The reported incidence rate of renal dysfunction after cardiopulmonary bypass (CPB) ranges from 3 to 35% and is considered multifactorial [1–4]. Postoperative oligoanuric renal failure has a mortality rate of 60–65% [5,6]. The contributing factors related to CPB are hypotension, non-pulsatile blood flow, renal hypoperfusion, hypothermia, haemolysis, alterations of the acid-base balance and the hormonal response caused by CPB ; postoperative reperfusion syndrome may increase these effects . Furthermore, these patients are frequently submitted to radiological contrast media or other nephrotoxic drugs before surgery and the influence of these factors in postoperative renal dysfunction has been described . Changes in plasma creatinine and creatinine clearance concentrations are widely used to define changes in renal function associated with CPB [8–11]. However, these measurements are of limited value in an acute clinical situation such as CPB because they measure already established changes.
Renal physicians use the clearances of protein, of diverse molecular weights, to assist in the establishment of possible sites of renal injury [12–15]. Increases in the excretion of albumin and immunoglobulin (IgG) indicate glomerular changes, while increased excretion of proteins of low molecular weight indicates tubular changes. We could not find any study of renal function in cardiac surgery during CPB measured by specific proteins after 1999 in the literature indexed by PubMed or MEDLINE. Although there are three previous published studies on small series of patients where CPB was correlated with low creatinine clearance  and high excretion of β-glucosaminidase (β-NAG)  and α1-microglobulin , no published study has yet defined a pattern of renal function during CPB using all the specific proteins.
The aim here was to characterize the pattern of changes following myocardial revascularization surgery with CPB using measures of plasma creatinine, creatinine clearance and of fractional excretion of sodium, albumin, IgG, α1-microglobulin and β-NAG excretion.
The study was approved by the Institutional Research and Ethics Committee of our hospital and informed written consent was obtained from all participants. Twenty consecutive patients (mean ± SD, range: 61.5 ± 9.7, 41–76yr) undergoing elective myocardial revascularization surgery and whose preoperative variables of renal function – plasma creatinine, creatinine clearance, fractional excretion of sodium, α1-microglobulin and β-NAG excretion – were within normal values and were included in the study.
Preoperative exclusion criteria were treatment with oral anticoagulant drugs within the 10 days before surgery, a left ventricular ejection fraction <50%, cardiac catheterization within the 24 h before surgery, liver transaminase concentrations more than twice standard values and treatment with loop diuretics. Postoperative exclusion criteria were prospectively agreed. These were postoperative sepsis, haemodynamic instability or the need for α-adrenergic agonists (i.e. norepinephrine, epinephrine or phenylephrine). Postoperative sepsis was defined as fever or leukocytosis with positive microbiological cultures. Haemodynamic instability was considered as present when the cardiac index (CI) was <2 L min−1 m−2 and/or the mean arterial pressure (MAP) was <40 mmHg when continuously maintained for >30 min.
Anaesthetic, CPB protocols and criteria for blood transfusion were the same in all patients. Anaesthesia was induced with propofol 1 mg kg−1, midazolam 0.1 mg kg−1, fentanyl 5 μg kg−1 (with atracurium 1 mg kg−1 for muscle relaxation) and maintained with propofol 3–5 mg kg−1 h−1, atracurium 0.5 mg kg−1 h−1 and fentanyl up to 20 μg kg−1. Radial artery and pulmonary artery catheters were inserted. If inotropic support was required, dobutamine 5–10μg kg−1 min−1 with nitroglycerin, if indicated, was administered. Neither furosemide nor any other diuretic was administered to the patients. Crystalloids 5mLkg−1 h−1 and gelatin polysuccinate (Gelafundina® (Braun Medical, Barcelona, Spain), gelatin: 4g, 100mL, 274mOsmL−1) at 1.5 mL kg−1 h−1 were given intravenously, and the rate increased according to the haemodynamic response. The maximum allowed amount of Gelafundina® was 500 mL in the perioperative period.
CPB was instituted using a Bard William Harvey® membrane oxygenator (Bard de España SA, Barcelona, Spain) primed with Viaflex® (Baxter, Valencia, Spain) (isotonic crystalloid, pH 7.4, 294 mOsm L−1), albumin 20% 50 mL, and mannitol 20% 0.5 g kg−1 for all patients. The flow was non-pulsatile at >1.9Lmin−1 m−2. Blood cardioplegia solution was used; haemodilution was calculated according to the estimated blood volume and the haematocrit; and moderate hypothermia was used (26–35°C). Blood transfusion criteria during CPB were haemoglobin <7gdL−1 and/or haematocrit <20%, and post-CPB haemoglobin <9gdL−1 and haematocrit <30%.
The following variables were recorded: age, gender, body surface area, concomitant diseases (diabetes mellitus, arterial hypertension), preoperative treatment (aspirin, calcium antagonists and angiotensinconverting enzyme inhibitors, duration of total CPB and aortic clamping, intra- and postoperative bleeding and transfusion, and length of stay in the intensive care unit (ICU) and in the hospital overall. Criteria for ICU discharge were haemodynamic and respiratory stability, although no patient was discharged during the night. When samples of blood and urine were taken, mean arterial pressure (MAP), central venous pressure (CVP), mean pulmonary artery pressure (MPAP), pulmonary artery occlusion pressure (PAOP), cardiac output (CO), cardiac index (CI), systemic vascular resistance (SVR) and pulmonary vascular resistance (PVR) were recorded. Post-CPB low cardiac output was defined as a CI < 1.6Lmin−1 m−2 with MAP < 60 mmHg in the immediate postoperative period. During the operation arterial pH, PaCO2 and PaO2 together with blood, rectal and oesophageal temperatures were also recorded.
The following renal variables were measured during and after operation: urine output (normal range 0.4–1.6 mL min−1); plasma creatinine (Crp) (5–13 mg L−1, 44.2–115 μmoL−1); creatinine clearance (by Cobas Mira S® analyser (Roche Diagnostics, Mannheim, Germany), 60–120 mL min−1); fractional excretion of sodium (NaFE) by an Eppendorf EIX 5056® photometer (Eppendorf, Hamburg, Germany) (0.75–2.5 mL 100 mL−1) ; urinary IgG (<4.6mgg−1 creatinine), albumin (5–20 mg g−1 creatinine), α1-microglobulin (<10 mg g−1 creatinine) and β-glucosaminidase (0–3.5 Ug−1 creatinine) were determined by a Behring® immunonephelometer (Dade Behring, Marburg, Germany), using:
where Nau is urine sodium, Cru is urine creatinine, Nap is plasma sodium, Crp is plasma creatinine;
where Protu is the IgG, albumin and α1-microglobulin in urine (mg g−1 creatinine) and β-NAG is the β-NAG in urine (U g−1 creatinine).
Measurements were made preoperatively, immediately before induction and before CPB, after the start of CPB and immediately after cessation of CPB and then at 1, 24, 72 h, 7 and 40 days postoperatively, this last recording being carried out during the postoperative check-up. Pre- and postoperative urine output was measured at 24 h intervals taking the total volume of urine output in order to calculate correctly the excretion of the different proteins, and venous blood samples were withdrawn at the end of each period. The intraoperative period was divided in pre-, intra- and post-CPB; the duration and urine output of each one were recorded and the corresponding venous blood samples for analysis were withdrawn approximately at the middle of each period.
All variables were expressed as mean ± SD or percentages. Statistical analysis for the changes of haemodynamic and urine variables with respect to baseline values was performed by a Wilcoxon's signed rank sum test. Pearson's correlation coefficient was used to correlate the renal and haemodynamic variables during the study and these variables with age, gender, concomitant hypertension and diabetes, preoperative treatment, duration of CPB and aortic clamping, postoperative bleeding and transfusion, and the length of stay in the ICU and hospital postoperation. Statistical significance was established at P < 0.05 for Wilcoxon's signed rank sum test and P < 0.01 (bilateral) for Pearson's correlation test.
All 20 patients completed the whole study and were followed until the 40th postoperative day. Pre-, intraand postoperative variables are shown in Tables 1 and 2. Two patients had uncomplicated preoperative diabetes mellitus treated with oral hypoglycaemic drugs and eight patients had controlled hypertension.
The haemodynamic variables recorded during the study are shown in Table 3. A significant increase in CVP and PCP and a significant fall in MAP and SVR in the immediate post-CPB period was observed, followed by a significant rise in cardiac index at 1 h after operation (P < 0.05), although all values were within the clinical normal range for this type of surgery. Dobutamine <7 μg kg−1 min−1 was administered to 14 patients after CPB and during the immediate postoperative period. Only two patients presented a period of haemodynamic instability requiring dobutamine at a rate >10μgkg−1 min−1 associated with dopamine at rates <5 μgkg−1 min−1 for >2h.
The renal variables recorded are represented in Figures 1 and 2. Although all renal variables showed an increase after the induction of anaesthesia, at the pre-CPB period they lacked statistical significance and values remained within the normal range. While plasma creatinine concentrations remained <115 μmol L−1 in all patients during the study, a significant increase in creatinine clearance during CPB was observed. The excretion of high molecular weight proteins IgG and albumin and fractional excretion of sodium were significantly increased during CPB, then showing a trend to decrease after CPB but still remaining high at 24 h postoperation (P < 0.05). The significant rise in the excretion of α1-microglobulin and in the excretion of the enzyme of renal origin, β-NAG observed from the CPB period, was more marked after CPB and in the postoperative period and was maintained until the seventh day postoperation (P < 0.05). On the 40th postoperative day, the excretion of albumin, IgG, β-NAG and α1 -microglobulin persisted for more than twice the preoperative values, but only in the case of β-NAG the elevation was statistically significant (P < 0.05).
During the first postoperative hour, the increase in α1 -microglobulin concentration was correlated with the increase in the excretion of β-NAG (P = 0.003, Pearson's correlation coefficient 0.637). On the seventh postoperative day, the increase in the excretion of albumin was correlated with the increase in the excretion of α1-microglobulin (P = 0.003, Pearson's correlation coefficient 0.708). On the 40th postoperative day, the increase in the excretion of IgG correlated with the increase in the excretion of α1 -microglobulin and β-NAG (P = 0.001, Pearson's correlation coefficient 0.955 and P = 0.003, Pearson's correlation coefficient 0.927, respectively). In addition, the elevation in the excretion of α1-microglobulin on the seventh postoperative day was significantly correlated with a low cardiac output post-CPB (P < 0.001, Pearson's correlation coefficient 0.856) and length of ICU stay (P < 0.001, Pearson's correlation coefficient 0.843). The elevation in the excretions of α1 -microglobulin and β-NAG on the 40th postoperative day were significantly correlated with preoperative diabetes mellitus (only two patients in this study, P < 0.001, Pearson's correlation coefficients 0.975 and 0.996, respectively). On the 40th postoperative day, the two diabetic patients presented values for β-NAG of 69 and 7 U g−1, respectively, and for α1-microglobulin of 60 and 19 mg g−1, respectively. The two patients that suffered from temporary haemodynamic instability after CPB had high concentrations of α1-microglobulin on the seventh postoperative day, 56 and 39 mg g−1, respectively.
We used specific measurements of glomerular and tubular function to assess renal function in cardiac surgery with CPB [12–15]. Although the patient sample studied was small, the homogeneity of the type of surgery, anaesthetic and CPB protocols and, above all, the normal preoperative renal function of the patients indicate that the renal dysfunction found here would characterize the pathophysiological effect of CPB. Patients with preoperative arterial hypertension and non-insulin-dependent diabetes – two conditions frequently associated with alterations in renal function – were not excluded owing to the difficulty in finding coronary surgical patients without these diseases, and in all cases the disease was uncomplicated and well controlled. All patients were included after a thorough preoperative assessment of renal function showing that renal variables were within normal limits.
Most studies on cardiac surgical patients have defined renal function during CPB according to the evolution of routine laboratory determinations: plasma creatinine, blood, and urinary nitrogen and creatinine clearance [8–11], but to assess the degree of renal damage, more specific indicators are necessary as the urine proteins are of different molecular weights. Only a few studies have used the more specific α1-microglobulin and β-NAG excretion to assess renal function during cardiac surgery with CPB [16,17]. We could find only one study where renal function, with or without CPB, was assessed for 48 h after operation during which creatinine clearance and albumin and β-NAG excretions were measured . The authors suggested that cardiac surgery without CPB caused less renal injury than surgery with CPB . Although indications of myocardial revascularization without CPB are becoming more common in many hospitals, all procedures were performed with CPB in our institution during the time in which our present study was conducted. Cardiac surgery without CPB with optimal renal perfusion, when indicated, could avoid renal dysfunction associated with CPB, but indications for procedures with CPB persist in many patients. The renal effects of moderate hypothermia, as used in this study, seem to lack clinical significance .
During CPB and after reperfusion, renal damage is usually preceded by a period of deficient renal perfusion whose effect is mainly at the tubular level . The pattern and amount of urine protein excretion may differentiate between glomerular and interstitial tubular diseases [19–22]. Under normal conditions, proteins should be almost absent from the urine; some 99% of protein filtered is reabsorbed in the proximal convoluted tubule. In glomerular disease, the glomerular filtration of proteins increases, as does their reabsorption at the proximal convoluted tubule. When this compensatory mechanism is saturated, proteins of high and low molecular weight appear in the urine. However, in interstitial tubular disease, glomerular protein filtration remains normal but the capacity of the proximal convoluted tubule for both protein reabsorption and protein metabolism by cellular lysosomes is reduced. This results in an increase in the urine of both low molecular weight plasma proteins and enzymes that denote damage of tubular origin. Thus, to define a pattern of renal function behaviour in myocardial revascularization surgery with CPB, we included measurements of different proteins in the urine to localize the level of renal dysfunction and damage.
The excretion of IgG indicates a structural alteration of the glomerular basal membrane, and albumin excretion indicates alteration in the surface pressure of the glomerular basal membrane and is commonly used in nephrology to follow-up glomerular nephropathies [23–25]. The urinary excretion of α1-microglobulin protein increases due to alterations at either the tubular or glomerular level when damage is extensive . β-NAG, a urinary enzyme of tubular origin, is an early and highly sensitive indicator of tubular cellular damage.
In the present study, a significant increase in creatinine clearance during CPB was observed while plasma creatinine concentration was normal. We have attributed this increase to haemodilution and the infusion of mannitol during the CPB period. There was a fivefold increase in urinary albumin excretion after the induction of anaesthesia and we could not account for this. However, the changes lacked statistical significance or clinical relevance since they remained within the range of normal values. We observed a pattern of renal dysfunction during and after CPB, reflected by increased excretion of IgG and albumin that persisted for the first 24 h after operation, thus denoting glomerular dysfunction. These changes were associated with a significant increase in the fractional excretion of sodium lasting 24 h.
In contrast, the increase in the excretion of proteins of low molecular weight, such as α1-microglobulin and β-NAG, remained at high levels at the seventh and 40th postoperative days, indicating a prolonged tubular alteration after CPB. An explanation might be that the proximal convoluted tubule is the portion of the nephron most sensitive to ischaemia . Prolonged tubular alteration after CPB could explain the decrease of the renal functional reserve, which has been attributed to this type of surgery up to 6 months after operation . The postoperative follow-up – standardized at the 40th day postoperation by our surgeons – allowed us to follow the patients until that day, although it was necessary to contact patients on the previous days to make sure that they had remembered the instructions given to them before hospital discharge to collect and bring their urine samples. These difficulties may explain that the studies on renal function after cardiac surgery with CPB in patients with normal function include observations made only during the postoperative period within hospital.
Our study showed that CPB in patients with normal preoperative renal function produced temporary glomerular changes and more persistent changes in the tubular portion of the nephron. We consider that these changes are the usual response of renal function to CPB in the absence of preoperative renal dysfunction or intraoperative renal insults other than CPB and probably lack long-term clinical significance. However, factors such as preoperative renal failure or dysfunction, administration of contrast media in the days before CPB, haemodynamic instability during or after CPB, and postoperative haemodynamic, metabolic or respiratory instability or sepsis may increase the likelihood of renal dysfunction that was observed in our study, thus explaining the high incidence of renal failure associated with this type of surgery [1,2,4].
Although correlation of postoperative tubular dysfunction with preoperative diabetes mellitus and low cardiac index was statistically significant in this study, no conclusions can be made since the patient sample was very small and only two patients with each condition were found. Nevertheless, it is interesting that renal dysfunction was not correlated with age or preoperative hypertension in this study. Other studies including large series of patients [27–31] have found a correlation between increases in creatinine plasma concentration and advanced age, diabetes mellitus, low cardiac output, prolonged times of CPB and aortic clamping, and profuse bleeding. Our results indicate that the plasma creatinine concentration, or its clearance, lacks the sensitivity to assess the renal changes provoked by CPB. Moreover, where a risk of renal damage is suspected or a patient has frank renal malfunction, the determination of the concentrations of urine specific proteins or enzymes will allow earlier detection of renal dysfunction, thus favouring closer control and application of measures to prevent further deterioration.
In conclusion, the use of CPB in myocardial revascularization surgery in patients with normal preoperative renal function is associated with a renal function pattern characterized by glomerular dysfunction during the first 24 h after operation and prolonged tubular dysfunction lasting until at least the 40th postoperative day.
The authors are grateful to the nurses of the Cardiovascular Diseases Institute and Biochemistry Laboratory for their co-operation in collecting and processing the samples, and to the patients for their contribution to medical progress.
1. Gailiunas P Jr, Chawla R, Lazarus JM, Cohn L, Sanders J, Merrill JP. Acute renal failure following cardiac operations. J Thorac Cardiovasc Surg
1980; 79: 241-243.
2. Hilberman L, Derby GC, Spencer RJ, Stinson EB. Sequential pathophysiological changes characterizing the progression from renal dysfunction to acute failure following cardiac operation. J Thorac Cardiovasc Surg
1980; 79: 838-844.
3. Hanna EL, Kilburg H, O'Donel JF, Lukacit G, Shields EP. Adult open heart hospital mortality rates. JAMA
1990; 264: 2768-2774.
4. Mazzarella V, Gallucci T, Tozzo C, et al.
Renal function in patients undergoing cardiopulmonary bypass
operations. J Thorac Cardiovasc Surg
1992; 104: 1625-1627.
5. Hilberman M, Myers BD, Carrie BJ, Derby G, Jamilsosn RL, Stinson EB. Acute renal failure following cardiac surgery. J Thorac Cardiovasc Surg
1979; 77: 880-888.
6. Hou SH, Bushinsky DA, Wish JB, Cohen JJ, Harrington JT. Hospital-acquired renal insufficiency. A prospective study. Am J Med
1983; 74: 243-248.
7. Rao V, Ivanof J, Welser RD, Christakis GT, David TE. Predictors of low cardiac output syndrome after coronary artery bypass
. J Thorac Cardiovasc Surg
1996; 112: 38-51.
8. Badner NH, Murkin JM, Lok P. Differences in pH management and pulsatile/non-pulsatile perfusion during cardiopulmonary bypass
do not influence renal function. Anesth Analg
1992; 75: 606-701.
9. Louagie YA, Gonzalez M, Collard E, et al. Does flow character of cardiopulmonary bypass
make a difference? J Thorac Cardiovasc Surg
1992; 104: 1628-1638.
10. Urzua J, Troncosco S, Bugedo G, et al.
Renal function and cardiopulmonary bypass
: effects of perfusion pressure. J Cardiovasc Anaesth
1992; 6: 309-312.
11. Bucci M, D'Ambrosio G, Cascino P, Pace Palatti V, Martines G. Modificazioni della funzionalita renale in soggetti sottoposti a by-pass aorto-coronarico in circolazione extracorporea. Rev Eur Sci Med Farmacol
1995; 17: 183-190.
12. Christensen EI, Rennke HG, Carone FA. Renal tubular uptake of protein: effects of molecular charge. Am J Physiol
1983; 244: F436-441.
13. Caliskan S, Hacibekiroglu M, Sever L, Ozbory G, Arisoy N. Urinary N
1996; 74: 401-404.
14. Everaert K, Delanghe J, Vande Wiele C, et al.
Urinary alpha 1-microglobulin detects uropathy. A prospective study in 483 urological patients. Clin Chem Lab Med
1998; 36: 309-315.
15. Loeb WF. The measurements of renal injury. Toxicol Pathol
1998; 26: 26-28.
16. Dehne MG, Boldt J, Heise D, Sablotzki A, Hempelmann G. Protein, α1
- und β2
Mikroglobulin als Nierenfunktionsmarker in der Herzchirurgie. Anaesthesist
1995; 44: 545-551.
17. Yokono S, Veki M, Wakano M, et al.
Urinary excretion of ulinastin and NAG after cardiopulmonary bypass
1997; 46: 388-392.
18. Ascione R, Lloyd CT, Underwood MJ, Gomes WJ, Angelini GD. On-pump versus off-pump coronary revascularization: evaluation of renal function. Ann Thorac Surg
1999; 68: 493-498.
19. Agihasli M, Raadhakrishmamuthy B, Jiang K, Bao W, Berenson GS. Urinary N-acetyl-D-glucosaminidase
changes in relation to age, sex, race and diastolic and systolic blood pressure in a young biracial population. The Bogalusa Heart Study. Am J Hypertens
1996; 9: 157-161.
20. Skrha J, Haas T, Sperl M, et al.
A six-year follow-up: the relationship between N
in relation to retinopathy. Diabetic Med
1991; 8: 817-821.
21. Galanti LM, Jamart J, Dell'Omo J, Docrier J. Comparison of urinary excretion of albumin, alpha 1-microglobulin and retinol-binding proteins in diabetic patients. Diabetes Metab
1996; 22: 324-330.
22. Kanwar YS, Liu ZZ, Kashihara W, Waltner EI. Current status of the structural and functional basis of glomerular filtration and proteinuria
. Semin Nephrol
1991; 11: 390-413.
23. Scandling JD, Black VM, Deen WM, et al.
Glomerular perm-selectivity in healthy and nephrotic humans. Adv Nephrol
1992; 21: 159-176.
24. Fujigaki Y, Nagase M, Kobayasi S, Hidaka S, Shimomura M, Hishida A. Intra-GBM side of the functional filtration barrier for endogenous proteins in rats. Kidney Int
1993; 43: 567-574.
25. Jung K, Becker S. Multiple forms of N
-acetyl-D-betaglucosaminidase of human urine: isolation, properties and the development of a practical approach of differentiation. Biomed Biochim Acta
1991; 50: 861-867.
26. Brezis M, Rosen S, Silva P, Epstein FH. Renal ischemia: a new prospective. Kidney Int
1984; 26: 375-383.
27. Ranucci M, Pavesi M, Mazza E, et al.
Risk factors for renal dysfunction after coronary surgery: the role of cardiopulmonary bypass
1994; 9: 19-26.
28. Wesselink RMJ, de Boer A, Morshuis WJ, Leusink JA. Cardio-pulmonary bypass time has important influence on mortality and morbidity. Eur J Cardiothorac Surg
1997; 11: 1141-1145.
29. Christenson JT, Schmuziger M, Maurice J, Simonet F, Velebit V. How safe is coronary bypass surgery in the elderly patients? Analysis of 111 patients aged 75 years or more and 2939 patients younger than 75 years undergoing coronary artery bypass
grafting in a private hospital. Coron Artery Dis
1994; 5: 169-174.
30. Lazar HL, Fitzgerald C, Gross S, Heeren T, Aldea GS, Shemin RJ. Determinants of length of stay after coronary artery bypass
graft surgery. Circulation
1995; 92 (Suppl 9): P1120-1124.
31. Mangano CM, Diamondstone LS, Ramsay JG, Aggarwal A, Herskowitz A, Mangano DT. Renal dysfunction after myocardial revascularization: risk factors, adverse outcomes, and hospital resource utilization. The Multicenter Study of Perioperative Ischemia Research Group. Ann Intern Med
1998; 128: 194-203.