Perioperative outcomes andcomplications of laparoscopicventral hernia repairJuan M. Perrone, MD, Nathaniel J. Soper, MD, J. Christopher Eagon, MD, Mary E. Klingensmith, MD,Rebecca L. Aft, MD, Margaret M. Frisella, RN, and L. Michael Brunt, MD, St. Louis, Mo

Background. Laparoscopic techniques are being used increasingly in the repair of ventral hernias andoffer the potential benefits of a shorter hospital stay, decreased wound complications, and possibly a lowerrecurrence rate. Despite good results from high-volume centers, significant complications may occur withthis approach and the morbidity of incisional hernia repair may be underestimated. The purpose of thisstudy was to review our experience with laparoscopic ventral hernia repair (LVHR) since its inception atour institution.Methods. Medical records of all patients who underwent LVHR at a single institution from May 2000through December 2003 were reviewed. Preoperative and postoperative variables including complicationswere analyzed. Follow-up evaluation was by office visit and phone survey with assessment of patientsatisfaction scores. Data are expressed as mean ± SD.Results. A total of 121 LVHR were performed in 116 patients (52 men, 64 women; mean age, 57 ± 13 y;mean body mass index, 35 ± 8). Hernias were recurrent in 35 cases (28.9%), with a mean of 1.4 priorrepairs (range, 1--7). The mean defect size was 109 ± 126 cm2 and the average mesh size used was 256 ±192 cm2. Operating time was 147 ± 45 minutes, and the hospital stay averaged 1.7 ± 1 days. Twelvecases (9.9%) were converted to open operation, most commonly because of extensive adhesions. Extensivelaparoscopic adhesiolysis was necessary in 29 cases (26.6%). Overall, perioperative complicationsoccurred in 33 cases (27.3%), 13 of which (39.3%) were persistent seromas. Major complications wereseen in 9 cases (7.4%). There were 4 enterotomies (3.3%): 3 occurred as a result of adhesiolysis and1 resulted from a trocar injury; 2 were detected intraoperatively and were converted to open operationand 2 presented postoperatively. One of these patients developed sepsis and died. Follow-up evaluationwas available for 83.6% of cases at a mean interval of 22 ± 16 months after repair. The herniarecurrence rate was 9.3% (9 cases) and was detected at a median of 6 months postoperatively. Theoverall patient satisfaction score was high at 4.3 ± 1.1 (scale, 1–5).Conclusions. Laparoscopic repair is effective for the vast majority of patients with primary or recurrentventral hernias and results in hernia recurrence rates of less than 10%, with high patient satisfactionscores. Although seroma is the most common complication, major morbidity occurred in 7.4% of thepatients in our series. Enterotomy is the most common serious complication and may result in sepsis anddeath. (Surgery 2005;138:708-16.)

From the Department of Surgery and Institute for Minimally Invasive Surgery, Washington University Schoolof Medicine, St. Louis, Mo

INCISIONAL HERNIA is one of the most common com-plications of abdominal operations and occursin 10% to 20% of patients undergoing midline

Presented at the 62nd Annual Meeting of the Central SurgicalAssociation, Tucson, Arizona, March 10-15, 2005.

Reprint requests: L. Michael Brunt, MD, Department of Sur-gery, Washington University School of Medicine, 660 S. EuclidAve., Campus Box 8109, St. Louis, MO 63110. E-mail:bruntm@wustl.edu.

0039-6060/$ - see front matter

� 2005 Mosby, Inc. All rights reserved.

doi:10.1016/j.surg.2005.06.054

708 SURGERY

laparotomy.1,2 Effective operative therapy for ven-tral and incisional hernias is problematic because10-year cumulative rates of recurrence are as highas 63% for suture repair and 32% formesh repair.3,4

As an alternative, laparoscopic techniques are beingused increasingly because of the potential benefitsof a shorter hospital stay, decreased wound compli-cations, and possibly a lower recurrence rate.5 Thelaparoscopic approach to incisional hernia repairuses the principles learned from conventionalrepairs. These include the use of a large mesh pros-thesis that overlaps the hernia defect onto severalcentimeters of normal fascia, and eliminating

SurgeryVolume 138, Number 4

Perrone et al 709

tension. Despite good results from high-volumecenters, significant complications may occur withthis approach and the morbidity of incisionalhernia repair in general may be underestimated.The purpose of this study, therefore, was to reviewthe experience with laparoscopic ventral herniarepair (LVHR) since its inception at our institutionwith an emphasis on complications.

PATIENTS AND METHODS

Patients. Medical records of all patients whounderwent LVHR at the Washington UniversityMedical Center/Barnes-Jewish Hospital from May2000 through December 2003 were reviewed ret-rospectively. The study was approved by the insti-tutional Human Studies Committee. Patients withventral hernias were selected for a laparoscopicapproach based on the preference of the operat-ing surgeon. LVHR was not attempted in cases ofloss of domain, strangulated intestine, in children,or in patients who could not tolerate generalanesthesia. Of the patients undergoing LVHR, 3(2.4%) had primary umbilical or epigastric herniasand the remainder had incisional or recurrentumbilical hernias. LVHR was performed by 5different surgeons with a mean of 24 ± 14 casesper surgeon (range, 6--41 cases).

For each patient, the following demographic,perioperative, and postoperative data were col-lected: age, sex, body mass index (BMI), numberof previous abdominal operations and herniarepairs, American Society of Anesthesiologists(ASA) classification, size of the fascial defect, sizeand type of the prosthetic mesh implanted, oper-ating time, length of hospital stay, perioperativeand postoperative complications, and hernia re-currences. Obesity was defined as a BMI of 30 orgreater and morbid obesity was defined as a BMI of40 or greater. Extensive adhesiolysis was defined asthat which took 45 minutes or more to complete.Patients converted to an open operation wereincluded in the analysis of perioperative resultsand complications as intent to treat but wereexcluded from the long-term follow-up analysis.Complications were graded according to the clas-sification system of Clavien et al6: grade 1 (minor),grade 2a (those that are life threatening withoutresidual disability and do not require invasive treat-ment), grade 2b (those that are life threateningwithout residual disability and require invasivetreatment), grade 3 (result in residual disability,eg, myocardial infarction), and grade 4 (deaths).

Operative technique. All patients received per-ioperative antibiotic therapy, most commonly witha first-generation cephalosporin. Patients were

positioned supine and, in most cases, bladderand gastric decompression was performed. Accessto the abdomen was obtained well away from thehernia, typically in either the subcostal area or thelateral abdomen lateral to the rectus sheath. In88% of cases a closed technique with a needle(Veress needle) was used for initial access. An opentechnique was used when a closed techniquefailed, usually because of adhesions. An angled(30�) 5-mm laparoscope was used.

After exploration of the abdomen, additionaltrocars were placed under direct visualization as farfrom the hernia defect and as lateral as possible.An adhesiolysis with minimal use of cautery thenwas performed to free the anterior abdominal wall.The falciform ligament was transected when nec-essary and the fatty tissue around it was removedcompletely in cases of upper abdominal hernias toallow adequate apposition of the mesh to thefascia. The hernia sac contents were reduced, butthe peritoneal sac was left in situ. The defect(s) wassized externally after decreasing the intra-abdom-inal pressure to 8 mm Hg or less. The margins ofthe defect then were measured externally on theabdominal wall in 2 dimensions to determine themaximum distances from the most cephalad tocaudad extent and between the lateral margins.For multiple defects, the distance between thesuperior-most defect and the inferior-most defectwas used to calculate the hernia defect size. Amesh size was selected to provide at least 3 cm ex-tension beyond the margins of the hernia defect.The surface areas of both the hernia defects andthe mesh were calculated by multiplying the maxi-mal 2-dimensional measurements obtained. Her-nias also were stratified into 3 defect sizes (small< 25 cm2, medium = 26-100 cm2, large > 100 cm2).

A variety of meshes were used for the repairsaccording to surgeon preference, most commonlypolytetrafluoroethylene or a polytetrafluoroethy-lene/polypropylene composite. The mesh wasintroduced through a 10-mm trocar site or, alter-natively, using a 2-cm incision over the defect. Themesh was anchored in place in a standard fashionusing nonabsorbable fixation sutures placed atequidistant points along the mesh. The circumfer-ence of the mesh then was stapled to the posteriorfascia at intervals of 1 cm. In some cases early inthe experience the mesh was secured with tacksonly without fixation sutures. The use of transfix-ion sutures and the number of sutures placedvaried according to surgeon preference.

Follow-up evaluation. Patients were examinedpostoperatively at 2 to 4 weeks, 3 to 6 months,and thereafter as clinically indicated. Follow-up

SurgeryOctober 2005

710 Perrone et al

evaluation was based on a combination of review ofclinic visits and phone interviews. Phone follow-upevaluation was performed by 2 reviewers using astandard questionnaire. Patients with any com-plaints possibly related to the hernia repair onphone follow-up evaluation were encouraged toreturn to the clinic for examination. Only patientswith a minimum 3-month clinical examinationfollow-up were included in the analysis of follow-up results. Patient satisfaction after operation wasassessed using a standard 5-point Likert scale.

Statistics. All proportional data were presentedas percentages and were analyzed using chi-squaretests. All continuous data are given as mean ± SDand were compared using a 2-tailed, unpairedStudent t test. Statistical calculations were com-pleted using statistical software (Graph Pad Prism3.0; San Diego, Calif). A P value of less than .05was considered significant.

RESULTS

Patient characteristics. The demographic fea-tures of the 116 patients undergoing LVHR arelisted in Table I. The mean patient age was 57 ± 13years with 10% more women than men. Sixty-sevenpercent of our patients had a BMI of 30 or greaterand 22%weremorbidly obese (BMI$ 40). The her-nias repaired were recurrent in 35 cases (28.9%)with a mean of 1.4 prior repairs (range, 1–7 previ-ous repairs). Five patients had a recurrence afterLVHR at our institution that was re-repaired lapa-roscopically for a total of 121 laparoscopic repairsin these 116 patients.

Operative results. Characteristics of the herniasrepaired are shown in Table II. The most commonhernia sites were midline and periumbilical. Themean defect size as measured at operation was115 ± 126 cm2. The average mesh size used was257 ± 191 cm2, for a mesh to defect ratio of 2.2.In 35 cases (28.9%), more than 1 hernia was foundat the time of operation, even when the abdominalexamination showed a single defect. Unless therewere separate remote defects, a single large pieceof mesh was used for the repair. However, 12 cases

Table I. Patient demographics

N 116Age, y (range) 57 ± 13 (26-93)Sex (male/female) 52/64BMI (range) 35 ± 8 (19-59)(ASA) 2.1 ± 0.4Patients with prior repairs 35 (28.9%)Mean number of prior repairs (range) 1.4 (1-7)

(9.9%) needed 2 pieces of mesh, which in eachcase was caused by separate defects. Of the 109procedures completed laparoscopically, the meshwas anchored with spiral tacks only and no transfix-ion sutures in 35 cases (32.1%). The primary vari-able in the use of transfixation sutures was surgeonpreference, ranging from 26% to 97% of cases.The mean number of transfixion sutures placedper repair was 5.6 ± 2 (range, 2–12).

The operating time was 143 ± 40minutes for first-time repairs and 157 ± 56 minutes for recurrenthernias (P = .15). Beforemesh placement, an exten-sive adhesiolysis (>45 min) was needed in 29patients (26.6%), accounting for the cases withthe longer operating time. Conversion to an openoperation was necessary in 12 cases (9.9%). Reasonsfor conversion were severe adhesions (8 patients),difficulty obtaining adequate fixation of the mesh(2 patients), and enterotomy (2 patients). The post-operative hospital stay averaged 1.7 ± 1 days (range,1-18 d; median, 1 d). One patient who had a pro-longed hospitalization of 55 days because of Can-dida sepsis was excluded as an outlier from thecalculation of the mean hospital stay. In evaluatingoutcomes in difficult groups, a trend toward longeroperating times and higher conversion and recur-rence rates was observed in morbidly obese patients(Table III). Patients with recurrent hernias also hada higher rate of conversion to open operation(17.4%) compared with repair of primary hernias(7%), but hernia recurrence rates were similar(9.4% primary vs 8% recurrent). However, thesedifferences were not statistically significant.

Table II. Hernia characteristics and perioperativeresults

N 121Hernia locationMidline 91 (75%)Periumbilical 11 (9.1%)Right upper quadrant 8 (6.6%)Right lower quadrant 4 (3.3%)Left upper quadrant 4 (3.3%)Left lower quadrant 3 (2.5%)

Hernia typePrimary umbilical/epigastric 3 (2.5%)Primary incisional 83 (68.6%)Recurrent umbilical 9 (7%)Recurrent incisional 26 (21.4%)Hernia size, cm2 115 ± 126Mesh size, cm2 257 ± 191Transfixion suture use 77 (69%)Operative time, min 146 ± 45Postoperative hospital stay, d 1.7 ± 1

SurgeryVolume 138, Number 4

Perrone et al 711

Complications. Overall, perioperative complica-tions occurred in 33 of 121 cases (27.3%) (TableIV). Most of these complications were grade1 and required no or minimal intervention. Se-roma at the site of the retained hernia sac wasthe most common consequence of LVHR, occur-ring in 24 cases (19.8%). In 13 patients (10.7%),the seroma persisted for more than 8 weeks post-operatively or created symptoms that warrantedintervention. One patient required multiple aspi-rations and placement of a percutaneous drainbefore the seroma resolved. No long-term compli-cations caused by seromas were observed, whetherthey were aspirated or not. Other minor complica-tions included prolonged ileus in 1 case (0.8%)and 3 cases (2.4%) of minor bleeding at a trocarinsertion site that required no specific treatment.

Hernia repair--related infections occurred in 6cases (5%). Two patients developed superficialcellulitis and 1 patient had a superficial trocar siteinfection, all of which responded to antibiotictherapy. Three other patients developed meshinfections that required removal of the mesh: 1 pa-tient developed an abscess beneath the mesh 4months postoperatively, 1 patient developed a sub-cutaneous abscess that led to mesh infection andremoval several months postoperatively, and 1 pa-tient developed Candida sepsis from an apparenturinary source that led to Candida peritonitis andmesh removal 13 days postoperatively and a pro-longed hospitalization of 55 days. Each of the pa-tients who required mesh removal ultimatelydeveloped a recurrent hernia.

Four patients (3.3%) had pain that persisted formore than 12 weeks. Each of these patients hadtransfixion sutures used in the repair and theirpain was usually at one of the suture sites. Patientswith prolonged pain were treated initially with anti-inflammatory medications but local injectionswere required in 2 cases and 1 patient required

Table III. Outcomes in nonobese compared withobese patients

Nonobese(BMI <30),n = 41

Obese(BMI 30 to <40),

n = 51

Morbidly obese(BMI >40),n = 29

BMI* 26 ± 3 33.7 ± 2.5 44.8 ± 5.9Operating

time, min138 ± 42 146 ± 50 156 ± 40

Complications 25% 28% 25%Conversions 7% 9% 14%Recurrences 10% 6% 16%

*P < .0001 between groups. All other comparisons are not significant.

removal of an anchoring suture with resolution ofpain.

Major complications (grade 2b or higher) wereseen in 9 patients (7.4%). Three of these compli-cations were infected meshes that required re-moval in all cases as noted earlier. Two patients(1.6%) had cardiac complications; 1 of them had amyocardial infarction that required catheterizationand placement of an internal automatic defibrilla-tor that was associated with a prolonged hospitalstay of 18 days. Another patient was readmitted7 days after an attempted laparoscopic repair thatwas converted to an open repair because of adhe-sions with an acute myocardial infarction and renalfailure from which recovery was uneventful.

Four enterotomies (3.3%) occurred from LVHR:1 was related to the initial access and the other 3resulted from the adhesiolysis. All 4 enterotomieswere in patients who had prior ventral herniarepairs, for an enterotomy rate of 11.4% in thatgroup compared with 0% in patients undergoingfirst-time repair of a ventral hernia (P < .01). In1 patient a through-and-through injury of thetransverse colon was made on entering the abdo-men with a 5-mm trocar. The colon injury was re-paired primarily after conversion to laparotomyand the hernia was repaired in an open fashionwithout mesh. In another patient an enterotomyoccurred during an extended adhesiolysis andwas treated by laparotomy and segmental ileal re-section with anastomosis. Because of spillage ofbowel contents, mesh repair was delayed until 6months later without further complications. Athird patient presented with an abscess and colocu-taneous fistula 3 weeks after a laparoscopic repairand extensive adhesiolysis. Subsequently, explora-tory laparotomy, mesh excision, adhesiolysis,

Table IV. Complications of LVHR in 121 cases

Type N (%)

Minor (grade I or IIa) 24 (19.8%)Prolonged or symptomatic seroma 13 (10.7%)Prolonged pain 4 (3.3%)Superficial surgical site infection 3 (2.5%)Trocar site bleed/hematoma 3 (2.5%)Prolonged ileus 1 (0.8%)

Major (grade IIb or higher)* 9 (7.4%)Deep surgical site infection

(mesh removal)3 (2.5%)

Cardiac (myocardial infarction) 2 (1.6%)Enterotomy 4 (3.3%)Death* 1 (0.8%)

*From enterotomy, sepsis, and multiorgan failure.

SurgeryOctober 2005

712 Perrone et al

Table V. Results of laparoscopic ventral hernia repair from recent single-institution series

Complications

Study N ConversionLength ofStay, d Recurrence

Follow-upevaluation, mo Overall Seroma Infection Enterotomy

LeBlanc et al8 100 4% 1 9% 51 14% 7% 2% 1%Ben-Haim et al9 100 7% 5 2% 19 24% 11% N/A 6%Carbajo et al10 270 0.3% 1.5 4.4% 44 25.3% 11.8% N/A 3.3%Rosen et al11 114 12% 1.8 17.7% 30 14% 4% 6% 1%Bencini and Sanchez12 64 6.2% 4.9 4.7% 19 26.5% 12.5% N/A 9.3%Moreno-Egea et al13 86 3.3% N/A 3.5% 42 23.2% 5.8% 1% 1%Ujiki et al14 100 3% 2 6% 3 23% 13% 4% 2%Present series 121 9.9% 1.7 9.3% 22 26.3% 10.7% 5% 3.3%Total/mean 955 5.0% 2.5 6.9% 31.4 22.4% 9.8% 3.6% 3.1%

N/A, Not available.

and transverse colectomy with anastomosis was re-quired. The abdominal wall was closed primarilyand the patient was without recurrence at the 12-month follow-up evaluation. Finally, 1 patient whowas discharged on postoperative day 1 tolerating adiet was readmitted to the hospital several hourslater with abdominal pain and evolving sepsis. Surgi-cal exploration showed a small-bowel injury and anischemic colon. Despite resection, the patient devel-oped multiorgan system failure and died 2 monthspostoperatively. All 3 patients whose enterotomiesresulted from adhesiolysis had prior hernia repairswith mesh and at the time of operation were foundto have dense adhesions to the mesh.

Follow-up and hernia recurrences. Follow-updata based on office visit and/or telephone inter-view were available for 97 of 116 patients (83.6%) ata mean interval of 22 ± 16 months after repair.Overall patient satisfaction score was high at 4.2 ±1.1 (scale, 1--5). In the 97 patients with a minimumof 3 months of follow-up (mean clinical examina-tion follow-up period, 14 mo), the hernia recur-rence rate was 9.3% (9 cases). When analyzed byhernia size, patients with large hernias had a trendtoward a higher recurrence rate (6 of 48 cases,12.5%) than did patients with small (1 of 34 cases,2.9%) or medium-size hernias (2 of 39 cases, 5.1%).However, these differences were not statisticallysignificant. Recurrences were detected at a medianof 6 months postoperatively. Because these patientswere all reoperated on at our institution, themechanism of failure could be assessed and mostcommonly was caused by separation of the meshfrom the posterior fascia that resulted in a herniabetween the fascia and the edge of the mesh.Although the numbers of patients are small, therecurrence rate among the group in which

transfixion sutures were used was 9.2% and forthe group in which tacks only were used was 8.5%(P = .85).

DISCUSSION

LVHR is being used increasingly in the manage-ment of patients with incisional hernia. The prin-ciples of the laparoscopic approach are similar tothe open Stoppa7 repair in which a large piece ofmesh is placed in a retrorectus position well beyondthe margins of the facial defect. However, withLVHR the mesh is placed 1 layer deeper (intraperi-toneal as opposed to above the posterior rectus fas-cia). Because the operation is performed through 3or 4 trocars, the need for a long incision and exten-sive fascial or skin flap dissection is eliminated.Therefore, when the laparoscopic technique isused, one of the most morbid aspects of ventralherniorrhaphy potentially is eliminated. Our expe-rience with 121 cases of LVHR depicts a low conver-sion rate, short hospital stay, and an acceptablecomplication and recurrence rate. The averagelength of stay in our patients was 1.7 days, and issimilar to what has been reported in previous series(Table V).8-14 Length of stay for open procedurestypically has ranged from 3 to 9 days.15-17

The laparoscopic approach appears to be effec-tive in complex patients, such as the obese andthose who have failed prior open repairs. Obese pa-tients especially may benefit because of the smallerwounds and, theoretically, decreased wound com-plications. Some investigators have found adverseoutcomes inmorbidly obese patients.5 In our series,a trend was seen toward longer operating times andhigher conversion and recurrence rates in the mor-bidly obese patients, but these differences were not

SurgeryVolume 138, Number 4

Perrone et al 713

statistically significant because of the small num-bers of patients. For patients undergoing repair ofa recurrent hernia, the average operative time wasonly 14 minutes longer than those who had a first-time LVHR; patients with recurrent hernias did,however, have a higher rate of conversion to openoperation.

The most common complication in our series of121 repairs was a symptomatic or prolonged se-roma, which occurred in 10.7% of patients. Se-romas are a common consequence of the LVHRtechnique because the hernia sac is not dissectedout and the hernia space is sequestered fromthe abdominal cavity by the mesh. In 1 series,18 sys-tematic ultrasound testing detected postoperativeseromas in 93% of patients. Because of the fre-quency with which this problem occurs, othershave questioned its classification as a complica-tion.19 Most of the seromas in our series weresmall, self-limited fluid collections that were man-aged expectantly and required aspiration in only6 cases (5% of repairs). Aspiration should be re-served for those that persist, are symptomatic, orfor which there is diagnostic uncertainty.

Despite the potential benefits of a reduction inwound complications from a laparoscopic ap-proach, 5% of patients in our series had infectionsrelated to the hernia repair including 3 cases thatrequired mesh removal. The patients who hadtheir mesh removed ultimately each redevelopeda hernia. To what extent these infections wererelated to the hernia repair technique as opposedto patient factors and associated comorbidities isunclear. However, perioperative measures to min-imize the risk of infection should include elimi-nating potential sources of infection beforeoperation, use of perioperative antibiotics, limit-ing the contact of mesh with the skin, and carefulattention to abdominal wall hygiene and skincare.

Prolonged postoperative pain greater than 6 to12 weeks after the operation has been reported in1.6% to 3.5% of patients undergoing LVHR5,16,20

and occurred in 3.3% of patients in our series. Ineach of our cases, the primary pain site was near1 or more suture anchor sites. Persistent painalso may precipitate further diagnostic evaluationincluding imaging to exclude other possiblecauses. Initial management should be conservativewith the administration of anti-inflammatory med-ications, application of ice to the affected area, andlocal injection of anesthetic and/or steroids. Re-moval of the offending suture may be necessaryfor persistent symptoms that do not respond toconservative treatment.

Enterotomy is a known complication of inci-sional hernia repair and has been reported tooccur in up to 5% in open series,2,7 and in 1% to9.3% with the laparoscopic approach.8-14 The 4 en-terotomies that occurred in our series were all inpatients undergoing repair of recurrent hernias,3 of whom had prior mesh repairs. These findingssuggest that patients with recurrent hernias, espe-cially if mesh has been placed previously, shouldbe approached cautiously. Enterotomies that werenot apparent at operation appear to be associatedwith the most serious morbidity and have led tomortality in our series and in other reports.13,18,20,21

If an enterotomy occurs during adhesiolysis, it maybe possible to repair it laparoscopically dependingon the nature and severity of the injury, althoughboth cases in our series required conversion toan open repair. Options for management of thehernia if an enterotomy occurs include abortingthe repair and primary suture repair or, if thattechnically is not possible, closure with an absorb-able material such as mesh (Vicryl Ethicon, Somer-ville, NJ) followed by repair of the hernia severalmonths later. The use of conventional prostheticmesh in the face of an enterotomy in generalshould be avoided because of the potential formesh infection. Alternatively, some groups haveused a bioremodelable material in a contaminatedsetting to try and achieve a durable repair withoutcompromising the infection risk.22 Because of thispotential risk and the management difficulties as-sociated with enterotomy during LVHR, patientsshould be counseled about this possibility andthe management options preoperatively. Regard-less, any patient who deviates from the expectedoutcome after LVHR with complaints of abdomi-nal pain, distension, or fever should be evaluatedaggressively for possible enterotomy, especially ifan extensive adhesiolysis has been performed.

Incisional hernias have the highest rate of recur-rence after repair of any abdominal-wall hernia.Even in the modern era, a recent prospectivecontrolled trial reported recurrence rates of 32%after open incisional hernia repair with mesh.3 Sev-eral factors may impact on the risk for incisionalhernia recurrence including the size and extent ofdefects, obesity, number of previous repairs, periop-erative complications, and other factors. Whetherthe laparoscopic approach results in a lower long-term recurrence rate currently is unclear, however,most laparoscopic series have reported recurrencerates in the 2% to 17% range.9-11 Potential advan-tages of the laparoscopic approach from the stand-point of adequacy of the hernia repair include theability to detect multiple or Swiss-cheese type defects

SurgeryOctober 2005

714 Perrone et al

(present in 29% of patients in our series) and broadanchorage of the mesh well beyond the margins ofthe fascial defect.

Despite these potential advantages, recurrencesdeveloped in 9.3% of cases in our study (notincluding the 3 cases who had infected meshremoved) and likely were related to a combinationof factors as noted earlier including patient factors,postoperative abdominal stressors, use of mesh thatwas too small, size of the hernia, inadequate fixa-tion of the mesh, or surgeon inexperience. One ofthe most critical technical points of the laparo-scopic repair that may impact the rate of recurrenceis the method of mesh fixation. A variety of tech-niques were used to anchor the mesh to theabdominal wall in our series owing to individualsurgeon preference. Tacks were used uniformly butthe use of transfixion sutures varied by surgeonfrom a low of 26% of cases to a high of 97%. In arecent experimental study from our institution,23

fixation sutures did not appear to result in greatermesh fixation strength when a combination meshof polytetrafluoroethylene and polypropylene witha rapidly incorporating anterior surface was used.However, others have suggested the importance offixation suture anchorage at 4- to 5-cm intervalscircumferentially around the mesh to minimizethe risk for mesh migration.5,8 The uncontrolledretrospective nature of this study limits any defini-tive conclusions about the preferred technique ofmesh fixation. Ultimately, prosthetic materials areneeded that will shrink or contract less while allow-ing firm incorporation into the abdominal wall butwithout a resultant increase in adhesions to theunderlying viscera to both decrease the hernia re-currence risk while reducing the need for extensivetransabdominal fixation.

CONCLUSIONS

In summary, this study shows that laparoscopicrepair of ventral and incisional hernias is associatedwith a low rate of conversion, a short hospital stay,and a recurrence rate of less than 10% over amedium-term follow-up period. Additionally, theprocedure appears to be effective in difficult groupssuch as obese patients and those who have failedprevious open repairs. Although seroma is themost common complication, major morbidity oc-curred in 7.4% of cases in our series. Enterotomyis the most common serious complication andmay result in sepsis and death if recognition or pre-sentation is delayed. Patients who have undergoneprevious ventral hernia repair, especially in the set-ting of prior mesh placement, appear to be at great-est risk for enterotomy.

The authors gratefully acknowledge the support ofthe Washington University Institute for Minimally Inva-sive Surgery in the completion of this study.

REFERENCES

1. Mudge M, Hugues LE. Incisional hernia: a 10-year pro-spective study of incidence and attitudes. Br J Surg1985;72:70-1.

2. Winslow ER, Fleshman JW, Birnbaum EH, Brunt LM.Wound complications of laparoscopic vs open colectomy.Surg Endosc 2002;16:1420-5.

3. Luijendijk RW, Hop WC, van den Tol P, et al. A comparisonof suture repair with mesh repair for incisional hernia.N Engl J Med 2000;343:392-8.

4. Burger JW, Luijendijk RW, Hop WC, Halm JA, VerdaasdonkEG, Jeekel J. Long-term follow-up of a randomized con-trolled trial of suture versus mesh repair of incisional her-nia. Ann Surg 2004;240:578-85.

5. Heniford BT, Park A, Ramshaw BJ, Voeller G. Laparoscopicrepair of ventral hernias: nine years� experience with 850consecutive hernias. Ann Surg 2003;238:391-400.

6. Clavien PA, Sanabria JR, Strasberg SM. Proposed classifica-tion of complications of surgery with examples of utilityin cholecystectomy. Surgery 1992;111:518-26.

7. Stoppa RE. The treatment of complicated groin and inci-sional hernias. World J Surg 1989;13:545-54.

8. LeBlanc KA, Booth WV, Whitaker JM, Bellanger DE. Lapa-roscopic incisional and ventral herniorrhaphy in 100 pa-tients. Am J Surg 2000;180:193-7.

9. Ben-Haim M, Kuriansky J, Tal R, Zmora O, Mintz Y, Rosin D,et al. Pitfalls and complications with laparoscopic intraperi-toneal expanded polytetrafluoroethylene patch repair ofpostoperative ventral hernia. Lessons from the first 100 con-secutive cases. Surg Endosc 2002;16:785-8.

10. Carbajo MA, Martin del Olmo JC, Blanco JI, Toledano M,de la Cuesta C, Ferreras C, et al. Laparoscopic approachto incisional hernia. Surg Endosc 2003;17:118-22.

11. Rosen M, Brody F, Ponsky J, Walsh RM, Rosenblatt S, Du-perier F, et al. Recurrence after laparoscopic ventral her-nia repair: a five-year experience. Surg Endosc 2003;17:123-8.

12. Bencini L, Sanchez LJ. Learning curve for laparoscopicventral hernia repair. Am J Surg 2004;187:378-82.

13. Moreno-Egea A, Torralba JA, Girela E, Corral M, Bento M,Cartagena J, et al. Immediate, early, and late morbiditywith laparoscopic ventral hernia repair and tolerance tocomposite mesh. Surg Laparosc Endosc Percutan Tech2004;14:130-5.

14. Ujiki MB, Weinberger J, Varghese TK, Murayama KM, JoehlRJ. One hundred consecutive ventral hernia repairs. Am JSurg 2004;188:593-7.

15. DeMaria EJ, Moss JM, Sugerman HJ. Laparoscopic intra-peritoneal polytetrafluoroethylene (PTFE) prostheticpatch of ventral hernia: prospective comparison to openprefascial polypropylene mesh repair. Surg Endosc 2000;14:326-9.

16. Park A, Birch DW, Lovrics P. Laparoscopic and open inci-sional hernia repair: a comparison study. Surgery 1998;124:816-22.

17. Holzman MD, Purut CM, Reintgen K, Eubanks S, PappasTN. Laparoscopic ventral and incisional hernioplasty. SurgEndosc 1997;11:32-5.

18. Berger D, Bientzle M, Muller A. Postoperative complica-tions after laparoscopic incisional hernia repair. Incidenceand treatment. Surg Endosc 2002;16:1720-3.

SurgeryVolume 138, Number 4

Perrone et al 715

19. Susmallian S, Gewurtz G, Ezri T, Charuzi I. Seroma after lap-aroscopic repair of hernia with PTFE patch: is it really acomplication? Hernia 2001;5:139-41.

20. Eid GM, Prince JM, Mattar SG, Hamad G, Ikrammudin S,Schauer PR. Medium-term follow-up confirms the safetyand durability of laparoscopic ventral hernia repair withPTFE. Surgery 2003;134:599-604.

21. Salameh JR, Sweeney JF, Graviss EA, et al. Laparoscopic ven-tral hernia repair during the learning curve. Hernia 2002;6:182-7.

22. Silverman RP, Singh NK, Li EN, Disa JJ, Girotto JA, Slezak S,et al. Restoring abdominal wall integrity in contaminatedtissue-deficient wounds using autologous fascia grafts. PlastReconstr Surg 2004;113:673-5.

23. Winslow ER, Diaz S, Desai K, Meininger T, Soper NJ,Klingensmith ME. Laparoscopic incisional hernia repairin a porcine model. What do transfixion sutures add?Surg Endosc 2004;18:529-35.

DISCUSSION

Dr W. Scott Melvin (Columbus, Ohio). Ventral herniaand specifically recurrent ventral hernia are relativelycommon and have a high failure rate and morbidityrate. Yet there is no consensus on optimal treatment.

Eighty-four percent of your patients had some follow-up. That is a pretty good number. You mentioned thatsome of these patients had phone follow-up and somehad a physical exam follow-up. How many were seen andactually examined---was a difference in hernia recur-rence rate in this group?

I was surprised to see that you used a Veress needle in88% of the patients for initial access. I wanted to know ifthat is a practice you continue today.

You discussed the various techniques of mesh fixa-tion. While you weren’t able to demonstrate a differenceas far as recurrence, you did mention some otherparameters such as pain and perioperative complica-tions. Is there a difference between mesh fixation withstaples alone or staple and transfixion sutures and whatwould you view as the optimal treatment?

Lastly, the size of the hernia seems to make asignificant difference in the rate of recurrence. Youdid not address this. Your average size of defect wasabout 115 centimeters yet the standard deviation was 126centimeters. So obviously there is a wide range of herniasize that exists. I would be interested in hearing theresults stratified to the size of the hernia repair.

You have demonstrated quite nicely that certainconditions, including recurrence and obesity, are signif-icant factors in the success of ventral hernia repair.Should we reevaluate our surgical dogma that all ventralhernias need to be repaired and stay away from patientswho we know that are going to be at high risk of failure?

Dr Juan M. Perrone. All patients had at least 3 monthsof clinical exam follow-up, and the mean clinical examfollow-up was 14 plus or minus 3 months.

We used a Veress needle because probably it allows usto use 5-millimeter trocars. It is usually possible to obtainan adequate insertion site for this needle, without majorrisk. There is also no evidence that you will not make abowel injury with an open technique.

We used suture fixation in about 70% of cases. We didcompare the recurrence rate in patients that had tacksalone versus tacks plus suture, and this difference wasnot statistically significant. In terms of pain, all 4 patientsthat had persistent pain in our series had transfixionsutures.

We did not look at the relationship between the sizeof the hernia and postoperative outcomes.

Dr Allan E. Siperstein (Cleveland, Ohio). I think youmake a very important point in this paper that is oftenunderappreciated. This is a sick group of patients, oftenundergoing reoperative surgery and with significantcomorbidities.

We, too, have reported our series of patients from theCleveland Clinic undergoing laparoscopic ventral herniarepair, and I want to see whether your conclusions mirrorours. We found it interesting that half of our recurrenceswere evident after the first year but half of the recurrencescontinued over time.

The other interesting thing we found was that inthose patients who underwent conversion to an openprocedure, half recurred. I am interested in what thatnumber was in your group of patients.

Dr L. Michael Brunt. I don’t think we specificallylooked at the recurrence rate in the patients that wereconverted, which was 9.9% of patients in our series. Idon’t have the data.

I would like to address a couple of the issues that Dr.Melvin raised.

One is that clinical exam follow-up was 14 months inthese patients, and we got some additional phone follow-up. But it is hard to get patients to come back. It wouldbe ideal to have everyone come back and have extendedlong-term follow-up. That is the only way that you trulyknow what the long-term outcomes are.

In terms of the initial access, you usually can’t domidline initial access because the patients have had priormidline incisions and an open insertion is more difficultin a lateral position than it is at the midline. That is oneof the reasons that we tend to choose a closed technique.As long as you have a quadrant of the abdomen that isfree and relatively undisturbed, we think it should besafe to obtain initial closed access there.

I don’t think there is any question that patients whohave transfixion sutures used are going to have morepain. It is usually easy to manage in a conservativefashion, first with anti-inflammatory drugs and then, ifthat doesn’t work, with local injections. Occasionally youmay need to remove the transfixion stitch. Most of uswho are doing this operation feel that the use oftransfixion sutures is important in order to providesecure fixation of the mesh and lower your recurrencerate.

Dr Alan B. Loren (Arlington Heights, Ill). The pa-tients that were opened were opened for significant in-juries. How would you handle a very minor small bowelinjury with minimal or no spillage? Some authors haveadvocated repairing the bowel injury laparoscopically,followed by completion of the hernia dissection withoutplacement of the mesh. They would then return to the

SurgeryOctober 2005

716 Perrone et al

operating room 1 to 2 weeks later to place the mesh.Have you ever done that and would you recommendthat approach?

Dr L. Michael Brunt. If you have a minor injury thatyou observed laparoscopically, I would repair it butwould not do anything else and would come back at alater interval and repair the hernia in an open fashion.

Even though the incidence of enterotomy is low, it ispotentially a disastrous complication, particularly if it is

not recognized at operation. Patients with a bowelinjury may present initially in the postoperative periodin a somewhat subtle fashion. If they have got a bigpiece of mesh, they may not have obvious peritonealsigns initially but they may complain of abdominalpain, have fever, or have evidence of associated organdysfunction. It takes a high index of suspicion andaggressive investigation to intervene in these patientsearly.