The Hemodynamic Effects of Propofol in Children with Congenital Heart Disease

Abstract

We studied the hemodynamic effects of propofol during elective cardiac catheterization in 30 children with congenital heart disease. Sixteen patients were without cardiac shunt (Group I), six had left-to-right cardiac shunt (Group II), and eight had right-to-left cardiac shunt (Group III). The mean (±SD) ages were 3.8 ± 3.1 yr (Group I), 3.2 ± 3.7 yr (Group II), and 1.0 ± 0.6 yr (Group III). After sedation and cardiac catheter insertion, hemodynamic data and oxygen consumption were measured before and after the administration of propofol (2-mg/kg bolus, 50- to 200-μg · kg⁻¹ · min⁻¹ infusion), and values were compared by using a paired t-test (significance: P < 0.05). After the propofol administration, systemic mean arterial pressure and systemic vascular resistance decreased significantly and systemic blood flow increased significantly in all patient groups; heart rate, pulmonary mean arterial pressure, and pulmonary vascular resistance were unchanged. Pulmonary to systemic resistance ratio increased (Group I, P = 0.005; Group II, P = 0.03; Group III, P = 0.10). In patients with cardiac shunt, propofol resulted in decreased left-to-right flow and increased right-to-left flow; the pulmonary to systemic flow ratio decreased significantly (Group II, P = 0.005; Group III, P = 0.01). Clinically relevant decreases in PaO₂ (P = 0.008) and SaO₂ (P = 0.01) occurred in Group III patients. We conclude that propofol can result in clinically important changes in cardiac shunt direction and flow.

Implications 

The principal hemodynamic effect of propofol in children with congenital heart defects is a decrease in systemic vascular resistance. In children with cardiac shunt, this results in a decrease in the ratio of pulmonary to systemic blood flow, and it can lead to arterial desaturation in patients with cyanotic heart disease.

The goals for anesthetic management of children undergoing cardiac catheterization include adequate analgesia, sedation, and immobility, with minimal depression of cardiac function and respiratory drive. Propofol has become increasingly popular as an anesthetic for cardiac catheterization in children (^1,2), primarily because it is an IV anesthetic with favorable pharmacokinetic and pharmacodynamic characteristics. It has rapid clearance and redistribution, which result in prompt, clear-headed return of orientation with a low incidence of nausea and vomiting (^3,4). Although propofol’s safety and efficacy in pediatric patients have been well demonstrated (^5–9), there has been little published about its use in children with congenital heart disease (^10–12). Propofol decreases systemic blood pressure (^6–9), and the hemodynamic changes induced by propofol may alter the information obtained during cardiac catheterization. In particular, the effects on the magnitude and direction of intracardiac shunt have not been reported in this population. We designed our study to investigate the hemodynamic consequences of propofol administered during diagnostic cardiac catheterization to children with congenital heart disease.

Methods

After institutional review board approval and informed parental consent, 31 children, ASA physical status II or III, aged 3 mo to 12 yr, who were scheduled for elective cardiac catheterization were enrolled. Patients requiring mechanical ventilation or IV inotropic support, patients with hepatic or renal dysfunction, and patients with suspected allergy to propofol were excluded.

All patients were monitored with a precordial stethoscope, electrocardiograph (lead II), a noninvasive blood pressure device (Dynamap; Critikon Inc., Tampa, FL), pulse oximeter (Nellcor Inc., Hayward, CA), and skin temperature probe. Depending on the anesthesiologist’s preference, children either were not premedicated or received oral midazolam (0.5 mg/kg) or rectal methohexital (30 mg/kg). An IV catheter was placed in a peripheral vein, and lactated Ringer’s solution was infused at maintenance rates. During the period before propofol administration, incremental IV boluses (2–5 mg/kg) of thiopental were administered to keep the patient appropriately sedated for the cardiac catheterization procedure. Care was taken to limit the total thiopental dose to the minimum necessary. In an attempt to reduce the influence of thiopental on hemodynamic measurements, data were not recorded within 10 min of a thiopental bolus, and thiopental was discontinued at least 15 min before propofol administration. All patients breathed spontaneously, and airway support was not provided. Supplemental oxygen was not provided until completion of the study. Catheters were inserted under local anesthesia by the cardiologist for pressure and oxygen saturation (SO₂) measurements from the vena cavae, left and right atria, left and right ventricles, pulmonary artery, and aorta. Blood for analysis of oxygen tension (PO₂), carbon dioxide tension (PCO₂), and pH was obtained from the femoral artery and from the site judged optimal for obtaining mixed venous samples (pulmonary artery for patients without cardiac shunt and superior vena cava for patients with shunt). Oxygen consumption (V̇O₂) was noted every minute and averaged over the period of hemodynamic measurements. After baseline measurements had been taken, propofol was administered in 0.5-mg/kg increments to a total dose of 2 mg/kg over a 5-min period. A propofol infusion (100 μg · kg⁻¹ · min⁻¹) was also initiated. The infusion was maintained for the duration of the catheterization procedure and adjusted (50–200 μg · kg⁻¹ · min⁻¹) depending on the patient’s clinical response. Approximately 3 min after the propofol loading dose, all pressure and oxygen saturation measurements were repeated, as were pulmonary artery/superior vena cava and femoral artery blood gas analyses.

Pulmonary blood flow (Qp) and systemic blood flow (Qs) were calculated by using the Fick equation; pulmonary vascular resistance (PVR) and systemic vascular resistance (SVR) were calculated using standard formulae (¹³).

Data are reported as mean ± SD. Based on the calculated ratio of Qp:Qs, children were separated into three groups: patients without cardiac shunt (Group I); patients with Qp:Qs ≥1 both before and after propofol administration (Group II); and patients with Qp:Qs <1 either before or after propofol administration (Group III). Demographic and selected pre-propofol baseline values were compared among the three groups of children by using the Kruskal-Wallis test. Within each group, values before and after the propofol administration were compared by using a paired Student’s t-test. Statistical significance was set at P <0.05.

Results

Thirty-one children were enrolled. One patient became apneic after propofol administration and was excluded because the study protocol was not completed. Patient demographics for the 30 children (14 female, 16 male) who completed the study are shown in Table 1. Two Group I and three Group III patients were taking chronic medication for congestive heart failure. Children in Group III were significantly younger and smaller than children in the other groups (Table 1); consequently, fewer of these patients received premedication (proportion premedicated in Group I 16 of 16, Group II 5 of 6, Group III 4 of 8). Group III patients suffered from cyanotic heart disease and had significantly lower baseline SaO₂ values than patients in Groups I and II (Table 2). Baseline hemoglobin values were also significantly higher for Group III patients (Group I 10.5 ± 1.1 g/dL, Group II 12.0 ± 1.9 g/dL, Group III 13.5 ± 2.0 g/dL; P = 0.003). The total thiopental dose was 13.1 ± 7.6 mg/kg for Group I (administered over 65 ± 18 min) 10.0 ± 3.7 mg/kg (over 60 ± 11 min) for Group II, and 12.8 ± 8.5 mg/kg (over 55 ± 9 min) for Group III. There was no difference among the three groups in the total doses (per kilogram body weight) of thiopental or propofol.

Table 1: Demographics of All Patients

Table 2: Arterial Blood Gas Analysis Data for All Patients

In all three patient groups, propofol administration was associated with significant decreases in mean systemic arterial pressure (SAP) and SVR and with a significant increase in Qs, whereas mean pulmonary artery pressure (PAP), PVR, and Qp did not change significantly (Table 3). These changes resulted in a significant increase in the pulmonary to systemic resistance ratio (Rp:Rs) in Groups I and II. Propofol administration did not alter heart rate or V̇O₂ in any of the three patient groups (Table 3).

Table 3: Hemodynamic and Oxygen Consumption Data for All Patients

Fourteen children had some form of cardiac shunt(s). All six patients in Group II were calculated to have Qp:Qs ≥1 (left-to-right shunt) both before and after the administration of propofol. Six of the eight patients in Group III were calculated to have Qp:Qs <1 (right-to-left shunt) both before and after propofol administration. For the remaining two patients in Group III (both with unrepaired tetralogy of Fallot), the baseline shunt was left-to-right (Qp:Qs ratios of 1.03 and 1.39) but became right-to-left (Qp:Qs ratios of 0.54 and 0.83, respectively) after propofol administration. Therefore, propofol was associated with reversal of shunt flow in these children.

The effects of propofol on shunt direction and flow for Groups II and III are illustrated in Figure 1. Regardless of the direction of shunt at baseline, the drug decreased left-to-right shunt and increased right-to-left shunt. Left-to-right flow decreased significantly (P = 0.045) in Group II, and right-to-left increased significantly (P = 0.006) in Group III. Consequently, Qp:Qs decreased significantly in both Group II (P = 0.005) and Group III (P = 0.01).

Figure 1: Calculated left-to-right and right-to-left cardiac shunt and calculated ratio of pulmonary to systemic blood flow for patients with cardiac shunt (Groups II and III).

In children with cyanotic heart disease (Group III), this decrease in Qp:Qs with propofol administration was associated with statistically significant decreases in PaO₂ and SaO₂. SaO₂ decreased by >10 percentage points in two patients.

Although mean values remained within normal limits, all patient groups experienced a small decrease in arterial blood pH and a small increase in PaCO₂. Changes in values attained statistical significance in Groups I and III (Table 2). The respiratory depressant effects of propofol became important in one patient, a 10-yr-old with dilated cardiomyopathy (without cardiac shunt) and pulmonary hypertension. A loading dose of propofol resulted in hypoventilation (arterial pH decreased from 7.37 to 7.26, and PaCO₂ increased from 45 to 60 mm Hg) and a marked increase in PVR (from 640 to 1040 dynes · s⁻¹ · cm⁻⁵). The patient was withdrawn from the study, and after tracheal intubation and assisted ventilation, PVR was restored to baseline values. The catheterization continued without further incident, and the patient’s recovery was uneventful.

Discussion

In this study of sedated children with congenital heart defects undergoing cardiac catheterization, propofol’s principal effect was a substantial reduction in SVR. Because PVR remained constant, the ratio between pulmonary and systemic resistance was increased. In children with cardiac shunt, the increased Rp:Rs led to diminished left-to-right shunt, increased right-to-left shunt, and a decrease in Qp:Qs. For patients with acyanotic heart defects, the propofol-induced reduction in Qp:Qs seemed well tolerated and could have been advantageous by diminishing pulmonary congestion. In contrast, some patients with cyanotic heart defects experienced reduced pulmonary blood flow, thereby increasing the risk of arterial oxygen desaturation. This is the first study to document the effects of propofol on the amount and direction of shunt flow in children with congenital heart disease.

In all patient groups, Qs increased after propofol administration. The 21% reduction in SVR in children without shunt lesions (Group I) was associated with a 9% increase in cardiac output, presumably due to an increase in stroke volume (as heart rate did not change).

The observation that propofol did not influence normal baseline PVR is consistent with data from a perfused lung model—propofol administration did not change pulmonary resistance when baseline PVR values were normal, but it reduced PVR when baseline values were increased (¹⁴). In our patients with increased baseline PVR, the effect of propofol was inconsistent. Further studies of the effect of propofol (with PaCO₂ held constant) are required in children with pulmonary hypertension.

Propofol (in combination with fentanyl) has been compared with ketamine anesthesia in 20 unpremedicated children with congenital heart disease undergoing cardiac catheterization (¹). Patients in that study had a broad spectrum of anatomic defects, including cardiac shunts. Those in the propofol group were more likely to experience a decrease in SAP of >20% (compared with baseline) during anesthetic induction. Heart rate was not significantly altered. These observations are consistent with our study. Of the 10 patients who received propofol, 4 developed arterial desaturations of >5 percentage points by pulse oximetry. All periods of desaturation occurred on induction, concomitant with a transient decrease in systemic blood pressure. However, the explanation for these desaturation episodes remains unclear, because SAP and heart rate were the only hemodynamic variables measured. In our study, small but statistically significant decreases in SaO₂ where observed in patients without shunt or with increased pulmonary flow (Groups I and II), which suggests that respiratory depression probably contributed to changes in oxygenation in some patients.

The cardiovascular effects of IV propofol induction in premedicated children without congenital heart disease have been reported (⁹). Propofol induction resulted in approximately a 30% reduction in SAP and a 15% reduction in SVR. Heart rate decreased 10%–20%, and stroke volume index increased by 12%. These changes are similar to those noted in our patients with congenital heart disease. Propofol had no effect on sinoatrial or atrioventricular node function in pediatric patients undergoing radiofrequency catheter ablation (¹²). Likewise, in our study, no alteration of heart rate or cardiac rhythm was noted.

Propofol’s cardiovascular effects have also been evaluated in adults undergoing cardiac surgery. Most such studies involved patients with ischemic heart disease and normal left ventricular function (^15–19). Propofol caused a 20%–40% decrease in blood pressure (^16,20), primarily via systemic vasodilation (^16,17,19). Heart rate was usually unchanged or decreased (²¹). Our study data are consistent with these observations. Propofol can produce rate- and dose-dependent myocardial depression (²²). The increase in cardiac index noted in our Group I patients suggests that children with congenital heart defects without baseline myocardial depression tolerate propofol’s effects well in the doses administered.

Propofol is likely to be used increasingly in children with congenital heart disease. It may become a preferred option for pediatric cardiac patients with good ventricular function in whom rapid recovery from anesthesia is desirable. Current practice at our institution is similar to that described by Reich (²), in that propofol is used as the primary anesthetic for most children undergoing cardiac catheterization. Like others (^1,23,24), we emphasize that propofol should only be administered to pediatric cardiac patients when the drug’s properties are appropriate for the child’s clinical situation. Propofol’s hemodynamic profile suggests caution in patients for whom systemic afterload reduction may be harmful (e.g., patients with severe aortic stenosis, hypertrophic obstructive cardiomyopathy) and in cyanotic patients whose pulmonary blood flow depends on the balance between systemic and pulmonary vascular resistance (e.g., patients with tetralogy of Fallot, hypoplastic left heart syndrome after the Norwood palliation). The potential for respiratory depression should be considered and precautions taken to avoid deleterious consequences, particularly in patients who may have pulmonary hypertension.

Propofol can alter the patient’s hemodynamic profile. Interpretation of hemodynamic data obtained during cardiac catheterization of children requires an awareness of the hemodynamic consequences of the anesthetic techniques used.

Midazolam and/or barbiturates were given before propofol because it was considered unethical (and technically difficult) to initiate a cardiac catheterization in awake young children. The duration of thiopental administration was approximately one hour, and the range in total dose (expressed as a rate) was moderate (0.17–0.23 mg · kg⁻¹ · min⁻¹). By studying each patient before and after propofol administration, we attempted to minimize the hemodynamic effects of sedatives. Nevertheless, prior administration of midazolam or thiopental may have influenced baseline hemodynamic variables and the response to propofol (²⁵). In addition, these drugs may have accentuated the respiratory depressant effects of propofol (²⁶).

In summary, we studied the hemodynamic effects of propofol in sedated children with congenital heart defects undergoing cardiac catheterization. Propofol resulted in a significant decrease in SAP and SVR and increase in Qs in all patients, whereas heart rate, PAP, PVR, and Qp remained unchanged. Qp:Qs decreased in patients with cardiac shunt, leading to further desaturation in patients with cyanotic heart disease (Qp:Qs <1). Awareness of propofol’s significant cardiorespiratory effects can facilitate the appropriate selection of anesthetics for children with congenital heart disease and can also aid in the interpretation of cardiac catheterization data obtained during propofol anesthesia.

We thank the cardiac catheterization laboratory staff for their cheerful support and Donald Baptiste for his expert and willing assistance.