64% (n = 26) developed $1 wound complication. The specific diagnoses included wound infection (n = 9; 23%), seroma (n = 5; 13%), nonhealing wound (n = 5; 13%), enterocutaneous fistula (n = 3; 8%), dehiscence (n = 2; 5%), lymphocele (n = 1; 3%), and hematoma (n = 1; 3%). All the enterocutaneous fistulae occurred in wounds left open at the initial operation. The average follow-up for the study population was 12 months (range, 1--39). There were 6 recurrent hernias diagnosed, and 8 patients developed eventration of the bioprosthesis so that combined, the recurrent hernia rate was 30% (n = 14). The majority of the recurrences were in the bridged group (n = 12; 60%) with an average of 11 months from the time of initial operation to the time of hernia recurrence. Two patients (8%) in the reinforced group had a recurrence. Seven of the 8 patients (88%) who experienced eventration of the bioprosthesis underwent a bridged repair. In contrast, when the bioprosthesis was used to reinforce the fascial repair, only 1 recurrence was attributed to eventration of the HADM. Only 1 of the recurrent hernias occurred in the 16 patients with an infected wound. There were 6 recurrent hernias among the 26 patients with a postoperative wound complication. Review of the operative notes revealed that in 10 of the 14 patients with a recurrent hernia, the surgeon commented on the type of suture technique used to attach the HADM to fascia. An interrupted technique used 60% of the time; in 40% of the repairs, a running suture technique was employed. Given the small sample size, this difference was not statistically significant. When the various data variables were analyzed for a correlation with hernia recurrence, none were identified except for the use of the HADM in a bridged fashion (Table IV). Mann-Whitney testing revealed no correlation between albumin levels or the number of previous abdominal operations with respect to hernia recurrence.
Table IV. Analysis of relationship of comorbidities and surgical technique with hernia recurrence Variable Hypertension Diabetes mellitus Coronary artery disease Wound contamination at time of repair ASA classification Component separation Closed versus open abdomen at time of repair Elective versus emergent operation Skin closure versus skin opened post repair Bridged versus reinforced repair
Pearson v2 value P P P P P P P
= = = = = = =
.608 .984 .523 .720 .768 .335 .523
P = .499 P = .984 P = .016
Table V. Analysis of relationship of comorbidities and surgical technique with eventration Variable Hypertension Diabetes mellitus Coronary artery disease Wound contamination at time of repair ASA classification Component separation Closed versus open abdomen at time of repair Elective versus emergent operation Skin closure versus skin opened post repair Bridged versus reinforced repair
Pearson v2 value P P P P P P P
= = = = = = =
.892 .814 .688 .523 .266 .001 .099
P = .228 P = .187 P = .05
A v2 analysis also revealed no correlation between patient comorbidities, wound contamination, ASA classification, placement of HADM in an open or closed abdominal wound, elective versus emergent nature of repair, or primary versus delayed closure of skin over the HADM and eventration (Table V). Performance of a fascial release via a component separation, however, did correlate with a greater incidence of eventration. In addition, a bridged repair was associated with a greater incidence of eventration when compared with the reinforced repair. Mann-Whitney testing again revealed no correlation between the albumin level and the number of previous abdominal operations with respect to eventration. A v2 analysis also revealed no correlation between patient comorbidities (hypertension, diabetes mellitus, coronary artery disease), wound contamination, ASA classification, elective versus emergent nature of the repair, or primary versus delayed closure of the
Surgery Volume 144, Number 4
skin over the HADM and perioperative wound complications. DISCUSSION Approximately 600,000 hernias are repaired annually in the United States alone, representing a significant cost to the health care system. More than 90,000 of these hernia repairs are performed for ventral and incisional hernias. Despite advances in surgical technique, the search continues for an ideal prosthetic material for the repair of abdominal wall defects. One of the most challenging problems for the general surgeon involves the repair of complex contaminated abdominal wall defects. In these situations, no ideal operative repair technique or prosthetic material for reconstruction of the fascial defect is available currently. There is near universal agreement that the use of nonabsorbable synthetic mesh is contraindicated in the setting of contaminated abdominal wall defects owing to an unacceptably high rate of perioperative complications including wound infection, postoperative adhesions, and the development of enterocutaneous fistulae.4-6 Absorbable mesh can be used in these situations to reestablish the integrity of the abdominal wall; however, this solution is temporary and predisposes the patient to multiple operations and a staged abdominal wall reconstruction to achieve an acceptable cosmetic and functional result. A number of autologous tissue mobilization and free tissue transfer techniques allow hernia closure without the use of prosthetic materials.7-14 These solutions also have their limitations and not only increase the cost of the patient hospital stay but also increase patient morbidity. This dilemma has led to an extensive search for alternative materials with which to achieve a tension-free repair in a single-stage operation in the setting of the contaminated wound. HADM has emerged in recent years as a suitable prosthesis for use in the setting of contaminated abdominal wall defects. Although its use has been described in the burn literature for the past 20 years as a tissue substitute for skin grafting, the emergence of HADM use for complex abdominal wall reconstructions is relatively recent. Despite the growing body of evidence that HADM is a suitable effective prosthetic for contaminated abdominal wall defects, there is still much debate regarding operative technique with respect to placement, postoperative wound complications, eventration, and hernia recurrence. This debate is partially due to limited follow-up.
Candage et al 707
This study demonstrates that large, contaminated or potentially contaminated abdominal wall defects can be closed in a single-stage, tension-free repair with the use of HADM. The patient demographics reveal that HADM is suitable for the high-risk patient population; 55% of the patients included in this study had clinically important comorbidities, including hypertension, diabetes, and coronary artery disease. This bioprosthesis demonstrates the ability to incorporate into healthy tissue in the setting of contaminated tissue without becoming infected. Its biologic properties, including its ability to support tissue regeneration with concomitant revascularization, repopulation with fibroblasts, collagen deposition, and eventual absorption and replacement with native fascia, permit its use in the contaminated surgical field. HADM, therefore, has a theoretic advantage over permanent synthetic prosthesis in the repair of infected fascial defects with an acceptable rate of hernia recurrence.15-17 In our series, 28% of the patients developed a recurrent hernia with long-term follow-up. These results are comparable to those of other series reported in the literature.2,3 One factor affecting the incidence of hernia recurrence is the technique used for the repair. Kim et al18 found that all hernia recurrences occurred when a running suture technique was used to secure the HADM to the fascia. As a result, they made a transition to the use of an interrupted suture technique. The results of our study differed because 60% of the patients with a hernia recurrence had the HADM sutured to the fascia using an interrupted technique, whereas 40% experienced a recurrence with the HADM sutured to the fascia in a running fashion; this difference, however, was not statistically significant. Bellows et al3 recommended a staged approach with the use of HADM in the setting of an open, infected abdominal wall based on observations that grossly contaminated wounds were associated with a greater rate of perioperative complications. As such, they have adopted a strategy of delayed primary closure of grossly contaminated wounds when clinically indicated to prevent a closed space infection; however, wound classification did not correlate with an increased risk of hernia recurrence. The results of our study support the contributions of Bellows et al because the wound classification was not associated with an increased risk of hernia recurrence. Bellows et al commented in their series that coronary artery disease and an ASA classification of 4 were also associated with a greater rate of perioperative complications
708 Candage et al
(wound dehiscence, graft infection, postoperative bleeding, evisceration, and death).3 The results of our study were not consistent with these previously reported data; we found no association between coronary artery disease, ASA classification, or wound contamination as a predictor of perioperative wound complications. Jin et al1 found that the only significant predictive factor for hernia recurrence with HADM was use of a bridging technique. The results of our study support this conclusion; we noted a 46% recurrence rate in the bridged group versus an 8% recurrence rate in those patients where the HADM was used to reinforce the fascial repair. Schuster et al19 placed HADM as an interposition graft in contaminated fields and reported a hernia recurrence rate of >80% when the skin was not closed over the graft. They concluded that the use of HADM is not justified unless primary wound closure can be achieved. In our series however, there was no correlation between primary versus delayed closure of the skin over the HADM and hernia recurrence. A unique complication of HADM is the development of laxity and stretching of the implant (eventration) after implantation. This complication has been noted by others.3 Gupta et al2 also note that 45% of the patients who underwent reconstruction with AlloDerm in their review developed eventration of the bioprosthesis at the repair site; however, they did not elaborate on the technique used to repair the fascial defect.2 Jin et al1 comment that, universally, the HADM was noted to develop eventration in their patients over time. In our series, only 8 (17%) patients who had a hernia repaired with HADM had eventration of the bioprosthesis occur. Importantly, the majority (88%) of these patients required the use of the HADM to bridge a fascial defect, and only 1 patient (12%) experienced eventration when the bioprosthesis was placed to reinforce a fascial repair. The results of our study support a growing body of evidence that eventration is a major problem associated with the use of HADM. This eventration may be due to a number of factors, including the inherent ability of the bioprosthesis to stretch over time as well as the manner in which the cadaveric skin is harvested. As mentioned in the study by Gupta et al,2 the dermatome harvest results in thin borders at the perimeter of the HADM, which is where the sutures are placed and most likely to disrupt from the thinner area of the HADM. Another proposed mechanism for eventration of HADM is that its collagen-like properties predispose it to degradation by bacterial
Surgery October 2008
collagenase in the setting of contaminated wounds. Although not proven, this could in theory explain the propensity of HADM for eventration as well as a propensity for hernia recurrence. Alternatively, the newly incorporated collagen that occurs after implantation could be stretched in a manner similar to the linea alba of a diastasis recti owing to excessive intraabdominal pressure, such as that seen in obese individuals. This study is limited by a number of factors including the inherent limitations of a retrospective review and the small number of patients in the bridged and reinforced groups. The HADM was used by a large number of surgeons (N = 13) with variable experience using this bioprosthesis. In addition, 32% of the patients in this study were followed up via telephone, leaving us without a clinical examination to verify the patients’ subjective sensation that the mesh had developed some degree of eventration or a recurrent hernia. As such, we were unable to exclude weight gain as a potential etiology for the patient’s subjective interpretation of an increase in abdominal girth. Furthermore, the average follow-up in patients was only about 12 months. Despite these limitations, the observation that 87% of the patients who developed a recurrent hernia (fascial defect or eventration) had HADM utilized to bridge a fascial defect within 1 year of implantation allows surgeons to inform patients of the high incidence of these complications. Based on a growing body of evidence, HADM is suitable for use in a bridging fashion with the understanding that it provides only a temporary means for restoration of abdominal wall integrity in contaminated abdominal fields. HADM seems to have no long-term durability and, in the absence of primary fascial reapproximation, a large percentage of patients will either go on to develop a recurrence of their hernia, eventration of the bioprosthesis, or both. As such, these patients may need an additional operation for implantation of a more durable prosthesis to reestablish the integrity of the abdominal wall. In the authors’ opinion, HADM is an excellent bioprosthesis for use in contaminated wounds. When used to bridge a fascial defect, one can expect a high recurrent hernia rate. In contrast, when used as a reinforcement for a primary fascial closure, the recurrent hernia rate seems to be comparable to that observed with other prostheses. A large, prospective, randomized trial is clearly needed to resolve many of the complex issues with regard to the use of HADM for repairing complex abdominal wall hernias.
Surgery Volume 144, Number 4
REFERENCES 1. Jin J, Rosen M, Blatnik J, et al. Use of acellular dermal matrix for complicated ventral hernia repair: does technique affect outcomes? J Am Coll Surg 2007;205:654-60. 2. Gupta A, Zahriya K, Mullens PL, et al. Ventral herniorrhaphy: experience with two different biosynthetic mesh materials, Surgisis and Alloderm. Hernia 2006;10:419-25. 3. Bellows C, Albo D, Berger D, et al. Abdominal wall repair using human acellular dermis. Am J Surg 2007;194:192-8. 4. Cobb WS, Harris JB, Lokey JS, et al. Incisional herniorrhaphy with intraperitoneal composite mesh: a report of 95 cases. Am Surg 2003;69:784-7. 5. Voyles CR, Richardson JD, Bland KI, et al. Emergency abdominal wall reconstruction with polypropylene mesh: short-term benefits versus long term complications. Ann Surg 1981;194:219-33. 6. Karakousis CP, Volpe C, Tanski J, et al. Use of a mesh for musculoaponeurotic defects of the abdominal wall in cancer surgery and the risk of bowel fistulas. J Am Coll Surg 1995;181:11-6. 7. Ramirez OM, Ruas E, Dellon AL. ‘‘Component separation’’ method for closure of abdominal-wall defects: an anatomical and clinical study. Plast Reconstr Surg 1990;6:519-26. 8. Thomas WO III, Parry SW, Rodning CB. Ventral/incisional abdominal herniorrhaphy by fascial partition/release. Plast Reconstr Surg 1993;91:1080-6. 9. Kuzbari R, Worseg A, Tairych G, et al. Sliding door technique for the repair of midline incisional hernias. Plast Reconstr Surg 1998;101:1235-42. 10. Lindsey JT. Abdominal wall partitioning (the accordion effect) for reconstruction of major defects: a retrospective review of 10 patients. Plast Reconstr Surg 2003;12:477–5. 11. Disa JJ, Goldberg NH, Carlton JM, et al. Restoring abdominal wall integrity in contaminated tissue-deficient wounds using autologous fascia grafts. Plast Reconstr Surg 1998;101:979-86. 12. DeFranzo AJ, Kingman GJ, Sterchi JM, et al. Rectus turnover flaps for the reconstruction of large midline abdominal wall defects. Ann Plast Surg 1996;37:18-23. 13. Williams JK, Carlson GW, deChalain T, Howell R, Coleman TT. Role of tensor fasciae latae in abdominal wall reconstruction. Plast Reconstr Surg 1998;101:713-8. 14. Brown DM, Sicard GA, Flye MW, et al. Closure of complex abdominal wall defects with bilateral rectus femoris flaps with fascial extensions. Surgery 1993;114:112-6. 15. Harper JR. Tissue regeneration using human acellular dermal matrix: a histological perspective. LifeCell; Clinical Monograph Series. Branchburg, NJ: LifeCell; 2005. 16. Menon NG, Rodriguez ED, Byrnes CK, et al. Revascularization of human acellular dermis in full-thickness abdominal wall reconstruction in the rabbit model. Ann Plast Surg 2003;50:523-7. 17. Silverman RP, Li EN, Holton LH, et al. Ventral hernia repair using allogeneic acellular dermal matrix in swine model. Hernia 2004;8:336-42. 18. Kim H, Bruen K, Vargo D. Acellular dermal matrix in the management of high risk abdominal wall defects. Am J Surg 2006;192:705-9. 19. Schuster R, Singh J, Safadi BY, Wren SM. The use of acellular dermal matrix for contaminated abdominal wall defects: wound status predicts success. Am J Surg 2006;192:594-7.
DISCUSSION Dr Gary C. Vitale (Louisville, Ky): This is a very good paper. It has got good numbers. And I think to summarize,
Candage et al 709
what is interesting for most of us is that many of their patients had mesh infection/peritonitis/enterocutaneous fistula before they were operated on, and there was a pretty interesting distribution between a mesh-reinforced closure in 26 patients versus bridging a defect in 20 patients. So that leads for some interesting comparisons. It is also interesting that none of the recurrences had wound infection and hernia recurrence was not associated with repair of the wound. Only 1 recurrence occurred in the patient with infected wound. I think this is very different than what you expect if you place proline mesh in these patients. So that is really a great accomplishment. I do have a few questions. First of all, do you think that the laparoscopic approach is better with biologics? With patients who have wound infection, is the placement of the mesh on top of the fascia worse than doing an underlay? In some of the cases we had, we have seen some absorption or dissolution of the mesh with wound infection on top of the fascia. And then, how much overlap is enough when placing these grafts? As you know, they are expensive, and you are sewing them together, it can be complex. How much do you really need to get? The other thing is that you emphasize a very high incidence of recurrence in bridging the defects, >80%. So should biologic mesh be used at all in that setting? I find it interesting that the reps are pushing that. ‘‘Oh, you can use it, big gaps, no problem,’’ but obviously that does not work. And if you could comment about that. Or, in fact, your follow-up was 12 months in the bridged group and only about 4 months in the reinforced group. So did that account for your difference and not in fact the difference in the technique? Component separation. We have used that with our pancreatitis patients. And you mention that you do have a few patients with component separation. Do you think that benefits the patient, or is the result going to be the same with or without it? Dr R. L. Candage (Maywood, Ill): In regard to the first question with the laparoscopic approach, at least at our institution I do not know that we have much data on these; all of these repairs that were performed were done open. From what I have reviewed in the literature, I have not seen much out there on laparoscopic repair that has really looked at the use of acellular dermal matrix in these types of contaminated wounds. With respect to placing the mesh on top of the fascia versus an underlay, I do not know that our sample size is sufficient to really break it down in terms of an onlay and underlay and sandwich repair, but there are other groups out there who reported in the American Journal of Surgery about a year ago actually looked at this variable and did find that those who had either an interposition or onlay placement of a HADM did have a higher incidence of hernia recurrence. But we did not look at that specifically in this series. In terms of the degree of overlap, in our series the degree of overlap was anywhere from 1 to 4 cm. Again we did not break it down in terms
710 Candage et al
of the exact amount of overlap in the fascia that we used to see if there was any evidence of increased recurrence with that degree of overlap. The follow-up is something interesting in terms of there was longer follow-up in the bridged group, which was on average about 12 months versus about 5 months in those who had undergone a reinforced repair. That may account for some of the difference. I think the majority of those who will experience a recurrence will do so in the first year. But in our series, at least in the bridged group, we found a recurrence as far out as 37 months and in those undergoing reinforced repair we found 1 out as far as 22 months now. So I think it emphasizes the point that these patients at least require very long-term follow-up for repair of these types of complex wounds with this particular type of mesh. In terms of using the biologic mesh to bridge a defect, I think the data are now starting to show that bridging these repairs probably is going to result in a higher rate of hernia recurrence. And whether or not it should be used at all is difficult to say. It certainly proves its usefulness with repair in infected wounds. In terms of component separation, there was about an equal number, I think 7 in the bridged group, 6 in the reinforced repair, who had undergone a component separation in the series. When we kind of looked at that subset of patients, component separation did not seem to make any difference with respect to them having a hernia recurrence. Dr L. Michael Brunt (St. Louis, Mo): I enjoyed the study and congratulate you on reporting this large series of patients. I would like some clarification about your follow-up. Because you mentioned in your slide you had phone follow-up. For following patients who have had a hernia defect repaired—examination follow-up in the office is extremely important. So can you clarify for us how many of your patients had follow-up with examination and what that link to follow-up was? Also, can you give us the percent recurrence rate in the patients who had bridged versus repairs that were over defects that were closed primarily? You showed the absolute numbers, but what was the percent recurrence? And then finally, you also showed that patients had about an average of 2 Alloderm sheets, but what were the sizes of those? What size defects were you actually bridging? Dr R. L. Candage (Maywood, Ill): To answer the first question in terms of the patient follow-up, again, we had an average of about 12 months for those undergoing the bridged repair and about 5 months for those undergoing the enforced repair. There were 15 patients that we contacted by telephone who had not been seen in the clinic in a while. So those patients we defined as having a hernia recurrence if they had either gone to another institution and had a hernia repair there, or if they had noticed a significant bulge since the last time they showed up at the clinic and they had just been sitting at home choosing not to do anything about it because they were not having any
Surgery October 2008
symptoms. Again, the time period for follow-up, the longest patient we have out for the bridged is about 37 months and for the reinforced repair right now it is about 22 months. In terms of the percent recurrence, the percent recurrence for bridged repair was about—I think the overall recurrence rate was around 28% for our study. In terms of the breakdown of the percentages, I think out of the 13 patients who did have a recurrence in the series, probably 80% were bridged and about 20% were in the reinforced group. So that would be the breakdown. And the size of the Alloderm, in terms of the sheets, all the sheets were thick. In terms of the size of the sheets, we did not look at that specifically. We did do a subset analysis on the patients in which the surgeon had commented on the size of the defect itself, the hernia defect. We did not find any association between the size of the hernia defect and recurrence. So that much I can say. But in terms of the actual breakdown for the size of the Alloderm mesh, I do not have. Dr Lawrence N. Diebel (Detroit, Mich): Excellent presentation. I just want to ask a question, because I just had a patient last week who came back with a recurrence after I used this stuff. I reoperated, and you can see the proline sutures very nicely there with nothing left in between. I said to the rep, ‘‘Your stuff dissolved,’’ and she said, and I quote, ‘‘No, it just well incorporated.’’ Why should this material handle differently as a bridge versus as a reinforcement? Is there something different about the way the physiologic properties of the wound on this material whether it is bridged? Or is it just that the tissue that you are using actually is sufficient to the patient’s own tissue and whether you left the Alloderm or not is irrelevant? And then the question is, based on my recent experience, why does this recur? Do they all eventually dissolve over time or not? Dr R. L. Candage (Maywood, Ill): To answer the second question first, the material is known to degrade over time. It is reabsorbed. There are a number of theories as to why that occurs. Some think that bacteria coagulate in the wound and in the setting of infection they contribute to the breakdown of this material. But it is eventually reabsorbed over time. I think that in terms of the recurrence in the tissue properties and things of that nature, and maybe a combination of both, I think intuitively it makes some sense that usually if you are going to use this as a reinforced repair you are probably dealing with somewhat smaller defects in the sense that you are able to completely reapproximate that fascia in those tissues and get them back together, whereas when you are dealing with it a bridged manner you are clearly dealing with a large abdominal defect and that may account for some of the reason that you see the recurrence rates such as they are, although again we could not find any direct correlation between the defect size and hernia recurrence per se. But I think that there is some play with that, and I think that some of it is just due to the inherent nature of
Surgery Volume 144, Number 4
the patient’s own tissue. If you can reapproximate that fascia, you are probably going to get a stronger repair. And I think that intuitively makes some sense compared with when you are not able to completely bring that fascia back together. Dr R. Matthew Walsh (Cleveland, Ohio): We do actually have some data from the clinic that are almost identical. I guess my question is, how actually are you going to use this? I am not clear about that. Because if you are going to go in and decide that you are going to use it as a bridge, are you really going to use the Alloderm? Or just to use your friend analogy, why not go with the cheapest date? Because if you look at the cost data, it is really quite prohibitive. Could you tell us
Candage et al 711
what your costs were versus what you actually got paid for some of these? Dr R. L. Candage (Maywood, Ill): That we did not look at—cost—specifically. I know the cost per sheet is expensive. I am not sure exactly what the numbers are, but I think it is like a 6-by-12 sheet or something, runs around $1,200 or something. So it is certainly not cheap. I think in the majority of the patients the decision whether or not to use was probably just made at the time of surgery. I think in the small number of patients that we were able to completely reapproximate the fascia, only about 13 patients or so, but I think in the majority of cases it ends up being made intraoperatively whether or not the attending surgeon of record decided to use the mesh or not.
Background. Surgeons continue to search for the ideal prosthetic material to repair complex abdominal wall hernias. Recently, a new biologic material was introduced into the surgeon’s arsenal. The purpose of this study is to review a single institution’s experience with the use of human acellular dermal matrix (HADM [AlloDerm]) for repair of hernias. Methods. This was a retrospective review of all patients who received HADM for repair of an abdominal wall hernia. Patient demographics, comorbidities, wound contamination, operative technique, complications, and hernia recurrence were analyzed. Results. Between May 2004 and October 2007, HADM was implanted in a total of 46 patients undergoing repair of a ventral hernia. The average age was 54 years (range, 26--77), with an average American Society of Anesthesiologists classification of 2.5 (range, 1--4). Indications for use of HADM included complex ventral hernia repair (n = 34), mesh infection/enterocutaneous fistula (n = 10), and peritonitis (n = 2). The incidences of comorbidities were hypertension in 47%, diabetes mellitus in 16%, and coronary artery disease in 11%. The majority (87%; n = 40) of the procedures were performed on an elective basis. Seventeen procedures were performed in contaminated wounds. The HADM was placed as reinforcement to the hernia repair in 26 patients and as a ‘‘bridge’’ between the fascial edges in 20 patients. The average follow-up was 12.1 months. Wound complications were frequent at 54%. There were 6 recurrent hernias and 8 patients with eventration of the bioprosthesis so that the recurrent hernia rate was 30%. None of the recurrences were associated with a postoperative wound infection. The majority (88%) of patients who developed eventration of the HADM had a repair using the bioprosthesis to ‘‘bridge’’ an abdominal wall defect. Hernia recurrence and eventration were not associated with use of HADM in a contaminated/infected wound. Conclusions. HADM is a suitable prosthesis for repair of complex and routine abdominal wall defects. This bioprosthesis can incorporate into contaminated tissue without becoming infected. Eventration occurs when HADM is utilized as a fascial replacement rather than as a reinforcement. (Surgery 2008;144:703-11.) From the Divisions of General Surgery and Plastic Surgery, Department of Surgery, Stritch School of Medicine, Loyola University Medical Center, Maywood, Ill
APPROXIMATELY 2 MILLION ABDOMINAL OPERATIONS are performed in the United States each year. The incidence of postoperative incisional hernia ranges from 3% to 20% and represents a tremendous financial burden to the health care system. As compared with primary repair, the introduction of prosthetic mesh appliances for the repair of elective ventral hernias has had a significant impact on reducing the rate of hernia recurrence. However, the use of synthetic mesh has its own set of
Accepted for publication June 19, 2008. Reprint requests: Fred A. Luchette, MD, Department of Surgery, 2160 South First Avenue, Maywood, IL 60153. E-mail: [email protected]. 0039-6060/$ - see front matter Ó 2008 Mosby, Inc. All rights reserved. doi:10.1016/j.surg.2008.06.018
complications, including bowel obstruction, seroma formation, fistula formation, and infection. There is near universal agreement that the use of permanent synthetic prosthesis is contraindicated in the setting of contaminated wounds owing to the high rate of infection of the mesh and recurrence of the hernia. Before the development of biologic substitutes, surgeons either implanted an absorbable synthetic mesh or allowed the wound to granulate in and heal by secondary intention with or without an overlying skin graft. Each of these situations predisposes the patient to a staged repair and adds considerable morbidity to what is already a daunting surgical procedure. The development of human acellular dermis (Alloderm; LifeCell Corporation, Branchburg, NJ) as well as other biologic substitutes, allows surgeons to achieve definitive abdominal wall reconstruction in a single staged procedure in the SURGERY 703
704 Candage et al
setting of contaminated abdominal wall defects. Human acellular dermal matrix (HADM) is a cadaveric dermal graft (bioprosthesis) in which the epidermis has been separated and all of the cellular elements in the dermis that can lead to tissue rejection and graft failure are removed. This biologic material is thought to act as a scaffold that permits tissue regeneration with revascularization and repopulation of cellular elements. The material is entirely degraded and replaced with host collagen over time after implantation. Although numerous reports in the literature demonstrate the safety and efficacy of HADM for use in the setting of contaminated ventral hernias, there are not enough data or long-term reports to support a universal consensus for its use. A few reports in the literature comment on the rate of recurrent hernia and an undesirable property of HADM described as laxity of the material or its ability to stretch over time. This unique characteristic is also described as eventration of the bioprosthesis and is considered to be a recurrent hernia.1-3 The purpose of this study was to identify factors that contribute to the rate of recurrence for hernias repaired with HADM specifically when used in a nonsterile surgical wound. METHODS The Loyola University Medical Center Institutional Review Board approved this study. All patients undergoing hernia repair with HADM from May 2004 through October of 2007 were reviewed. Patient demographics including age, gender, comorbidities, American Society of Anesthesiologists (ASA) classification, preoperative plasma concentration of albumin, indication for repair, and number of previous abdominal operations were recorded. Perioperative data included wound classification (clean/contaminated), operative technique, skin closure status, elective versus emergent nature of the procedure, and placement of the mesh in an open versus closed abdomen at the time of operation. For study purposes, surgical wounds were considered to be contaminated actively in the following clinical settings: in the presence of an infected synthetic mesh, enterocutaneous fistula with hernia, feculent peritonitis (perforation or anastomotic leak), necrotizing fasciitis, and intraabdominal abscess. Operative wounds were considered to be potentially contaminated at the time of HADM implantation with any of these scenarios: prior history of feculent peritonitis, iatrogenic enterotomy with contamination during an elective hernia repair, peristomal hernia
Surgery October 2008
repair, incarcerated hernia without bowel necrosis or perforation, bowel resection performed at the time of the hernia repair, and fascial dehiscence with evisceration. Complex ventral hernias were defined as recurrent incisional/ventral hernias with or without removal of noninfected mesh, or repair of an initial incisional/ventral hernia with performance of another clean procedure at the time of operation (e.g., nephrectomy). Operative techniques. Various sizes of thick (2 mm) sheets of acellular cadaveric dermis were reconstituted at room temperature in normal saline solution for 20 to 30 minutes before implantation. The material was prestretched before its use and placed in the wound with an appropriate amount of tension. Thirteen surgeons utilized HADM to repair hernias. Details of the operative procedure were determined by a review of the operative report and by direct conversation with the attending surgeon of record where appropriate. Before implantation of the mesh, all wounds were first debrided of foreign body and any nonviable tissue and then irrigated copiously. The HADM was secured to the fascia with one of two techniques---continuous running suture or interrupted mattress sutures using permanent suture material. The running technique attached the bioprosthesis to the posterior rectus sheath fascia around the perimeter of the HADM. The interrupted method used horizontal mattress sutures to secure the bioprosthesis to the fascia using horizontal mattress sutures with the knots tied in the subcutaneous tissue. Placement of the HADM was classified as either bridging a fascial defect or reinforcing the fascial repair. In the bridged group, the bioprosthesis was placed in an intraperitoneal position with a variable degree of fascial underlay (1--4 cm) and then ‘‘bridged’’ a defect in the abdominal wall fascia. For a reinforced repair, the HADM was placed as an underlay, onlay, or a sandwich technique before and/or after primary reapproximation of the fascia. The use of a component separation to facilitate primary closure of the fascia was also recorded. If the hernia defect was larger than the size of a single piece of HADM, individual sheets were sewn together with the use of permanent monofilament polypropylene sutures in an interrupted or running fashion according to surgeon preference. If the incision was left open to avoid a wound infection, the wounds were managed with either vacuum-assisted closure therapy or with 0.9% NaCl-moistened dressings. After discharge, clinic records were reviewed for the presence of wound complications, hernia recurrence, or the presence of eventration. Patients
Candage et al 705
Surgery Volume 144, Number 4
who had not had a clinic visit since July 2007 were contacted by telephone. Although not ideal, as a substitute for a focused clinical examination, the patient was asked if there were any changes in their medical condition with regard to the most recent hernia repair. Patients who reported a subjective increase in their abdominal girth and/or change in waist size were defined as having eventration of the HADM. Patients who noted a significant bulge with pain or who underwent a subsequent procedure to repair a hernia after their initial operation at our institution were defined as having a hernia recurrence. Statistical analysis was performed on the nominal data using chi square analysis and MannWhitney test. Inquiries included the correlation between hernia recurrence and each of the following: wound infection, ASA classification, albumin level, type of surgical repair, patient comorbidities, primary wound closure, and status of the abdomen (open vs closed) at the time of operation. v2 analysis was also used to determine whether there was a correlation between the above variables and eventration of the mesh. Statistical significance was defined as P < .05. RESULTS A total of 53 patients underwent a hernia repair with HADM during the study period. Indications for the use of HADM included complex ventral hernia repair, mesh infection/enterocutaneous fistula with recurrent hernia, umbilical hernia, peritonitis, peristomal hernia, incarcerated femoral hernia, and inguinal hernia (Table I). Patients who underwent umbilical, peristomal, femoral, or inguinal hernia repairs were excluded from the final data analysis. There were 2 deaths during the immediate postoperative period. One patient death was the result of a cerebrovascular event 48 hours after hernia repair. Ten days later, the family decided to withdraw care. The second death was the result of a myocardial infarction with cardiac arrest on postoperative day 1. Forty-eight hours later, the family decided to withdraw care. Given the short time interval from hernia repair to patient death, we were unable to infer any meaningful outcome data for the study and these patients were excluded from the data analysis. Accordingly, there were 46 patients for final data analysis. A total of 15 (32%) patients were contacted by telephone for additional clinical information. Patient demographics are listed in Table II. HADM was utilized in 40 elective operations and in an emergent setting in 6 patients. Seventeen
Table I. Indications for use of human acellular cadaveric dermis Indications
Number of patients
Complex ventral hernia repair Mesh infection/enterocutaneous fistula Umbilical hernia Peritonitis Peristomal hernia Incarcerated femoral hernia Incarcerated inguinal hernia
34 10 3 2 2 1 1
Table II. Patient characteristics ASA classification No. of previous abdominal operations No. of sheets of HADM Albumin Age (yrs)
Mean
Median
2.8 3.0
3.0 3.0
1.7 3.0 56.5
1.0
patients underwent abdominal wall reconstruction in a contaminated field. An additional 13 patients undergoing an elective hernia repair were repaired with HADM owing to a history of significant intraabdominal contamination, open abdomen for damage control in trauma, or enteric contamination. Thirty-eight patients had HADM placed in the setting of a closed abdomen and the remaining 8 patients had HADM placed in an open abdominal wound. A total of 20 patients underwent a bridged repair; 26 patients were repaired using the HADM to reinforce the fascial closure. Patient demographics in both the bridged and reinforced group were similar (Table III). On review of the operative record, the operating surgeon commented on whether the HADM was sutured in running or interrupted fashion in 42 of the 46 patients. The majority of patients (n = 34; 81%) had the HADM sutured using an interrupted technique. Component separation was also performed in 7 patients (27%) in the bridged group and 6 patients (30%) in the reinforced group to facilitate approximation of the fascia. It is reasonable to infer that those defects requiring component separation and a bridged repair were large defects with considerable loss of domain. In contrast, if a reinforced repair was performed with primary fascial reapproximation, the original hernia was likely a smaller defect.
706 Candage et al
Surgery October 2008
Table III. Patient characteristics stratified by placement of HADM
ASA classification No. of previous abdominal operations No. of sheets of HADM Albumin Age
Bridged (n = 20)
Reinforced (n = 26)
2.9 ± 0.5 3.0 ± 2.1
2.6 ± 0.7 2.9 ± 1.7
2.0 ± 1.3 2.9 ± 0.9 54.7 ± 11.2
1.2 ± 0.5 3.1 ± 0.8 58.7 ± 15.4
Values are presented as means ± SD.
Of the 39 patients with a primary closure of the initial surgical wound, 64% (n = 26) developed $1 wound complication. The specific diagnoses included wound infection (n = 9; 23%), seroma (n = 5; 13%), nonhealing wound (n = 5; 13%), enterocutaneous fistula (n = 3; 8%), dehiscence (n = 2; 5%), lymphocele (n = 1; 3%), and hematoma (n = 1; 3%). All the enterocutaneous fistulae occurred in wounds left open at the initial operation. The average follow-up for the study population was 12 months (range, 1--39). There were 6 recurrent hernias diagnosed, and 8 patients developed eventration of the bioprosthesis so that combined, the recurrent hernia rate was 30% (n = 14). The majority of the recurrences were in the bridged group (n = 12; 60%) with an average of 11 months from the time of initial operation to the time of hernia recurrence. Two patients (8%) in the reinforced group had a recurrence. Seven of the 8 patients (88%) who experienced eventration of the bioprosthesis underwent a bridged repair. In contrast, when the bioprosthesis was used to reinforce the fascial repair, only 1 recurrence was attributed to eventration of the HADM. Only 1 of the recurrent hernias occurred in the 16 patients with an infected wound. There were 6 recurrent hernias among the 26 patients with a postoperative wound complication. Review of the operative notes revealed that in 10 of the 14 patients with a recurrent hernia, the surgeon commented on the type of suture technique used to attach the HADM to fascia. An interrupted technique used 60% of the time; in 40% of the repairs, a running suture technique was employed. Given the small sample size, this difference was not statistically significant. When the various data variables were analyzed for a correlation with hernia recurrence, none were identified except for the use of the HADM in a bridged fashion (Table IV). Mann-Whitney testing revealed no correlation between albumin levels or the number of previous abdominal operations with respect to hernia recurrence.
Table IV. Analysis of relationship of comorbidities and surgical technique with hernia recurrence Variable Hypertension Diabetes mellitus Coronary artery disease Wound contamination at time of repair ASA classification Component separation Closed versus open abdomen at time of repair Elective versus emergent operation Skin closure versus skin opened post repair Bridged versus reinforced repair
Pearson v2 value P P P P P P P
= = = = = = =
.608 .984 .523 .720 .768 .335 .523
P = .499 P = .984 P = .016
Table V. Analysis of relationship of comorbidities and surgical technique with eventration Variable Hypertension Diabetes mellitus Coronary artery disease Wound contamination at time of repair ASA classification Component separation Closed versus open abdomen at time of repair Elective versus emergent operation Skin closure versus skin opened post repair Bridged versus reinforced repair
Pearson v2 value P P P P P P P
= = = = = = =
.892 .814 .688 .523 .266 .001 .099
P = .228 P = .187 P = .05
A v2 analysis also revealed no correlation between patient comorbidities, wound contamination, ASA classification, placement of HADM in an open or closed abdominal wound, elective versus emergent nature of repair, or primary versus delayed closure of skin over the HADM and eventration (Table V). Performance of a fascial release via a component separation, however, did correlate with a greater incidence of eventration. In addition, a bridged repair was associated with a greater incidence of eventration when compared with the reinforced repair. Mann-Whitney testing again revealed no correlation between the albumin level and the number of previous abdominal operations with respect to eventration. A v2 analysis also revealed no correlation between patient comorbidities (hypertension, diabetes mellitus, coronary artery disease), wound contamination, ASA classification, elective versus emergent nature of the repair, or primary versus delayed closure of the
Surgery Volume 144, Number 4
skin over the HADM and perioperative wound complications. DISCUSSION Approximately 600,000 hernias are repaired annually in the United States alone, representing a significant cost to the health care system. More than 90,000 of these hernia repairs are performed for ventral and incisional hernias. Despite advances in surgical technique, the search continues for an ideal prosthetic material for the repair of abdominal wall defects. One of the most challenging problems for the general surgeon involves the repair of complex contaminated abdominal wall defects. In these situations, no ideal operative repair technique or prosthetic material for reconstruction of the fascial defect is available currently. There is near universal agreement that the use of nonabsorbable synthetic mesh is contraindicated in the setting of contaminated abdominal wall defects owing to an unacceptably high rate of perioperative complications including wound infection, postoperative adhesions, and the development of enterocutaneous fistulae.4-6 Absorbable mesh can be used in these situations to reestablish the integrity of the abdominal wall; however, this solution is temporary and predisposes the patient to multiple operations and a staged abdominal wall reconstruction to achieve an acceptable cosmetic and functional result. A number of autologous tissue mobilization and free tissue transfer techniques allow hernia closure without the use of prosthetic materials.7-14 These solutions also have their limitations and not only increase the cost of the patient hospital stay but also increase patient morbidity. This dilemma has led to an extensive search for alternative materials with which to achieve a tension-free repair in a single-stage operation in the setting of the contaminated wound. HADM has emerged in recent years as a suitable prosthesis for use in the setting of contaminated abdominal wall defects. Although its use has been described in the burn literature for the past 20 years as a tissue substitute for skin grafting, the emergence of HADM use for complex abdominal wall reconstructions is relatively recent. Despite the growing body of evidence that HADM is a suitable effective prosthetic for contaminated abdominal wall defects, there is still much debate regarding operative technique with respect to placement, postoperative wound complications, eventration, and hernia recurrence. This debate is partially due to limited follow-up.
Candage et al 707
This study demonstrates that large, contaminated or potentially contaminated abdominal wall defects can be closed in a single-stage, tension-free repair with the use of HADM. The patient demographics reveal that HADM is suitable for the high-risk patient population; 55% of the patients included in this study had clinically important comorbidities, including hypertension, diabetes, and coronary artery disease. This bioprosthesis demonstrates the ability to incorporate into healthy tissue in the setting of contaminated tissue without becoming infected. Its biologic properties, including its ability to support tissue regeneration with concomitant revascularization, repopulation with fibroblasts, collagen deposition, and eventual absorption and replacement with native fascia, permit its use in the contaminated surgical field. HADM, therefore, has a theoretic advantage over permanent synthetic prosthesis in the repair of infected fascial defects with an acceptable rate of hernia recurrence.15-17 In our series, 28% of the patients developed a recurrent hernia with long-term follow-up. These results are comparable to those of other series reported in the literature.2,3 One factor affecting the incidence of hernia recurrence is the technique used for the repair. Kim et al18 found that all hernia recurrences occurred when a running suture technique was used to secure the HADM to the fascia. As a result, they made a transition to the use of an interrupted suture technique. The results of our study differed because 60% of the patients with a hernia recurrence had the HADM sutured to the fascia using an interrupted technique, whereas 40% experienced a recurrence with the HADM sutured to the fascia in a running fashion; this difference, however, was not statistically significant. Bellows et al3 recommended a staged approach with the use of HADM in the setting of an open, infected abdominal wall based on observations that grossly contaminated wounds were associated with a greater rate of perioperative complications. As such, they have adopted a strategy of delayed primary closure of grossly contaminated wounds when clinically indicated to prevent a closed space infection; however, wound classification did not correlate with an increased risk of hernia recurrence. The results of our study support the contributions of Bellows et al because the wound classification was not associated with an increased risk of hernia recurrence. Bellows et al commented in their series that coronary artery disease and an ASA classification of 4 were also associated with a greater rate of perioperative complications
708 Candage et al
(wound dehiscence, graft infection, postoperative bleeding, evisceration, and death).3 The results of our study were not consistent with these previously reported data; we found no association between coronary artery disease, ASA classification, or wound contamination as a predictor of perioperative wound complications. Jin et al1 found that the only significant predictive factor for hernia recurrence with HADM was use of a bridging technique. The results of our study support this conclusion; we noted a 46% recurrence rate in the bridged group versus an 8% recurrence rate in those patients where the HADM was used to reinforce the fascial repair. Schuster et al19 placed HADM as an interposition graft in contaminated fields and reported a hernia recurrence rate of >80% when the skin was not closed over the graft. They concluded that the use of HADM is not justified unless primary wound closure can be achieved. In our series however, there was no correlation between primary versus delayed closure of the skin over the HADM and hernia recurrence. A unique complication of HADM is the development of laxity and stretching of the implant (eventration) after implantation. This complication has been noted by others.3 Gupta et al2 also note that 45% of the patients who underwent reconstruction with AlloDerm in their review developed eventration of the bioprosthesis at the repair site; however, they did not elaborate on the technique used to repair the fascial defect.2 Jin et al1 comment that, universally, the HADM was noted to develop eventration in their patients over time. In our series, only 8 (17%) patients who had a hernia repaired with HADM had eventration of the bioprosthesis occur. Importantly, the majority (88%) of these patients required the use of the HADM to bridge a fascial defect, and only 1 patient (12%) experienced eventration when the bioprosthesis was placed to reinforce a fascial repair. The results of our study support a growing body of evidence that eventration is a major problem associated with the use of HADM. This eventration may be due to a number of factors, including the inherent ability of the bioprosthesis to stretch over time as well as the manner in which the cadaveric skin is harvested. As mentioned in the study by Gupta et al,2 the dermatome harvest results in thin borders at the perimeter of the HADM, which is where the sutures are placed and most likely to disrupt from the thinner area of the HADM. Another proposed mechanism for eventration of HADM is that its collagen-like properties predispose it to degradation by bacterial
Surgery October 2008
collagenase in the setting of contaminated wounds. Although not proven, this could in theory explain the propensity of HADM for eventration as well as a propensity for hernia recurrence. Alternatively, the newly incorporated collagen that occurs after implantation could be stretched in a manner similar to the linea alba of a diastasis recti owing to excessive intraabdominal pressure, such as that seen in obese individuals. This study is limited by a number of factors including the inherent limitations of a retrospective review and the small number of patients in the bridged and reinforced groups. The HADM was used by a large number of surgeons (N = 13) with variable experience using this bioprosthesis. In addition, 32% of the patients in this study were followed up via telephone, leaving us without a clinical examination to verify the patients’ subjective sensation that the mesh had developed some degree of eventration or a recurrent hernia. As such, we were unable to exclude weight gain as a potential etiology for the patient’s subjective interpretation of an increase in abdominal girth. Furthermore, the average follow-up in patients was only about 12 months. Despite these limitations, the observation that 87% of the patients who developed a recurrent hernia (fascial defect or eventration) had HADM utilized to bridge a fascial defect within 1 year of implantation allows surgeons to inform patients of the high incidence of these complications. Based on a growing body of evidence, HADM is suitable for use in a bridging fashion with the understanding that it provides only a temporary means for restoration of abdominal wall integrity in contaminated abdominal fields. HADM seems to have no long-term durability and, in the absence of primary fascial reapproximation, a large percentage of patients will either go on to develop a recurrence of their hernia, eventration of the bioprosthesis, or both. As such, these patients may need an additional operation for implantation of a more durable prosthesis to reestablish the integrity of the abdominal wall. In the authors’ opinion, HADM is an excellent bioprosthesis for use in contaminated wounds. When used to bridge a fascial defect, one can expect a high recurrent hernia rate. In contrast, when used as a reinforcement for a primary fascial closure, the recurrent hernia rate seems to be comparable to that observed with other prostheses. A large, prospective, randomized trial is clearly needed to resolve many of the complex issues with regard to the use of HADM for repairing complex abdominal wall hernias.
Surgery Volume 144, Number 4
REFERENCES 1. Jin J, Rosen M, Blatnik J, et al. Use of acellular dermal matrix for complicated ventral hernia repair: does technique affect outcomes? J Am Coll Surg 2007;205:654-60. 2. Gupta A, Zahriya K, Mullens PL, et al. Ventral herniorrhaphy: experience with two different biosynthetic mesh materials, Surgisis and Alloderm. Hernia 2006;10:419-25. 3. Bellows C, Albo D, Berger D, et al. Abdominal wall repair using human acellular dermis. Am J Surg 2007;194:192-8. 4. Cobb WS, Harris JB, Lokey JS, et al. Incisional herniorrhaphy with intraperitoneal composite mesh: a report of 95 cases. Am Surg 2003;69:784-7. 5. Voyles CR, Richardson JD, Bland KI, et al. Emergency abdominal wall reconstruction with polypropylene mesh: short-term benefits versus long term complications. Ann Surg 1981;194:219-33. 6. Karakousis CP, Volpe C, Tanski J, et al. Use of a mesh for musculoaponeurotic defects of the abdominal wall in cancer surgery and the risk of bowel fistulas. J Am Coll Surg 1995;181:11-6. 7. Ramirez OM, Ruas E, Dellon AL. ‘‘Component separation’’ method for closure of abdominal-wall defects: an anatomical and clinical study. Plast Reconstr Surg 1990;6:519-26. 8. Thomas WO III, Parry SW, Rodning CB. Ventral/incisional abdominal herniorrhaphy by fascial partition/release. Plast Reconstr Surg 1993;91:1080-6. 9. Kuzbari R, Worseg A, Tairych G, et al. Sliding door technique for the repair of midline incisional hernias. Plast Reconstr Surg 1998;101:1235-42. 10. Lindsey JT. Abdominal wall partitioning (the accordion effect) for reconstruction of major defects: a retrospective review of 10 patients. Plast Reconstr Surg 2003;12:477–5. 11. Disa JJ, Goldberg NH, Carlton JM, et al. Restoring abdominal wall integrity in contaminated tissue-deficient wounds using autologous fascia grafts. Plast Reconstr Surg 1998;101:979-86. 12. DeFranzo AJ, Kingman GJ, Sterchi JM, et al. Rectus turnover flaps for the reconstruction of large midline abdominal wall defects. Ann Plast Surg 1996;37:18-23. 13. Williams JK, Carlson GW, deChalain T, Howell R, Coleman TT. Role of tensor fasciae latae in abdominal wall reconstruction. Plast Reconstr Surg 1998;101:713-8. 14. Brown DM, Sicard GA, Flye MW, et al. Closure of complex abdominal wall defects with bilateral rectus femoris flaps with fascial extensions. Surgery 1993;114:112-6. 15. Harper JR. Tissue regeneration using human acellular dermal matrix: a histological perspective. LifeCell; Clinical Monograph Series. Branchburg, NJ: LifeCell; 2005. 16. Menon NG, Rodriguez ED, Byrnes CK, et al. Revascularization of human acellular dermis in full-thickness abdominal wall reconstruction in the rabbit model. Ann Plast Surg 2003;50:523-7. 17. Silverman RP, Li EN, Holton LH, et al. Ventral hernia repair using allogeneic acellular dermal matrix in swine model. Hernia 2004;8:336-42. 18. Kim H, Bruen K, Vargo D. Acellular dermal matrix in the management of high risk abdominal wall defects. Am J Surg 2006;192:705-9. 19. Schuster R, Singh J, Safadi BY, Wren SM. The use of acellular dermal matrix for contaminated abdominal wall defects: wound status predicts success. Am J Surg 2006;192:594-7.
DISCUSSION Dr Gary C. Vitale (Louisville, Ky): This is a very good paper. It has got good numbers. And I think to summarize,
Candage et al 709
what is interesting for most of us is that many of their patients had mesh infection/peritonitis/enterocutaneous fistula before they were operated on, and there was a pretty interesting distribution between a mesh-reinforced closure in 26 patients versus bridging a defect in 20 patients. So that leads for some interesting comparisons. It is also interesting that none of the recurrences had wound infection and hernia recurrence was not associated with repair of the wound. Only 1 recurrence occurred in the patient with infected wound. I think this is very different than what you expect if you place proline mesh in these patients. So that is really a great accomplishment. I do have a few questions. First of all, do you think that the laparoscopic approach is better with biologics? With patients who have wound infection, is the placement of the mesh on top of the fascia worse than doing an underlay? In some of the cases we had, we have seen some absorption or dissolution of the mesh with wound infection on top of the fascia. And then, how much overlap is enough when placing these grafts? As you know, they are expensive, and you are sewing them together, it can be complex. How much do you really need to get? The other thing is that you emphasize a very high incidence of recurrence in bridging the defects, >80%. So should biologic mesh be used at all in that setting? I find it interesting that the reps are pushing that. ‘‘Oh, you can use it, big gaps, no problem,’’ but obviously that does not work. And if you could comment about that. Or, in fact, your follow-up was 12 months in the bridged group and only about 4 months in the reinforced group. So did that account for your difference and not in fact the difference in the technique? Component separation. We have used that with our pancreatitis patients. And you mention that you do have a few patients with component separation. Do you think that benefits the patient, or is the result going to be the same with or without it? Dr R. L. Candage (Maywood, Ill): In regard to the first question with the laparoscopic approach, at least at our institution I do not know that we have much data on these; all of these repairs that were performed were done open. From what I have reviewed in the literature, I have not seen much out there on laparoscopic repair that has really looked at the use of acellular dermal matrix in these types of contaminated wounds. With respect to placing the mesh on top of the fascia versus an underlay, I do not know that our sample size is sufficient to really break it down in terms of an onlay and underlay and sandwich repair, but there are other groups out there who reported in the American Journal of Surgery about a year ago actually looked at this variable and did find that those who had either an interposition or onlay placement of a HADM did have a higher incidence of hernia recurrence. But we did not look at that specifically in this series. In terms of the degree of overlap, in our series the degree of overlap was anywhere from 1 to 4 cm. Again we did not break it down in terms
710 Candage et al
of the exact amount of overlap in the fascia that we used to see if there was any evidence of increased recurrence with that degree of overlap. The follow-up is something interesting in terms of there was longer follow-up in the bridged group, which was on average about 12 months versus about 5 months in those who had undergone a reinforced repair. That may account for some of the difference. I think the majority of those who will experience a recurrence will do so in the first year. But in our series, at least in the bridged group, we found a recurrence as far out as 37 months and in those undergoing reinforced repair we found 1 out as far as 22 months now. So I think it emphasizes the point that these patients at least require very long-term follow-up for repair of these types of complex wounds with this particular type of mesh. In terms of using the biologic mesh to bridge a defect, I think the data are now starting to show that bridging these repairs probably is going to result in a higher rate of hernia recurrence. And whether or not it should be used at all is difficult to say. It certainly proves its usefulness with repair in infected wounds. In terms of component separation, there was about an equal number, I think 7 in the bridged group, 6 in the reinforced repair, who had undergone a component separation in the series. When we kind of looked at that subset of patients, component separation did not seem to make any difference with respect to them having a hernia recurrence. Dr L. Michael Brunt (St. Louis, Mo): I enjoyed the study and congratulate you on reporting this large series of patients. I would like some clarification about your follow-up. Because you mentioned in your slide you had phone follow-up. For following patients who have had a hernia defect repaired—examination follow-up in the office is extremely important. So can you clarify for us how many of your patients had follow-up with examination and what that link to follow-up was? Also, can you give us the percent recurrence rate in the patients who had bridged versus repairs that were over defects that were closed primarily? You showed the absolute numbers, but what was the percent recurrence? And then finally, you also showed that patients had about an average of 2 Alloderm sheets, but what were the sizes of those? What size defects were you actually bridging? Dr R. L. Candage (Maywood, Ill): To answer the first question in terms of the patient follow-up, again, we had an average of about 12 months for those undergoing the bridged repair and about 5 months for those undergoing the enforced repair. There were 15 patients that we contacted by telephone who had not been seen in the clinic in a while. So those patients we defined as having a hernia recurrence if they had either gone to another institution and had a hernia repair there, or if they had noticed a significant bulge since the last time they showed up at the clinic and they had just been sitting at home choosing not to do anything about it because they were not having any
Surgery October 2008
symptoms. Again, the time period for follow-up, the longest patient we have out for the bridged is about 37 months and for the reinforced repair right now it is about 22 months. In terms of the percent recurrence, the percent recurrence for bridged repair was about—I think the overall recurrence rate was around 28% for our study. In terms of the breakdown of the percentages, I think out of the 13 patients who did have a recurrence in the series, probably 80% were bridged and about 20% were in the reinforced group. So that would be the breakdown. And the size of the Alloderm, in terms of the sheets, all the sheets were thick. In terms of the size of the sheets, we did not look at that specifically. We did do a subset analysis on the patients in which the surgeon had commented on the size of the defect itself, the hernia defect. We did not find any association between the size of the hernia defect and recurrence. So that much I can say. But in terms of the actual breakdown for the size of the Alloderm mesh, I do not have. Dr Lawrence N. Diebel (Detroit, Mich): Excellent presentation. I just want to ask a question, because I just had a patient last week who came back with a recurrence after I used this stuff. I reoperated, and you can see the proline sutures very nicely there with nothing left in between. I said to the rep, ‘‘Your stuff dissolved,’’ and she said, and I quote, ‘‘No, it just well incorporated.’’ Why should this material handle differently as a bridge versus as a reinforcement? Is there something different about the way the physiologic properties of the wound on this material whether it is bridged? Or is it just that the tissue that you are using actually is sufficient to the patient’s own tissue and whether you left the Alloderm or not is irrelevant? And then the question is, based on my recent experience, why does this recur? Do they all eventually dissolve over time or not? Dr R. L. Candage (Maywood, Ill): To answer the second question first, the material is known to degrade over time. It is reabsorbed. There are a number of theories as to why that occurs. Some think that bacteria coagulate in the wound and in the setting of infection they contribute to the breakdown of this material. But it is eventually reabsorbed over time. I think that in terms of the recurrence in the tissue properties and things of that nature, and maybe a combination of both, I think intuitively it makes some sense that usually if you are going to use this as a reinforced repair you are probably dealing with somewhat smaller defects in the sense that you are able to completely reapproximate that fascia in those tissues and get them back together, whereas when you are dealing with it a bridged manner you are clearly dealing with a large abdominal defect and that may account for some of the reason that you see the recurrence rates such as they are, although again we could not find any direct correlation between the defect size and hernia recurrence per se. But I think that there is some play with that, and I think that some of it is just due to the inherent nature of
Surgery Volume 144, Number 4
the patient’s own tissue. If you can reapproximate that fascia, you are probably going to get a stronger repair. And I think that intuitively makes some sense compared with when you are not able to completely bring that fascia back together. Dr R. Matthew Walsh (Cleveland, Ohio): We do actually have some data from the clinic that are almost identical. I guess my question is, how actually are you going to use this? I am not clear about that. Because if you are going to go in and decide that you are going to use it as a bridge, are you really going to use the Alloderm? Or just to use your friend analogy, why not go with the cheapest date? Because if you look at the cost data, it is really quite prohibitive. Could you tell us
Candage et al 711
what your costs were versus what you actually got paid for some of these? Dr R. L. Candage (Maywood, Ill): That we did not look at—cost—specifically. I know the cost per sheet is expensive. I am not sure exactly what the numbers are, but I think it is like a 6-by-12 sheet or something, runs around $1,200 or something. So it is certainly not cheap. I think in the majority of the patients the decision whether or not to use was probably just made at the time of surgery. I think in the small number of patients that we were able to completely reapproximate the fascia, only about 13 patients or so, but I think in the majority of cases it ends up being made intraoperatively whether or not the attending surgeon of record decided to use the mesh or not.