Introduction

Despite long-term survival benefits, the use of BITA grafts for myocardial revascularization is still limited [1—3]. This is commonly linked to the concern of many surgeons about poor wound healing that restricts appropriate indications [4, 5]. These wound healing complications have been attributed to excessive devascularization of anterior chest wall inherent in BITA harvesting. As a result, BITA grafts have only been used in selected patients with no comorbidities in order to minimize sternal complications. However, this strict selection would deprive too many patients from good long-term patency rate after BITA grafting, in particular those suffering from diabetes. This study aimed to analyze the risk factors of SWI complicating BITA grafting.

Methods

Patient population

We performed a retrospective comparative cohort study between January 2018 and March 2023. There were patients with double and triple vessel disease who underwent isolated coronary bypass surgery using skeletonized BITA grafts. Patients who underwent emergency surgeries and those who needed concomitant procedures other than CABG, as well as those requiring only a single graft were not included. Patients with unavailable perioperative data were excluded.

According to the guidelines of the Centers for Disease Control and Prevention, SWIs were defined and classified as superficial or deep based on visual evidence of mediastinitis, chest instability, fever associated with purulent mediastinal drainage or identification of organism from culture of mediastinal tissue or fluid [6—8]. Superficial SWIs were defined as infection involving the skin and subcutaneous tissue of surgical wound. Deep SWIs were defined by combination of sternal dehiscence and infection involving retrosternal area and mediastinitis.

Patients were divided into two groups: diabetics (Group D) and non-diabetics (Group ND) both receiving BITA grafts. We have opted for skeletonized ITAs harvesting as a revascularization strategy in our routine practice. The outcomes of patients were reviewed retrospectively and risk factor analysis was performed.

Management of sternal infections

Prior to surgery, a nasal swab is routinely taken from all patients to identify those who are carriers of staphylococcus aureus. Decontamination with viatris 2% fusidic acid is indicated if the result is positive and surgery is deferred until the swab is negative. The day before, all patients shaved and washed with 4% chlorhexidine. Patients with sternal infection were admitted, and a sample from surgical site was taken and sent for bacteriological analysis. If a germ is isolated, antibiotic treatment is initiated and adapted to culture results.

For superficial SWIs, debridement and directed healing with approximations associated with daily dressing changes were carried out in case of small skin dehiscence. Regarding deep SWIs, vacuum-assisted closure was followed by secondary suturing if there was extensive tissue loss with sternum stability (Figure on a colored sticker). Otherwise, surgical removal of infected tissue with sternal refixation was performed.

a — deep sternal wound infection after coronary artery bypass grafting using bilateral internal thoracic artery; b — deep SWI after the first VAC procedure; c — wound bed prior to reconstruction following multiple courses of VAC therapy.

Data collection

The data from all patients who underwent CABG were collected from medical records. All data were recorded and tabulated in retrospective database.

Preoperative variables included age, sex, BMI, diabetes, hypertension, dyslipidemia, smoking history, serum creatinine, need for dialysis, chronic obstructive pulmonary disease (COPD), EuroScore II, left ventricular ejection fraction (EF). Intraoperative data included aortic cross-clamp (ACC) time, cardiopulmonary bypass (CPB) time, number of distal anastomoses. Postoperative data included ventilation time, blood transfusion, chest reopening, postoperative events, ICU and hospital stay.

Statistical analysis

Continuous variables were expressed as means and standard deviation. Categorical variables were presented as counts with percentages and analyzed using the Fisher exact test and the chi-squared test. Comparisons of several means were carried out using the ANOVA parametric analysis of variance. The association between two numerical variables was tested using the Pearson’s correlation coefficient. Univariate analysis was performed to identify the relationships between SWIs and pre-, intra- and postoperative variables. The probability of a patient having a SWI was determined using logistic regression analysis. The model selection was done with backward stepwise method or input method. A p-value <0.05 was significant.

Results

A total of 243 patients who underwent CABG were enrolled; 134 patients were diabetic (Group D), 109 — non—diabetic (Group ND). Morbid obesity was the main epidemiological factor significantly associated with SWI in diabetic and non-diabetic populations. Similarly, additional bypass and prolonged hospital stay were intra- and postoperative factors of SWIs with significant association in diabetic and non-diabetic patients. Demographic characteristics of both groups are summarized in Table 1, 2.

Table 1. Clinical characteristics of patients in the ND group with and without SWI

Note. BMI — Body Mass Index, AMI — Acute myocardial infarction, EF — Ejection Fraction, ECMO — Extracorporeal membrane oxygenation, ARDS — Acute respiratory distress syndrome, MI — Myocardial infarction

Table 2. Clinical characteristics of patients in the D group with and without SWI

The group D enrolled 134 patients (55.1%) including 110 (82%) men. Mean age was 60.4±8.8 years. Of these, 25.4% ones developed SWI (14.2% with superficial SWI and 11.2% with deep SWI). In contrast, the group ND comprised 109 patients (44.8%) including 98 (89.9%) men. Mean age was 60.2±8.5 years. Of these, 16.5% of patients had SWI (13.8% superficial SWI, 2.8% deep SWI). Overall SWI risk was insignificantly higher in the diabetic group (25.4% in group D vs 16.5% in group ND; p=0.094). However, deep SWI was more common in diabetic patients (p=0.012). The difference between groups ND and D is presented in Table 3.

Table 3. SWI in diabetic and non-diabetic patients

In-hospital mortality associated with SWIs after BITA grafting was 5.6% and 8.8% in groups ND and D, respectively. There was no statistical difference between patients who died with or without SWI in group ND (5.6% vs 4.4% respectively; p=0.830). In contrast, 30-day mortality was higher in patients without SWI than in those with SWI in group D, but not being significant (12% vs 8.8% respectively; p=0.612).

Discussion

Although the use of arterial conduits confers better survival benefit after CABG, BITA is not adopted as a primary practice by many surgeons. This is due to conflicting results, especially regarding poor wound healing associated with increased morbidity, mortality and healthcare costs [9].

SWIs have been considered to be serious and potentially life-threatening complications following BITA grafting. Recent data have shown that the incidence of SWI following BITA grafting is estimated to be between 2—3%, without significant difference between patients with single ITA grafting and those with bilateral ITA grafting [10]. The risk factors of SWI following BITA grafting can be broadly categorized as patient-related and surgery-related.

Patient-related risk factors include advanced age, obesity, diabetes, COPD, renal insufficiency, and immunosuppression. Previous studies have shown that patients with these comorbidities are at higher risk of SWI after BITA grafting [11]. Diabetic patients are more likely to have SWI due to several physiologic derangements. However, this eventuality has been greatly reduced by skeletonized harvesting technique that preserves ITA collaterals to the chest wall [9—14]. The use of skeletonized BITA grafts may potentially reduce the risk of deep SWI [15, 16]. Qiang Ji et al. concluded that skeletonized BITA harvesting compared to skeletonized unilateral ITA harvesting was not associated with higher risk of deep SWI regardless of whether the merger was diabetic or non-diabetic. Patients who received a skeletonized BITA graft had similar surgical mortality and major postoperative morbidity as matched patients who received a skeletonized SITA graft [16]. Thomas et al. reported that skeletonization of BITA grafts was protective factor against deep SWI. They noted a 28% reduction in deep SWI with skeletonized BITA compared to non-skeletonized BITA, but without significant difference [17]. Some studies showed that skeletonized BITA improved perioperative survival compared to pedicled BITA [18]. A study by Dos Santos Filho et al. comparing sternal blood flow after skeletonized bilateral harvesting has shown better flow compared to pedicled approach [19]. Moreover, preventive measures for diabetes control with preoperative glycosylated hemoglobin<7% reduced the risk of SWI following BITA grafting in diabetic population [20]. In the present study, 25.4% of diabetic patients developed SWI compared to 16.5% of non-diabetic patients (p=0.094). However, deep SWI was more common in diabetic patients receiving BITA grafting (p=0.012).

Overweight and morbid obesity were implicated in development of sternal wound complications after BITA grafting. Excessive fat in sternal area may facilitate bacterial colonization and delay healing. It is therefore recommended to lose weight before surgery and to wear chest belts afterwards in patients with BMI≥29 kg/m² [21]. In our series, morbid obesity was associated with 18 (52.9%) SWI in the group D and 11 (61.1%) in the group ND, whereas 14 (41.2%) patients with overweight developed SWI in the group D and seven others (38.9%) developed SWI in the group ND.

Elderly patients have difficulty coughing and expelling bronchial secretions, which can lead to respiratory infections requiring non-invasive ventilation or even reintubation and prolonged mechanical ventilation. Such circumstances progressively cause sternal dehiscence that can induce bacteremia and wall infection.

Surgery-related risk factors include prolonged surgery time, reopening for bleeding, sternal closure technique, and prolonged ICU-stay. Prolonged CPB and redo sternotomy for bleeding increase bacterial exposure. During reopening, retrosternal hematoma and clots surrounding the heart can be sources of infections. Therefore, hemostasis disorders should be corrected in initial hours before coagulation factors are consumed and making surgical hemostasis necessary. Total arterial grafting associated with new techniques of myocardial revascularization, in particular sequential bypass, has enabled to reduce the number of proximal anastomosis and therefore CPB time. In our study, total CPB time and reopening were not statistically associated with SWI in both groups.

Patients with severe left ventricular dysfunction usually stay for longer period in intensive care units with higher risk of line infection. Transfer to general wards must be done as early as possible to reduce the risk of SWI.

Vacuum-assisted closure (VAC) is a new technique for deep SWI with stable sternum and no need for redo surgery. VAC was primarily designed for the treatment of pressure ulcers or chronic debilitating wounds. Recently, advanced experience of VAC treatment in cardiac surgery demonstrates superior results over conventional treatment of sternal infection. The controlled use of subatmospheric pressure on wounds has dual benefits. It contributes to wound edge closure, tissue growth and better healing through lower swelling and bacterial contamination [22]. Thus, VAC therapy reduced the need for surgical interventions and redo surgeries for persistent infections [23].

This study has some limitations. Firstly, it adhered to retrospective and observational design, as well as single-center nature. Secondly, aspects of bones including osteoporosis, ischemia, failure to follow antisepsis procedures, faulty sternotomy and rewiring, as well as excessive use of electric scalpel favoring SWI, were not systematically studied. Furthermore, this did not evaluate the contribution of potentially relevant factors such as causative pathogens and antibiotic prophylaxis to SWI risk. Thirdly, it is conceivable that overall count of SWI might have been underestimated because some patients in a debilitated clinical state died in intensive care unit postoperatively due to myocardial infraction, pneumonia, or cerebral infraction. Consequently, we presumed that patients who died had no SWI. Finally, ITA harvesting technique and timing of surgical interventions are surgeon-dependent, thereby lacking uniformity across all patient cases.

Conclusion

Deep SWI has long been considered as a devastating complication following BITA grafting and associated with higher morbidity and mortality. Although the present study was conducted on a relatively small sample of patients, it showed satisfactory results of isolated CABG with higher rate of BITA grafting and low incidence of deep SWI and in-hospital mortality. These findings are encouraging and support routine use of BITA grafting even in patients with comorbidities.

Funding

This study did not receive any grant from funding agencies in the public, commercial or not-for-profit sectors.