The effects of hypothermia on human left ventricular contractile function during cardiac surgery

Patient details

Ten patients undergoing elective coronary artery bypass grafting (CABG) were recruited. All had triple-vessel disease but good LV function (>50% ejection fraction) and no hypertension, diabetes, atrial fibrillation, left main stenosis, unstable angina, recent infarct (<2 months) or previous cardiac surgery.

All patients received identical premedication (temazepam, 10 to 20 mg; ranitidine, 150 mg; and metoclopramide, 10 mg orally) and induction (fentanyl, 15 μg/kg; etomidate, 0.1 mg/kg; and pancuronium, 0.1 mg/kg). A central venous line, right radial arterial line and a urinary catheter were inserted. In addition, a Swan-Ganz catheter was utilized to allow measurement of cardiac output and the measurement of core blood temperature.

Maintenance anesthesia (fentanyl and propofol) and bypass techniques were standardized in all patients. A maximum heat exchanger temperature gradient of 10°C was used for cooling/warming while on cardiopulmonary bypass. After sternotomy and conduit procurement, pericardiotomy was performed and the patient was prepared for CPB.

The conductance catheter

The conductance method of determining LV volume is based on the measurement of blood conductivity within the LV. The technique has been described in detail elsewhere (9,10). Briefly the Millar conductance catheter (Millar Instruments, Houston, Texas) has 12 electrodes arranged with an interelectrode distance of 1.0 cm. The catheter is inserted into the LV along its long axis. A conditioner/processor applies a current between the most distal and proximal electrodes and quantifies the conductances (reciprocal of resistances) between the electrode pairs. The sum of these signals is used to obtain an expression for the total volume of the LV. The catheter also contains a solid-state pressure transducer. Hence, the Millar conductance catheter allows online assessment of pressure and volume. A Leycom Sigma-5 signal conditioner processor (Cardio-dynamics; Rÿnsburg, Netherlands) was used to process volume signals. Analysis of data was performed using a customized software package. End-systolic elastance (Ees) was evaluated from the end-systolic pressure-volume relation (ESPVR) using an iterative method as described by Sagawa et al. (11). The stroke-work end-diastolic volume relationship (12)is the ratio of stroke work to end-diastolic volume (EDV) at a fixed EDV, where stroke work is calculated from the area bounded by the pressure-volume (P-V) loop. Time to reach peak ejection rate (tPER) is defined as the time between the instant when the aortic valve opens (i.e., at dP/dt_max) and the instant when ejection reaches its peak rate (i.e., dV/dt_min). It is related to the time to peak elastance (T_max).

Study protocol

The conductance catheter was inserted via the aortic root, and its position was verified by observing the volume signals from each segment. End-expiratory thermodilution cardiac output was measured at this point to calibrate the catheter for stroke volume. The resistivity of the blood was measured, and a series of steady-state P-V loops were obtained at end expiration. Parallel conductance effect was measured by injection of 10 ml of 5% saline into the pulmonary artery port of the Swan-Ganz catheter. Patients were then assigned to one of two groups of five patients. All measurements were made following the temporary discontinuation of CPB.

In the first group (HR100), right atrial pacing was established at a fixed rate of 100, to minimize variation in HR with temperature and to simulate the postoperative period. A series of loops at differing preloads were then generated by brief occlusion of the inferior vena cava at end-expiration. Following this, CPB was instituted and the patient re-warmed to 37°C. The blood resistivity measurement was then repeated, and a further series of pressure-volume loops were generated. The patient was then core-cooled on CPB. At 35, 33 and 31°C (by Swan-Ganz thermistor and myocardial temperature probe), the data collection was repeated including repeat resistivity estimation at each temperature. Heart rate was maintained at 100 beats/min throughout. In the second group (HR80/120), the relationship between HR and temperature was investigated further. Right atrial pacing was established. Following this, CPB was instituted and the patient was re-warmed to 37°C. After warming, the patient was weaned from CPB, the blood resistivity was measured, and runs of pressure volume loops were generated at this temperature at atrially paced HRs of 80 and 120 beats/min. The patient was then core-cooled on CPB to 33°C. The sequence of measurements was then repeated at the two different HRs.

Following the final measurement in each group, CPB was reinstituted, cooling was continued to 28°C, and CABG was performed. Each patient received three bypass grafts. Mean CPB time was 88 ± 11 mins. All patients survived the study and suffered no complications attributable to the conductance catheter. One patient died after an acute myocardial ischemic event postoperatively.

Ethical committee approval

Approval was granted for this study by the South Birmingham Local Research Ethics Committee. All patients gave informed, written consent.

Statistical analysis

Statistical analysis was performed using one-way analysis of variance (ANOVA) with multiple comparison (Tukey-Kramer) with the ARCUS software (Cambridge Software Publishing, Cambridge, UK) package, on a standard personal computer. Statistical significance was assumed when p < 0.05. In group 1 (HR100), p values were calculated by standard ANOVA to compare all results for each index of contractility and by ANOVA with multiple comparison to compare consecutive temperatures (i.e., 37 vs. 35, 35 vs. 33). In group 2 (HR80-120), p values were calculated by using the paired ttest.

Group 1 (HR100)

In one patient the study was terminated at 33°C due to atrial fibrillation. At a paced HR of 100 beats/min, there was a progressive fall in Ees (Figs. 1⇓and 2) ⇓and stroke work at constant volume (Fig. 3) with cooling. The tPER (Fig. 4) increased progressively as the temperature was reduced.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

Pressure volume loops from a patient demonstrating a fall in contractility at constant paced heart rate (100 beats/min), with cooling (A= 37°C; B= 33°C). Group 1.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

Change in normalized end systolic elastance from baseline at constant paced heart rate (100 beats/min) with cooling. Group 1. ∗∗p < 0.0001.

• Download figure
• Open in new tab
• Download powerpoint

Figure 3

Change in normalized stroke work (SW) -end diastolic volume (EDV) relationship from baseline at constant paced heart rate (100 beats/min) with cooling. Group 1. ∗∗p < 0.05.

• Download figure
• Open in new tab
• Download powerpoint

Figure 4

Change in normalized time to peak ejection rate from baseline at constant paced heart rate (100 beats/min) with cooling. Group 1. ∗∗p < 0.05.

Group 2 (HR80/120)

Since Ees appeared the most sensitive indicator of change in contractility in group 1, only this index was assessed in group 2. At 37°C, Ees increased with increasing HR (Fig. 5). Mean contractility at 120 beats/min was 205.9 ± 13.2% of baseline (p = 0.0021). However, at 33°C, this effect was reversed with a significant fall in Ees at an HR of 120 beats/min (mean 53.7 ± 13.2%) compared to an HR of 80 beats/min (100%, p = 0.0014) (Fig. 6). No consistent changes in the volume intercept, V₀, were demonstrated in either group.

• Download figure
• Open in new tab
• Download powerpoint

Figure 5

Change in normalized end systolic elastance with changing heart rate at 37°C. Group 2.

• Download figure
• Open in new tab
• Download powerpoint

Figure 6

Change in normalized end systolic elastance with changing heart rate at 33°C. Group 2.

Mechanism

Myocardial contractility is dependent on myosin and actin crossbridge formation and is sensitive to the concentration of calcium ([Ca²⁺]) prevailing within the cell. The Ca²⁺is dependent on both the duration of the action potential and the activity of adenosine triphosphate–dependent ion exchange pumps within the cell wall and the sarcoplasmic reticulum. Since enzymatic activity is directly related to temperature, the effect of cooling is therefore dependent on the relative thermal sensitivity of each enzymatic reaction step. Consequently the effects of hypothermia on systolic function cannot easily be predicted. If the negative effects of decreasing temperature are relatively greater on myosin/actin crossbridge formation, then contractility will fall. If hypothermia exerts a greater relative inhibitory effect on Ca²⁺ion exchange activity and causes a rise in intracellular Ca²⁺, this may lead to increased contractility. Although Ruf et al. (17)demonstrated that the reduction in actin/myosin crossbridge cycling rate in a human muscle preparation was directly proportional to temperature, Henderson et al. (4)showed that the force of contraction was increased with decreasing temperature down to 29°C in rat myofibrils, suggesting that Ca²⁺handling is the more temperature sensitive of the two processes. Variations in Ca²⁺handling with temperature, causing an increase in the intracellular level of Ca²⁺, have been demonstrated in vitro at the level of the Ca²⁺channel, in guinea pigs (6)and dogs (18), the Ca²⁺/Na⁺exchange pump in humans (19)and in the sarcoplasmic reticulum Ca²⁺pump, Ca²⁺/Na⁺exchanger, sarcolemmal Ca²⁺pump and the mitochondrial Ca²⁺uniporter in small mammals (20).

In addition, changes in Ca²⁺handling are likely to have an effect on the timings of the contractile cycle. This is because the rate of release and sequestration of Ca²⁺directly determine the rate of interaction of myosin and actin (21). As a result of this, hypothermia has been shown to cause a prolongation of the time to peak tension (TPT) (4,22).

Implications

An explanation that unites the effects of hypothermia on myosin/actin and Ca²⁺handling and explains the aforementioned apparently contradictory findings is that even if the rate of tension development (dP/dT) is decreased by hypothermia because of myosin/actin effects, then because the duration of contraction is increased (TPT or T_max), maximum inotropy may well be increased (22,23).

Hence, the increased inotropy seen with hypothermia in these earlier experiments may be due to an alteration in the duration of contraction. This positive inotropic effect is likely to be sensitive to changes in HR. In both load independent experiments (15,16)in which cooling increased contractility, HR was either allowed to vary (15)or the dog was paced at a normal resting HR (16). However, in a study performed in hypothermic (30°C) isolated rabbit hearts, Mattheussen et al. (5)showed that the increased contractility noted at low HR (30 beats/min) was reversed, and reduced contractility was noted when HR was increased to 90 beats/min by pacing.

Our results confirm the findings of these animal studies. The interaction may occur because of an increase in maximal elastance (E_max) and T_maxwith cooling. Hence although there may be an increase in E_max, this is at the expense of an increase in the time taken to reach maximal contraction. In the circumstances of a high HR, E_maxmay not be reached, resulting in a decrease in the amount of available work (13).

Clinical implications

The postoperative management of hypothermic cardiac surgery patients would appear to represent a compromise between HR and contractility. Care should be taken when increasing a hypothermic patient’s HR either through pacing or the use of chronotropic agents.

Experimental limitations

Pressure-volume analysis with the conductance catheter does not give a perfect measure of LV kinetics. This is due to, for example, problems with the nonlinearity of P-V relations at the extreme ends of the physiologic range (24,25)and the difficulty in measuring absolute volumes (9,26). However, it does represent the best technique currently available for human in vivo studies, particularly in the situation where patients act as their own controls. A potential limitation of this study was the fact that the patients were cooled on bypass. Although Wood et al. (27)have shown that cooling on bypass does not affect catecholamine levels in children, the period of time on bypass itself may have a detrimental effect on myocardial function. Knowledge of changes in cardiac function in the early stages of bypass is rather limited (28,29). However, data from Pavlides et al. (30)imply that little change occurs in cardiac function during CPB used to support coronary angioplasty. A further potential limitation relates to possible changes in parallel conductance with changing temperature. Although resistivity was checked prior to each contractility assessment, pre-set time constraints to ensure patient safety precluded repeated measurement of parallel conductance, which was assessed only at normothermia. It is therefore possible that the corrected calculation of absolute volume at lower temperature was inaccurate (Fig. 1). Nevertheless, we believe that the conclusions drawn remain valid, as Ees, the main measure of contractility used in this study, is determined by relative changes in volume rather than the absolute volume. The conductance catheter is thought to provide a more accurate determination of relative rather than absolute volume, even at normothermia (31).

In summary, our results confirm that contractility falls with temperature if HR is artificially maintained. Increasing HR at low temperature causes a fall in contractility. These findings are consistent with differential thermal sensitivities of the two prime elements of contractility: actin-myosin crossbridge formation and intracellular Ca²⁺handling. These effects may be clinically important in unstable postsurgical patients. Optimization of postoperative cardiac function in hypothermic patients would appear to represent a compromise between contractile status and HR.