Significance:

Complex surgery of the ascending thoracic aorta and the aortic arch requires use of cardiopulmonary bypass (CPB) with moderate or profound total body hypothermia and circulatory arrest. As expected, coagulation disorder can be common and severe requiring multiple transfusions of blood product, which, in turn, is associated with increased risk of morbidity and mortality. Over the years, many techniques have been employed to ameliorate CPB- related bleeding with some agents such as aprotinin proving to be efficacious but centered in controversy either because of high cost or due to safety concerns.

Aprotinin is a serine protease inhibitor, intended to limit blood loss in patients undergoing surgery, was approved for use 13 years ago, and now administered in more than 250,000 surgery patients per year. Although its mechanism of action is not entirely known, it is shown to interfere with contact activation, preserve platelet function, inhibit fibrinolysis, and having some anti-inflammatory effects. Multiple studies have shown its superiority for reducing blood loss over other antifibrinolytics while other studies showing safety concerns regarding early graft closure causing myocardial infarction (following coronary bypass surgery), renal failure, vascular thrombosis, and death. Of note, several other studies have addressed such safety concerns- finding that the risk of the aforementioned vascular complications are comparable to results utilizing placebo or other antifibrinolytics such as aminocaproic acid. For example, the associated renal dysfunction was attributed to a competitive excretion process involving creatinine. That is, aprotinin competes with creatinine for elimination by the kidneys such that the apparent renal dysfunction is transient and a false marker of renal injury. However, the issue of renal injury has recently re-emerged as a serious problem in a multicenter study on patients undergoing coronary revascularization, whereby more than two fold increase in postoperative renal dysfunction as well as a nearly two fold rise in postoperative stroke was reported.

Much of the controversy and many of the conflicting results arise from either inadequate information (i.e., the definitive prospective study has not been done), from inadequately powered studies for the intended outcome, inclusion of multi-institution and multi-surgeon outcomes, or from studies involving patients with advanced ischemic heart disease whose ventricular function may have been compromised perioperatively. To overcome the mentioned limitations, and to better characterize to postoperative renal outcomes, we propose to investigate the safety of aprotinin in a single institution, single surgeon, and very large homogenous surgical cohort who underwent surgical repair of the ascending and aortic arch with use of hypothermic CPB.

Methods:

This will be a retrospective review of data collected between 1998 and present (nearly 900 patients) on adult patients having had ascending aorta and or aortic arch surgery at the NewYork Presbyterian Hospital. Data on demographics, preoperative, intraoperative, and postoperative clinical events will be extracted from the patient's electronic record and recorded on a database (Variables to be collected are presented in Appendix 1). A renal event will be comprised of renal dysfunction; defined by a serum creatinine level of at least 2.0 mg/dL (177 µmol/L), accompanied by an increase of at least 0.70 mg/dL (62 µmol/L) from baseline or renal failure; defined need for renal replacement therapy. Creatinine values and creatinine clearance will be recorded/calculated preoperatively and postoperatively until hospital discharge.

Statistical Analysis Include analysis for association between potential preoperative, intraoperative, and early postoperative variables with postoperative renal dysfunction by univariate analysis (student's t test and Fisher's exact test). The variables identified by univariate testing (p< 0.15) will be entered into a multivariate logistic regression. Significance will be set a P < 0.05 and odds ratio will be calculated where applicable.