Methods

Top Methods Results Discussion Author & Article Info References

Case Report

A 55-year-old white woman with idiopathic dilated cardiomyopathy had an angiographically measured ejection fraction of 19% and had New York Heart Association functional class III symptomatic congestive heart failure. She was referred to Ohio State University Hospital for further management of congestive heart failure. As part of her evaluation, she had hemodynamic assessment of her response to vasodilator administration using a protocol that was reviewed and approved by the Human Use Review Board of The Ohio State University; the patient provided informed consent. The patient had received nitroglycerin ointment, 1 inch every 8 hours, and had received furosemide for management of heart failure symptoms before admission. In accordance with the protocol, nitrates and diuretic agents were withheld on the evening before the hemodynamic evaluation.

Hemodynamic Evaluation

A 6-fr Millar catheter (Millar Instruments, Houston, Texas) with a micromanometer pressure transducer was inserted through the femoral artery using aseptic techniques and was positioned in the central aorta for high-fidelity measurement of central aortic pressure. A balloon-tipped catheter with a thermistor was inserted through the subclavian vein and positioned in the pulmonary artery for monitoring pulmonary artery pressures and thermodilution cardiac output. Heart rate and rhythm were monitored using standard electrocardiographic leads. After catheter insertion, equilibration of the hemodynamic state (defined as less than 10% variation in three consecutive measurements of cardiac output, pulmonary arterial pressure, and aortic pressure) was established. Throughout the hemodynamic evaluation, a mean heart rate of 97 beats/min and an average blood pressure of 119/67 mm Hg (mean blood pressure, 83 mm Hg) were seen.

Electrocardiographic results taken at 4-minute intervals and simultaneous measurements of aortic pressure were recorded on frequency-modulation tape for generation of the power-density spectra of heart rate and blood pressure variability as measures of autonomic tone [4]. These signals were subsequently digitized at a sampling rate of 512 Hz/channel using a 12-bit analog-to-digital converter installed in a microcomputer. From the digitized spectra, the power-density spectra of heart rate and blood pressure variability were generated as previously reported by our laboratory [4]. The power-density spectrum represents the magnitude of variability of a signal (heart rate or blood pressure), shown on the vertical axis, that is attributable to a given frequency, shown on the horizontal axis (Figure 1). In this way, the frequencies at which heart rate and blood pressure oscillate are determined.

View larger version (17K): [in this window] [in a new window]   Figure 1. Power-density spectrum of heart rate variability in a patient with congestive heart failure. A. Heart rate variability in the absence of pulsus alternans. The spectrum is characteristic of patients who have congestive heart failure with an absence of heart rate variability at frequencies greater than 0.1 Hz. The high-frequency peak in heart rate variability noted during pulsus alternans (panel B) is absent. B. Heart rate variability during a period of pulsus alternans. In contrast to the usual spectrum in patients with congestive heart failure, a prominent peak occurred in high-frequency heart rate variability (compare with panel A). C. Systolic blood pressure during pulsus alternans. A high-frequency peak in blood pressure variability coincides with the peak in heart rate variability shown in panel B. This high-frequency peak in blood pressure variability corresponds to the two-beat cycle of alternating high- and low-amplitude systolic pressure waveforms that define pulsus alternans (see Figure 2). BPM2 = beats/min2; CHF = congestive heart failure.

View larger version (14K): [in this window] [in a new window]   Figure 2. Electrocardiogram (top) and central aortic pressure recording (bottom) obtained immediately before and during the onset of pulsus alternans. The recordings show that heart rate alternans appears to follow, rather than precede, pulsus alternans. A relatively stable electrocardiographic (ECG) cycle length is maintained before the onset of pulsus alternans (cycle length indicated above crossbars on the ECG recording). Pulsus alternans is initiated with a high-amplitude aortic pressure wave (arrow) with subsequent alternating variations in ECG cycle length (heart rate alternans) that are of a magnitude greater than that of previous beat-to-beat variations in cycle duration.

Results

Top Methods Results Discussion Author & Article Info References

The electrocardiogram recorded during pulsus alternans Figure 2 shows that a sinus mechanism was maintained throughout this period and that, accordingly, the alternation in blood pressure was not simply the mechanical response to an atrial or ventricular bigeminal rhythm. That the high-frequency variation in heart rate was strictly associated with pulsus alternans was confirmed by spectral analysis of heart rate variability on a segment of heart rate data acquired during a period in which pulsus alternans was not observed. The spectrum shown in Figure 1, panel A typifies the usual spectrum of heart rate variability observed in congestive heart failure [3, 4] and shows the disappearance of heart rate alternans in the absence of pulsus alternans.

Discussion

Top Methods Results Discussion Author & Article Info References

Although pulsus alternans has long been noted as a manifestation of severe ventricular failure [1, 2], our report describes "heart rate alternans," which is a previously unrecognized manifestation of pulsus alternans. Electrical alternans, consisting of beat-to-beat changes in the amplitude or configuration of the QRS complex, sometimes accompanies pulsus alternans [1]. However, to our knowledge, a beat-to-beat variation in heart period recorded on the electrocardiogram has not been previously reported. Indeed, such a rapid oscillation in instantaneous heart rate is uncharacteristic for patients with ventricular failure; these patients typically have an absence of heart rate variability in frequency ranges greater than 0.1 Hz Figure 1 [3, 4]. This inability to produce rapid variations in heart rate has been ascribed to a reduction of parasympathetic tone, which, under normal conditions, is responsible for modulating rapid fluctuations in heart rate [3, 4].

Our patient had an uncharacteristic beat-to-beat alternation in heart period (heart rate alternans) that appears to have been a response to beat-to-beat changes in the amplitude of systolic blood pressure (pulsus alternans) and that constitutes a new mechanism for high-frequency oscillations in heart rate. The mechanism for these observations in a single patient is speculative. However, the reduced baroreflex sensitivity that is typical for patients with congestive heart failure is analogous to a low-gain feedback control system in which uncontrolled oscillations in signals typically occur [6]. Therefore, if this mechanism pertains in this setting, an oscillation in blood pressure may produce an oscillation in heart rate that persists because of low-gain feedback control. Indeed, Figure 2 shows the initiation of pulsus alternans in our patient and indicates that a fluctuation in blood pressure preceded alterations in heart rate. This finding suggests that the variation in heart rate was initiated by a change in blood pressure and was not due to a primary alteration in cardiac cycle length or even a cyclic alteration in a higher-order control mechanism that would simultaneously affect heart rate and blood pressure.

Our report shows that in patients with congestive heart failure, heart period may vary in a beat-to-beat fashion coincident with the beat-to-beat alternations in systolic blood pressure that define pulsus alternans. This heart rate alternans, therefore, is a new element in the spectrum of findings that constitute cardiac alternans [1] and defines an important exception to the observation that high-frequency heart rate variability is substantially attenuated in patients with congestive heart failure. Furthermore, this type of high-frequency heart rate oscillation has not been previously reported. This heart rate alternans differs from the classic peaks that arise from respiratory-linked changes in vagal tone. We used a spectral analysis technique that is extremely sensitive to changes in heart period variability; thus, the heart period oscillations that we were able to record may not have otherwise been clinically evident. This may account for the fact that heart rate alternans had not been described until we used this relatively new technique to analyze physiologic signals. Because our observation was made in a single patient, it remains to be determined whether heart rate variability is a consistent, concomitant finding in patients with pulsus alternans and whether heart rate alternans occurs in other disease states that also have abnormal autonomic and baroreflex circulatory control mechanisms.