Topical application of autologous blood products (PRP) during surgical closure following coronary artery bypass grafting
Entrance
Objective: Surgical site wound complications are associated with increased cost, morbidity, and mortality after cardiothoracic surgery.
Recent publications advocate the administration of autologous blood components (PRP, PRF, l-PRP, i-PRF, etc.) as an adjunct to surgical closure during various surgical procedures. The aim of this study was to evaluate the safety and efficacy of autologous platelet-rich and platelet-poor plasma administration to the sternal closure and saphenous vein collection site during closure after a coronary artery bypass graft.
Patients and methods: A retrospective analysis of 1446 consecutive coronary artery bypass grafting procedures by two surgeons was performed. A group of patients to whom platelet-rich and platelet-poor plasma was applied topically during closure of chest and leg surgical incisions was compared with a population of patients receiving standard-of-care therapy. Forty covariates were collected for each patient included in the study. Trend scoring was used to adjust for the underlying imbalance. Asymptotic logistic regression and precise statistical methods were used to determine the effect of autologous blood application on infection and drainage of sternal and leg wounds.
Results: One thousand one hundred and twenty-eight patients, of whom 571 were treated compared to 557 control patients, had sufficient data to be included in the final analysis. No treatment-related side effects were recorded, and the administration process did not significantly affect the operative time.
Conclusion: This retrospective analysis of a series of consecutive patients undergoing a coronary artery bypass grafting procedure concluded that administration of platelet-rich and platelet-poor plasma significantly reduced the incidence of chest wound infection, chest drainage, and leg wound drainage. This new therapy deserves further research.
1. Introduction
Post-operative wound disorders, particularly surgical site infection, are associated with increased morbidity, mortality, and costs following cardiothoracic surgery [1-2]. Sternal infection (mediastinitis) has been reported to occur in up to 20% of cases, but most studies report an incidence rate of 1-2% [3]. The overall wound discomfort rate for chest incision following cardiothoracic surgery is 8-10% [4,5]. Leg incision for saphenous vein harvest may also be the site of significant postoperative wound complications. The trend towards endoscopic removal of the saphenous vein has reduced complication rates, but endoscopic removal of the saphenous vein still has an infection rate of 2-4% and an overall wound complication rate of 6-8% [6-8]. The high morbidity, mortality, and costs associated with these postoperative wound conditions provide advances that can reduce postoperative wound complications that are attractive to cardiothoracic surgeons. Topical application of platelet-rich plasma (PRP) and platelet-poor plasma (PPP) together with a coagulation agent (typically bovine thrombin) has been advocated for many indications [9-11]. The purpose of PRP application is to accelerate the healing process with the effect of high cytokine concentrations released during platelet degranulation. Platelet-derived growth factor (PDGF), epidermal growth factor (EGF), vascular endothelial growth factor, and transforming growth factor beta (TGFb) are examples of cytokines that have been shown to be present at concentrated levels in PRP [12]. These high cytokine levels are hypothesized to cause an accelerated healing response at the application site. Produced as a by-product during the centrifugation process used to produce the platelet-rich product, PPP has been advocated as a tissue adhesive for topical hemostasis during surgical closure [11]. Two recent publications have reported improved postoperative outcomes following topical application of PRP and PPP in combination with bovine thrombin to the chest incision and vein collection site during surgical closure [13–15]. Englert et al. showed a trend towards reductions in chest incision pain, leg incision pain, and measurable bruising in a randomized, blinded study of a total of 30 patients [13]. Trowbridge et al. compared 382 patients receiving platelet-rich plasma with a nonrandomized concurrent control group of 948 patients and a historical control group of 929 patients [14]. This analysis showed a significant reduction in the rate of superficial and deep chest wound infections in the platelet-rich plasma group compared to both concurrent and historical controls. These studies support the adjunctive use of autologous blood components in cardiothoracic procedures and warrant further research into the safety and efficacy of this technique.
The following study is a retrospective analysis of 1446 consecutive coronary artery bypass graft (CABG) procedures. A population receiving topical application of autologous platelet-rich and platelet-poor plasma to the sternal closure and saphenous vein collection site was compared with a control group. Forty surgical measures were compared between the two groups to assess the safety of treatment and to determine the relationships between treatment and various outcome measures. Propensity scoring techniques were used to address bias between the treatment and control groups, thus increasing the ultimate value of this analysis.
​
2. Materials and methods
2.1. surgical method
The procedures included in this analysis were performed by two surgeons at six different surgical centers. All patients were given preoperative intravenous antibiotics 30-60 minutes before the procedure. Open surgical methods were used for all median sternotomies, and a standard wiring technique was used for sternal closure. An endoscopic approach was used for saphenous vein harvesting. In the treatment population, 55 cc of whole blood was drawn immediately before surgery and 5 cc of the anticoagulant citrate dextrose was mixed with formula A. For patients requiring saphenous vein harvest, an additional 55 cc of blood was drawn in the same way. Anticoagulated blood was then processed using the GPSTM II platelet concentrate system (Biomet Biologics, Warsaw, IN). After a 15-minute centrifuge cycle, 5-7 cc of PRP and 25-30 cc of PPP were obtained per 60 cc disposable. Topical bovine thrombin was diluted with 10% calcium chloride at a concentration of 1000 U/ml and applied 1:10 with the desired blood component using a dual spray device (Micromedics, Eagan, MN). PRP was sprayed onto the exposed sternal edges of the chest wound and subcutaneous tissue, with PPP applied to each tissue layer during closure. In the leg area, PRP was applied along the graft retrieval area using a 30 cm long double cannula (Micromedics).
​
2.2. Statistical analysis
All patients who underwent CABG between October 2000 and September 2005 were included in the study. Records of forty operative covariates and procedural adverse events were collected for each patient. Cases with serious morbidities (multi-organ failure, cancer, etc.) beyond the scope of the cardiothoracic procedure were excluded from the final analysis to allow for a more heterogeneous patient population. The relationship between ABP administration and outcomes (chest infection and drainage, leg infection and drainage) was examined using Fisher's exact test. In addition, trend scoring was used to adjust for potential fundamental imbalance.
Because propensity scoring requires complete completion of covariates, missing covariates were replaced with the overall mean for that variable. Although this method of data management artificially reduces the variance of a trend score, it does not affect the ranking of the score used in the final analysis. After reducing the selection bias, trend scores were generated for outcome measures using observed covariates. Asymptotic model-based parameter estimations were used to estimate parameters for binary outcomes [20]. The normality of the count results was tested using the Shapiro-Walk statistic. These variables were also examined for compliance with a Poisson distribution. In cases where normality was rejected, the evaluation of count variables used a Poisson regression model.
​
3. Results
A total of 1128 patients were included in the analysis. Autologous blood products (APP) were administered to 571 patients and 557 patients served as controls (ABP not administered). Data were collected between February 2002 and September 2005 for ABP patients and between October 2000 and February 2005 for control patients. Forty covariates were included in the final analysis (Table 1). Analysis of leg infection and drainage rates used a subset of patients harvested from the saphenous vein (nABP = 560, nccontrol = 456). There was an incidence of sternal infection (0.18%) in the ABP group compared to 11 cases (1.98%) in the control group. There was significant drainage from the sternum in 3 cases (0.53%) of ABP compared to 30 cases (5.39%) in the control group. For the leg vein collection site, there were no reported cases of infection and 61 (10.89%) cases of excessive drainage in the ABP group, compared to 3 (0.66%) cases of surgical site infection and 212 (48.4%) cases of excessive leg drainage (Table 2 ). After trend scoring, it can be concluded that ABP application reduces the probability of chest wound infection by 93%, chest drainage by 96% and leg wound drainage by 88% (Table 3). Since there was no leg infection (zero) in the ABP group, there was no significant difference between the groups in terms of leg infection.