World Journal of Surgery

, Volume 37, Issue 3, pp 538–544

Lateral Incisional Hernia Repair by the Retromuscular Approach with Polyester Standard Mesh: Topographic Considerations and Long-term Follow-up of 61 Consecutive Patients


    • Department of General, Digestive and Metabolic Surgery, Ambroise Paré Teaching HospitalUniversité Versailles Saint-Quentin, Assistance Publique - Hôpitaux de Paris
  • Tigran Poghosyan
    • Department of General, Digestive and Metabolic Surgery, Ambroise Paré Teaching HospitalUniversité Versailles Saint-Quentin, Assistance Publique - Hôpitaux de Paris
  • Nicola Corigliano
    • Department of General, Digestive and Metabolic Surgery, Ambroise Paré Teaching HospitalUniversité Versailles Saint-Quentin, Assistance Publique - Hôpitaux de Paris
  • Guillaume Canard
    • Department of General, Digestive and Metabolic Surgery, Ambroise Paré Teaching HospitalUniversité Versailles Saint-Quentin, Assistance Publique - Hôpitaux de Paris
  • Stephane Servajean
    • Department of General, Digestive and Metabolic Surgery, Ambroise Paré Teaching HospitalUniversité Versailles Saint-Quentin, Assistance Publique - Hôpitaux de Paris
  • Jean-Luc Bouillot
    • Department of General, Digestive and Metabolic Surgery, Ambroise Paré Teaching HospitalUniversité Versailles Saint-Quentin, Assistance Publique - Hôpitaux de Paris

DOI: 10.1007/s00268-012-1857-9

Cite this article as:
Veyrie, N., Poghosyan, T., Corigliano, N. et al. World J Surg (2013) 37: 538. doi:10.1007/s00268-012-1857-9



Because of the lack of published data and the relative rarity of lateral incisional hernia (LIH), their repair remains a major challenge for surgeons. The aim of the present study was to evaluate the outcome of LIH treated by the retromuscular approach (RMA) with a polyester standard mesh.


Sixty-one patients were treated between June 2000 and November 2007 in an academic tertiary referral center using one standardized surgical technique and one type of mesh. Lumbar incisional hernia was excluded. All data were prospectively culled. The early complications and recurrence rates were evaluated.


There were 14 (23 %) subcostal, 12 (19.6 %) flank, and 35 (57.4 %) iliac fossa LIH. The mean patient age was 57 years, and 60 % were male. The average width of the defect was 7.6 cm and the overall defect size averaged 56 cm². Seventeen patients (28 %) had had previous LIH repair. Ten patients had double hernia locations (midline and lateral) repaired simultaneously. The average operative time and hospital stay were 136 min and 7 days, respectively. The early complications rate was 18 %. Four patients required reoperation. There were no mesh infections. The median follow-up was 47 months (range: 1–125 months). Recurrence was observed in three patients (4.9 %).


LIH repair by RMA with a polyester heavyweight mesh proves to be a safe treatment with a moderate complication rate and a low infection rate, even in the treatment of large or multifocal parietal defects.


The term lateral incisional hernia (LIH) can be used to define any postoperative abdominal wall hernia outside of the midline incision. Anatomically, such hernias are limited by the 12th rib, the erector spinae, the iliac crest, and the lateral edge of the rectus muscle. Because of the absence of an appropriate classification and the relative rarity of this incisional hernia, authors publishing large series report LIH together with midline incisional hernia, without classifying the LIH according to topographic localization, e.g. iliac fossa, subcostal, flank, or lumbar incisional hernia [1, 2]. Recently, a new classification has been proposed by the European Hernia Society based on topography and width of the parietal defect, but this classification is rarely used [3].

The occurrence of LIH following laparotomy is much less frequent than is midline incisional hernia, ranging from 1 to 4 % versus 14 to 19 % [46]. The most plausible explanation lies in the topographic considerations of lateral incisions (i.e., transverse or oblique), which frequently follow the direction of the segmental blood vessels and nerves, thereby avoiding parietal devascularization and denerevation. In parallel, over the past two decades an increase of 8 % in incisional hernia on the port site was reported secondary to laparoscopic procedures [7, 8]. Few studies have been published, and to date there is neither sufficient evidence-based data nor guidelines concerning the management of lateral abdominal wall incisional hernias [9, 10]. Because of high recurrence rates following simple herniorrhaphy (exceeding 50 %) the use of a non-absorbable prosthesis in the repair of all types of hernia in elective surgery is currently accepted by all investigators [11, 12]. The different topographic variations of these defects and the various repair techniques used make it impossible to establish a gold standard procedure (i.e., type of prosthesis or insertion site) for the repair of this type of incisional hernia.

The aim of the present study was to evaluate the early morbidity and recurrence rate in a large cohort of LIH repaired by the retromuscular approach (RMA) with a polyester standard (heavyweight) prosthesis.

Patients and methods

After Institutional Review Board approval, a prospective database was created. All patients underwent operation in the same university teaching hospital, by four surgeons, each having at least 10 years of experience. Data of 440 patients operated on for ventral hernia in our department between June 2000 and November 2007 were culled: 61 patients were treated for LIH by RMA, including 10 with double hernias (midline and lateral) repaired simultaneously. The study was designed and retrospective analysis of the records was performed.

Parietal defects were classified according to their location: subcostal, flank, or iliac fossa. Lumbar incisional hernia repair was excluded from this study because a different surgical procedure is used by our team for the repair of this type of defect.

For each patient, demographic, perioperative, and postoperative data were collected. Inclusion criteria were patient age greater than 18 years and first or recurrent LIH measuring more than 2 cm in major axis.

Noninclusion criteria were patients undergoing ventral hernia mesh repair by other procedures (open or laparoscopic), midline abdominal hernia repair, simple herniorraphy (<2 cm in major axis), and preoperative infectious contexts.

All complications occurring within the first 30 postoperative days were considered early. Hematoma was defined as a parietal postoperative collection of blood with drain bleeding, sudden development of a parietal mass postoperatively, or confirmed by computed tomography (CT) scan. Seroma was defined as a subcutaneous uninfected collection of fluid persisting beyond 2 weeks and confirmed by CT scan. Mesh infection was defined as an infection affecting the periprosthetic space below the external oblique muscle and proved by positive bacteriological sampling.

An abdominopelvic CT scan was performed preoperatively; for patients with a history of pain in the right hypochondrium, hepatobiliary ultrasonography was also performed in search of gallstones.

Operative technique

All patients underwent operation under general anesthesia by the same standardized technique, depending of the localization. For prosthetic reinforcement, a polyester prosthesis (Mersilene) was used. A single dose of intravenous broad-spectrum antibiotic (cephalosporin) was administrated to all patients before skin incision. After incision of pre-existing scar and examination of the anterior fascia of the external oblique muscle to identify all parietal defects, the hernia sac was dissected and resected. Perioperative measurement of the hernia neck was carried out to calculate the surface of defect. The plane between the external and internal oblique muscles was found, and dissection was performed depending on the location and size of the defect.

For the subcostal LIH, dissection was carried up to the costal arch where the internal oblique muscle was detached from the ribs disclosing the external fascia of the transverse muscle which inserts into the posterior side of the ribs. Dissection was continued medially to the rectus muscle, and the homolateral rectus sheaths were then opened up to the xyphoid appendix. Anterior and posterior aponeuroses were respected. Dissection was completed externally with at least a 5 cm lateral margin.

For the flank LIH, the dissection was performed between the external and internal oblique muscles up to the costal arch and down to the Cooper’s ligament if necessary. Medially, the sheath of the rectus was opened, respecting the anterior and posterior aponeuroses.

For the iliac fossa LIH, the dissection was performed between the external and internal oblique muscles down to the Cooper’s ligament. Medially, the sheath of the rectus was open, respecting the anterior and posterior aponeuroses. Dissection was then completed down to the linea arcuata, at a point 5 cm below the umbilicus and continued in a preperitoneal direction to the Cooper’s ligament then upward and externally, with at least a 5 cm lateral margin.

The edges of the posterior layer including the internal oblique and transverse muscles were closed by absorbable suture. The polyester mesh was placed in the submuscular space in order to cover the whole area of the dissection, extending at least 5 cm above and below the fascial defect. It was then fixed to solid tissue with non-absorbable sutures: Cooper’s ligaments and/or under the xyphoid process at the linea alba. In case of a very large defect, the mesh was also fixed at the anterosuperior iliac crest by transfixing non-absorbable suture. Finally, the mesh was fixed to the musculofascial and aponeurotic wall by multiple absorbable sutures.

The edges of the external oblique muscle were closed with continuous absorbable sutures followed by subcutaneous suture. Suction drains were systematically placed in preprosthetic and subcutaneous spaces and were removed when producing <10 cc/24 h. However, all drains were removed before day 6 regardless of output. An abdominal contention belt was routinely applied to diminish postoperative pain. ASA3 patients were systematically monitored in the Surgical Intensive Care Unit (SICU) for the first postoperative 48 h.


After hospital discharge, patients were routinely reviewed at 1 month, 1 year, and beyond if necessary. Follow-up consultation consisted of a thorough clinical examination, and an abdominopelvic CT scan was performed for either obese patients or in cases of suspected hernia recurrence. Beyond the first year of follow-up, all patients and their primary care doctors were regularly contacted by phone or mail for updating. 5 years after surgery, all patients were contacted and asked to undergo an abdominopelvic CT scan followed by a consultation with the surgeon. The main end-point for follow-up was considered the last documented contact between the patient and his surgeon.

Statistical analysis

Data were prospectively collected in an Excel database (Microsoft Office Excel 2007, Microsoft, Redmond, WA). Descriptive analysis including means (±standard deviation) and proportions were recorded. For data measured on interval scale, Student’s t-test was used to compare different groups. Univariate analysis was performed for nominal data, applying the χ2 test and Fisher’s exact test when appropriate. A value of P ≤ 0.05was considered significant. All analyses were carried out with SPSS PASW statistical analysis version 18.0.0 software (PASW 18, Statistical Package for Social Science).



Of the 61 patients in this study, 60% (n = 37) were male. The mean age was 57 years (±12), and the mean BMI was 28 kg/m² (±5). Preoperative chronic bronchopulmonary co-morbidity and diabetes were observed in 20 and 15 % of patients, respectively. Six patients (10 %) were organ transplanted (five renal and one hepatic). According to the American Society of Anesthesiologists (ASA) score, 12 patients (20 %) were classified ASA 1, 36 patients (59 %) ASA 2, and 13 patients (21 %) ASA 3. Hernia characteristics are listed in Table 1. Among 61 patients, the parietal defect was subcostal in 23 % (n = 14), into the flank in 19.6 % (n = 12), and at the iliac fossa in 57.4 % (n = 35). An abdominopelvic CT scan was performed preoperatively for 56 patients (91.8 %). Seventeen patients (28 %) had one or more previously failed hernia repairs: 9 patients (14.7 %) had one repair, 6 patients (9.8 %) had two repairs, and 3 patients (3.2 %) had three previous failed hernia repairs. Nineteen patients (31.1 %) had multifocal lateral localizations. Ten patients had double hernia localization (midline and lateral) treated simultaneously, and five patients underwent other synchronous procedures together with the LIH repair: 3 cholecystectomies, 1 hiatal hernia repair by Nissen fundoplication, and 1 thoracic wall incisional hernia repair. All 61 patients underwent mesh repair, and there was no potential septic contamination such as intraoperative intestinal injury or suppuration on a pre-existing prosthesis. Drainage was performed in 54 patients (88.5 %). The average operative time was 136 (±69 min). Additional data concerning subgroups are shown in Table 2.
Table 1

Incisional hernia characteristics

Width, mean (SD), cm

7.6 (4.6)

Surface, mean (SD), cm²

56 (60)

Multifocal lateral incisional hernia, n (%)

19 (31)

Double location (midline and lateral), n (%)

10 (16)

Previous hernia repair, n (%)

17 (28)

Table 2

Subgroup characteristics of subcostal, flank, and iliac fossa lateral incisional hernias (LIH)


Subcostal (n = 14)

Flank (n = 12)

Iliac fossa (n = 35)

BMI, mean (SD), kg/m2

31.4 (1.5)

27.8 (2.2)

26.6 (0.6)

Surface, mean (SD), cm2

107 (24)

40 (12.7)

40 (6)

Width, mean (SD), cm

11.7 (5.1)

5.9 (4.19)

6.6 (3.5)

Multifocal LIH, n (%)

8 (42.1)

4 (21.1)

7 (36.8)

Double location (midline and lateral), n (%)

3 (21)

3 (25)

4 (11)

Previous hernia repair, n (%)

5 (35.7)

4 (33.3)

8 (22.8)

Operative time, mean (SD), min

202 (25)

105 (9)

121 (8)

Hospital stay, mean (SD), days

9 (1.3)

5.8 (0.8)

6.3 (0.5)

BMI body mass index

Postoperative course

Eleven patients (18 %) were hospitalized in the SICU. Three patients required blood transfusions. There was no postoperative mortality, and overall early morbidity, including medical and surgical complications, was observed in 11 patients (18 %). Concerning surgical complications, four patients developed hematoma requiring surgical management in three cases, two patients developed seroma treated by needle aspiration with favorable outcomes, and one patient required the removal of the sutures fixing the prosthesis to the posterior plane due to the significant cruralgia secondary to the suture in the lumbar muscle. There was no mesh infection. Concerning medical complications, three patients developed pneumonia and another had a pulmonary embolism. The mean hospital stay was 7 (±3.5) days.

In the univariate analysis, the BMI and defect size were significantly more important for patients who developed early complications, as were the mean operative time and length of hospital stay (Table 3).
Table 3

Univariate analysis of the risk factors for early complications after LIH repair


With early complications (n = 11)

Without early complications (n = 50)

P value

Age, mean (±SD), years

57 ± 2.2

55 ± 1.8


BMI, mean (±SD), kg/m²

31.6 ± 2.7

27 ± 0.55


Defect size, mean (±SD), cm²

100 ± 33

46 ± 5.5


Cardiovascular co-morbidity n (%)

4 (36)

26 (52)


Pulmonary co-morbidity n (%)

3 (27)

9 (18)


Previous hernia repair n (%)

4 (36)

13 (26)


Operative time, mean (±SD), min

188 ± 37

125 ± 6


Hospital stay, mean (±SD), days

9.3 ± 2

6.3 ± 0.4


Recurrence, n (%)


3 (6)


Median follow-up was 47 months (range: 1–125 months). Eleven patients (18 %) were lost to follow-up during the first postoperative year. Among patients reviewed 5 years after surgery, 17 underwent a CT scan. Two patients died during the study period: one patient from myocardial infarction and second from cancer-related causes.


The overall recurrence rate was 4.9 % (n = 3): 1 in the subcostal LIH subgroup and 2 in the iliac fossa LIH subgroup. The average time to recurrence was 19 months (range: 11–36 months). None of the patients with recurrences had developed early postoperative complications. There was no difference in term of BMI between the patients with and without recurrence (28.6 [± 2.8] vs 27.9 [±5.5]; P = 0.8). However, a significant difference was found in hernia recurrence among patients with previously failed hernia repair and those who had not (21 versus 0 %; P = 0.018).

Two patients with recurrences underwent reoperation: one patient was treated by the same surgical procedure and the second patient had a total intraperitoneal composite mesh. No recurrence was observed among reoperated patients.


Very few studies devoted to the management of lateral abdominal wall defects have been published [9, 10]. Because of a small prevalence (14 % of all parietal repairs over 7 years in our department), some authors integrate these defects into their series without specifying their localizations and their specific results. We found only one study analyzing specifically the results of LIH repair by laparoscopic approach; therefore the possibility of comparison remains limited [9]. Contrary to retromuscular mesh repair of midline incisional hernia, which is a relatively standardized technique, LIH repair by RMA is still debated. Due to the presence of several muscle layers, the choice of dissection to prepare the space for mesh insertion remains unclear. Excellent knowledge of the abdominal wall anatomy is obviously crucial. The lateral abdominal wall is composed of (1) the external oblique muscle, which has its origin on the anterior face of the 5th to 12th ribs and follows a mediocaudal direction; (2) the internal oblique muscle, which originates from the iliac crest and follows a medioproximal direction; (3) the transversus muscle, which takes its origin from the posterior face of the lower six ribs, the lumbodorsal fascia and the iliac crest, which is composed of fibers that follow horizontally. The blood supply of the lateral abdominal wall is provided by transverse segmental arteries arising from the aorta. Innervation consists of ventral branches of the 5th to 12th thoracic nerves, and the iliohypogastric and ilioinguinal nerves, which like the blood vessels, follow transversely. Our approach to repair of LIH was based on the necessity to restore a normal abdominal wall anatomy. In our surgical procedure, the dissection was performed between the external oblique and internal oblique muscles, and the choice to insert the prosthesis between these layers was conditioned by the fact that all important structures providing blood supply and innervation are located between the internal oblique and transversus muscles. Our approach is in accordance with that of Stumpf et al. [10], who in a recent work investigating the best position to introduce the mesh in fresh-frozen human cadavers, concluded that to avoid any nerve dysfunction (i.e., leading to pain or functional alteration of the lateral abdominal wall), dissection between the internal oblique and transversus muscles should not be used.

In parallel, the recent development of laparoscopic incisional hernia repair with encouraging results has added another alternative in the choice of surgical approach [9]. Compared with the laparoscopic approach for the treatment of LIH, Moreno-Egea et al. [9] reported in a cohort of 53 patients an average operative time of 60 min, while it was 136 min in our study. However, 31 % of our patients had miltifocal defects versus 9 % for Moreno-Egea et al. [9]. In addition, 25 % of our patients underwent other synchronous procedures, prolonging the average operative time.

The laparoscopic approach seems seductive in terms of shortened hospital stay, which was confirmed by a recent meta-analysis [13]. Thus, the hospital stay was 2.7 days for Moreno-Egea et al. [9] versus 7 days for our patients. But when comparing the patients’ characteristics we found that there were more co-morbidities in our patients (i.e., chronic bronchopulmonary disease (20 %) and diabetes (15 %) versus 13 and 5.7 %, respectively, for Moreno-Egea et al. [9]). Moreover, 10 % of our patients had organ transplantation and 15 % had chronic renal failure.

In contrast, extensive adhesiolysis and viscerolysis are frequently necessary during laparoscopic surgery for LIH, which could potentially increase postoperative morbidity associated with visceral or parenchymal organ injuries [13]. We have not encountered such problems during surgery, but this rate was 7.5 % when the laparoscopic approach was used [9].

We found that patients with early complications had more important defect size and BMI, and this unsurprisingly explained longer operative times and hospital stays. A similar pattern was observed when comparing different subgroups, particularly in patients with subcostal LIH, which commonly were larger defects more difficult to correct in terms of surgical technique.

Concerning the early surgical complications, the only published study on the treatment of LIH by the laparoscopic approach reported a 26 % early complication rate with predominance of hematoma (17 vs 6.5 % in our patients) [9]. Concerning seroma, the rates were similar. The results of the present study are notable for the absence of prosthetic infection, which is the most feared complication requiring prolonged hospital stay, long antibiotic treatment, and, frequently, mesh removal. In our opinion, this positive result in our series reflects some technical details, such as hernia sac systematic complete resection and extensive drainage of the subcutaneous and prosthetic spaces. At present there is not consensus concerning the value of drainage after incisional hernia repair [14]. Some authors performed it with encouraging results [15, 16], while others avoided it, fearing the increase of mesh infection risk [17]. Drainage was performed in 88.5 % of our patients. Again in our opinion, adequate drainage of the prosthetic and subcutaneous spaces allows optimal evacuation of fluids, decreasing the risk of fluid collection, particularly in patients with large parietal defects that require extensive tissue dissection and in cases of recurrence repair with previous dissection of the abdominal wall. Twenty-eight percent of our patients had previous hernia repair, and in 36 % of our patients the width of the parietal defect was >10 cm, with an average of 12.6 cm. Nevertheless, we agree that prolonged drainage could theoretically increase the probability of mesh infection. Therefore all drains were removed systematically by day 6, regardless of daily output volume. Our use of drainage is supported by the low rate of seroma and mesh infection reported in our previous study of patients treated by retromuscular mesh repair for midline incisional hernia in which we employed the same approach to drainage [16].

Other factors that may have decreased the risk of early morbidity are our standardized protocol, the implementation of rigorous antiseptic measures, and an aggressive therapeutic approach to postoperative hematomas.

In our study we observed three hernia recurrences (4.9 %). The risk factors increasing recurrence rate, such as previous hernia repair, obesity, chronic steroid use, obstructive pulmonary disease, previous abdominal aortic aneurysm and aortoiliaic occlusive disease operations, and especially postoperative wound infection, have been identified by several authors [1820]. All our patients with recurrence had previous hernia repair, including two patients who had twice undergone LIH repair, and one had obstructive pulmonary disease. In terms of localization, there were one subcostal and two iliac fossa recurrences. In all cases the recurrence was lateral, probably caused by inadequate mesh overlap and potentiated by risk factors specific to each patient. However, small sample sizes of subgroups do not allow us to draw conclusions about the influence of topographic localization (subcostal, flank, or iliac fossa) on the recurrence rate.

When compared, our patients’ overall recurrence rate was more important but comparable to results obtained after laparoscopic management (4.9 vs 3.7 %) [9].

Apart from the fundamental surgical and preventive principles discussed above, all of which are applied by most authors in the repair of any type of hernia, other debates have recently emerged. With the development of new technologies and advanced histological findings, it appears that the type of mesh employed and its characteristics (structure, porosity, absorbable component, weight) could play a considerable role in the solidity of repair. In the worldwide literature, the three materials most often reported in parietal repair are polypropylene, polyester, and expanded polytetrafluoroethylene based meshes. Some investigators have demonstrated that pain, seroma, abdominal wall functional impairments, and persisting infection are related to the small pore size of heavyweight meshes [2123]. To improve the quality of these materials some authors have used a new type of prosthesis called large pore-sized, lightweight meshes [2123]. These are characterized by an absorbable component that reduces the weight of the material and an increased diameter of pores, which allows rapid colonization by fibroblasts and collagen fiber secretion. Still other investigators have proposed the addition of a bactericidal property by adding antibiotics to the mesh surface [24]. Even though the functional outcomes reported with the use of large pore-sized, lightweight meshes are encouraging, a recent meta-analysis could not determine their superiority compared to “classic” heavyweight meshes [25].

The main bias of the present study is the lack of follow-up of our patients, 11 of whom (18 %) were lost during the first postoperative year. Another limitation of this study was the impossibility to perform a multivariate analysis to identify the independent risk factors of recurrence because of the low number of recurrences observed.

In conclusion, LIH repair by RMA using a polyester heavyweight mesh is a safe approach with moderate complication rate, low incidence of infection, and an acceptable recurrence rate, even in large or multifocal parietal defects. Small sample size of subgroups did not allow us to draw any conclusions concerning influence of topography of LIH on recurrence. Further larger prospective and comparative studies reporting different approaches are needed to determine the influence of surgical technique on the recurrence rate in treatment of LIH.

Copyright information

© Société Internationale de Chirurgie 2012