The incidence of incisional hernia after aortic aneurysm is not higher than after benign colorectal interventions

Background Abdominal aortic aneurysms (AAA) have most probably an inflammatory origin, whereby the elastica is the layer actually involved. In the past, collagen weackness was supposed to be the shared cause of both, AAA and incisional hernias. Since the development of new techniques of closure of the abdominal wall over the last decade, collagen deficency seems to play only a secondary etiologic role. Objectives The aim of the study was to investigate whether the incidence of incisional hernia following laparotomy due to AAA differs from that of colorectal interventions. Material and methods This was a retrospective control matched cohort study. After screening of 403 patients with colorectal interventions and 96 patients with AAA, 27 and 72 patients, respectively were included. The match criteria for inclusion of patients with colorectal interventions were: age, benign underlying disease and median xiphopubic laparotomy. The primary endpoint was the incidence of an incisional hernia. The secondary endpoints were the risk profile, length of stay in the intensive care unit and postoperative complications. Data analysis was carried in the consecutive collective from 2006 to 2008. Results In the group with AAA the mean follow-up was 34.5±18.1 months and in the group with colorectal interventions 35.7±21.4 months. The incidence of incisional hernias showed no significant differences between the two groups. In the AAA group 10 patients (13.8%) developed an incisional hernia in contrast to 7 patients in the colorectal intervention group (25.9%). Conclusions In our collective patients with AAA did not show an increased incidence of incisional hernia in comparison to patients with colorectal interventions with comparable size of the laparotomy access and age. The quality of closure of the abdominal wall seems to be an important factor for the prevention of incisional hernia.


Introduction
The development of incisional hernia following abdominal surgeryis anundesired and, depending on the findings, high-risk event. Its incidence ranges between 2% and 15% in the literature [1][2][3]. There are a number of theories on the etiology of incisional hernia, with the techniques of abdominal closure and collagen metabolism being the two most important. The latter is based on the assumption that the development of incisional hernia is triggered by a disproportion of type I collagen to type III [4,5]. It was also long suspected that aortic aneurysms were caused by a collagen defect [6] and that these patients were at particularly high risk for developing incisional hernia; however, according to the current state of knowledge on the development of aortic aneurysms, collagen metabolism is no longer considered to be a major triggering factor. Thus, from a clinical perspective, the hypothesis can be pos- tulated that the incidence of incisional hernia following medial xiphopubic laparotomy is not higher in abdominal aortic aneurysm patients than in matched controls with benign colorectal disease.
This hypothesis is based on the following rationales: a) Incisional hernias develop differently depending on the incision and need to be differentially classified and grouped. Medial xiphopubic laparotomy is performed for abdominal aortic surgery. Therefore, it is essential that this patient group be compared exclusively with a control group that has also undergone medial xiphopubic laparotomy; patients serving as the control group should have no recognizable endogenous risk factors for incisional hernia [7]. b) From an historical perspective, it is assumed that the pathogenesis of abdominal aortic aneurysms (AAA) is based on impaired collagen production, which, at the same time as being an endogenous risk factor, also increases the risk of incisional hernias [6]. c) The pathogenesis of AAA remains unclear. Novel experimental investigations show that it is probably not caused by a collagen defect as previ-ously assumed [8] but arises instead from the Tunica intima. d) As a result, the historical assumption that the increased incidence of incisional hernia in aortic aneurysm patients is due to a collagen defect needs to be reviewed, and the incisional hernia incidence in this patient group can possibly be attributed to some other causality.

» Collagen deficency is
probably not the etiology of abdominal aortic aneurysms In order to answer the working hypothesis, this article examines the following questions: a) Are there differences in the incidence of incisional hernia at 1 year (primary endpoint) between the two groups? b) What are the differences between the two groups in terms of the secondary endpoints: operating time, intensive care stay, postoperative complications, and inpatient stay? c) What perspectives result from answering the study question in terms of further clinical and experimental research? In order to ensure comparability of the two groups, the following inclusion criteria were used to achieve a valid matched control. The inclusion criteria were: age >57 and <77 years (MW ± SD of age in the AAA group), medial xiphopubic laparotomy (European Hernia Society [EHS] classification: M1 + M2 + M3 + M4 + M5), and benign resected tissue on histology [9].
A total of 96 cases were evaluated in the aortic aneurysm group. Of the 24 cases that were excluded, 17 had already died by the time the study was conducted and no post-operative follow-up had been documented. In a further two cases, no followup dates could be determined despite attempts to contact these patients both in writing and by telephone. The remaining five cases that were excluded consisted of: one case that was reported twice, one case declared to be an aortic aneurysm but which was treated with a colorectal procedure, one case in which an aortic aneurysm and the colon were operated on in a single-stage approach and two cases in which the indications for aortic aneurysm repair using stent grafts were established. This left a total of 72 complete records for evaluation.
The colorectal procedure group comprised 403 potential cases. Once the inclusion criteria and the abovementioned age matching (> 57 and < 77 years) had been applied, 63 cases remained. Of these 63 cases, 20 had already died by the time the study was conducted and no post-operative follow-up had been documented. In a further nine cases, no follow-up date could be determined despite attempts to contact these patients both in writing and by telephone. A further seven cases were excluded as they had not undergone primary abdominal wall closure. This left a total of 27 complete records for evaluation (. Table 1; . Fig. 1).

Preoperative data
The surgical indications in the aortic aneurysm group were symptomatic aortic aneurysm or aneurysm measuring over 5 cm in diameter and showing annual progressive growth of over 0.5 cm in diameter. Patients in the colorectal group were selected on the basis of the appropriate surgical procedure, verification in the surgical report that a medial xiphopubic laparotomy had indeed been performed, and exclusion of oncological disease.
The risk factors for the development of incisional hernia were determined. Preoperative risk factors included: male gender, obesity with a body mass index (BMI) of over 30 kg/m 2 , nicotine use, diabetes mellitus, hypertension, kidney failure, cortisone use, lung disease, preoperative hemoglobin below 10 g/dl, preoperative albumin below 3.5 g/dl, a known pre-existing hernia, known cancer, collagen disorders, emergency surgical procedures, known previous surgery, as well as a second procedure within 1 month. Pulmonary diseases included chronic obstructive pulmonary disease (COPD), chronic bronchitis, and bronchial asthma. Postoperative risk factors were divided into two categories: particular features of the postoperative course and postoperative complications [2,10].

Medial laparotomy and abdominal closure in both patient groups
Laparotomy was performed using a skin incision around the left side of the umbilicus from below the xiphoid process to immediately above the pubic symphysis; incision of subcutaneous tissue and sparing exposure of the linea alba; dividing the linea alba with a scalpel and opening the peritoneum.
Abdominal wall closure was performed using a continuous absorbable monofilament 1-USP loop suture while maintaining a suture-to-wound length ratio of 4:1 [11]. At least two loops were used: starting in the suprapubic region, the first was used as a continuous suture to the navel, and the second, beginning in the subxiphoid region, also as a continuous suture to the umbilicus, where the two loops were knotted together. After rinsing the subcutis with saline solution, a subcutaneous redon drain was placed and the skin incision closed using polypropylene USP 3-0 Donati stitches. The precise length or distance from the xiphoid process to the symphysis was not routinely measured.

Primary outcome
The occurrence of incisional hernia was defined as the primary endpoint. The follow-up date was determined from the previous medical report and any available diagnostic computed tomography (CT) images were also evaluated in the diagnosis. The date of the CT scan was set as the follow-up date. If no postoperative Abstract Background. Abdominal aortic aneurysms (AAA) have most probably an inflammatory origin, whereby the elastica is the layer actually involved. In the past, collagen weackness was supposed to be the shared cause of both, AAA and incisional hernias. Since the development of new techniques of closure of the abdominal wall over the last decade, collagen deficency seems to play only a secondary etiologic role. Objectives. The aim of the study was to investigate whether the incidence of incisional hernia following laparotomy due to AAA differs from that of colorectal interventions. Material and methods. This was a retrospective control matched cohort study. After screening of 403 patients with colorectal interventions and 96 patients with AAA, 27 and 72 patients, respectively were included. The match criteria for inclusion of patients with colorectal interventions were: age, benign underlying disease and median xiphopubic laparotomy. The primary endpoint was the incidence of an incisional hernia. The secondary endpoints were the risk profile, length of stay in the intensive care unit and postoperative complications. Data analysis was carried in the consecutive collective from 2006 to 2008. Results. In the group with AAA the mean follow-up was 34.5±18.1 months and in the group with colorectal interventions 35.7±21.4 months. The incidence of incisional hernias showed no significant differences between the two groups. In the AAA group 10 patients (13.8%) developed an incisional hernia in contrast to 7 patients in the colorectal intervention group (25.9%). Conclusions. In our collective patients with AAA did not show an increased incidence of incisional hernia in comparison to patients with colorectal interventions with comparable size of the laparotomy access and age. The quality of closure of the abdominal wall seems to be an important factor for the prevention of incisional hernia.

Secondary outcomes
Postoperative complications were divided into two categories. Postoperative risk factors were assigned to an "intensive medicine" category and a "non-intensive medicine" category. The former included wound infections, catecholamine ther-apy, relaparotomy, artificial ventilation, and the length of stay on the intensive care unit. Non-intensive medical complications included infections, seroma, and secondary wound healing. These complications were classified according to the Clavien-Dindo classification [12].

Follow-up examination
Follow-up examinations of patients after AAA repair were planned for 3, 6, and 12 months as standard at the Surgical Department of the University Hospital Würzburg. Following benign colorectal procedures, patients were examined at 1 year. Patients who did not attend follow-up at the University Hospital Würzburg were contacted in writing  [12].No significant differences were seen between the two groups.No postoperative classification could be determined for one patient.This is not shown for reasons of clarity and asked to complete a questionnaire. Beyond the 12-month period, the latest clinical outcome of medial abdominal wall closure was noted.

Outcomes
Outcomes were collected for 69 patients from the prospective electronic medical records (51/18 AAA/colorectal, respectively). Current CT images of the abdomen were available for 10/7 patients, respectively; these were used to objectively investigate the clinical diagnosis "no incisional hernia" or "incisional hernia". A total of 30 patients were contacted via questionnaire. Mean follow-up in the aortic aneurysm group was 34.5 ± 18.1 months and 35.7 ± 21.4 months in the colorectal procedure group. No significant difference could be seen between the two groups in terms of mean follow-up in months. . Table 2 shows the age and gender distribution between the patient groups. There were significantly more females in the colorectal procedure group compared with the aortic aneurysm group.   There was no significant difference be-tween the two patient groups in terms of the incidence of incisional hernia. Analysis of the sex distribution in relation to the incidence rate of incisional hernia revealed no significant difference between male and female patients (. Table 4).
No significant correlation was seen between risk factors and the occurrence of incisional hernia in aortic aneurysm patients (. Table 5).
. Table 6 presents the risk factors for the occurrence of incisional hernia fol-lowing colorectal resection. Here again, there is nosignificantcorrelation between riskfactors and the incidence ofincisional hernia (. Table 6).

Secondary outcomes
Postoperative complications are shown in . Table 7. None of the postoperative complications in either group affected the development of incisional hernia. . Fig. 2 shows the distribution of postoperative complications according to severity in line with the CDC [12]; these were also comparably distributed in both cohorts. No significant differences were seen in terms of the incidence of incisional hernia and postoperative non-intensive medical complications. The relative risk of developing an incisional hernia in the case of complications requiring revision was increased by 34%, without, however, reaching significance in this collective.

Key messages of the study
The matched-control comparison revealed no differences between the two groups in relation to the primary endpoint, i. e., the development of incisional hernia. The incidence of incisional hernia following aortic aneurysm surgery tended to be even lower compared with the colorectal group. According to Clavien-Dindo score, complications in both groups were comparable. In neither group was it possible to establish and influence of the postoperative course and postoperative complications on the development of incisional hernia. These results confirm earlier results obtained by Israelsson [11], who found that patients do not have a greater risk for incisional hernia following aortic aneurysm repair than patients following colorectal procedures with accesses of similar dimensions.

Evaluation and critical review of own results
The surgical technique of abdominal wall closure is of crucial importance for the postoperative incidence of incisional hernia. In this study, an incidence rate of 13.9% was seen in the aortic aneurysm group. Incidence rates of between 6% [11] and 35% in aortic aneurysm patients are reported in the literature [11,13,14].
The not significantly higher rate of incisional hernia in the colorectal group, as well as the longer stay in the intensive care unit, can probably be accounted for by underlying diseases, such as sigmoid diverticulitis in this cohort. The most important weaknesses of this study are its retrospective data collection and the low patient number; however, it will not become easier to include patients undergoing open aortic aneurysm repair in studies in the future, given that endovas-cular techniques have become standard in the last 10 years. For this reason, the present study is of value despite its low patient number. What is of particular relevance about this study is that, for the first time, incisional hernia incidence was investigated while taking into account the anatomy of access, i. e., not only medial morphology, but also the xiphopubic extension of the incision was an inclusion criterion. These criteria can be found in the Würzburg incisional hernia classification [7] and the European Hernia Society (EHS) classification [9]. All cases of shorter medial access were eliminated in the control group, resulting in a highly homogeneous and comparable study population.

Explanatory models for the development of incisional hernia and aortic aneurysm
Collagen I and III play a particularly important role in scar formation: collagen III is metabolized primarily in the early phase of scar formation, collagen I in the late phase [15]. Collagen I is essential for scar stability. It is known that hernia patients have a lower collagen I:III ratio in favor of predominantly high collagen III [4,5,16]. An increased incidence of incisional hernia has also been observed in hereditary diseases involving collagen defects [17][18][19][20]. Matrix metalloproteases (MMP) play a crucial role in the alteration of collagens. They are particularly important in the regulation of growth and apoptotic processes [21,22]. There appears to be significant overexpression of MMP-2 in incisional hernia [23]. Since collagen synthesis can be suppressed by inflammatory processes, such as postoperative sepsis, the functionality of the various MMP and collagens may be impaired by the wound milieu [24]. Current models of the development of aortic aneurysm work on the premise that the pathogenesis of aortic aneurysms is not based on a collagen defect as previously assumed [6]: an elastase defect has been postulated [8], while more recent experimental studies also focused on inflammation as the cause of aortic wall weakening during the process of aneurysm development [25]. Ultimately, the molecular etiology of aneurysm development remains unclear; however, the collagen weakness assumed in the past to be the cause of AAA is highly unlikely according to current knowledge.

Taking a careful approach to abdominal wall closure
One possible factor that affects the low incidence of incisional hernia following aortic surgery in the authors' department is related to the fact that the vascular surgeons there have broad general surgical training. The importance of meticulous abdominal wall closure cannot be disregarded in this context. Evidence shows that maintaining a suture length-wound   [26,27], using a continuous suture with slowly absorbable or non-absorbable suture material [26,[28][29][30][31][32][33][34][35][36][37], maintaining a distance from needle puncture to wound edge of at least 1 cm [30], as well as distances of less than 1 cm between two punctures [31], can reduce the incidence of incisional hernia. Ever since the publication of works by Israelsson in the mid 1990s, it has been the accepted wisdom that less can be more: by using smaller needles and small tissue bites, incisional hernia incidence in medial laparotomy following aortic aneurysm repair can be reduced from 21% to 10% (p < 0.01), and to as little as 5.2% in selected patients [11,[31][32][33].

Is prophylactic mesh placement beneficial in AAA?
The surgical management of hernias has moved away from the conventional suture technique in recent years in favor of repair using mesh [35]. According to these positive results, it has been postulated that maybe prophylactic mesh placement could significantly reduce the incidence of incisional hernias in highrisk patients, for example, in obese patients. A recent meta-analysis of prophylactic mesh placement in the setting of open bariatric procedures also revealed a positive effect for mesh when the available cohort studies were included (OR 0.39; CI 95% 0.13-0.68; p = 0.004); however, if one considers only the four randomized controlled trials (RCT)included in the meta-analysis, prophylactic mesh placement conferred no benefit (OR 0.52; CI 95% 0.24-1.16; p = 0.104) [36]. That particular meta-analysis did not include the recently published RCT conducted by Muysoms et al. Here, the incidence of incisional hernia at 2 years (with approximately 10% loss to follow-up) in the control group (continuous suture using the large-bites technique) was 28% and in the prophylactic mesh group (retromus-cular mesh placement) 0% (p < 0.001); however, retromuscular mesh placement is considerably more complex and the advantages of suture techniques are not exhausted with the large-bites control group [37]. This study needs to be re-interpreted and re-weighted in the light of the small-stitches technique.
In summary, it must be concluded that according to current evidence, the meticulousness of abdominal wall closure based on Israelsson's recommendations [11,32] appears to be the most appropriate approach irrespective of the surgical indications.