The Utility of the Omentum Flap for Complex Intrathoracic Problems

Abstract

Introduction: Omentum flap is a viable reconstructive option for complex chest wall and mediastinal reconstruction. The impact of vasoconstrictors and the laminar pattern of blood flow associated with left ventricular assist devices (LVADs) on the outcomes of reconstructions has not been thoroughly evaluated. Methods: A retrospective review of all patients who underwent chest wall or mediastinal reconstruction using pedicled omentum flaps between 2003 and 2019. Results: Forty patients (60% males) underwent chest wall or mediastinal reconstruction using a pedicled omentum flap at a mean age of 58 years. The median follow-up was 24.3 months. The most common indication was the reconstruction of anterior chest wall/sternal defects (n = 16), followed by coverage of repaired bronchopleural fistula (n = 6), osteoradionecrosis of the anterolateral chest wall (n = 5), reconstruction of anterior/lateral chest wall following oncologic resections (n = 5), coverage of replaced infected LVAD (n = 4), and coverage of exposed/replaced aortic root vascular grafts (n = 4). Vasoconstrictors were used in 26 patients (65%). Eight flaps had partial necrosis, and none of the flaps had complete necrosis. There was no difference in flap complication rates in patients who received vasoconstrictors during the case compared to those who did not (P = 1.0). Thirteen (33%) flaps were skin grafted at a median of 13 days with 100% skin graft viability. Abdominal incisional hernia developed in 8 patients. In patients with LVADs, the omentum remained viable during the follow-up period. Conclusion: The ability of the omentum to easily reach various regions in the chest and the low failure rate make this flap a reliable reconstructive method.

Introduction

The omentum flap is a versatile reconstructive option that can be used to treat various soft tissue defects in the trunk, pelvis, and thorax and can also be used as a free flap for the management of remote defects and lymphedema. The use of the omentum as either a pedicled ^{1 -5} or free flap ⁶ for chest wall reconstruction has been described for several indications. ^{1 -5,7 -22} Most of the available reports are limited to case reports and small case series. In addition, the frequent occurrence of hypotensive episodes either during surgery or postoperatively in this group of patients who have a high prevalence of medical morbidities and cardiovascular disease may put the transferred omentum in a state of inadequate perfusion either due to prolonged hypotensive episodes or due to the effect vasoconstrictor infusions used to normalize these episodes. A small percentage of congestive heart failure patients have left ventricular assist devices (LVADs) (Figure 1) that are used either as destination therapy or a bridge to heart transplantation. The laminar, non-pulsatile nature of blood flow produced by LVADs raises a question about the outcomes of pedicle flaps in this setting, due to the concern of potential inadequate perfusion secondary to a non-physiologic blood flow pattern. ^19,20 In order to get further insight into the utility and versatility of omentum flap and to better understand the effect of vasoconstrictors and LVAD circulation on the viability of the transferred omentum, we review our experience with chest wall and mediastinal reconstruction using pedicled omentum flaps. In addition, we discuss case examples and technical tips to maximize reach and enhance versatility (Figures 2 and Supplemental Videos 1-4).

Figure 1.

This presents 3 different patients, and the photos are divided in part 1, part 2, and part 3. Part 1: (1A) A 62-year-old male patient with heart failure secondary to nonischemic dilated cardiomyopathy who underwent left ventricular assist device (LVAD) placement and postoperatively developed anterior mediastinitis with abscess and sternal osteomyelitis due to methicillin-resistant Staphylococcus aureus. After sternotomy and multiple debridements, the omentum flap was used to provide coverage of the infected, exposed LVAD. (1B) Omentum flap harvested based on the right gastroepiploic vessels. (1C) Omentum flap after the inset. Part 2: (2A) Exposed aortic graft in a 29-year-old male patient with infected aortic root and ascending hemiarch replacement graft. After debridement, a pedicled omentum flap was tunnelled into the mediastinum for coverage and space obliteration (see Supplemental Videos), and the soft tissues were closed with pectoralis flaps, and the omentum flap filling the partial sternotomy space and mediastinal cavity. (2B) Axial (left) and Sagittal (right) computed tomography (CT) 3 months postoperatively. The CT shows the omentum flap (O), obliterating sternotomy defect and periaortic cavity, and bilateral pectoralis flap (P) closure. Part 3: A 64-year-old male with mediastinal abscess after aortic root replacement. After left partial chest wall and pectoralis resection and multiple debridements (11 surgical debridements prior flap coverage), an omentum flap was performed to fill the cavity.

Figure 2.

A, A 50-year-old patient who underwent right extrapleural pneumonectomy for mesothelioma. He presented with bronchopleural fistula (arrow), empyema, and thoracotomy wound complications. B, The omentum flap was passed into the right pleural cavity through a right diaphragmatic tunnel over the right lobe of the liver.

Material and Methods

After institutional review board approval, the electronic medical records of all patients who underwent thoracic reconstruction using pedicled omentum flaps between July 2003 and January 2019 were reviewed. Descriptive statistics are reported in the form of mean ± SD, median (interquartile range [IQR]), or counts (percentages) as applicable. All tests were 2 sided, and P values <.05 were considered statistically significant.

Results

Patient Demographics

Forty patients (60% males) underwent chest wall or mediastinal reconstruction using a pedicled omentum flap at a mean ± SD age of 58 ± 15.8 years. The median follow-up was 24.3 months (IQR: 3.3, 49.7). Significant medical comorbidities were present in 32 (80%). The mean ± SD body mass index was 29 ± 5.2, and 20 (50%) patients had a history of tobacco use (Table 1).

Table 1.

Patient Demographics.

Variable	N = 40 (100%)
Male	24 (60%)
Female	16 (40%)
Mean age (years)a	57.7
BMI	29 ± SD 5.2
Comorbidities
Diabetes	11 (27.5%)
Hypertension	21 (52.5%)
Chronic heart failure	9 (22.5%)
Coronary artery disease	19 (47.5%)
COPD	7 (17.5%)
Chronic kidney disease	6 (15%)
Hyperlipidemia	12 (30%)
Smoking history	20 (50%)

Abbreviations: BMI, body mass index; COPD, chronic obstructive pulmonary disease.

^a Standard deviation: 15.8.

Indications

Omentum flaps were performed as the primary reconstructive method in 17 (42.5%) patients as an adjunct to other flaps in 10 (25%) (pectoralis major = 8, latissimus dorsi = 1, serratus anterior = 1, intercostal = 1) and as salvage of unsuccessful reconstruction in 13 (32.5%) (pectoralis major = 9, latissimus dorsi = 3, rectus abdominis = 2, vertical rectus abdominis myocutaneous (VRAM) = 1, and free anterolateral thigh (ALT) flap = 1). The most common indication was reconstruction of anterior chest wall/sternal defects secondary to sternal complications—sternal osteomyelitis, mediastinitis, and sternal dehiscence (n = 16), followed by coverage of repaired bronchopleural fistula (n = 6), osteoradionecrosis of the anterolateral chest wall (n = 5), reconstruction of anterior/lateral chest wall following oncologic resections (n = 5), coverage of replaced infected LVAD (n = 4), coverage of exposed/replaced aortic root vascular grafts (n = 4).

Flap Harvest and Effect of Vasoconstrictors

The omentum was harvested via a laparotomy in 39 patients and laparoscopically in 1 patient. This flap was based on the right gastroepiploic artery in 35 patients, the left gastroepiploic artery in 2 patients, and in 3 patients the flap has harvested based on both (right and left) gastroepiploic arteries. None of the flaps had complete necrosis. Ten (25%) patients underwent a single operation for omentum flap with a clinical satisfactory outcome. Thirty (75%) patients had planned return to the operating room (ROR) allowing omental flap inspection. Out of these 30 patients, 8 (27%) patients had partial flap necrosis. Vasoconstrictors were used in 26 patients (65%) (intraoperatively in 11, postoperatively in 4, and both in 11). Partial necrosis of the omentum occurred in 26% (5/19) of patients who required vasoconstrictors, and it occurred in 27% (3/11) of patients who did not require vasoconstrictors (P = 1.0). No association between vasoconstrictors and partial flap necrosis was identified. Thirteen (33%) flaps were skin grafted at a median of 13 days (IQR: 5. 5, 26. 5) with 100% skin graft viability.

Omentum Flaps in Patients With LVADs

Four flaps were used for coverage of replaced infected LVAD. In patients with LVADs, the omentum remained clinically viable with no evidence of recurrent infection during the follow-up period. Patients’ intraoperative characteristics are described in the Table 2.

Table 2.

Intraoperative Characteristics.

Variable	N = 40 (100%)
Surgical approach
Primary reconstructive option	17 (42.5%)
Adjunct to other flaps	10 (25%)
After failed reconstructions	13 (32.5%)
Indication
Cardiac/mediastinal complication	24 (60%)
Anterior chest defects secondary to sternal complicationsa	16 (40%)
Coverage of aortic root graft	4 (10%) 4 (10%)
Coverage of LVAD	4 (10%) 4 (10%)
Musculoskeletal resection	10 (25%)
Osteoradionecrosis	5 (12.5%)
Oncologic resection	5 (12.5%)
Pulmonary complication (bronchopleural fistula)	6 (15%)
Pedicle used
Right gastroepiploic artery	35 (87.5%)
Left gastroepiploic artery	2 (5%)
Both	3 (7.5%)
Vasoconstrictors
Yes	26 (65%)
Intraoperative	11 (27.5%)
Postoperative	4 (10%)
Both	11 (27.5%)
No	14 (35%)

Abbreviation: LVAD, left ventricular assist device.

^a Sternal osteomyelitis, mediastinitis and sternal dehiscence.

Postoperative Complications

Postoperative bleeding requiring ROR developed in 6 (15%) patients. Bleeding occurred at 0, 1, 1, 7, 7, and 70 days postoperatively. One patient was bleeding from the omentum surface, and the remaining 4 patients had chest wall hematomas related to the debridement. Three patients were on therapeutic doses of anticoagulation medications at the time of bleeding.

Ventral hernia developed in 8 patients, and 5 of them underwent surgical repair. In 4 of these patients, the omentum had been transferred to the anterior chest wall for wounds secondary to sternal osteomyelitis. In 3 of these cases, the omentum was brought out through the abdominal fascia and skin; in the fourth patient, the omentum was passed to the anterior chest through a subxiphoid mediastinal window.

In 2 patients, the omentum was transferred intrathoracically for bronchopleural fistula coverage. In one of these cases, the omentum was transferred through an opening in the diaphragm, and in the other, the omentum was passed through the abdominal wall fascia and tunnelled subcutaneously to the right chest cavity via the thoracotomy incision.

One patient had anterior chest wall defect secondary to chest wall sarcoma resection, and the omentum was transferred through an opening made in the Gore-Tex mesh that was used for Gore-Tex reconstruction. The ventral hernia in this patient occurred in the abdominal portion of the incision and was repaired primarily.

One other patient had a ventral hernia in the lower aspect of the sternal incision. The patient had a previous history of sternal dehiscence and underwent chest wall reconstruction with pectoralis flap associated with omentum flap, which was transferred through an opening in the diaphragm.

One patient died during hospital admission due to acute cardiogenic shock as a result of severe tricuspid valve regurgitation and recurrent constrictive pericarditis. Patients’ outcomes are described in Table 3.

Table 3.

Postoperative Outcomes.

Variable	N = 40 (100%)
Follow-up (months)a	24.3
Flap survival	40 (100%)
Postoperative complication
Partial flap necrosis	8
With vasoconstrictors	5
Without vasoconstrictors	3
Bleeding requiring return to OR	6 (15%)
Ventral hernia	8 (20%)
Perioperative mortality	1 (2.5%)

Abbreviation: OR, operating room.

^a Interquartile range: 3.3 to 49.7.

Discussion

The omentum flap is a versatile reconstructive tool for a variety of complex chest wall and mediastinal defects. ²³ The dual blood supply from the 2 gastroepiploic arteries allows this flap to be based on either vascular pedicle, enhancing the flap reach to different thoracic regions. ⁶ The abundance of lymphatic vessels and nodes makes it a suitable reconstructive option in anatomic locations and defects where an active infection exists or has been recently treated. The omentum is a well-vascularized, pliable tissue with an almost fluid consistency that allows it to fill complex 3D wounds and cavities.

In this study, the omentum was used as a primary reconstructive option, as an adjunct to other reconstruction, and as a salvage option for failed previous reconstructions. When the omentum is used as a primary reconstructive option, local options utilizing chest wall musculature have usually been deemed as inadequate, either due to the defect size or location. In these situations, the omentum provides ample tissue for reconstruction and resurfacing. The omentum was also used as adjunct for other reconstructive options. In our study, there were 2 main reasons for using the omentum as adjunct to other reconstructive options: location and size of the area needing coverage. There are certain locations in the chest wall cavity which are difficult to adequately cover with local chest wall muscles flaps such as the serratus anterior, latissimus dorsi, and pectoralis major such as the most inferior portion of the sternum because the arc of rotation hinders adequate reach. Moreover, the chest wall musculature can be inadequate to obliterate large intrathoracic cavities or difficult to reach middle mediastinal pathology. In these circumstances, the omentum is an ideal option to obliterate dead space, resurface, and provide stable soft tissue coverage of the lower sternal/upper abdominal region. The large surface area of the omentum coupled with the ease of shaping and molding make it an ideal coverage and resurfacing option for exposed vascular grafts and LVADs. The abundance of lymphatic channels and nodes in the omentum may be of theoretic benefit in infected or colonized wounds.

Harvesting the omentum involves violation of another body cavity with the added risk of bowel injury, injury to the surrounding abdominal viscera, gastric outlet obstruction, and development of abdominal wall hernia. No case of bowel injury or other significant intra-abdominal complication was reported in our series. Nevertheless, the potential risk of significant adverse events related to omentum flap harvest should be considered when deciding on the appropriate reconstructive option for chest wounds, and this is the likely reason most plastic surgeons avoid using the omentum as a first-line option. The most commonly encountered adverse events related to this procedure seem to be related to the abdominal wall; with 8 patients (20%) developed an abdominal wall hernia and 5 of these required surgical repair, previous studies have shown a hernia rate ranging from 7% to 32%. ^{23 -28} Harvesting the omentum laparoscopically may reduce this rate. ^23,29 Passing the omentum from the abdominal cavity to the chest wall or cavity requires creation of a connection between these 2 body cavities. Any connection between these cavities is nonanatomic and has the potential for intra-abdominal contents to herniate through the defect and around the transferred omentum. The occurrence of a hernia did not seem to be related to any specific tunnelling method or location; however, it is difficult to draw conclusions from reviewing only 8 cases of these hernias in this complex population. Development of a hernia in these patients is likely multifactorial and related to body habitus, wound configuration, site of passage, size of the tunnel, patient activity, and anchoring techniques around the defect. The tunnel through which the omentum is passed should be adequate enough to prevent mechanical compression of the pedicle and should be wide enough to allow for postoperative swelling and avoid vascular compromise of the transferred omentum. There is no “ideal tunnel size,” and the decision of how wide the tunnel needs to be is at the discretion of the surgeon. Although Kolbenschlag et al identified that herniation affects patients’ quality of life by limiting mobility and causing pain, ²³ the risks of having a necrotic omentum within the mediastinum or chest cavity far exceed the risk of subsequently developing an abdominal wall hernia and likely factor into intraoperative decision making in regards to the tunnel size when passing the omentum.

Bleeding requiring either transfusion or ROR occurred in about 15% of the cases. The large surface area of the well vascularized omentum and its fragile vasculature may be risk factors for postoperative bleeding; however, only 1 patient was found to have bled from the omentum, with the remaining cases likely related to the thoracic debridement. The concomitant administration of anticoagulant medications is also likely contributory.

The Impact of the Laminar, Nonpulsatile Blood Flow of LVADs on the Viability of the Omentum

A large number of patients are affected with heart failure annually and require LVAD placement as a bridge to heart transplant or as destination therapy. When these devices become infected, it is a challenge to treat the resulting wounds and eradicate the infection. Failure ultimately results in significant morbidity and mortality. Source control along with debridement and antibiotics can control the infection, but there may also be large defects or spaces adjacent to the device that would benefit from soft tissue obliteration or coverage. In these situations, the omentum is an excellent choice due to its ability to conform to complex spaces, as well as its excellent vascularity, volume, and theoretical immunological properties.

The absence of pulsatile flow in patients with LVADs coupled with vasoconstrictor infusions poses an intraoperative challenge to the plastic surgeon when trying to assess the presence of an adequate pedicle with the handheld Doppler. In our series, the omentum was used to cover replaced, infected LVADs in 4 patients. The omentum remained clinically viable in all patients. Recent cases series have also reported successful salvage of LVAD with omentum flaps. ^2,7,19,20,30

The Effect of Vasoconstrictors on Omentum Viability

The high prevalence of medical comorbidities, especially cardiac comorbidities, in our patient population increases the likelihood of requiring vasoconstrictors during the majority of chest wall and mediastinal reconstructions. We were unable to find an association between partial omentum flap loss and the use of vasoconstrictors. The low metabolic rate associated with the high concentration of fat cells in the omentum could explain this ability of the omentum to remain viable in the midst of relative hypoperfusion associated with vasoconstrictors. We could not find other studies evaluating the effect of vasoconstrictors on omentum viability, and larger series will be required to challenge or support our findings.

Partial flap loss occurred in 8 patients; 27% of those patients who had reoperation and had the omentum flap were evaluated. The majority of the flap was viable in these patients, and minimal debridement was required to remove the devitalized tissues. For this reason, we prefer to delay skin grafting to a later date (average of 13 days). This usually allows enough time for areas with marginal perfusion to demarcate. Following this interval, patients are taken to the operating room for excisional debridement of devitalized tissues—when present—and split thickness skin grafting.

There were several limitations to this study. The retrospective collection of data was dependent on the accuracy and completeness of the medical charts and operative reports. Our results are from a single academic medical centre, and indications for performing omentum flaps are dependent on the surgical practice pattern at our institution, and these indications may not represent the national practice pattern. However, we seek to critically appraise our results, question the potential impact of vasoconstrictors and LVADs flow pattern on omentum viability, and highlight the utility and technical aspects of thoracic reconstruction with the omentum flap to help address knowledge gaps in the subject and provide results against which we can compare future outcomes.