Approval for this retrospective case control study was obtained from the Institutional Review Board (IRB) of the Brigham and Women's Hospital/Harvard Medical School. Written informed consent was waived for this retrospective study.

A total of 454 patients, 50 years old and older (mean ± 68 years) with normal renal function preoperatively (creatinine < 1.5 mg/dL), who had undergone heart surgery at Brigham and Women's Hospital between January 1995 and December 2005 were retrospectively evaluated. Out of those patients we selected all patients who had received an IABP within 24 hours prior to or during surgery. Patients with preoperative renal insufficiency (creatinine ≥1.5 mg/dl) and TEE imaging-windows which were inadequate to demonstrate the severity of DTA atheroma, as well as patients who received IABP support after leaving the operating room, were excluded from the study population. Matching for the remaining 454 patients was then attempted according to the degree of DTA atheroma, performed surgical procedure, age, gender and left ventricular ejection fraction (LVEF) to one of four groups: -IABP/-DTA atheroma, +IABP/-DTA atheroma, -IABP/+DTA atheroma, +IABP/+DTA atheroma. Patients with DTA atheroma were slightly older and less likely to be male.

Demographics and historical data were obtained from the computerized institutional database and included age, gender, history of hypertension necessitating treatment with antihypertensive medications, diabetes mellitus, congestive heart failure, hypercholesterolemia, medications including Angiotensin-converting ezyme (ACE) -inhibitors, Aspirin (ASA) and 3-hydroxy-3-methyl-glutaryl (HMG) -CoA reductase inhibitors and use of aminocaprioc acid or aprotinin during surgery. Preoperative co-morbidities were defined according to the Society of Thoracic Surgeons (STS) standard definitions. Congestive heart failure (CHF) was defined as either paroxysmal nocturnal dyspnea, dyspnea on exertion due to CHF, a chest x-ray showing pulmonary congestion or pedal edema and dyspnea and receiving diuretics or digoxin within two weeks prior to the surgical procedure. Use of ASA, ACE-inhibitors or statins was defined as whether the patient received the medication within 24 hours prior to surgery. Diabetes mellitus indicated whether the patient had a history of diabetes, regardless of duration or need for anti-diabetic agents. Dyslipidemia was referred to as hypercholesterolemia. Criteria included the documentation of total cholesterol > 200 mg/dl or LDL > 130 mg/dl or HDL < 30 mg/dl or admission cholesterol > 200 mg/dl or triglycerides > 150 mg/dl. The criteria for hypertension included a documented history of hypertension diagnosed and treated with medication/diet and/or exercise, blood pressure > 140 systolic or > 90 diastolic on at least two occasions or currently on antihypertensive medication.

Perioperative AKI was defined according to the criteria of The Society of Thoracic Surgeons (STS) as an increase in creatinine > 2.0 (mg/dl), a 100% increase from the most recent preoperative creatinine level, or need for dialysis [¹⁵]. In-hospital mortality was defined as death occurring during the hospitalization in which the original cardiac surgical intervention was performed, according to STS standards [¹⁵].

The institutional intraoperative TEE database was used to identify cardiac surgical patients in whom imaging windows of the descending thoracic aorta were obtained. Off-line analysis of the recorded study was performed, and the degree of DTA atheroma was independently assessed by two attending cardiac anesthesiologists certified in perioperative TEE. Data from the thoracic aorta were extracted from the comprehensive TEE examination that is performed as part of routine care in applicable cardiac surgery patients. The following views were reviewed in accordance with the published Society of Cardiovascular Anesthesiologists/American Society of Echocardiography guidelines:

Descending Aorta:

1. Midesophageal descending aorta short axis at 0 degrees

2. Midesophageal descending aorta short axis at 90 degrees.

The severity of DTA atheroma was graded according to the practice guidelines for perioperative transesophageal echocardiography. A report by the American Society of Anesthesiologists and the Society of Cardiovascular Anesthesiologists Task Force on Transesophageal Echocardiography defines the grades as follows: [¹⁶] Grade 1 = no calcification, Grade 2 = intimal thickening ≤1 mm, Grade 3 = calcific protrusion 1 to 5 mm into the aortic lumen, Grade 4 = calcific protrusion > 5 mm into aortic lumen, Grade 5 = mobile atheroma. For this study, we defined significant DTA atheroma as the presence of an intraluminal plaque > 1 mm (that is, Grades 3 to 5 as described above) [¹⁷].

Descriptive statistics were obtained on patient characteristics including demographic data, past medical history and in-hospital data. For bivariate analysis the Fishers exact test for the categorical variables and the non-parametric Mann-Whitney U test for the continuous variables were employed. Multivariate analysis was performed using logistic regression. Covariates adjusted for included all the variables that had a P ³0.2 or that were felt to be clinically significant confounders. The Wald statistic was used to define the significance of the regression coefficients. A backward stepwise (likelihood ratio) technique was used for the multivariate analysis. The OR, P values and 95% CI were obtained from this. A P value of 0.05 was used to reject the null hypothesis (that the regression coefficient equates to zero). The statistical analysis was performed using JMP (Cary, NC, USA) and SPSS (Chicago, IL, USA).

Demographic data did not differ significantly between groups (Table 1) except for patients in the +DTA group who were slightly older and less likely to be men. Diabetes mellitus was more prevalent in the patient groups with calcification of the DTA, but this difference was not statistically significant. The use of preoperative drug therapy was not significantly different between groups (Table 1). Surgical procedures performed were not statistically different between groups (Table 2). The use of antifibrinolytics such as amicar (ε-aminocaproic acid) or aprotinin did not differ between groups and was not an independent predictor of AKI or death in our patient population (Table 2).

The total incidence of AKI was 6.4% within the study population. The analysis showed that 3.9% of the patients without IABP support developed AKI compared with 8.8% of the patients with IABP. The incidence and the ORs of AKI within groups were as follows: -DTA atheroma/-IABP: 1.7%, -DTA atheroma/+IABP: 5.1%, +DTA atheroma/-IABP: 5.5%, and +DTA atheroma/+IABP: 12.6%. Using Fisher's exact test, patients with the combination of IABP and DTA atheroma (+IABP/+DTA atheroma) had a significantly increased risk of developing postoperative AKI (P = 0.0016, relative risk increase 7.062, CI = 1.64 to 30.4) when compared to patients without IABP and without DTA atheroma (-IABP/-DTA atheroma). Patients with either IABP or DTA atheroma showed an increase in AKI postoperatively, which was not statistically significant when compared to either group (Figure 1).

In the univariate analysis, patients who developed postoperative AKI were more likely to be older than 65 years (P = 0.008); have a history of CHF (P = 0.026), undergoing a valve procedure (P = 0.012) and have DTA calcification (P = 0.003) or an IABP (P = 0.002) (Table 3). Multivariate logistic regression controlling for confounders such as patient demographics, preoperative risk factors, medications and intraoperative variables demonstrated a history of CHF (P = 0.026, OR = 2.81, CI = 1.13 to 6.97), the use of IABP (P = 0.015, OR = 3.04, CI = 1.24 to 7.45) and DTA calcification (P = 0.002, OR = 4.13, CI = 1.66 to 10.30) as independent predictors for the development of AKI in our patient population (Table 4). An analysis of effects (the chi-square test statistics and associated P-values) indicates that IABP and calcification are significantly associated with increased postoperative AKI individually, but not the interaction between the two.

The incidence of in-hospital mortality among all patients was 11.3% and included 6.6% of patients without an IABP versus 16.2% of patients with IABP support. The incidence of in-hospital mortality by Group was: -DTA atheroma/-IABP: 3.4%, -DTA atheroma/+IABP: 8.5%, +DTA atheroma/-IABP: 6.4%, and +DTA atheroma/+IABP: 20.2%.

Using Fisher's exact test, patients with the combination of IABP and DTA atheroma (+IABP/+DTA atheroma) had a significantly increased risk of increased postoperative mortality (P = 0.0001, relative risk increase 5.7, CI = 2.04 to 15.9) when compared to patients without IABP and without DTA atheroma (-IABP/-DTA atheroma) (Figure 2).

In the univariate analysis, patients who died were more likely to be older than 65 years (P = 0.008), have a history of CHF (P = 0.01), receiving either aminocaproic acid (P = 0.04) or aprotinin (P = 0.041) and have DTA calcification (P = 0.04) or an IABP (P = 0.036) (Table 3). Multivariate logistic regression controlling for confounders such as patient demographics, preoperative risk factors, medications and intraoperative variables showed age (P = 0.015, OR = 3.53, CI = 1.28 to 9.75), perfusion time (P = 0.002, OR = 1.01, CI = 1.00 to 1.01), the use of an IABP (P = 0.016, OR = 2.61, CI = 1.19 to 5.70) and DTA calcification (P = 0.012, OR = 2.69, CI = 1.24 to 5.83) to be independent risk factors for mortality (Table 5).

An analysis of effects (the chi-square test statistics and associated P-values) indicates that IABP and calcification are significantly associated with in-hospital mortality individually, but not the interaction between the two.

ACE, angiotension converting enzyme inhibitors; ASA, aspirin; CHF, congestive heart failure; DM, diabetes mellitus; HCL, hypercholesterolemia; HTN, hypertension; LVEF, left ventricular ejection fraction; Statins, HMGCoA reductase inhibitors.

^aincludes primary and reoperation; ^{b, c}primary, ReOp and double valves; ^dLVA repair, heart transplant, TMLR, Ascending Aorta repair, VSD repair.

CABG, coronary artery bypass graft; CPB time, cardio-pulmonary bypass time; CX time, cross-clamp time; SD, standard deviation

* statistically significant (P < 0.05)

ACE, angiotension converting enzyme inhibitors; ASA, aspirin; CHF, congestive heart failure; CX time, cross-clamp time; DM, diabetes mellitus; DTA, descending thoracic aorta; HTN, hypertension; IABP, intra-aortic balloon pump; LV dysfunction, left ventricular dysfunction; LVEF, left ventricular ejection fraction; Statins, HMGCoA reuctase inhibitors.

CHF, congestive heart failure; CI, confidence interval; DTA, descending thoracic aorta; IABP, intra-aortic balloon pump; OR, odds ratio.

CI, confidence interval; DTA, descending thoracic aorta; IABP, intra-aortic balloon pump; OR, odds ratio.