Author + information
- Received January 14, 1997
- Revision received June 24, 1997
- Accepted June 26, 1997
- Published online October 1, 1997.
- David M Clark, MDA,
- Vance J Plumb, MD, FACCA,
- Andrew E Epstein, MD, FACCA and
- G.Neal Kay, MD, FACCA,* ()
- ↵*Dr. G. Neal Kay, Division of Cardiovascular Disease, Department of Medicine, 321J Tinsley Harrison Tower, University of Alabama at Birmingham, Birmingham, Alabama 35294.
Objectives. The aim of this study was to determine the independent hemodynamic effects of an irregular sequence of ventricular cycle lengths in patients with atrial fibrillation (AF).
Background. Atrial fibrillation may reduce cardiac output by several possible mechanisms, including loss of the atrial contribution to left ventricular filling, valvular regurgitation, increased ventricular rate or irregular RR intervals. This study was designed to evaluate the effects of an irregular RR interval, independent of the average ventricular rate, on cardiac hemodynamic data during AF.
Methods. Sixteen patients with AF were studied invasively. During intrinsically conducted AF (mean rate 102 ± 22 beats/min), the right ventricular apex electrogram was recorded onto frequency-modulated (FM) tape. After atrioventricular node ablation, the right ventricular apex was stimulated in three pacing modes in randomized sequence: 1) VVI at 60 beats/min; 2) VVI at the same average rate as during intrinsically conducted AF (102 ± 22 beats/min); and 3) during VVT pacing in which the pacemaker was triggered by playback of the FM tape recording of the right ventricular apex electrogram previously recorded during intrinsically conducted AF (VVT 102 ± 22 beats/min).
Results. Compared with VVI pacing at the same average rate, an irregular sequence of RR intervals decreased cardiac output (4.4 ± 1.6 vs. 5.2 ± 2.4 liters/min, p < 0.01), increased pulmonary capillary wedge pressure (17 ± 7 vs. 14 ± 6 mm Hg, p < 0.002) and increased right atrial pressure (10 ± 6 vs. 8 ± 4 mm Hg, p < 0.05).
Conclusions. An irregular sequence of RR intervals produces adverse hemodynamic consequences that are independent of heart rate.
Atrial fibrillation (AT) may have several detrimental effects on cardiac hemodynamic data [1–3], including loss of the atrial contribution to ventricular filling with a reduction in end-diastolic pressure and volume in the left and right ventricles , an increase in the mean diastolic pressure in the atria , a reduced interval for passive diastolic filling [5–7]and possible atrioventricular valvular regurgitation . The relative contribution of these factors is likely to vary between patients, and may differ with changes in ventricular rate. Although much is known regarding these determinants of cardiac function, very little is known about the independent contribution that irregularity in the ventricular cycle length has on hemodynamic data. Naito et al. reported that cardiac output was reduced with an irregular as compared with a regular paced ventricular rhythm in dogs with AF. Whether similar findings can be demonstrated in patients with AF has not been reported. Although catheter ablative techniques modify the rate and rhythm for patients with rapid ventricular rates during AF [8–13], the independent effects of slowing or regularizing the ventricular rate on resultant hemodynamic data are unknown. The present study was designed to test the hypothesis that an irregular sequence of ventricular cycle lengths results in hemodynamic deterioration as compared with a regular ventricular rhythm at the identical average rate in patients with AF.
1.1 Study group.
The study group consisted of 16 highly symptomatic patients with paroxysmal or chronic AF who were referred for catheter ablation of the atrioventricular conduction system and permanent pacemaker implantation. Informed, written consent for the study protocol, which had been approved by the Institutional Review Board for Research Involving Human Subjects at the University of Alabama at Birmingham, was obtained from all subjects. Exclusion criteria included failure to obtain informed consent, uncompensated congestive heart failure, unstable angina pectoris or recent myocardial infarction.
1.2 Study protocol.
Patients were studied in the fasting state while sedated with intravenous meperidine and midazolam. A permanent pacemaker was implanted in all subjects through the cephalic or subclavian veins using standard techniques. After permanent pacemaker implantation, a 6F quadripolar catheter was advanced from the right femoral vein and positioned at the right ventricular apex, and a 7F deflectable-tip catheter was advanced from the femoral vein to the right atrial (RA) appendage under fluoroscopic guidance. A 7.5F pulmonary artery catheter capable of continuously measuring oxyhemoglobin saturation by three wavelengths of light (Opticath, Abbot Laboratories) was positioned in the right pulmonary artery. The oximetric catheter was calibrated before insertion and used for continuous on-line recording of the pulmonary artery oxyhemoglobin saturation (Svo2) (Oximetrix-3 SO2/CO, Abbot Laboratories). The pulmonary artery pressure, pulmonary capillary wedge pressure (PCWP) and RA pressure were continuously monitored. Systemic arterial oxyhemoglobin saturation (Sao2) was recorded with a continuous transcutaneous oxygen saturation monitor (Nellcor). Systemic blood pressure was measured with an automatic arm sphygmomanometer rather than invasively to minimize the risks associated with the study protocol. Oxygen consumption (V̇o2) was measured continuously with a Medical Graphics CPX Metabolic Cart (Medical Graphics Corporation) using a vacuum-controlled plastic bubble. Hemoglobin concentration was measured immediately before the study. The Sao2and Svo2measurements were used in conjunction with the measured V̇o2and hemoglobin concentration to calculate cardiac output using the Fick principle .
1.3 Experimental conditions.
For all subjects with sinus rhythm at the beginning of the study, cardiac output, systemic arterial pressure, pulmonary artery pressure, PCWP, RA pressure, pulmonary and systemic vascular resistance, arterial and pulmonary artery oxygen content and V̇o2were measured with the patient in a resting and sedated state. After recording of baseline hemodynamic data, AF was induced with burst atrial pacing from the RA appendage in those patients initially in sinus rhythm. The hemodynamic data were allowed to stabilize for at least 5 min during continuous AF. During at least 5 min of AF, a bipolar electrogram was recorded from the right ventricular apex catheter onto frequency-modulated (FM) tape using an analog data recorder (model 71, TEAC, Japan). The mean ventricular cycle length during intrinsically conducted AF was determined from the recorded right ventricular electrograms.
The atrioventricular conduction system was then ablated with radiofrequency current in all subjects using standard techniques [8, 15]. After ablation, hemodynamic data were measured with the permanent pacemaker programmed to each of three experimental conditions in random sequence with a balanced block design: 1) ventricular demand (VVI) pacing from the right ventricular apex at 60 beats/min (a regular ventricular cycle length of 1,000 ms); 2) VVI pacing from the right ventricular apex at the identical mean ventricular rate as recorded during intrinsically conducted AF (a regular, shorter cycle length); and 3) ventricular triggered (VVT) pacing in which the right ventricle was paced by the permanent pacemaker triggered by playback of the FM tape recording that had been made during intrinsically conducted AF before ablation (an irregular sequence of cycle lengths). Thus, during VVT pacing, the right ventricle was paced at the identical sequence of cycle lengths as observed during intrinsically conducted AF.
1.4 Echocardiographic data.
An echocardiogram (ATL Ultramark 9) was recorded during each of the five experimental conditions from the parasternal long-axis, four-chamber long axis and apical two-chamber long-axis views to qualitatively assess mitral and tricuspid valvular regurgitation. The severity of valvular regurgitation was assessed by color flow Doppler imaging by averaging the degree of regurgitation over 10 consecutive cycles [16, 17]. Left atrial size, left ventricular end-diastolic dimension, left ventricular septal wall thickness and left ventricular ejection fraction (measured by the Simpson rule) were measured at the beginning of the study. The left ventricular ejection fraction was averaged over 10 consecutive beats to account for beat to beat variation. Patients in whom adequate windows could not be obtained were excluded from echocardiographic analysis. Images were reviewed off-line by a reviewer who had no knowledge of the clinical data and the hypothesis of the study.
1.5 Data analysis.
Continuous variables were expressed as mean value ± SD. Comparisons of each hemodynamic variable across the experimental conditions were made with one-way analysis of variance with the Bonferroni correction for multiple comparisons. Paired hemodynamic data comparing regular and irregular ventricular pacing at the same average rate were analyzed using the paired Student ttests.
2.1 Study group.
The demographic data of the study group are shown in Table 1. There were 11 women and 5 men with medically refractory AF (paroxysmal in 8 patients and chronic in 8 patients). The patients’ mean age was 66.1 ± 9.1 years (range 43 to 81). The average number of failed antiarrhythmic drugs used before catheter ablation was 3.6 ± 1.5. All of the patients with paroxysmal AF were in sinus rhythm at the onset of the study protocol. The hemodynamic data for each of the five experimental conditions of the study protocol are summarized in Table 2.
2.2 Hemodynamic effects of a regular versus an irregular ventricular rhythm.
Fig. 1demonstrates representative electrocardiographic and pulmonary artery pressure tracings from Patient 2 during intrinsically conducted AF (top panel) and during VVT pacing with the pacemaker triggered by playback of the right ventricular electrogram recorded during AF (bottom panel). The tracings were matched with respect to time and show the identical sequence of ventricular cycle lengths. In this example, post-extrasystolic potentiation in the pulmonary artery pressure tracing appears to be more marked during VVT tracing than during intrinsically conducted AF. This was not a consistent finding, however. Fig. 2demonstrates similar tracings from the same individual during VVI pacing at the same average ventricular rate as recorded during intrinsically conducted AF (top panel) and at a cycle length of 1,000 ms (bottom panel).
The cardiac output was significantly greater during regular VVI pacing than during VVT pacing triggered by playback of the intrinsically conducted ventricular electrogram recorded in AF despite an identical average cycle length (5.2 ± 2.4 vs. 4.4 ± 1.6 liters/min, p < 0.01) (Fig. 3A). Twelve of the 16 patients had a reduction in cardiac output during VVT pacing as compared with VVI pacing at the identical average cycle length (587 ms) (Fig. 3B). Pulmonary capillary wedge pressure was lower during normal sinus rhythm than during the other four experimental conditions (Fig. 4A). Pulmonary capillary wedge pressure was also lower during regular VVI pacing than during irregular VVT pacing at the identical average ventricular cycle length (14 ± 6.5 vs. 17 ± 7.0 mm Hg, p = 0.002). In 14 of 16 patients, PCWP was lower with regular VVI pacing than with an irregular paced rhythm at the same average cycle length (Fig. 4B). The RA pressure during each experimental condition is shown in Fig. 5A. The RA pressure was significantly higher during irregular VVT pacing than during regular VVI pacing at the same rate (9.6 ± 6.1 vs. 7.6 ± 3.6 mm Hg, p = 0.04) (Fig. 5B). The highest Svo2was observed during sinus rhythm, whereas the lowest was observed during VVT pacing at the identical ventricular cycle length sequence as recorded during AF (Fig. 6A). The Svo2was higher during regular VVI pacing than during irregular VVT pacing at the identical average rate in 12 of 16 patients, with no difference in 2 patients (Fig. 6B).
The hemodynamic data during intrinsically conducted AF were generally superior to those during VVT pacing. The mean cardiac output was 5.4 ± 2.4 liters/min during AF as compared with 4.4 ± 1.6 liters/min with VVT pacing (p < 0.01). Likewise, PCWP was lower during AF compared with VVT pacing (15 ± 5 mm Hg vs. 17 ± 7 mm Hg, p < 0.02).
2.3 Echocardiographic observations.
In six patients, echocardiography was not performed because of poor acoustic windows. The mean baseline left ventricular ejection fraction was 53 ± 13% (n = 10). The mean left ventricular end-diastolic dimension was 42 ± 8.3 mm, with a posterior wall thickness of 10.1 ± 2.1 mm. The mean left atrial dimension was 36 ± 3.3 mm. The degree of mitral regurgitation was graded by analysis of the color Doppler recordings as 0 or 1 in all patients during intrinsically conducted AF. The severity of the mitral regurgitation increased in only one patient (from grade 1 to grade 2) with a change from regular VVI pacing to irregular VVT pacing at the identical rate, with no significant change in the other subjects.
3.1 Independent effects of an irregular ventricular rate on hemodynamic data.
The hemodynamic consequences of AF have been actively studied since the first description of this arrhythmia by Jolly and Ritchie in 1910 . Multiple mechanisms have been proposed to explain the adverse hemodynamic consequences during AF, including an increase in heart rate, loss of atrioventricular synchrony, irregularity in the ventricular rhythm, valvular regurgitation and neurohormonal effects [1–3]. A previous canine study suggested that an irregular ventricular rhythm had an adverse effect on cardiac output that was independent of the ventricular rate . To our knowledge, the present study is the first prospective human experiment that has examined the independent effects of heart rate irregularity on cardiac function in patients with AF. Consistent with canine studies, an irregular ventricular rhythm does indeed have adverse hemodynamic consequences that are independent of rate.
3.2 Previous studies.
Lau et al. reported that the pulmonary artery pressure and PCWP during AF are increased as compared with atrial pacing at the same average rate in 10 patients with paroxysmal AF . The present study also demonstrated that sinus rhythm was associated with the highest cardiac output of any of the five experimental conditions, highlighting the importance of atrioventricular synchrony.
Greenfield et al. reported a weak but positive correlation between the immediately preceding RR interval and stroke volume in patients with chronic AF . The duration of left ventricular ejection was directly proportional to the preceding RR interval, whereas the pre-ejection period was shorter and the stroke volume larger with a longer preceding cycle length. A negative correlation was found between the stroke volume of a particular beat and the second preceding beat. The relation between cycle length and stroke volume was significantly stronger when the two preceding RR intervals were considered. A pulse deficit—beats that did not open the aortic valve—was recorded with either a very short cycle or a short cycle preceded by a long cycle.
The mechanisms responsible for the reduction in cardiac output during an irregular sequence of paced cycle lengths as compared with regular pacing at the same average rate are incompletely understood. However, the Starling mechanism relating myofiber length to the strength of ventricular contraction and the force–interval relation are two likely mechanisms. In the absence of valvular obstruction, passive filling of the left ventricle occurs during the first half of diastole at normal rest heart rates. Previous studies have suggested that lengthening of the diastolic interval beyond ∼700 ms does not appreciably increase the rest left ventricular end-diastolic volume, whereas shortening of the diastolic interval to <500 ms impairs left ventricular filling and stroke volume .
Gosselink et al. have demonstrated that the left ventricular ejection fraction is inversely related to the preceding RR interval during AF by the use of radionuclide techniques. As expected, the left ventricular ejection fraction was directly related to the end-diastolic volume. For any given end-diastolic volume, the longer the preceding RR interval, the greater the left ventricular ejection fraction. This relation was made more complex if the two preceding RR intervals were considered. For example, the influence of the penultimate RR interval (the pre-preceding interval) on the left ventricular ejection fraction was opposite that of the immediately preceding interval. In other words, a long pre-preceding RR interval has an effect to decrease the left ventricular ejection fraction. A short pre-preceding interval augments the stroke volume of the second successive beat. This observation suggests that post-extrasystolic potentiation is one likely mechanism for the effect of a short pre-preceding interval on left ventricular systolic performance.
However, Herbert demonstrated in a retrospective review of hemodynamic data in patients with chronic AF that the cardiac index was inversely correlated with the RR variability produced by short cycles but not with the standard deviation of all RR intervals. Thus, short RR intervals decreased cardiac output more than long RR intervals increased cardiac output. The correlation between the percentage of short cycle lengths and a decrease in cardiac index was more marked for patients with an average ventricular rate >75 beats/min than for those with a slower average rate. Therefore, a short RR interval appears to have a negative inotropic effect that is independent of ventricular filling. Radionuclide studies also suggest that left ventricular function in AF is dependent on multiple preceding cycle lengths. Short RR intervals have been demonstrated to produce mechanical restitution, where impaired diastolic filling (and decreased inotropy) affect the next beat by reducing both end-diastolic pressure and volume. In contrast, long preceding RR intervals allow for potentiation in which the enhanced filling and release of calcium from the sarcoplasmic reticulum augment the stroke volume of the subsequent contraction . However, the beneficial effects of potentiation decay over each successive beat, such that a long interval may not balance the adverse hemodynamic consequences of a short interval. Because of the persistent adverse effects of short RR intervals on ventricular performance, a regular RR interval (in which short intervals are completely eliminated) produces a higher average stroke volume and cardiac output than an irregular ventricular rhythm at the same average rate. This may be more true for patients with higher average heart rates who are more likely to be considered for atrioventricular node ablation than for patients with well controlled ventricular rates.
In addition to examining the effects of ventricular irregularity, the present study demonstrates that ventricular activation by the normal His-Purkinje system provides superior hemodynamic effects to stimulation at the right ventricular apex. These observations may lend support to the utility of pacing from a site near the His bundle.
3.3 Study limitations.
There are several limitations of this study that should be emphasized. First, this study compared the hemodynamic data of an irregular to a regular paced ventricular rhythm. Because pacing from the right ventricular apex produced adverse hemodynamic effects of its own, it is uncertain whether these results also apply to a narrow QRS complex or to pacing at an alternative site (such as the His bundle or interventricular septum). Nevertheless, the great care taken to exactly reproduce the sequence of ventricular cycle lengths during intrinsically conducted AF suggests that these results may be relevant to patients with AF. Second, although the sequence of pacing modes was randomized after atrioventricular node ablation, comparison with the preablation hemodynamic data could not be randomized. Third, these results may not apply to patients with more severely impaired left ventricular systolic or diastolic function. Finally, these results were observed during an acute hemodynamic study and may not apply to more long-term conditions in which adaptive mechanisms may reduce these effects. The differences in the hemodynamic response to AF in patients with established AF and those patients with paroxysmal AF were not studied.
An irregular sequence of ventricular cycle lengths reduces cardiac output and increases PCWP as compared with a regular rhythm at the same average rate.
We thank Consuella Mays and Gilbert Perry, MD for technical expertise in acquiring and analyzing echocardiograms.
☆ This study was supported in part by a grant from Medtronic, Inc., Minneapolis, Minnesota. Dr. Clark is the recipient of an Institutional National Service Award (T32-HL-07703) from the Heart, Lung, and Blood Institute, National Institutes of Health, Bethesda, Maryland.
- atrial fibrillation
- frequency modulated
- pulmonary capillary wedge pressure
- right atrium, right atrial
- systemic arterial oxyhemoglobin saturation
- pulmonary artery oxyhemoglobin saturation
- oxygen consumption
- ventricular demand pacing
- ventricular triggered pacing
- Received January 14, 1997.
- Revision received June 24, 1997.
- Accepted June 26, 1997.
- The American College of Cardiology
- Morris JJ,
- Entman M,
- North WC,
- Kong Y,
- McIntosh H
- Shapiro W,
- Klein G
- Resnekov L,
- McDonald L
- Naito M,
- David D,
- Michelson EL,
- Schaffenburg M,
- Dreifus LS
- Executive Committee of the Percutaneous Cardiac Mapping and Ablation Registry,
- Scheinman MM,
- Evans-Bell T
- Feld GK,
- Fleck RP,
- Fujimura O,
- Prothro DL,
- Bahnson TD,
- Ibarra M
- Fick A
- Langberg JJ,
- Chin MC,
- Rosenqvist M,
- et al.
- Helmcke F,
- Nanda NC,
- Hsiung MC,
- et al.
- Jolly WA,
- Ritchie WJ
- Lau CP,
- Leung WH,
- Wong CK,
- Cheng CH
- Gosselink AT,
- Blanksma PK,
- Crijns HJGM,
- et al.
- Samet P