Omentoplasty in Deep Sternal Wound Infection

Patients and Methods

Between March 2009 and December 2012, 3,600 sternotomy procedures were performed at two general hospitals in Tehran as part of coronary artery bypass graft (CABG) operations. During this interval, 40 consecutive cases of DSWI were identified in adult patients, an incidence of approximately 1.1%. Patients with DSSSI were aged 51 to 72 y (mean 63 y), with a gender distribution of 63% males and 37% females.

All DSWI met the criteria of the U.S. Centers for Disease Control and Prevention [3]. Thus, diagnosis was based on at least one of the following: (1) A bacterium was isolated from mediastinal tissue or fluid; (2) evidence of mediastinitis was seen during operation; (3) sternal instability or fever was present, and there was purulent discharge from the mediastinum.

Surgical technique

After hemodynamic stabilization, the patient was taken to the operating room for a single-stage surgical debridement and immediate OF transposition. The incision was re-opened by removing all sternal wires and subcutaneous sutures. Samples of mediastinal tissue and fluid were harvested for culture and susceptibility testing. Necrotic tissue and infected sternum were excised to bleeding tissue (Fig. 1A). The mediastinal incision and pericardial cavity were then irrigated with warm saline povidone–iodine solution.

OPEN IN VIEWER

FIG. 1.

Surgical technique. (A) Sternum has been excised and debrided to healthy tissue with omentum brought to incision. (B) Omentum has completely filled site with no tension. (C) Final appearance of primarily closed incision with one drain in place.

A midline laparotomy (±10 cm) was performed. The omentum was detached from the transverse mesocolon and 10% right gastroepiploic artery on a pedicle, which is larger and has more epiploic branches and therefore generally is used as a nutrient vessel. After ligation of one gastroepiploic artery, the branches to the stomach from the gastroepiploic arcade were divided as far as necessary to obtain a suitable pedicle for omentoplasty. Compression or kinking of the vascular pedicle was prevented. An incisional hernia with sufficient diameter was left in the subxiphoid region for transit of the pedicle flap. The omentum was then placed tension free in the sternal defect and fixed with sutures at the edges (Fig. 1B). No sternal closure was performed. Following insertion of a closed-suction drain, the skin was closed primary by interrupted nylon sutures (Fig. 1C). All patients received antibiotic and thrombosis prophylaxis during the procedure.

Results

All 40 patients with DSWI were treated by this single-stage OF transposition procedure. Thirty-four of the studied patients (85%) were found to have DSWI within 1 mo of their cardiac surgery, whereas the six other patients (15%) had their infection diagnosed between 30 d and 45 d after the cardiac operation. No patient presented earlier than 2 wks of the original operation. Thus, our patients had Type IIIB or Type V infections (see Table 1).

Underlying diseases included diabetes mellitus in 13 patients (32.5%), hypertension in seven (17.5%), renal dysfunction in three (7.5%), and history of stroke in two patients (5%). Thirty-two patients (80%) were current or ex-smokers at the time of DSWI diagnosis (Table 2).

The mean operative time for the OF procedure was 90±16.8 min (range 70–135 min). Site cultures demonstrated positive results in 28 patients. In 22 cases (78.57%), the pathogens were gram-positive, of which 14 (63.6%) were methicillin-resistant Staphylococcus aureus (MRSA) (Table 3).

There were three intra-operative complications (7.5%). Bleeding from the right ventricle, occurred in one patient (2.5%), but and direct suture was performed without the need for cardiopulmonary bypass. Two patients (5%) had rupture of the right ventricle when the sternal incision was reopened, which resulted in fatal bleeding.

Post-operative complications occurred in 10 patients (25%) and consisted of supra-ventricular dysrhythmia in three (7.5%), pneumonia in two (5%), urinary tract infection in 2 (5%), gastrointestinal bleeding in one (2.5%), and recurrent infections within 60 d of primary omentoplasty in two (5%). No patient developed any sign of OF necrosis. The median length of post-operative ICU and hospital stays were 4 d (range 1–6 d) and 7 d (range 5–14 d), respectively. At the time of discharge, all patients had normal cardiac and mental status.

Discussion

Deep sternal wound infections after sternotomy for cardiac and thoracic surgery are important causes of potentially lethal complications and added cost. The risk factors linked to such infections are numerous and can be divided in two groups: Host-linked and surgery-linked. Host-intrinsic risk factors are increasing age, obesity, cigarette smoking, and glucocorticoid therapy, but the major risk factors are nasal carriage of S. aureus and diabetes mellitus. Surgery-linked factors are the duration of surgery, post-operative bleeding, re-operation, re-wiring, use of an intra-aortic balloon pump, and, especially, the use of the internal mammary arteries bilaterally in CABG.

Continuous antibiotic irrigation and debridement has been the primary therapeutic option for years, allowing secondary re-fixation of the sternum. For patients in whom this treatment modality fails, Lee et al. [5] described the technique of sternal excision and transposition of the omentum, combined with primary closure of the incision. In the last 20 y, the use of a pectoralis major MF [6] has led to a further reduction of the mortality rate in patients suffering from DSWI [7,8]. In other words, repair of sternal defects compromised with infection can be classified as follows [9]: (1) Primary closure in combination with irrigation; (2) local negative-pressure therapy, also known as vacuum-assisted closure (VAC) therapy [10,11]; (3) VAC therapy followed by definitive closure or sternal re-fixation using sternal plating [12]; (4) primary or delayed flap closure by the use of regional myocutaneous flaps (myoplasty), such as a pectoralis major, rectus abdominis, or latissimus dorsi flap; or (5) transposition of the greater omentum, with or without split-thickness skin grafting [13].

There is no consensus as to which of these options is superior. Several publications recommend omentoplasty for reconstruction of chest wall defects after tumor resection or post-sternotomy mediastinitis after CABG or as a final solution when MFs have failed [9]. The main goal of this study was to evaluate the outcome of several cases in which OF transposition was used for repair of DSWIs.

Given its rich vascularity and angiogenic and immunogenic features, the greater omentum seems ideal for the repair of infected sternotomy incisions [14]. Lopez-Monjardin et al. [15] concluded that mediastinal reconstruction using an OF in patients with mediastinitis secondary to open heart surgery is associated with fewer infectious complications than using a pectoralis major flap. Also, reconstruction of a chest wall defect after full-thickness resection for local recurrence or primary malignant neoplasm of the chest wall can be successful with pedicled omentoplasty and split-thickness skin grafting with absorbable implants [16].

Nowadays, pedicled omentoplasty is reserved for situations in which myocutaneous flap coverage has failed or where myoplasty is not sufficient. Use of musculocutaneous flaps also depends on the vitality and the condition of the regional muscles. By not making use of the omentum, the patient can be spared a laparotomy. Besides, the greater omentum is not always available, for example, when a patient already has undergone abdominal surgery.

Whether the left or the right gastroepiploic artery should be used for the pedicle of the OF remains a matter of debate. The right gastroepiploic artery was used in our series because it is the larger. Although the left gastroepiploic vessel is closer to the sternal defect, in our experience, the omentum had sufficient bulk to cover the sternal defect without the need for further lengthening.

Another issue of discussion is the way the OF has to be guided to the thoracic cavity. We chose the transdiaphragmatic route, because some authors have said that it reduces operative time and trauma without increasing morbidity [17].

Publications favoring MFs far outnumber those on OFs. These same publications, especially those advocating multiple MFs, often emphasize the same potential complications of an OF without relating these to the severity and incidence of the complications of MF reconstruction.

Several studies investigated a possible association between the flap type and the mortality rate. López-Monjardin et al. [15] reported a significantly higher sepsis-related mortality rate in their MF group (p<0.05). All sepsis-related deaths reported in the studies of Semper et al. [18] and Ivert et al. [19] followed bilateral pectoralis MF advancement. Re-exploration for either flap necrosis or recurrent sternal infection was reported more frequently after MF reconstruction, with an incidence ranging from 3% to 17.5% [20]. In their well-balanced, direct comparative study, Milano et al. [21] found a 19% higher incidence of re-exploration after MF reconstruction. Caution is advised in interpreting these results: Although re-infection could be an indication of flap failure, it also can result from inadequate debridement prior to flap reconstruction. Hematomas or seromas requiring drainage occurred almost exclusively after MF reconstruction, with a mean incidence of 10.6% (range 6.5%–15%) [20].

A recent systematic review by van Wingerden et al. [22] concluded that taking all limitations of this analysis into consideration, MF reconstructions fail to show an obvious survival advantage over reconstructions with an OF. Furthermore, the level of evidence does not support adherence to the belief in the superiority of an MF over OFs.

Pectoral MF advancement offers benefit in terms of ease and expeditious creation in carefully selected cases (with narrow, mainly cranial, sternal defects following adequate debridement). For larger defects, or in more extensive disease, current best evidence obliges surgeons not simply to rely on preference when selecting reconstruction with MF rather than OF. Use of the omentum instead of a complex MF reconstruction warrants serious consideration. Harvesting the omentum laparoscopically, rather than through a laparotomy, may tip the scales. The case series of reconstruction with a laparoscopically harvested OF currently are too small for comparative purposes.

In our study, the risk of intra-abdominal infection following OF transposition was eliminated by approaching the flaps through a separate abdominal incision, meticulous prevention of cross-infection–contamination from the sternal incision, and mobilization of the flaps with great care. From our knowledge, inadequate debridement of sternal bone and cartilage could cause an incisional sinus tract; thus, we strongly recommend aggressive debridement, especially of the lower third of the sternum and xyphoid cartilage, which are the most common sites of sinus tract formation.

This approach would not be appropriate for patients having previous abdominal surgery, a severe low cardiac output state, or malnutrition, which are usual causes for the inability of the OF to seal effectively and facilitate the delivery of antibiotics to the infected space.

In conclusion, we have treated all DSWI patients with a single-stage OF transposition after early diagnosis and adequate debridement. Our study documents favorable results during both in-hospital and long-term follow-up. Although the limitation of our study is the small number of patients, it supports the feasibility and efficacy of OF transposition, which is recommended as a single-stage treatment for DSWI. Confounding factors that may influence both the choice of flap and the outcome should be elucidated by a future randomized study.