Background

Acute kidney insufficiency (AKI) post isolated coronary artery bypass grafting surgery (iCABG) is a serious complication associated with increased morbidity and mortality [1,2,3,4,5,6]. Despite the advancements achieved in cardiac surgery and postoperative care, the incidence of AKI is increasing [6]. The reported incidence of AKI after cardiac surgery ranged between 2.4 and 40%; this variability is attributed to different AKI definitions and patients’ characteristics among studies [7,8,9,10]. AKI increased hospital mortality by 10–30%, and it can be as high as 40–60% if continuous renal replacement therapy (CRRT) was required [11, 12].

The aim of this study was to identify the potential risk factors of postoperative AKI in patients undergoing iCABG in order to implement measures to reduce its incidence in high-risk patients.

Methods

Study design and data collection

This is a retrospective cohort study performed at the Cardiothoracic Surgery Unit - Liverpool Hospital, Sydney, New South Wales, Australia. Data were retrieved from the prospectively maintained database collected at the time of surgery using the questionnaire of the Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) database. An “opt- out” consent and information sheet were utilized. This study was approved by the Sydney South West Area Health Service Human Research Ethics Committee (Western Zone, HREC reference number: HE17/015)

The definition of new AKI and grouping

The last serum creatinine prior to cardiac surgery and the highest postoperative serum creatinine prior to discharge were used to define the stage of AKI. AKI stages were defined by Acute Kidney Injury Network (AKIN) definitions as follows: stage 1: 50% increase in serum creatinine from baseline, stage 2: a 2-fold increase in serum creatinine, and stage 3: a 3-fold increase in serum creatinine or a new dialysis-dependent renal failure.

Because of the small proportion of patients with new AKI, we merged stages 1, 2, and 3 in one group (AKI group), and patients with normal postoperative serum creatinine were designated as No AKI group.

Patients

All iCABG patients operated between January 2008 to December 2015 were included in the study (n = 1914). Patients with new AKI (n = 101) were compared to patients with No AKI (n = 1813). Patients who had off-pump CABG or concomitant valve surgery with CABG were excluded to create homogenous cohort of patients. Factors investigated included age; gender; comorbidities; perioperative baseline data including the clinical status; operative data including cardiopulmonary bypass, aortic cross clamp times; and need for intra-aortic balloon pump (IABP); and postoperative data which could be associated with the development of postoperative AKI.

Statistical analysis

Continuous variables were presented as mean ± standard deviation (SD) and categorical data as frequencies and percent. A chi-square test was used to investigate the association between the presence of AKI and the categorical risk factors. Student’s t test and the Wilcoxon rank-sum test were used to compare continuous variables between AKI and No AKI groups. For the dichotomized outcome variable (AKI presence/absence), multivariable logistic regression analysis was implemented to investigate the independent predictors of postoperative AKI. All analyses were performed using SAS statistical software (SAS v9.4; SAS Institute, Cary, North Carolina, USA), and p value < 0.05 was considered statistically significant.

Results

Patients’ characteristics and operative data

Incidence of AKI post iCABG was 5.3% (n = 101/1914 patients) during the study period. Comparison of the patients’ demographic and perioperative characteristics between patients with and without postoperative AKI is shown in Table 1.

Table 1 Comparisons of demographic and clinical variables in patients with and without postoperative AKI in isolated CABG cohort
Full size table

Patients who had postoperative AKI were significantly older (p < 0.0001), smoker (p = 0.04), hypertensives (p = 0.0008), and diabetics (p < 0.0001) and had more peripheral vascular disease (p = 0.0003), had recent MI (p < 0.0001), had congestive heart failure symptoms (NYHA III & IV) (p < 0.0001), had severe left ventricular dysfunction (p < 0.0001), and had cardiogenic shock (p < 0.0001).

Postoperative data

Postoperatively, the patients in the AKI group had more cardiac support with IABP (p < 0.0001), inotropic support, blood transfusion (p < 0.0001), and more frequent surgical re-exploration (p < 0.0001) (Table 2).

Table 2 Comparison of postoperative variables between the AKI group and No AKI group
Full size table

Predictors of AKI

A multivariable logistic regression analysis model was constructed using variables presented in Table 3, and it identified the following risk factors as predictors of postoperative AKI after iCABG (Table 3): age above 80 years (OR 4.27; 95% CI 1.6–11.81), diabetes mellitus (OR 1.96; 95% CI 1.073–3.65), peripheral vascular disease (OR 2.55; 95% CI 1.16–5.59), severe congestive cardiac symptoms (NYHA III & IV) (OR 1.9; 95% CI 1.26–2.92), surgical re-exploration (OR 2.49; 95% CI 1.07–5.76), postoperative red blood cell (RBC) transfusion (OR 4.93; 95% CI 2.25–10.80), and postoperative low cardiac output syndrome (OR 2.85; 95% CI 1.49–5.46).

Table 3 Multivariable analysis of predictors of postoperative acute kidney insufficiency (AKI) in isolated coronary artery bypass grafting
Full size table

Discussion

In this study, the incidence of AKI was 5.2% which is higher than the expected for the isolated CABG but lies within the reported range (2.4–40%) of AKI after cardiac surgery [1, 5, 6, 8, 10, 13,14,15,16]. The higher incidence of AKI in our study could be explained by merging patients with different AKIN stages into one stage. The variation in the incidence of AKI in literature is likely due to two reasons: a lack of universal definition for postoperative AKI and difference in patient population among studies. Warren and colleagues defined AKI as the rise in creatinine 25% from baseline or an absolute increase of > 0.5 mg/dL [6]. Other studies used different equations to calculate the effective glomerular filtration rate (EGFR) and then defined the AKI according to the change in the pre- and postoperative EGFR [10]. AKI definition in other studies was based on the calculated creatinine clearance [8]. In a large cohort, Chertow and associates defined AKI as the deterioration of the renal function sufficient to require dialysis within 30 days after surgery [13]. The difference in the demographics and operative variations of the studied populations affected the occurrence of postoperative AKI; previous studies included the results of valvular and combined cardiac surgeries which are more complex than iCABG [5, 7,8,9,10,11,12,13, 17]. Despite the advances in cardiac surgery and postoperative care in the last decades, the incidence of postoperative AKI is increasing which may be explained by the increased number of high-risk patients undergoing iCABG in recent decades making them vulnerable to postoperative AKI [1, 6].

The pathogenesis of AKI after iCABG is still not fully understood and likely multifactorial. In general, it requires combinations of risk factors as well as several intraoperative or postoperative insults. The independent risk factors for AKI identified in this cohort were age above 80 years, diabetes mellitus, peripheral vascular disease, and the presence of congestive heart failure symptoms. Similar findings were reported in our previous study and other studies [18,19,20]. Unsurprisingly, the combination of elderly, diabetics, and patients with vascular disease predisposed patient to AKI with minimal insult. These are non-modifiable multi-systemic risk factors which affected both heart, kidney and neurovascular systems. In our institute, we did not use any preoperative risk stratification algorithms [9, 21, 22] to predict and identify high-risk patients who were vulnerable to have AKI and could benefit from peri- or intraoperative renal protective strategies. Huen and their colleagues concluded in their reviews that these scoring systems require more validation across huge cohorts in different centers before being adopted [9].

Operative factor predicting AKI in this cohort was emergency surgery. Patients who suffer from inadequate renal perfusion due to acute coronary syndrome and cardiogenic shock are often associated with postoperative low cardiac output syndrome which causes renal hypoperfusion especially to vulnerable kidneys. Additionally, operating urgently on a patient just loaded by coronary angiogram contrast dye adds more insult to those vulnerable kidneys. In recent series [4, 23, 24], prolonged cardiopulmonary bypass was associated with postoperative AKI, a finding which was not confirmed in our study as the univariable analysis revealed no difference in CPB and ischemic time between AKI and No AKI groups.

We found that surgical re-exploration due to postoperative bleeding or cardiac tamponade were the most significant postoperative predictors of AKI post iCABG; this result was also reported by others [10, 25]. We assumed that postoperative surgical re-exploration exacerbated other risk factors that add more insult to the renal perfusion as hemodynamic instability, anemia, blood transfusion, and the use of vasopressors. Postoperative bleeding may also initiate multiple episodes of renal hypoperfusion due to either hypotension or cardiac tamponade. Loor and Ranucci and their colleagues found anemia to be independent predictor of postoperative AKI [26, 27]. Anemia decreases renal oxygen delivery, impairs hemostasis and enhances oxidative stress contributing to the risk of renal insult. Anemia and blood transfusion had a synergistic effect on postoperative AKI, as there is an increasing risk of postoperative AKI in transfused anemic patient compared to non-transfuse anemic patient which was attributed to the increased risk of post-transfusion oxidative injury in anemic patients with abnormal iron metabolism [28, 29].

This study found that the transfusion of RBCs postoperatively was an independent predictor of AKI. We believed that blood transfusion was a consequence of postoperative bleeding and many studies reported that stored RBC transfusions contributed to organ dysfunction in borderline patients. Stored RBCs have no ability to generate nitric osxide, increasing its adhesive power to the vascular endothelium which may result in impairment of tissue oxygen delivery, increasing oxygen oxidative stress and coagulation cascade which may contribute to the establishment of postoperative AKI [28,29,30].

Study limitations

This study was a retrospective and observational study, and therefore, the conclusion drawn should be considered according to these constraints. Because of the inclusion of on-pump isolated CABG only, our results cannot be generalized to other populations; additionally, the small number of events compared to the variables included in the logistic regression model added more limitation to this study. The effect of unknown confounders on the observed associations between the risk factors and AKI, such as contrast used in coronary angiogram and CT scan, perioperative and postoperative use of nephrotoxic medication, hydration state or perioperative anemia, intraoperative hypotension, the effect of CPB, and intraoperative hematocrit level, cannot be ruled out. It is difficult to implicate or conclude a temporal relationship between AKI and iCABG and whether all AKI events in iCABG were due to correlation and not a causation

Conclusion

This study identified the predictors of acute kidney insufficiency after isolated CABG. Older patients with non-modifiable comorbidities such as diabetes mellitus, peripheral vascular disease, and symptomatic heart failure could be identified preoperatively and should have accurate risk-based assessments to give the patients and their families accurate risk prediction; moreover, measures can be put in place by the heart team to decrease the risk of postoperative AKI. Postoperative predictors can be mitigated by preoperative optimization of hemoglobin levels, coagulation profiles, and proper surgical hemostasis.