The purpose of defibrillation is to rapidly restore blood flow and tissue perfusion following ventricular fibrillation (VF) and shock delivery. We tested the hypotheses that 1) a series of 1-ms pulses of various amplitudes delivered before the defibrillation shock can improve hemodynamics following the shock, and 2) this hemodynamic improvement is due to stimulation of cardiac or thoracic sympathetic nerves. Ten anesthetized pigs received a burst of either 15 or 30 1-ms pulses (0.1–10 A in strength) during VF, after which defibrillation was performed. ECG, arterial blood pressure, and left ventricular (LV) pressure were recorded. Defibrillation shocks and burst pulses were delivered from a right ventricular coil electrode to superior vena cava coil and left chest wall electrodes. Sympathetic blockade was induced with 1 mg/kg timolol and trials were repeated. The first half of this protocol was repeated in two animals that were pretreated with reserpine. Heart rate (HR) after 1-, 2-, 5-, and 10-A pulses was significantly higher than after control shocks without preceding pulse therapy. Mean and peak LV pressure measurements increased 38 and 72%, respectively, following shocks preceded by 5- and 10-A pulses compared with shocks preceded by no burst pulses. Mean and peak arterial pressures increased 36 and 43%, respectively, following shocks preceded by 5- and 10-A pulses compared with shocks preceded by no burst pulses. After β-blockade, HR, mean and peak arterial pressures, and mean LV pressure were not significantly different after pulses of any strength compared with control shocks. LV peak pressure following the 10-A pulses was significantly higher than with no burst pulses but was significantly lower than the response to the 10-A pulses delivered without β-blockade. HR, mean and peak arterial pressures, and mean and peak LV pressure responses after 15 or 30 5- or 10-A pulses were similar to the responses to the same pulses after β-blockade. We conclude that a burst of 15–30 1-ms pulses delivered during VF can increase HR, arterial pressure, and LV pressure following defibrillation. β-Blockade or reserpine pretreatment prevents most of this postshock increase in HR, arterial pressure, and LV pressure.
- ventricular fibrillation
the goal of defibrillation is to rapidly restore blood flow and tissue perfusion after an episode of ventricular fibrillation (VF) by restoring a regular rhythm. Both the shock itself (19) as well as the preceding VF (1, 14) can contribute to depressed myocardial function following successful defibrillation. If VF continues long enough [∼120 s in dogs (18)], pulseless electrical activity can result and the subject will die if hemodynamic support is not started. The mechanism for this depressed myocardial function is not well understood.
Electrical therapy can be used to improve cardiac contractility in several ways. Postextrasystolic potentiation causes the contraction that follows a premature stimulus to be augmented compared with baseline or the contraction associated with the premature stimulus (6). Field stimulation that alters the transmembrane potential without initiating a new action potential can either increase or decrease cardiac contractility depending on the timing and polarization of the pulse (20, 21). The activation sequence of patients with congestive heart failure and a prolonged QRS complex can be altered by pacing from one or more sites to increase the synchrony of contraction and improve cardiac function (2, 10, 13).
The autonomic nervous system can be stimulated to either increase or decrease cardiac contractility (15). Postganglionic fibers in sympathetic cardiac nerves are distributed over the heart (8). Extracardiac stimulation of these nerves increases both heart rate (HR) and the strength of cardiac contraction (16). Field stimulation of intrinsic cardiac nerves in isolated myocardium also can increase cardiac contractility (3).
We tested the hypotheses that 1) a series of 1-ms pulses of various amplitudes delivered during fibrillation before defibrillation can improve hemodynamics following the shock, and 2) this improvement in hemodynamics following the shock is due to stimulation of cardiac or thoracic sympathetic nerves. To test these hypotheses, we examined the ability of these pulse trains to change hemodynamics following a defibrillation shock before and after β-blockade and following reserpine pretreatment.
All procedures were approved by the Institutional Animal Care and Use Committee at the University of Alabama (Birmingham, AL). Furthermore, all preoperative and operative care for animals complied with section 6 of the Animal Welfare Act of 1989 and adhered to the principles outlined in the National Institutes of Health Guide for the Care and Use of Laboratory Animals (NIH Pub. No. 85-23).
Ten pigs (body wt 25–30 kg) were premedicated with Telazol (4.4 mg/kg), xylazine (2.2 mg/kg), and atropine (0.04 mg/kg). Animals were intubated with a cuffed endotracheal tube and were ventilated at a rate of 10–15 ml·kg–1·min–1. Anesthesia was maintained using 1–3% isoflurane. A 7-Fr hemostatic sheath was placed in the left femoral artery, and an 8-Fr hemostatic sheath was placed in the right femoral artery and left jugular vein. Both the left and right jugular veins were isolated. A 5-Fr high-fidelity pressure catheter (Millar Instruments) was advanced in a retrograde fashion across the aortic valve into the left ventricle. A 3-Fr high-fidelity pressure catheter was placed via a femoral sheath into the abdominal aorta. A transvenous defibrillation catheter (Endotak model DSP, Guidant) with a 6-cm distal coil, a 7-cm proximal coil, and a 12-cm interelectrode spacing was advanced into the right ventricular (RV) apex. A defibrillation can electrode was placed under the skin of the anterior left chest wall overlying the second intercostal space. The distal defibrillation electrode was the anode for all burst stimulation and the first phase of the biphasic defibrillation shock. The proximal defibrillation catheter electrode connected to the can electrode served as the cathode.
A defibrillation threshold was determined using a three-reversal up-down protocol using a log(0.1)-J step size (11). VF was induced using 60-Hz current applied to the RV apex. Defibrillation shocks were delivered after 10 s of VF. An external cardiovertor-defibrillator (VENTAK ECD, Guidant) was used to deliver all defibrillation shocks. The defibrillator was triggered by a pulse generator (WPI). All subsequent defibrillation shocks were delivered at ∼1.5× the energy of the measured threshold. If a shock failed to defibrillate the heart, the trial was discarded and repeated.
The animal was challenged with 10 μg·kg–1·min–1 dobutamine, a β-agonist (17), for 4 min. HR and arterial blood pressure were recorded before and at the end of the dobutamine infusion. After the dobutamine challenge, a 20-min washout period was observed before any testing was performed.
Pulse therapy. Before defibrillation-shock delivery, a series of therapy pulses were delivered using a linear power amplifier (Guidant). Burst stimulation was delivered through the defibrillation electrodes. Burst stimulation pulses were 1 ms in duration. Six pulse strengths were used: 0.1, 0.3, 1, 2, 5, and 10 A. Either 15 or 30 pulses were delivered. Time between pulses was 30 ms for 0.1- and 0.3-A pulses and 10 ms for all others. A 3-s delay was allowed between the burst therapy and the defibrillation shock. A total of 12 different pulse therapies was tested. Trial order was randomized across each animal. A control trial with no burst stimulation before defibrillation was performed before any burst-therapy trials and after every fourth therapy trial.
β-Blockade. After the initial 16-burst stimulation and control trials, the animal was given timolol (1 mg/kg). After 10 min, the dobutamine challenge was repeated. The dobutamine challenge was given to test the effectiveness of β-blockade. The initial dobutamine challenge was given as a control. After a 20-min washout period for the dobutamine, the 12 burst-therapy trials and 4 control trials were repeated.
At the end of the study, the animal was killed with KCl while still anesthetized. The heart was removed, weighed, and examined grossly.
Data collection and analysis. Lead II ECG and the two pressure signals were collected with an eight-channel data-acquisition system (Windaq, Dataq) at a minimum sampling rate of 500 samples/s. Data were analyzed using Matlab (MathWorks). The investigator performing the analysis was blinded to the amplitude of the burst stimulation at the time of analysis. Values are given as means ± SD.
Repeated-measures ANOVA was used to compare mean values (SPSS). When the repeated-measures ANOVA showed a significant difference, a Student-Newman-Keuls post hoc test was used to determine differences between the response to each burst-stimulation level and the response during control episodes without burst stimulation. A P value ≤0.05 was considered significant.
In two additional animals, reserpine (1 mg/kg) was given 24 h before the study to chemically sympathectomize them. The animals subsequently underwent the control portion of the protocol for part I including defibrillation-threshold determination, dobutamine challenge, and pulse-therapy testing. After pulse-therapy testing, the animals were challenged with tyramine (1 mg/kg) to test the effectiveness of the chemical sympathectomy.
Animal body weight was 34 ± 3 kg, and heart weight was 160 ± 20 g. Mean defibrillation threshold was 488 ± 55 V. Mean defibrillation shock strength delivered during pulse-therapy testing was 732 ± 89 V. VF was induced an average of 44 times per animal. Baseline hemodynamics did not change significantly over the duration of each half of the protocol. HR values and hemodynamic responses to the four control episodes in the first half of the study before the animal received timolol were not significantly different from one another. Gross pathological examination at the end of each study showed no unusual changes in any of the hearts studied.
Pulse therapy before β-blockade. For control shocks, HR did not change significantly over the 30 s following the shock from the baseline value of 110 ± 17 beats/min (Fig. 1). For shocks preceded by pulse therapy, HR was significantly higher following the 1-, 2-, 5-, and 10-A pulses at 1, 10, and 20 s after the shock compared to baseline but was not different from baseline at 30 s. There was a trend toward a greater increase in HR with 30 pulses than with 15 pulses, but this difference did not reach significance.
For control shocks, systolic left ventricular (LV) pressure decreased from 130 ± 20 mmHg at baseline to 51 ± 10 mmHg at 1 s after the shock. By 30 s, systolic LV pressure had increased to 91 ± 12 mmHg, which was still significantly different than the baseline pressure (Fig. 2). For shocks with pulse therapy, systolic LV pressure was significantly higher for pulse strengths of 2, 5, and 10 A than for the control shocks. At 30 s after the shock, systolic LV pressure was still significantly higher than for control shocks. Again there was a trend toward a greater increase in systolic LV pressure for 30 than for 15 therapy pulses, but this did not reach significance.
For control shocks, LV mean pressure dropped from 42 ± 9 at baseline to 18 ± 5 mmHg at 1 s following the shock (Fig. 2). Mean LV pressure was significantly higher for either 15 or 30 pulses at pulse strengths of 2, 5, and 10 A than for control episodes. LV mean pressure did not significantly increase with 30 pulses compared with 15 pulses.
For control shocks, systolic arterial pressure decreased from 115 ± 17 at baseline to 51 ± 14 mmHg at 1 s following the shock. By 30 s, systolic arterial pressure had increased to 87 ± 10 mmHg (Fig. 3). Systolic arterial blood pressure was significantly increased over baseline for 2-, 5-, and 10-A pulses at 1, 10, 20, and 30 s. There was no significant difference in the systolic arterial pressure response to 15 and 30 pulses.
For control shocks, mean arterial pressure decreased from 84 ± 14 to 45 ± 16 mmHg at 1 s following the shock. By 30 s, mean arterial pressure had increased to 73 ± 13 mmHg (Fig. 3). For shocks with pulse therapy, mean arterial blood pressure was significantly higher for either 15 or 30 pulses for pulse strengths of 5 and 10 A than for control episodes at 1, 10, 20, and 30 s. There was no significant difference in the response of mean arterial pressure to 15 and 30 pulses.
Dobutamine test. Before any burst-stimulation testing, dobutamine changed HR from 110 ± 12 to 134 ± 18 beats/min and systolic blood pressure from 112 ± 24 to 158 ± 24 mmHg (P ≤ 0.05). After β-blocker administration, dobutamine changed HR from 107 ± 17 to 111 ± 17 mmHg and systolic blood pressure from 106 ± 18 to 113 ± 22 mmHg [P = not significant (NS)].
Pulses during β-blockade. β-Blockade caused a significant decrease in HR from 114 ± 17 to 107 ± 12 beats/min. Systolic blood pressure was decreased from 115 ± 17 to 100 ± 19 mmHg, and mean arterial blood pressure was lowered from 86 ± 13 to 70 ± 15 mmHg. Systolic LV pressure was decreased from 132 ± 21 to 116 ± 16 mmHg, and mean LV pressure was lowered from 43 ± 9to36 ± 8 mmHg. All of these changes were significant. HR and hemodynamic response to the four control episodes in the second half of the study after the animal received timolol were not significantly different from one another.
After β-blockade, HR measurements 1, 10, 20, or 30 s following defibrillation for control episodes did not change significantly from the 106 ± 10 beats/min value before fibrillation induction (baseline) or 1, 10, 20, or 30 s following the shock. None of the pulse therapy changed HR from the control values (Fig. 4).
For control episodes, systolic LV pressure decreased from a baseline of 116 ± 16 to 66 ± 15 mmHg at 1 s following the shock. By 30 s, systolic LV pressure had increased to 113 ± 17 mmHg (Fig. 5). Only the 10-A pulses increased systolic LV pressure above control shock values. There was no significant difference in the response to 15 and 30 pulses.
For control episodes, mean LV pressure decreased from a baseline value of 42 ± 9 to 33 ± 10 mmHg at 1 s following the shock. By 30 s, mean LV pressure had increased to 38 ± 10 mmHg (Fig. 5). None of the pulse therapies changed mean LV pressure significantly differently than for control episodes.
For control episodes, systolic arterial pressure dropped from 98 ± 15 at baseline to 58 ± 15 mmHg at 1 s after the shock. By 30 s, systolic blood pressure had returned to 98 ± 15 mmHg (Fig. 6). None of the pulse therapies changed systolic arterial pressure significantly differently than for control episodes.
For control episodes, mean arterial pressure decreased from 84 ± 13 to 45 ± 16 mmHg at 1 s following the shock. By 30 s following the shock, mean arterial pressure had returned to 73 ± 15 mmHg (Fig. 6). None of the pulse therapies changed mean arterial pressure significantly from the control episode values.
For the two animals that received reserpine, dobutamine increased HR by 40% and systolic blood pressure by 80%. The defibrillation threshold was 452 ± 113.
We limited our analysis to control episodes and either 15 or 30 pulses of 5 or 10 A. These burst-stimulation combinations showed the greatest changes in HR and hemodynamics in part I before timolol administration. HR increased ∼10% with pulse therapy compared with control episodes in the 30 s following the shock. Systolic arterial pressure, mean arterial pressure, systolic LV pressures, and mean LV pressure also rose ∼10% with pulse therapy compared with controls. These results are similar to those observed when timolol was given to the animal in part I before pulse therapy at these pulse strengths and numbers.
The findings in this paper are twofold. First, pulse stimulation that is too weak to defibrillate, when given during VF before the defibrillation shock, increases HR, arterial pressure, and LV pressure following successful defibrillation compared to that following defibrillation without prior stimulation. Second, the effect on postshock cardiac hemodynamics can be blocked by β-blockers or by reserpine pretreatment, which suggests that the pulses delivered during VF stimulate the sympathetic nervous system.
This study showed that after 10 s of VF and defibrillation with a shock at a strength 1.5 times the defibrillation threshold, LV and arterial pressures dropped transiently followed by recovery that was still not complete by 30 s after defibrillation. This decrease in hemodynamics following the shock is thought to be primarily due to the episode of VF rather than to the shock itself. Panegrau and Abboud (12) showed that similar to our results, following 15–30 s of fibrillation and a 400 W-s capacitor discharge defibrillator shock delivered to the chest wall, HR and mean arterial pressure were significantly lower than at baseline. Over the next 2–3 min, HR returned to baseline. However, at 1 min following the shock, mean arterial pressure had overshot baseline. It subsequently returned to baseline over the next 2–3 min. In contrast, when the same shock was delivered to the chest wall during sinus rhythm, changes in hemodynamics were small and not statistically significant with the exception of minimal reductions in mean arterial pressure. Kerber et al. (9) showed that there was no significant change in HR and aortic mean pressure following shocks of up to 100 J delivered to the epicardial surface or following shocks up to 460 J delivered to the chest wall during sinus rhythm. When shocks of the same strength were delivered following 10–15 s of fibrillation, HR and mean arterial pressure transiently decreased. Park et al. (14) have shown in humans that there is a negative logarithmic relationship between the duration of VF and the return of systolic arterial pressure following defibrillation. Longer periods of VF were followed by longer periods of arterial pressure recovery.
This study shows that the decrease in cardiac hemodynamics can be moderated or reversed by delivering a series of short pulses during VF before the defibrillation shock. Furthermore, the change in postshock HR and hemodynamics caused by the pulse therapy is abolished either by sympathetic blockade by acute timolol treatment or chemical sympathectomy by pretreatment with reserpine. It is well known that sympathetic nervous system stimulation can alter cardiac contractility and hemodynamics. Postganglionic sympathetic fibers arise from the dorsal sympathetic trunk and travel to the heart and other thoracic viscera through cardiac nerves (7). Direct electrical stimulation of the thoracic anterior roots T1-T5 can elicit large changes in blood pressure, LV pressure, and HR (15). Stimulation of left-sided cardiac efferent nerves augments ventricular contractile force, LV pressure, and arterial pressure with minor changes in HR (5). Right-sided stimulation tends to cause a relatively greater increase in HR (8). Our results showed increases in both HR and LV and arterial pressures, which suggests that we stimulated all or most of the sympathetic nervous system in the heart.
The pulse train that we used was very similar to pulse trains that have been used to increase cardiac contractility in isolated tissue preparations. Blinks (3) has shown that field stimulation of isolated tissue with “strong pulses of electric current” increased cardiac contractility. Isolated tissue was field stimulated at a field strength of ∼5 V/cm with 1- to 2-ms pulses spaced 10-ms apart during the refractory period. Graded responses in contractility were produced by delivering various numbers of pulses per unit time to muscles paced at regular intervals. Similar to the study presented here, Blinks showed that either bathing the isolated tissue with propanolol or taking the tissue from animals which had previously been treated with reserpine eliminated the increase in contractility induced by the field stimulation. Brady et al. (4) performed similar studies and showed similar results. They showed that 0.5-ms pulses were “optimal” for increasing cardiac contraction and that after stimulation, the potentiation of contraction approached a maximum with a rate constant of ∼10 s.
Although several studies have shown that autonomic tone can modulate the defibrillation threshold, we are not aware of any previous study that has examined the effects of β-blockade on hemodynamics following the shock. This study showed that a β-blocker did not change the hemodynamic response following defibrillation compared to the response without β-blockade.
Limitations. There are several limitations to our study. Testing in the presence of β-blockade always occurred after testing without β-blockade. This was necessary because of the half-life of the β-blocker timolol. The fact that the hemodynamic responses during control episodes without β-blocker were not significantly different from one another and the hemodynamic responses during control episodes with β-blocker were not significantly different suggests that the order of testing did not affect the results. Because the high dose of timolol in this study probably completely blocked the β-receptors, further studies are needed to determine the effect of pulse therapy under partial β-blockade. Also, we chose to use the nonselective β-blocker timolol rather than a more-selective β-blocker to best test our hypothesis that sympathetic nerve stimulation was the mechanism through which the burst stimulation was increasing arterial and LV pressures. Further studies are necessary to determine the effects of the burst stimulation following defibrillation in more clinically relevant situations. We did not vary pulse duration and timing between pulses. As noted above, the values for pulse duration and timing between pulses were near the optimal values determined in isolated tissue. Additional studies are needed to determine the optimal pulse duration and timing during VF. Furthermore, we only studied HR and hemodynamic changes soon after the defibrillation shock. Studies are needed to determine the long-term effects on cardiac function of sympathetic surge following defibrillation.
Testing is also needed to determine the proarrhythmic effects of burst stimulation if the burst is inadvertently delivered during a nonshockable rhythm rather than during VF. Thirty pulses extend over a period of 300 ms. If these pulses are delivered during sinus rhythm, it is highly likely that one or more of the pulses will be delivered during the T wave and induce VF. This outcome is somewhat mitigated, because the burst therapy would be delivered by an implantable defibrillator and so if VF was inadvertently induced, it could be quickly reversed. Studies are also needed to determine the effects of burst stimulation on the defibrillation threshold.
This work is supported in part by National Heart, Lung, and Blood Institute Grants HL-67961, HL-63775, and HL-42760 and grants from Guidant and Medtronic Physio-Control.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2003 by the American Physiological Society