Beta-blockade causes a reduction in the frequency spectrum of VF but improves resuscitation outcome: A potential limitation of quantitative waveform measures

Abstract

IntroductionMethods to identify appropriate treatments for the various stages of ventricular fibrillation (VF) involve differentiating groups of subjects who will respond to defibrillation with return of spontaneous circulation (ROSC) and those who require other therapies (e.g., CPR, drugs) prior to defibrillation. The use of quantitative waveform measures (QWM) which measure the frequency and fractal dimension of the VF electrocardiogram have shown success in predicting response to defibrillatory shock in animal models. Patients in cardiac arrest are often taking medications affecting adrenergic activity such as the beta blocker metoprolol and the combined alpha and beta blocker, labetalol. How this exposure might alter the QWM and ROSC rates is not known. HyothesisWe sought to determine how pretreatment with adrenergic agents alters two QWM measures, the amplitude spectrum area (AMSA) and the detrended fluctuation analysis (DFA). We also examined how these medications alter the probability of ROSC after shock. MethodsA swine model of ischemically induced VF cardiac arrest was used in which metoprolol and labetalol were administered prior to VF onset. 30 swine were randomly assigned to three groups of 10; control, metoprolol and labetalol. They were anesthetized, intubated and given the appropriate study drug. A balloon catheter was placed in the LAD coronary artery and inflated until VF occurred. ECG was recorded at 1000Hz for 7min of untreated VF. Closed chest compressions were then begun and after 1min a 200J shock was delivered. Resuscitation was continued with repeat defibrillation shocks as indicated for 15min or until ROSC was achieved (defined: systolic BP>60 for 10min). The Fourier frequency spectra, AMSA and DFA measures from VF onset to 7min were calculated using custom MATLAB routines. The QWM were compared over the electrical and circulatory phases of VF using generalized estimating equations. The rates of ROSC in the three groups were compared using relative risk measures. ResultsAll 10 control animals fibrillated after coronary occlusion, 8 metoprolol and 7 labetalol animals fibrillated.The frequency spectrum in metoprolol treated animals demonstrated a reduction in mean frequencies from 1 to 3min (electrical phase) and from 3 to 7min (circulatory phase). Labetalol produced an even greater reduction in frequencies in these intervals. The decline in AMSA was similar in all three groups over the first 3min. From 3 to 7min the metoprolol group was significantly lower than the control group (p<0.001) and the labetalol group was lower still (p<0.001). The DFA demonstrated little difference between the control and metoprolol groups, but showed a linear increase over 7min in the labetalol group (p<0.001 vs compared to control and metoprolol groups).ROSC was noted in 2/10 in the control group, 7/8 in metoprolol group and 2/7 in the labetalol group. The frequentist analysis of ROSC showed a relative risk (RR) of ROSC of 4.4 when comparing control to metoprolol animals and 1.4 comparing control to labetalol animals. DiscussionMetoprolol results in a reduction in frequencies in the Fourier spectrum of VF as compared with controls. There is a further decrease in frequencies with labetalol. The AMSA reflects this reduction in frequencies with lower AMSA values from 3 to 7min of VF. The DFA demonstrates consistent changes with labetalol treated animals over the 7min, but the metoprolol treated animals do not differ from the controls. The marked improvement in ROSC seen with metoprolol (RR 4.4) is unexpected and is not seen in labetalol treated animals. Adrenergic blockade prior to VF induction affects quantitative measures of the VF waveform and may limit the ability of such measures to predict downtime or defibrillation outcome.