It is essential to ensure haemodynamic stability during cardiac anaesthesia. Postoperative hypertension is a common event after coronary artery bypass graft (CABG) surgery [1,2] and may be associated with subendocardial ischaemia [3,4], bleeding  and cerebrovascular haemorrhage [5,6]. Sodium nitroprusside and nitroglycerin are short-acting vasodilators commonly used intravenously in the management of postoperative hypertension due to high systemic vascular resistance [6-8]. However, these drugs have disadvantages including tachyphylaxis, reflex tachycardia and rebound hypertension. Another drawback of both sodium nitroprusside and nitroglycerin is a lack of vascular selectivity; in addition to causing arterial dilatation, they also produce a variable degree of venodilatation [8-12].
Other drugs available for acute arterial pressure control during cardiac surgery include calcium channel antagonists of the dihydropyridine class such as nifedipine, nicardipine and isradipine [13,14]. These drugs, as opposed to the nitrovasodilators sodium nitroprusside and nitroglycerin, have a relatively long plasma half-life [14-16] and hence prolonged duration of effect.
Clevidipine is a new ultrashort-acting, specific arterial vasodilator of the dihydropyridine (calcium antagonist) class specifically developed for the acute reduction and control of arterial pressure in the cardiac surgical setting [17,18]. It has a high clearance and small volume of distribution resulting in an extremely short plasma half-life . When clevidipine was compared with sodium nitroprusside in the control of blood pressure in patients after cardiac surgery, it was found that the systemic vascular resistance was lower and the stroke volume higher in those patients who received clevidipine .
The aims of the present study were to compare clevidipine and sodium nitroprusside in a double-blind fashion in respect of their efficacy in controlling arterial pressure, the number of dose-rate adjustments to do so (here named 'ease of use') and the haemodynamic changes they produced when used to control arterial pressure in sedated post-CABG surgery patients.
This study involved two specialist centres in the UK. After the approval of Local Research Ethics Committees and informed written patient consent, 39 patients scheduled for elective CABG were enrolled into the study. Exclusion criteria included preoperative administration of calcium channel blocking agents (within 6 h) or clonidine, infusion of nitroglycerin (≥50 μg min−1) and the need for inotrope support, phosphodiesterase inhibitors or intra-aortic balloon counterpulsation. Exclusion criteria were: patients who had an acute myocardial infarction within 24 h of the study; an ejection fraction <0.35; pre-existing left ventricular bundle branch block or ventricular pacing; supine heart rate ≥110 beats min−1; a history of cerebrovascular disease in the preceding 6 months; and clinically significant renal or hepatic disease.
A randomized double-blind study design was used. Patients were randomly allocated to one of the two different treatment regimens: (1) clevidipine (active, 0.3 mg mL−1)/sodium nitroprusside placebo (5% dextrose) or (2) sodium nitroprusside (active, 0.5 mg mL−1)/clevidipine placebo (Intralipid®). Randomization was done in blocks of four. The drugs were administered using a double-dummy technique as clevidipine was presented as a white lipid emulsion and sodium nitroprusside is a clear colourless solution. Intralipid® was used as the clevidipine placebo, and both the sodium nitroprusside and sodium nitroprusside-placebo solutions were covered in foil to maintain blinding and to protect the sodium nitroprusside from light. Each patient therefore received a white emulsion and foil-covered solution. The pharmacist at each hospital received an open randomization list and prepared the drugs.
Metoprolol 50 mg was given to patients on chronic oral β-adrenoceptor blocking therapy on the morning of surgery instead of their usual medication. Patients were premedicated with intramuscular morphine (7.5-10 mg) and hyoscine (0.2-0.3 mg). Anaesthesia was induced with fentanyl (10-15 μg kg−1), midazolam (up to 0.1 mg kg−1) etomidate (up to 20 mg) and maintained with either isoflurane (1%) or propofol at 3 mg kg−1 h−1; pancuronium (0.15 mg kg−1) was used for neuromuscular block. An arterial cannula and continuous thermodilution pulmonary artery catheter (Edwards, Santa Ana, CA, USA) were sited before surgery, enabling continuous monitoring of systemic, pulmonary and central venous pressures.
Postoperatively, a standardized sedation regimen was used with a bolus of fentanyl (up to 5 μg kg−1) given at the end of surgery followed by a propofol infusion (up to 2 mg kg−1 h−1). After surgery, 30 patients developed a mean arterial pressure (MAP) >90 mmHg for 10 min and were randomized to the study. The baseline haemodynamics were recorded in these 10 min. Of the 30 hypertensive patients, 15 received an infusion of clevidipine and 15 received an infusion of sodium nitroprusside to control the MAP within a target window of 70-80 mmHg for the following 3 h.
The study drug and placebo were infused with Braun Perfusor® pumps (B. Braun, Melsungen, Germany) at an initial rate of 0.06 mL kg−1 h−1. This dose-rate was intended to result in equipotent doses of clevidipine and sodium nitroprusside (clevidipine at 0.3 μg kg−1 min−1, sodium nitroprusside at 0.5 μg kg−1 min−1), although this proved to be an incorrect estimate (see Results). The dose-rates were adjusted to maintain the MAP between 70 and 80 mmHg. The infusion rates of the active drug and placebo were adjusted simultaneously, and the protocol allowed the drug infusions to be stopped and restarted as necessary. If the MAP > 105 mmHg, despite reaching the maximum rate of infusion (defined as the lesser of 100 mL h−1 or 1.2 mL kg−1 min−1), the patient concerned was transferred to open treatment. If MAP fell to <65 mmHg, the infusion of the study drug was stopped.
A Computerised Automated Data Acquisition and Monitoring system (ADAM) was developed to allow MAP and heart rate to be downloaded automatically at 1 min intervals from a Hewlett Packard Viridia monitor (Agilent Technologies, Hewlett Packard, Palo Alto, CA, USA) or a Datex AS3 monitor (Datex-Ohmeda, Datex-Engstrom Division, Helsinki, Finland) onto a laptop computer during the 3 h study period. In addition, the infusion rates of the study drugs were downloaded from the Braun Perfusor® pumps. The ADAM system was developed and operated according to the US Food and Drug Agency (FDA) guidelines . A sample printout generated from the ADAM is shown in Fig. 1, illustrating changes in MAP, with arrows indicating the times when the infusion rate was changed.
The efficacy was inversely related to the area under the pressure-time curve (AUCMAP mmHg min) when the MAP fell outside (above or below) the target pressure window (70-80 mmHg). AUCMAP is presented normalized per hour (AUCMAP mmHg min h−1) and therefore integrates both the magnitude and duration of the MAP falling outside the target pressures. A greater total AUCMAP h−1 means that the MAP was outside the target range for a longer period. Total AUC is the sum of the AUC above and below the predetermined target ranges. A large AUCMAP above the target MAP range indicated inadequate MAP control and poor efficacy, while an AUCMAP below the target MAP range probably indicated overdosing. The stability of heart rate was determined by calculating the area under the heart rate-time curve (AUCHR beats min−1 h−1) for the period when the heart rate was above or below 10% of the baseline pretreatment value for each patient. Similar calculations of mean AUCHR indicate the extent of heart rate increase (AUCHR above) and heart rate reduction (AUCHR below) compared with the target heart rate range. A higher AUCHR h−1 indicates that the heart rate increased by >10% from baseline for a longer period.
The number of dose-rate adjustments needed to control the MAP was recorded using the ADAM system and used as a measure of the 'ease of use' of the drug.
Central venous pressure (CVP, mmHg), pulmonary artery pressure (PAP, mmHg) and pulmonary capillary wedge pressure (PCWP, mmHg) were measured with the pulmonary artery catheter. Cardiac output (CO, L min−1) was measured with the help of a continuous thermodilution technique (Edwards). Stroke volume (SV, mL), systemic vascular resistance (SVR, dyne s cm−5) and pulmonary vascular resistance (PVR, dyne s cm−5) were calculated using standard formulae. These haemodynamic variables were arbitrarily recorded at 10 min before the start of infusion (baseline) and at 10, 30, 60, 90, 120, 150 and 180 min after the start of the infusion.
Postoperative myocardial ischaemia, during and after infusion of the study drugs, was assessed with a Holter monitor ECG technique (Tracker 2®; Reynolds Medical Ltd, Hertford, UK) using two extremity leads. The ischaemic burden - defined as the area of ST depression, measured between the baseline ST level (reference points between P and Q, at the junction and 80 ms after the J-point) and the level of ST-segment depression during ischaemia, versus time - was calculated and expressed (mV min h−1) for each patient during the 12 h from arrival in the intensive acre unit. Troponin I was measured preoperatively, 24 h and 5 days after the start of the infusion of the study drug as a measure of myocardial ischaemia (the reference value in non-cardiac surgery patients was 0-0.3 ng mL−1).
The total fluid input (crystalloid, colloid and autotransfused blood from the chest drains) and fluid output (urine and chest drains) was recorded over 3 h periods (0-3, >3-6, >6-9, >9-12 h).
The efficacy variables, AUCMAP and AUCHR, during the 3 h infusion were normalized for the hour and analysed after log transformation using a two-way analysis of variance with treatment and centre as factors. Descriptive statistics for the two treatment groups are presented. The estimate, 95% confidence interval and corresponding P values for the ratio between clevidipine and sodium nitroprusside are given.
The number of interventions - measured by the mean number of dose-rate adjustments per hour - was calculated for each patient and analysed by a two-way analysis of variance with treatment and centre as factors. Descriptive statistics for the two treatment groups are presented. The estimate, 95% confidence interval and corresponding P values for the ratio between clevidipine and sodium nitroprusside are given. Since no empirical data were available, 15 evaluable patients per treatment group in the statistical analysis, with a power of 80% and significance level of 5%, allowed a coefficient of variation (CV) of 0.9 for the number of dose adjustments.
A two-way ANOVA with treatment and centre as factors and baseline was used for all haemodynamic variables. Descriptive statistics and P values are presented both for the percentage changes from baseline in the two treatment groups and for the difference between clevidipine and sodium nitroprusside. Differences in fluid input and output were compared between the two groups using a two-way analysis of variance, again with treatment and centre as factors.
Data are presented as mean ± SD. All data were analysed using SAS 6.12 and Microsoft Excel® 7.0 (Microsoft, Redmond, WA, USA).
Thirty-nine patients were enrolled in the study. Thirty developed postoperative hypertension with a MAP > 90 mmHg for at least 10 min. Fifteen were randomly allocated to receive clevidipine and 15 to receive sodium nitroprusside. Characteristics of the treated patients are shown in Table 1.
There was no statistically significant difference in AUCMAP, which is inversely related to the efficacy between clevidipine and sodium nitroprusside. The AUCMAP in each group are shown in Table 2. No difference was seen in the total AUCMAP between clevidipine and sodium nitroprusside, and similarly there was no difference when the AUCMAP was measured only for MAP above or below the target range.
The mean dose-rates of clevidipine and sodium nitroprusside needed to control arterial pressure throughout the study were 0.76 ± 0.83 and 0.58 ± 0.54 μg kg−1 min−1 for clevidipine and sodium nitroprusside, respectively. There were no statistically significant differences in the number of dose-rate adjustments between the two groups, both when calculated for the initial 30 min and for the remaining 150 min of the drug infusion. The total number of dose-rate adjustments for the clevidipine group for the first 30 min was 8.54 ± 7.35 and the last 150 min was 2.40 ± 2.13 compared with 9.08 ± 5.56 at 30 min and 2.13 ± 0.84 at 150 min for the sodium nitroprusside group.
The study infusion was terminated in two patients. Arterial pressure control in one of the patients in the clevidipine group was inadequate, so the study drug infusion was terminated and open treatment was commenced. The other patient withdrawn from the study was in the sodium nitroprusside group. This patient had a low MAP due to a severe fall in arterial pressure despite a low sodium nitroprusside infusion rate. The AUCMAP of this patient was therefore high but is still included in the data analysis.
The effect on heart rate was also studied, and a greater increase in heart rate was observed when sodium nitroprusside was used to lower the arterial pressure than when clevidipine was used, resulting in a significantly (P < 0.001) larger AUCHR for sodium nitroprusside than for clevidipine (Table 2).
The baseline haemodynamic data for the clevidipine and sodium nitroprusside groups determined during the 10 min before the start of the study drug infusion were similar (Table 3). In both groups, the SVR was reduced compared with baseline, but only in the sodium nitroprusside group were SV, CVP and PAP significantly reduced and heart rate increased (P < 0.05). There was no significant change from baseline in these parameters during clevidipine infusion.
The total input of fluid in the 12 h postoperative period was significantly greater in the sodium nitroprusside group than in the clevidipine group (P < 0.05). The volumes given (including autotransfusion) in the first 3 and 12 h in the intensive care unit were 962 ± 396 and 2262 ± 1090 mL in the clevidipine group and 1126 ± 416 and 3060 ± 1184 mL in the sodium nitroprusside group.
Troponin I (ng mL−1) was measured and the ischaemic burden calculated for each patient from an ambulatory electrocardiograph. At pre-entry, 24 h after the start of the study drug infusion and 5 days later, the medians (range) of troponin I levels for clevidipine-treated patients were 0.1 (0.1-1.4), 1.5 (0.5-14.9) and 0.1 (0.1-6.4) ng mL−1, respectively, compared with the corresponding values obtained with sodium nitroprusside treatment, which were 0.1 (0.1-2.5), 1.1 (0.1-14.0) and 0.1 (0.1-1.3) ng mL−1, respectively. The results from the ambulatory electrocardiogram showed no significant differences between the two groups.
Clevidipine is an ultrashort-acting calcium channel antagonist  used in the treatment of hypertension after coronary surgery [19,22]. In the present study, we compared the efficacy of clevidipine when used as an infusion to control arterial pressure in patients after CABG surgery with the commonly used vasodilator sodium nitroprusside. There was no statistically significant difference in efficacy seen between clevidipine and sodium nitroprusside in controlling arterial pressure.
Hypertension has been increasingly recognized as a frequent complication of CABG surgery and has been reported to occur in 30-50% of cases . Patients therefore often require vasodilators to control arterial pressure during and after surgery. The level to which arterial pressure can be safely allowed to increase in the perioperative setting has been debated [1,17]. One recent survey of practice in the perioperative care of 2000 cardiac surgical patients showed that treatment in intensive care is initiated at an average of 104 mmHg and that MAP is maintained at 80 mmHg . In the present study, arterial pressure increased postoperatively to >90 mmHg in 30/39 recruited patients, justifying pharmacological intervention. According to the above-mentioned survey , nitroglycerin and sodium nitroprusside are the most commonly used vasodilators in cardiac surgery in Europe and the USA, and sodium nitroprusside was therefore chosen as the agent with which to compare clevidipine.
The efficacy of each drug was evaluated by measuring AUCMAP. This gives an inverse numerical measure of how well arterial pressure was controlled over a certain period. There was no observed difference in AUCMAP between the two groups in this study. In all but one patient in each group, MAP was controlled at doses within the recommended range selected for this study. The ease of use of both drugs was estimated by counting the number of interventions required to reach the effective rate of infusion, and was shown to be similar. With both drugs, the number of dose adjustments in the first 30 min was higher than in the following 150 min. This was expected as the drug is titrated to a required effect for a particular patient. The number of dose adjustments is dependent on the efficacy of the drug, its concentration in the infused solution and the experience of the operator. However, the concentration of the clevidipine solution was lower than that of sodium nitroprusside and the recommended dose-rates proved to be non-equipotent. Therefore, the number of dose adjustments might be an inadequate measure of ease of use in this study as a given dose-rate adjustment will result in a smaller dose change in the clevidipine group than in the sodium nitroprusside group. For this reason, the standard clevidipine concentration was set at 0.5 mg mL−1 in the further clinical development of the drug.
Since there was no statistically significant difference between the AUCMAP and the number of dose adjustments, it is reasonable to conclude that the efficacy of the two drugs is indeed similar. While a difference may have been found with a larger sample size and stricter dose recommendation, it is unlikely that such a difference would have been clinically relevant.
Haemodynamic changes produced by clevidipine were also studied and compared with those of sodium nitroprusside. The increase in heart rate, measured as AUCHR seen in the sodium nitroprusside group, was significantly greater than that observed in the clevidipine group. Compensatory tachycardia is a well-recognized side-effect of sodium nitroprusside, and the lack of increase in heart rate seen with clevidipine may confer an advantage in that it minimizes the increased myocardial oxygen consumption in patients at risk of myocardial ischaemia in the immediate postoperative period.
We observed significantly less change in CVP with clevidipine than with sodium nitroprusside, which might explain the decrease in SV from baseline measurements seen with sodium nitroprusside but not with clevidipine. These effects of sodium nitroprusside may be attributable to venodilatation. Although significant differences between the groups in the recorded haemodynamic variables were few in this parallel group study, the compiled haemodynamic pattern after clevidipine appears to differ from that of sodium nitroprusside. The present results thus confirm those of a previous crossover study in which the haemodynamic effects of clevidipine were compared with those of sodium nitroprusside during 10 min administration to sedated cardiac surgical patients in the postoperative period . It is likely that the difference in the effect between clevidipine and sodium nitroprusside is a result of the different site of action of the two drugs. While sodium nitroprusside has a combined action on arterial and venous vascular smooth muscle, clevidipine is a selective arteriolar vasodilator . There was minimal change in the preload with clevidipine, suggesting that the SV, and hence CO, is less likely to decrease.
The postoperative fluid input and output (urine and drain losses) were measured throughout the 12 h period in the intensive care unit. The sodium nitroprusside group demonstrated a greater need of fluid replacement than the patients receiving clevidipine. The difference may again be related to the different sites of action. As already mentioned, sodium nitroprusside acts on both arteriolar and venous smooth muscle, and as venodilatation causes greater expansion of the circulatory system, a larger circulatory volume is therefore required to maintain the haemodynamics. Most calcium antagonists, on the other hand, have a natriuretic and diuretic action , and patients with essential hypertension receiving clevidipine have been reported to increase urine production , suggesting that clevidipine may also cause diuresis in the postoperative setting following cardiac surgery. However, there was no difference in the urine output between the sodium nitroprusside and clevidipine groups.
In a previous study in which nimodipine, also a dihydropyridine calcium antagonist, was used in patients undergoing cardiac valve replacement, increased bleeding was reported as an unexpected finding . In the present study, the extent of blood drainage did not differ in the two groups, although the small sample size may have prevented such an effect from being detected. However, results compiled from several clevidipine studies do not indicate any tendency towards increased bleeding [19,21,22].
In conclusion, the efficacy and ease of use of clevidipine and sodium nitroprusside in the control of arterial pressure in post-CABG patients were identical in this study. The differences seen in the haemodynamic profile of the two drugs can be attributed to the (greater) venodilatation caused by sodium nitroprusside, which produces a decrease in preload (CVP, PCWP) and consequent fall in SV and CO, with a compensatory increase in heart rate. Clevidipine controlled arterial pressure without causing as much change in heart rate as sodium nitroprusside. Unlike sodium nitroprusside, which has a combined action on the arterial and venous vasculature, clevidipine is a selective arterial vasodilator with little effect on the preload, implying that SV, CO and heart rate will remain stable during administration.
This study was sponsored by AstraZeneca R & D Mölndal, Sweden. Å. J.-M. and M. N. are employees of AstraZeneca.
1. Vuylsteke A, Feneck RO, Jolin-Mellgård Å, et al.
Perioperative blood pressure control: a prospective survey of patient management in cardiac surgery. J Cardiothorac Vasc Anesth
2. Mora CT, Dudek C, Torjman MC, White PF. The effects of anesthetic technique on the hemodynamic response and recovery profile in coronary revascularization patients. Anesth Analg
3. Estafanous FG, Tarazi RC, Viljoen JF, el-Tawil MY. Systemic hypertension following myocardial revascularization. Am Heart J
4. Mangano DT, Siliciano D, Hollenberg M, et al.
Postoperative myocardial ischemia. Therapeutic trials using intensive analgesia following surgery. Anesthesiology
5. Viljoen JF, Estafanous FG, Tarazi RC. Acute hypertension immediately after coronary artery surgery. J Thorac Cardiovasc Surg
6. Roberts AJ, Niarchos AP, Subramanian VA, et al.
Systemic hypertension associated with coronary artery bypass surgery. Predisposing factors, hemodynamic characteristics, humoral profile, and treatment. J Thorac Cardiovasc Surg
7. Gray RJ. Postcardiac surgical hypertension. J Cardiothoracic Anesth
8. Tinker JH, Michenfelder JD. Sodium nitroprusside: pharmacology, toxicology and therapeutics. Anesthesiology
9. Kaplan JA, Finlayson DC, Woodward S. Vasodilator therapy after cardiac surgery: a review of the efficacy and toxicity of nitroglycerin and nitroprusside. Can Anaesth Soc J
10. Amaranath L, Kellermeyer WF Jr. Tachyphylaxis to sodium nitroprusside. Anesthesiology
11. Khambatta HJ, Stone GJ, Khan E. Hypertension during anesthesia on discontinuation of sodium nitroprusside-induced hypotension. Anesthesiology
12. Benumof JL. Hypoxic pulmonary vasoconstriction and infusion of sodium nitroprusside. Anesthesiology
13. Van Zwieten PA, van Wezel HB. Antihypertensive drug treatment in the perioperative period. J Cardiothorac Vasc Anesth
14. Van Wezel HB, Koolen JJ, Visser CA, et al.
The efficacy of nicardipine and nitroprusside in preventing poststernotomy hypertension. J Cardiothorac Anesth
15. Nathan HJ, Laganière S, Dubé L, et al.
Intravenous nifedipine to treat hypertension after coronary artery revascularization surgery. A comparison with sodium nitroprusside. Anesth Analg
16. Fitton A, Benfield P. Isradipine: a review of its pharmacodynamic and pharmacokinetic properties, and therapeutic use in cardiovascular disease. Drugs
17. Levy JH, Huraux C, Nordlander M. Treatment of perioperative hypertension. In: Epstein M, ed. Calcium Antagonists in Clinical Medicine.
Philadelphia, USA: Hanley & Belfus, 1997: 345-358.
18. Ericsson H, Fakt C, Hoglund L, et al.
Pharmacokinetics and pharmacodynamics of clevidipine in healthy volunteers after intravenous infusion. Eur J Clin Pharmacol
19. Kieler-Jensen N, Jolin-Mellgård Å, Nordlander M, Ricksten S-E. Coronary and systemic hemodynamic effects of clevidipine, an ultra-short-acting calcium antagonist, for treatment of hypertension after coronary artery surgery. Acta Anaesthesiol Scand
20. FDA Guidance for Industry: Computerized Systems Used in Clinical Trials.
Washington, DC, USA: Department of Health and Human Services, Food and Drug Administration, April 1999.
21. Ericsson H. The pharmacokinetics and pharmacodynamics of clevidipine, a new ultra-short-acting calcium antagonist, in different animal species and man. Dissertation, Uppsala University, Sweden, 1999.
22. Bailey JM, Lu W, Levy JH, et al.
Clevidipine in adult cardiac surgical patients. A dose finding study. Anesthesiology
23. Estafanous FG, Tarazi RC. Systemic arterial hypertension associated with cardiac surgery. Am J Cardiol
24. Epstein M, Loutzenhiser R. Calcium Antagonists and the Kidney.
Philadelphia, USA: Hanley & Belfus, 1990.
25. Schwieler JH, Ericsson H, Löfdahl P, Thulin T, Kahan T. Circulatory effects and human pharmacology of clevidipine, a novel ultra short acting and vascular selective calcium antagonist, in hypertensive humans. J Cardiovasc Pharmacol
26. Wagenknecht LE, Furberg CD, Hammon JW, Legault C, Troost BT. Surgical bleeding: unexpected effect of a calcium antagonist. BMJ