| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2002;74:378-383
© 2002 The Society of Thoracic Surgeons
a Department of Cardiac Surgery, Tor Vergata University of Rome, Rome, Italy
b 2nd University of Naples, Naples, Italy
Accepted for publication April 21, 2002.
* Address reprint requests to Dr Nardi, Tor Vergata University of Rome, European Hospital, Via Portuense 700, 00149 Rome, Italy
Abstract
Background. The incremental surgical risk caused by different categories of renal failure is not well defined.
Methods. Data from 159 patients with moderate to end-stage renal dysfunction, who had consecutive operations using cardiopulmonary bypass, were included in a multivariate analysis of morbidity and survival. Ninety-nine patients had preoperative serum creatinine levels (PSCL) of 1.9 to 2.5 mg/dL (moderate), 36 had PSCL higher than 2.5 mg/dL and were not dialysis dependent (severe), and 24 required chronic dialysis (end-stage dysfunction).
Results. Operative mortality was 4% with moderate dysfunction and compared favorably with 16.7% in severe and 8% in end-stage dysfunction (p < 0.05). Independent predictors of death were severe non-dialysis-dependent renal dysfunction (p < 0.05), diabetes (p < 0.05), and cardiopulmonary bypass time (p < 0.01). Severe renal dysfunction (p < 0.01) and diabetes (p < 0.01) also predicted pulmonary and neurologic morbidity. Freedom from late death at 4 years was 82% ± 5% with moderate, 49% ± 10% with severe, and 60% ± 10% with end-stage dysfunction (p < 0.01). Time to late death was adversely affected by severe (p < 0.05) and end-stage dysfunction (p < 0.01). Persistent improvement of symptoms was observed in all subgroups.
Conclusions. Satisfactory early and late surgical outcomes may be expected in patients with moderate renal failure, but outcomes are often poor with severe non-dialysis-dependent and end-stage renal dysfunction.
It is well known that end-stage renal failure adversely affects early and long-term results of cardiac surgery [1–4]. Studies in non-dialysis-dependent renal patients are not numerous, but also demonstrate the adverse influence of renal dysfunction on surgical outcomes [5–8]. Very few attempts have been made to precisely assess the influence of various degrees of renal dysfunction on surgical results. In a retrospective investigation, Durmaz and associates [7] analyzed the influence of mild to end-stage dysfunction on the early and medium-term results of cardiac surgery, and observed poor results in non-dialysis-dependent patients with preoperative creatinine levels greater than 2.5 mg/dL and in patients supported by dialysis. More recently, Weerasinghe and colleagues [8] included data from coronary patients with normal function and with mild to moderate dysfunction in a multivariate analysis model. They found that early surgical mortality and morbidity increased significantly with even mild elevation of the preoperative creatinine level, and more with moderately elevated levels. An accurate assessment of early and late surgical results seems therefore particularly useful in patients with moderate, severe, and end-stage renal disease, as it could contribute to better establish the expected survival in these high-risk patients. The purpose of our investigation was to delineate the influence of those categories of renal dysfunction on operative mortality, postoperative complications, and late survival, including in multivariable models data from patients with more than mild chronic renal dysfunction consecutively operated on with cardiopulmonary bypass at our institution.
Material and methods
One hundred fifty-nine patients with chronic renal failure and a preoperative serum creatinine level (PSCL) of or greater than 1.9 mg/dL were selected from 4,820 patients consecutively operated on with cardiopulmonary bypass from April 1994 to March 2001. They were divided into three groups, according to the severity of renal dysfunction: moderate dysfunction, 99 patients with PSCL of 1.9 to 2.5 mg/dL; severe dysfunction, 36 non-dialysis-dependent patients with PSCL greater than 2.5 mg/dL; and end-stage dysfunction, 24 patients having long-term dialysis. The preoperative and operative characteristics of patients are summarized in Tables 1 and 2. In patients with end-stage dysfunction, hemodialysis was performed approximately 24 hours before the operation and then resumed on the second postoperative day. In the other patients, administration of 3 µg/Kg/min of dopamine was started at induction of anesthesia and continued during the operation and the stay in the intensive care unit. Cardiopulmonary bypass was instituted by means of a Sorin Monolyth-Pro (Sorin Biomedica; Turin, Italy) or Capiox (Terumo Cardiovascular System; Borken, Germany) membrane oxygenator and a Stöckert roller pump (Stöckert Instrumente; Munich, Germany), after administration of approximately 300 IU/Kg of heparin. Albumin solution or bank blood was added as required, to maintain least flow rates of 2.4 L/min/m2, mean arterial pressure of 80 mm Hg, and hematocrit level of 24%. Operations were performed with antegrade cold blood cardioplegia, moderate hypothermia (26°C), and topical cooling (95 patients, 60%), or tepid blood cardioplegia and mild hypothermia (34°C) (62 patients, 39%). Deep hypothermic (20°C) circulatory arrest was performed in 2 patients with aneurysm of the aortic arch, 1 with moderate and 1 with end-stage renal dysfunction. Intraoperative continuous ultrafiltration was performed in 11 non-dialysis-dependent patients with preoperative serum potassium levels above 5.2 meq/L. Heparin was reversed by protamine sulphate. Patients routinely spent the first and second postoperative days in the intensive care unit. In particular, intravenous fluids included 3 µg/Kg/min of dopamine, 1.5 to 2 mL/Kg/h of 5% dextrose in water, colloids, and/or whole blood to optimize preload, keeping the hemoglobin level greater than 11 g/dL. In non-dialysis-dependent patients, administration of frusemide, 20 to 40 mg/h, was started if urine output decreased below 0.7 to 0.8 mL/Kg/h and then discontinued when output was sustained above 1 mL/Kg/h. A systolic arterial pressure at or above 120 mm Hg was considered adequate for optimal tissue perfusion. Mechanical ventilation was usually discontinued the morning of the first postoperative day.
|
|
Statistical analysis
Analysis was performed with the SPSS statistical software version 10.1 for Windows (SPSS Inc., Chicago, IL). Continuous data were compared by using the Student’s t test or one-way analysis of variance (ANOVA) with the Scheffe F-test for posthoc comparisons, as required; the Kruskal-Wallis and the Wilcoxon test were used when data were not normally distributed. The 
2 or Fisher’s exact tests were used for categorical data. The association of potential risk factors with operative mortality and morbidity was assessed by univariate analysis; factors with a p value less than or equal to 0.2 were included in a stepwise logistic regression analysis model. Twenty variables were selected for the univariate and multivariate analyses, including age, gender, previous myocardial infarction, smoking habit, comorbid diseases (arterial hypertension, diabetes mellitus, peripheral vascular disease, cerebrovascular disease, chronic obstructive pulmonary disease, hyperlipidemia), Canadian Cardiovascular Society (CCS) grade of angina, New York Heart Association (NYHA) functional class, left ventricular ejection fraction, type of operation (coronary, valvular, thoracic aorta), type of procedure (isolated, associated), type of cardioplegia (cold, tepid), and cardiopulmonary bypass and aortic cross-clamp times. Distribution of the potential risk factors in the categories of renal failure is shown in Tables 1 and 2. Overall survival and freedom from late death were expressed as mean values plus or minus 1 standard error and computed by using the Kaplan-Meier method; the log-rank test was used to compare survival estimates among subgroups, and the Cox proportional hazards method was used to evaluate the influence of variables on time to death. All other values in the analysis are expressed as mean plus or minus standard deviation of the mean.
Results
Minor differences were noted among renal groups; in particular, older age (p < 0.05) and higher NYHA functional class (p < 0.05) were detected in moderate renal failure (Table 1). As expected, preoperative anemia was more severe in patients having dialysis than in non-dialysis-dependent patients (p < 0.001).
Operative mortality
The operative mortality rate in the 36 non-dialysis-dependent patients with severe renal dysfunction was 16.7% (n = 6) and compared unfavorably with the 4% rate (n = 4) in 99 patients with moderate renal dysfunction (p = 0.02) and the 8.3% rate (n = 2) in 24 patients with end-stage dysfunction. Causes of death were cardiac in 4 patients (33%), mesenteric ischemia in 4 (33%), sepsis in 1 patient with mediastinitis (8%), neurologic damage in 1 (8%), gastrointestinal hemorrhage in 1 (8%), and postoperative hemorrhage due to intravascular disseminated coagulopathy in 1 (8%). Of the 2 patients operated on for arch replacement, 1 with severe renal dysfunction died due to neurologic damage, after a combined procedure. On univariate analysis, several factors were associated with hospital death with a p value less than 0.1, as shown in Table 3.
Variables were then analyzed in a multivariate model, showing that severe renal dysfunction (PSCL > 2.5 mg/dL) in non-dialysis-dependent patients (p < 0.05), diabetes mellitus (p < 0.05), and cardiopulmonary bypass time (p < 0.01) were the independent predictors of mortality. The mortality rate was 16.1% (n = 5) in 31 patients undergoing associated procedures, and 5.5% (n = 7) in 128 patients undergoing isolated procedures (p = 0.06); in the latter group, mortality was 4.9% (n = 5) in 103 coronary patients, 5.9% (n = 1) in 17 valvular patients, and 12.5% (n = 1) in 8 patients having thoracic aorta replacement (p = 0.9). Of note, end-stage renal dysfunction was not associated with an increased risk of in-hospital death.
|
|
Late survival
Thirty-two patients (20%) died late. Causes of late death were cardiac in 10 patients (31%), sudden death in 9 (28%), renal in 4 (12.5%), malignancy in 4 (12.5%), neurologic in 2 (6.3%), pulmonary in 2 (6.3%), and unknown in 1 (3.1%). Overall, 4-year survival in the three categories of renal failure was 79% ± 5% in moderate, 41% ± 12% in severe, and 55% ± 14% in end-stage renal dysfunction (p < 0.001). The 4-year freedom from late death was 82% ± 5%, 49% ± 10%, and 60% ± 10%, respectively (p < 0.01) (Fig 1).
Coronary bypass patients were the largest group in this series; therefore, survival curves of these patients were computed separately and confirmed the high risk of late death in the categories of severe and end-stage renal dysfunction (Fig 2).
On multivariate analysis, preoperative severe renal dysfunction (odds ratio 2.1, 95% CI 1.2 to 2.7, p = 0.014) and end-stage dysfunction (odds ratio 4.1, CI 1.6 to 10.4, p = 0.003) affected the time to late death.
|
|
In the present study, we investigated the influence of various degrees of renal dysfunction on early and late surgical outcomes. Investigations on results of cardiac surgery in different categories of renal patients have not been numerous (Table 5). Weerasinghe and colleagues [8] reported a hospital mortality rate of 7.6% in 66 coronary patients with mild dysfunction, increasing to 18.5% in 54 patients with more severe non-dialysis-dependent dysfunction. Durmaz and associates [7] observed hospital mortality rates of 11.8% in 93 cardiac patients with mild to moderate renal failure, 12.5% in 8 patients on chronic dialysis, and 33% in 18 non-dialysis-dependent patients with creatinine levels greater than 2.5 (3.3 ± 1) mg/dL; the rate of complications was 21%, 75%, and 44% respectively. Both studies performed a multivariable analysis of several potential risk factors, including the severity of renal dysfunction, and showed a significant incremental risk of death caused by increasing preoperative serum creatinine levels. Our investigation included patients with at least moderate renal dysfunction; nevertheless, as already observed by Durmaz and coauthors [7], it found that the hospital mortality rate was higher in non-dialysis-dependent patients with severe renal failure (16.7%, PSCL of 3.5 ± 1.3 mg/dL) than in patients with moderate dysfunction (4%, PSCL of 2.1 ± 0.2 mg/dL) or those having chronic dialysis (8%). Multivariate analysis confirmed that severe non-dialysis-dependent renal dysfunction was an independent predictor of adverse surgical outcomes. It is well known that chronic renal failure affects the function of every organ system and that several abnormalities may be present in asymptomatic patients without reduced urine output but advanced impairment of kidney function. In that category, the surgical stress may compromise systems’ function still further, leading to the high hospital mortality. It seems, therefore, that additional perioperative measures are urgently required in that category of renal failure during the perioperative period; for example, the benefit of early postoperative diafiltration is currently under investigation at our institution. On the other hand, alterations in the composition of intracellular and extracellular fluids, effects of retained toxic metabolites, and many other abnormalities are presumably less critical in patients with less elevated creatinine levels (< 2.5 mg/100 mL in our series), or are efficiently reversed in patients having hemodialysis. These considerations might well explain the relatively low surgical risk we observed in the latter categories of renal failure.
|
Analyzing the influence of various degree of renal dysfunction on late mortality, Durmaz and associates [7] found that the mortality rate at 34.8 months of follow-up was lower in 62 patients with mild and moderate disease (6.1%) than in 7 non-dialysis-dependent patients with severe renal failure (25%) and in 12 patients having chronic dialysis (14.3%) (Table 5). The sample sizes were, however, quite small. In the present series of patients with at least moderate renal dysfunction, late mortality rates were 15.8%, 33.3%, and 31.8% in moderate, severe, and end-stage renal failure, respectively (Table 5). On multivariate analysis, both severe and end-stage renal dysfunction were predictors of late death in our series. Of note, we found a high late mortality in end-stage renal dysfunction, despite the relatively low early mortality observed in that morbid condition.
In conclusion, an incremental risk of death is caused by different degrees of renal failure in patients with more than mild renal dysfunction, (PSCL higher than 1.9 mg/dL). Surgical outcomes seem satisfactory in patients with moderate renal dysfunction (creatinine level of 1.9 to 2.5 mg/dL in our series). Postoperative results may be poor in more severe renal failure, particularly in non-dialysis-dependent patients with a preoperative creatinine level higher than 2.5 mg/dL and other associated risk factors, such as diabetes mellitus or complex procedures requiring a prolonged cardiopulmonary bypass time.
References
This article has been cited by other articles:
|
|
 |
|
  J. H. Khan, E. A. Davis, L. S. Dean, M. J. Huff, N. Y. Khan, and A. Rehman The Role of Elective Perioperative Dialysis in Nondialysis Renal Failure Patients Ann. Thorac. Surg., April 1, 2009; 87(4): 1085 - 1089. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Filsoufi, J. Chikwe, J. G. Castillo, P. B. Rahmanian, J. Vassalotti, and D. H. Adams Prosthesis type has minimal impact on survival after valve surgery in patients with moderate to end-stage renal failure Nephrol. Dial. Transplant., November 1, 2008; 23(11): 3613 - 3621. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 D. Ledoux, M. Monchi, J.-P. Chapelle, and P. Damas Cystatin C blood level as a risk factor for death after heart surgery Eur. Heart J., August 1, 2007; 28(15): 1848 - 1853. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 M. P. Devbhandari, A. J. Duncan, A. D. Grayson, B. M. Fabri, D. J.M. Keenan, B. Bridgewater, M. T. Jones, J. Au, and On behalf of the North West Quality Improvement Pr Effect of risk-adjusted, non-dialysis-dependent renal dysfunction on mortality and morbidity following coronary artery bypass surgery: a multi-centre study. Eur. J. Cardiothorac. Surg., June 1, 2006; 29(6): 964 - 970. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
V. Chan, W.R. E. Jamieson, A. G. Fleisher, D. Denmark, F. Chan, and E. Germann Valve Replacement Surgery in End-Stage Renal Failure: Mechanical Prostheses Versus Bioprostheses Ann. Thorac. Surg., March 1, 2006; 81(3): 857 - 862. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. Witczak, A. Hartmann, and J. L. Svennevig Multiple Risk Assessment of Cardiovascular Surgery in Chronic Renal Failure Patients Ann. Thorac. Surg., April 1, 2005; 79(4): 1297 - 1302. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
P. E. Antunes, D. Prieto, J. F. de Oliveira, and M. J. Antunes Renal dysfunction after myocardial revascularization Eur. J. Cardiothorac. Surg., April 1, 2004; 25(4): 597 - 604. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
