Acute Renal Failure in Coronary Artery Bypass Surgery: Independent Effect of Cardiopulmonary Bypass

Transcription

1 Acute Renal Failure in Coronary Artery Bypass Surgery: Independent Effect of Cardiopulmonary Bypass Mark I. Stallwood, FRCA, Antony D. Grayson, BS, Keith Mills, FRCA, and Nigel D. Scawn, FRCA Departments of Cardiothoracic Anaesthesia, and Research and Development, The Cardiothoracic Centre, Liverpool, United Kingdom Background. Acute renal failure after cardiac surgery is associated with a high morbidity and mortality, particularly when associated with hemodialysis. The aim of the study was to investigate whether the use of cardiopulmonary bypass increased the risk of developing acute renal failure. Methods. The 2199 consecutive patients undergoing isolated coronary artery bypass grafting between January 2000 and March 2002 were retrospectively analyzed. Patients with significant preoperative renal dysfunction (preoperative serum creatinine > 200 mol/l) were excluded. A multivariate logistic regression model was constructed to identify independent risk factors for the development of acute renal failure. Results. In the study, 53 patients (2.4%) developed acute renal failure before hospital discharge. The crude incidences of acute renal failure for isolated coronary artery bypass grafting in the on- and off- pump groups were 2.9% and 1.4%, respectively (p 0.031). There were 1483 patients who underwent on-pump surgery whereas 716 patients were in the off-pump group. The two groups were broadly comparable on many variables. The offpump group were slightly younger on average (63.6 versus 64.9 years old [p 0.017]), but had more angina class IV patients (39.5% versus 28.9% [p < 0.001]) and a greater proportion of redo surgery (4.1% versus 1.6% [p < 0.001]). The on-pump group had more patients with three-vessel disease (82.8% versus 74.3% [p < 0.001]). The logistic regression model identified use of cardiopulmonary bypass as an independent risk factor for the development of acute renal failure (odds ratio 2.64 [95% confidence intervals 1.27 to 5.45]). Other independent predictors of acute renal failure were preoperative creatinine levels, diabetes, emergency operations, increasing age, increasing body mass index, and peripheral vascular disease. Conclusions. Cardiopulmonary bypass is associated with significantly increased risk of acute renal failure following isolated coronary artery bypass surgery. (Ann Thorac Surg 2004;77:968 72) 2004 by The Society of Thoracic Surgeons Accepted for publication Sept 8, Address reprint requests to Mr Grayson, The Cardiothoracic Centre- Liverpool, Thomas Dr, Liverpool L14 3PE, UK; tony.grayson@ ctc.nhs.uk. Postoperative acute renal failure (ARF) remains a serious complication of coronary artery bypass graft (CABG) surgery and is associated with significant increases in morbidity and mortality; particularly if there is a requirement for hemodialysis [1, 2]. Depending on the population studied and definitions employed, rates of occurrence have been reported between 1% and 30%. Quoted risks of mortality similarly range between 7% and 38% [1 6]. Avoidance of cardiopulmonary bypass (CPB) by using beating heart off-pump techniques has been suggested to reduce renal damage perioperatively and result in lower rates of postoperative renal dysfunction. However, studies conducted previously in this field have provided conflicting evidence to support this hypothesis [7, 8]. With off-pump surgery forming an increasing proportion of our workload, we therefore sought to analyze rates of renal failure following CABG surgery at our own institution and the effect of CPB. Material and Methods Patient Population and Data Data were prospectively collected on a total of 2199 consecutive patients undergoing isolated CABG surgery between January 1, 2000 and March 31, 2002 at the Cardiothoracic Centre, Liverpool, United Kingdom. We excluded patients with a preoperative serum creatinine level greater than 200 mol/l or a previous history of renal dysfunction. Data collection methods and definitions have been described in detail previously [9]. The different surgical techniques used in our patients, offpump and on-pump have already been published [10]. Off-pump surgery was performed consecutively by two of the seven surgeons at our institution, which represents their entire coronary revascularization series during the study period. Any conversions to cardiopulmonary by by The Society of Thoracic Surgeons /04/$30.00 Published by Elsevier Inc doi: /j.athoracsur

2 Ann Thorac Surg STALLWOOD ET AL 2004;77: EFFECT OF CPB ON ARF Table 1. Patient and Disease Characteristics Based on Procedure Performed Off-Pump (n 716) On-Pump (n 1483) p Value Age at operation (years) 63.6 (57.4 to 69.9) 64.9 (58.8 to 70.4) Female sex (%) Body mass index (kg/m 2 ) 27.9 (25.7 to 30.7) 27.6 (25.3 to 30.4) NYHA class IV (%) Angina class IV (%) Previous MI (%) Current smoker (%) Serum creatinine ( mol/l) 100 (90 to 111) 100 (90 to 110) 0.66 Creatinine clearance (ml/min) 73.4 (57.8 to 88.7) 72.5 (59 to 87.6) 0.54 Diabetes (%) Hypercholesterolaemia (%) Hypertension (%) Peripheral vascular disease (%) Respiratory disease (%) Ejection fraction 30% (%) Three-vessel disease (%) Left main stenosis (%) Cardiogenic shock (%) Intravenous nitroglycerin (%) Intraaortic balloon pump (%) Inotrope support (%) Prior cardiac surgery (%) Emergent procedure (%) CARDIOVASCULAR Continuous variables are indicated as median with 25 th and 75 th percentiles. Categorical variables are depicted as a percentage. MI myocardial infarction; NYHA New York Heart Association functional class. pass were classified as off-pump patients on the basis of intention to treat analysis. Preoperative data were collected during the patient s admission as part of routine clinical practice on the variables listed in Table 1. Creatinine clearance was calculated using the Cockroft and Gault formula: Creatinine clearance (140 age) weight (kg)/(serum creatinine 72 [ 0.85 for women]). Outcome variables collected included ARF, in-hospital mortality, and length of postoperative hospital stay. ARF was defined as patients with a postoperative serum creatinine level greater than 200 mol/l or patients requiring dialysis therapy or hemofiltration before hospital discharge. In-hospital mortality was defined as death within the same hospital admission regardless of cause. All patients transferred from the base hospital to another hospital were followed up to confirm their status at discharge. Statistical Methods Due to the non-normality of continuous variables they are shown as median with 25 th and 75 th percentiles. Categorical variables are shown as a percentage. Comparisons were made with Wilcoxon rank-sum tests and 2 tests as appropriate. Multivariable forward stepwise logistic regression analysis was used to identify independent risk factors for ARF [11]. All preoperative variables within Table 1 were offered as potential risk factors to the logistic regression model, including the use of cardiopulmonary bypass. The C statistic (equivalent to the area under the receiver operating characteristic curve) and the Lemeshow-Hosmer goodness-of-fit statistic were calculated to assess the performance and calibration of the models, respectively [11]. A C statistic of greater than 0.7 indicates a reasonable ability to discriminate between patients who developed ARF and those who did not. For the Lemeshow-Hosmer goodness-of-fit statistic the predicted risks of individual patients were rank-ordered and divided into deciles. Within each decile of estimated risk, the number of patients with ARF predicted was accumulated against the number of observed ARF, a p value of greater than 0.05 indicates acceptable calibration of the model. In all cases a p value less than 0.05 was considered significant. All statistical analysis was performed retrospectively with SAS for Windows Version 8.2. Results Overall, of the 2199 patients in the study, 53 (2.4%) patients developed acute renal failure following CABG before discharge from hospital. Thirty-four patients (1.5%) developed ARF without requiring dialysis, compared with 19 patients (0.9%) who developed ARF that required dialysis support. Those patients classified as having ARF with a postoperative serum creatinine level greater than 200 mol/l without requiring dialysis, had

3 970 STALLWOOD ET AL Ann Thorac Surg EFFECT OF CPB ON ARF 2004;77: Table 2. Independent Risk Factors for Acute Renal Failure Coefficient SE Adjusted Odds Ratio (95% CI) p Value Preoperative serum creatinine ( mol/l) a ( ) Diabetes ( ) Emergency procedure ( ) Age at operation (years) a ( ) Use of CPB ( ) Body mass index (kg/m 2 ) a ( ) Peripheral vascular disease ( ) Intercept a For every additional unit. Calculation of predicted risk using patient data and logistic regression coefficients: calculate the odds of acute renal failure exp( [ preoperative serum creatinine (continuous value)] [ diabetes] [ emergency surgery] [ age at operation (continuous value)] [ use of CPB] [ body mass index (continuous value)] [ peripheral vascular disease]). Predicted risk of acute renal failure as a percentage [odds/(1 odds)] 100. CI confidence interval; CPB cardiopulmonary bypass; SE standard error. an average preoperative serum creatinine of 83 mol/l (minimum 32 mol/l and maximum 112 mol/l). Of the study population, 716 patients (32.6%) received off-pump surgery, whereas 1483 patients (67.4%) patients were performed on-pump. The incidence of ARF following off-pump surgery was 1.4% compared with 2.9% for on-pump patients (p 0.031). Table 1 lists patient and disease characteristics based on procedure performed. Off-pump patients were younger (p 0.017) and had less extensive coronary artery disease (p 0.001). Additionally, off-pump patients were more likely to have a higher severity of angina (p 0.001) and had undergone prior cardiac surgery (p 0.001). The results of the multivariate forward stepwise logistic regression analysis found that use of cardiopulmonary bypass was an independent risk factor for the development of postoperative ARF (odds ratio 2.64, 95% confidence intervals 1.28 to 5.45; p 0.009). Other independent predictors of ARF were increased preoperative serum creatinine levels, diabetes, emergency surgery, increasing age and body mass index, and peripheral vascular disease. These variables are summarized in Table 2, with their regression coefficient, adjusted odds ratios and p values. Preoperative serum creatinine levels were a much stronger predictor of ARF compared to creatinine clearance. The discriminatory ability of the logistic model, as measured by the C statistic, was The Lemeshow-Hosmer goodness-of-fit statistic across deciles of risk was not statistically significant (p 0.689). An example of the calculation of predicted risk for an individual patient is described in Table 2. The incidence of in-hospital mortality following isolated off-pump CABG for patients with ARF was 30.0%, compared to 0.9% for patients without ARF (p 0.001). The median postoperative length of stay in survivors following isolated off-pump CABG for patients with ARF was 12 days (25 th and 75 th percentiles 7 to 21), compared with 7 days (25 th and 75 th percentiles 6 to 8) for patients without ARF (p 0.001). The incidence of in-hospital mortality following isolated on-pump CABG for patients with ARF was 20.9%, compared with 0.9% for patients without ARF (p 0.001). The median postoperative length of stay in survivors following isolated on-pump CABG for patients with ARF was 14 days (25 th and 75 th percentiles 10 to 21), compared to 8 days (25 th and 75 th percentiles 7 to 9) for patients without ARF (p 0.001). Comment Renal failure after cardiac surgery has a significant influence on postoperative morbidity and mortality. The benchmark study performed by Chertow and colleagues [1] reported a mortality of 63% at 30 days for those patients requiring hemodialysis compared with 4.3% for those without renal dysfunction. Although the starting point of that study was more than 15 years ago, more recent work by Mangano and associates [3] identified similar mortality rates. Our study had an overall rate of renal failure of 2.4%, with 0.9% requiring renal replacement therapy; rates which are comparable to other studies of equivalent size. The resultant impact on mortality for both on-pump and off-pump patients is also comparable to that reported elsewhere [2, 3, 12, 13]. Previous work at our institution [14], and this study have concurred with the predictive factors associated with the development of renal dysfunction identified elsewhere, ie, increased preoperative serum creatinine, diabetes, nonelective surgery, and duration of CPB. Crudely, the on-pump group revealed double the rate of renal failure to that demonstrated in the off-pump group (2.9% versus 1.4%). In the past, the counterbalance to this difference in crude rates has been that the offpump surgery tended to be reserved for the more straightforward patients. However, as the sophistication and reliability of stabilization devices has improved, and surgical experience with the technique has increased, so the range of patients for whom off-pump surgery is

4 Ann Thorac Surg STALLWOOD ET AL 2004;77: EFFECT OF CPB ON ARF offered has broadened. Previous work has demonstrated the comparable safety of the technique in those with poor left ventricular function and in the elderly [15 17]. The patient characteristics of our study groups were comparable on many fronts. There were similar Parsonnet scores, similar rates of diabetes, preoperative creatinine levels, left main stem stenoses, and emergency operations. There were some differences in patient characteristics with the off-pump group being slightly younger, but had greater rates of class IV angina and a higher percentage of patients with prior cardiac surgery. The on-pump group had a higher percentage of patients with three-vessel disease. Multivariate stepwise logistic regression confirmed the use of CPB as an independent risk factor for the development of ARF, even after accounting for age. Extent of coronary disease was not a risk factor for postoperative development of ARF. Interestingly, even at normal levels of preoperative creatinine, there was risk associated with each additional micromole per liter. There is no uniting mechanism explaining renal failure associated with cardiac surgery, rather a complex interplay of a number of related factors [1 3, 5]. Previous work has used algorithms to stratify individual risk, and has described clinical variables as exerting their effects in four areas. First, there are factors relating to occult renal ischemia caused largely by arteriosclerosis and exacerbated by perioperative reduction of cardiac output, hypotension, and resultant hypoperfusion. Second, the kidneys may be damaged by exogenous nephrotoxins such as aminoglycoside antibiotics, diuretics, or radiologic contrast media. Third, endogenous nephrotoxins may be released (eg, myoglobin, free radicals, or proinflammatory cytokines such as interleukin-8, interleukin-1, and tumor necrosis factor- ). Last, there may be a background of reduced renal reserve as assessed by preoperative estimates of creatinine clearance. Baseline renal impairment has been consistently identified as a major predictor for perioperative renal failure and thus we excluded those with a preoperative creatinine level in excess of 200 mol/l in order to concentrate on those with de novo renal dysfunction following surgery. Previous work has identified several factors associated with increased risk of developing ARF, as outlined previously. The majority of these factors are patient related, and beyond our control; an exception being the duration of CPB. The advent of off-pump techniques in the 1990s has lead to investigation as to whether the avoidance of CPB altogether could be renoprotective. Use of beating heart techniques means the maintenance of pulsatile flow and no exposure to an extracorporeal circuit, with an anticipated reduction in the inflammatory cytokine response that this would entail. It also means normothermia and a decreased requirement for vasoconstrictor administration to maintain target mean arterial pressures [8]. However, off-pump techniques are more technically demanding, there has been previous concern over anastamotic quality, and the contortion of the heart for the lateral and 971 posterior vessels may cause outflow tract obstruction and low cardiac output [18]. Previous retrospective analyses have failed to demonstrate a significant renoprotective benefit from using off-pump techniques, although the studies by Gamoso [8] and Zamvar [12] had relatively small numbers of offpump patients in them. The study reported by Wallace and colleagues [13] in Ohio was of comparable size to our study and indicated rates of ARF of 12% and 6%, respectively, for on-pump and off-pump surgery. However, the on-pump group had a higher percentage of high-risk patients, such as diabetics, patients with poor left ventricular function, and reoperations [13]. In our study the differences between the risk factors of the on-pump and off-pump group appeared less marked. Randomized controlled trials in this area have looked beyond serum urea and creatinine to more sensitive biochemical markers of renal function. Several studies have demonstrated rises in microalbuminuria and retinol-binding protein (early markers of glomerular damage), and increases in urinary n-acetylglucoaminidase, an indicator of tubular damage. These have been associated with decreases in measured creatinine clearance and have been demonstrated to occur with both onpump and off-pump surgery. Some studies have reported significantly lower levels of disturbance with off-pump techniques, but this has not been a universal finding. Most of these biochemical markers approximate to baseline within 48 to 72 hours, and there were no associated reports of overt renal failure. The clinical significance of these findings remains unclear, and it is worth noting that the sample size was small in these studies [7, 19 21]. Our study reports on a relatively recent population undergoing isolated CABG surgery with a large sample size. A substantial proportion of our work is performed off-pump, and the population on which this technique is utilized is becoming similar to the population receiving CABG surgery as a whole. However, the use of retrospective data provides limitations to the conclusions drawn. Multivariate retrospective analysis is no substitute for a properly conducted randomized controlled trial, but given the relatively low incidence of renal failure (2.4%), the numbers required to give sufficient power to the study statistically would render its conduct unlikely. The low event rate is another limitation, with only 53 events recorded, which may not be sufficient for an accurate prediction [22]. Usually a ratio of 1 risk factor for every 10 events is considered acceptable. In our analysis CPB use was identified as the fifth risk factor from 53 events. Perhaps one of the most important limitations is inextricable confounding [23] caused by the off-pump procedures being performed completely by two of the seven surgeons. Even after careful application of multivariate logistic regression analysis, it remains difficult to distinguish between a surgeon and a treatment difference. However, historical results of the two surgeons when they performed on-pump CABG, with regards to postoperative ARF, were similar to their colleagues. Unfortunately, our database does not record peak postoperative creatinine if more than 200 mol/l. Using CARDIOVASCULAR

5 972 STALLWOOD ET AL Ann Thorac Surg EFFECT OF CPB ON ARF 2004;77: a threshold definition of more than 200 mol/l of postoperative serum creatinine for renal failure may imply that some patients had a trivial baseline increase in serum creatinine resulting in a classification of renal failure. However, all our patients classified as postoperative renal failure, without requiring dialysis, had a baseline serum creatinine increase of 88 mol/l or more. This is comparable with the findings of Mangano and colleagues [3], who regarded anyone with a serum creatinine increase of 62 mol/l or more over baseline as having clinically significant renal failure. Using the levels of serum creatinine as the sole marker of renal function is also a limitation. Previous studies have utilized preoperative and peak postoperative creatinine levels to calculate a change in creatinine clearance by the Cockroft-Gault equation, which was an option we could not use due to the nature of the database recording of peak postoperative level outlined previously. Although measurements of patients urinary microalbumin, retinal-binding protein, or n-acetylglucoaminidase might give a more detailed picture of renal insult, in a study of this size it is largely unrealistic. Another potential limitation is the variation between individual anesthetists techniques and preferences in our study. Use of drugs with renal effects has not been regulated in any way, and although there is little strong evidence for a renoprotective effect for drugs such as dopamine and mannitol, they may be another confounding variable for which we have not accounted. In summary, we have demonstrated that the incidence of renal failure following isolated CABG surgery is doubled by using CPB at our institution. Other independent risk factors for the development of ARF were increased preoperative creatinine levels, increasing age and body mass index, emergency surgery, diabetes, and peripheral vascular disease. We would like to acknowledge the cooperation of all the Consultant Cardiac Surgeons at the Cardiothoracic Center- Liverpool: John A. C. Chalmers, Walid C. Dihmis, Brian M. Fabri, Elaine M. Griffiths, Neeraj K. Mediratta, D. Mark Pullan, and Abbas Rashid. We would also like to thank Janet Deane, who maintains the quality and ensures completeness of data collected in our Cardiac Surgery Registry. References 1. Chertow GM, Lazarus JM, Christiansen CL, et al. Preoperative renal risk stratification. Circulation 1997;95: Conlon PJ, Stafford-Smith M, White WD, et al. Acute renal failure following cardiac surgery. Nephrol Dial Transplant 1999;14: Mangano CM, Diamondstone LS, Ramsay JG, Aggarwal A, Herskowitz A, Mangano DT. Renal dysfunction after myocardial revascularistion: risk factors, adverse outcomes, and hospital resource utilization. The Multicenter Study of Perioperative Ischaemia Research Group. Ann Intern Med 1998; 128: Andersson LG, Ekroth R, Bratteby LE, Hallhagen S, Wesslen O. Acute renal failure after coronary surgery a study of incidence and risk factors in 2009 consecutive patients. J Thorac Cardiovasc Surg 1993;41: Corwin HL, Sprague SM, DeLaria GA, Norusis MJ. Acute renal failure associated with cardiac operations: a casecontrol study. J Thorac Cardiovasc Surg 1989;98: Mangos GJ, Brown MA, Chan WY, Horton D, Trew P, Whitworth JA. Acute renal failure following cardiac surgery: incidence, outcomes and risk factors. Aust NZ J Med 1995; 25: Ascione R, Lloyd CT, Underwood MJ, Gomes WJ, Angelini GD. On-pump versus off-pump coronary revascularisation: evaluation of renal function. Ann Thorac Surg 1999;68: Gamoso MG, Phillips-Bute B, Landolfo KP, Newman MF, Stafford-Smith M. Off-pump versus on-pump coronary artery bypass surgery and post-operative renal dysfunction. Anaesth Analg 2000;91: Wynne-Jones K, Jackson M, Grotte G, Bridgewater B. On behalf of the north west regional cardiac surgery audit steering group. Limitations of the Parsonnet score for measuring risk stratified mortality in the north west of England. Heart 2000;84: Patel NC, Deodhar AP, Grayson AD, et al. Neurological outcomes in coronary surgery: independent effect of avoiding cardiopulmonary bypass. Ann Thorac Surg 2002;74: Hosmer D, Lemeshow S. Applied logistic regression. New York, NY: John Wiley & Sons Inc, Zamvar VY, Khan NU, Madhavan A, Kulatilake N, Butchart EG. Clinical outcomes in coronary artery bypass graft surgery: comparing off-pump and on-pump techniques. Heart Surg Forum 2002;5: Wallace LK, Leventhal ML, Starr NJ. Renal dysfunction following off-pump versus on-pump coronary artery bypass grafting. Anaesth Analg 2001;92:SCA Grayson AD, Khater M, Jackson M, Fox MA. Valvular heart operation is an independent risk factor for acute renal failure. Ann Thorac Surg 2003;75: Shennib H, Endo M, Benhamed O, Morin JF. Surgical revascularisation in patients with poor left ventricular function. On- or off-pump? Ann Thorac Surg 2002;74:S Al-Ruzzeh S, George S, Yacoub M, Amrani M. The clinical outcome of off-pump coronary artery bypass surgery in the elderly patients. Eur J Cardiothoracic Surg 2001;20: Hoff SJ, Ball SK, Coltharp WH, Glassford DM Jr, Lea JW 4th, Petracek MR. Coronary artery bypass in patients 80 years and older. Is off-pump the operation of choice? Ann Thorac Surg 2002;76:S Patel NC, Grayson AD, Jackson M, et al., on behalf of the North West Quality Improvement Programme in Cardiac Interventions. The effect off-pump coronary artery bypass surgery on in-hospital mortality and morbidity. Eur J Cardiothorac Surg 2002;22: Tang AT, Knott J, Nanson J, Hsu J, Haw MP, Ohri SK. A prospective randomised study to evaluate the renoprotective action of beating heart coronary surgery in low risk patients. Eur J Cardiothoracic Surg 2002;22: Loef BG, Epema AH, Navis G, Ebels T, Van Oeveren W, Henning RH. Off-pump coronary revascularization attenuates transient renal compared with on-pump coronary revascularization. Chest 2002;121: Gerritsen WB, van Boven WJ, Driessen AH, Haas FJ, Aarts LP. Off-pump versus on-pump coronary artery bypass grafting: oxidative stress and renal function. Eur J Cardiothoracic Surg 2001;20: Concato J, Feinstein AR, Holford TR. The risk of determining risk with multivariable models. Ann Intern Med 1993;118: Blackstone EH. Comparing apples and oranges. J Thorac Cardiovasc Surg 2002;123:8 15.