Malignant Vagotonia Due to Selective Baroreflex Failure
Abstract Baroreflex failure is characterized by dramatic fluctuations of sympathetic activity and paroxysms of hypertension and tachycardia. In contrast, unopposed parasympathetic activity has not been described in patients with baroreflex failure because of concurrent parasympathetic denervation of the heart. We describe the unusual case of a patient with baroreflex failure in a setting of preserved parasympathetic control of HR manifesting episodes of severe bradycardia and asystole. Thus, parasympathetic control of the HR may be intact in occasional patients with baroreflex failure. Patients with this selective baroreflex failure require a unique therapeutic strategy for the control of disease manifestations.
Sustained baroreflex failure occurs after bilateral damage to the carotid and aortic baroreceptors and has been observed in animals1 2 and in humans.3 4 5 In humans, baroreflex failure is a rare disorder, which is usually related to neck surgery, neck irradiation therapy for malignant tumors, brain stem infarction, or bilateral carotid body tumors.3 4 5 6 7 8 The most common symptoms of baroreflex failure are dramatic fluctuations of sympathetic activity with paroxysms of arterial hypertension, tachycardia, flushing, and headache resembling the symptoms of pheochromocytoma.3 4 5 6 7 These manifestations are presumably caused by an inability to buffer supramedullary input to cardiovascular control centers in the brain stem.5 9
Theoretically, patients with a lesion involving only the afferent arc of the baroreflex (selective baroreflex failure) should display not only undamped sympathetic discharge but also undamped parasympathetic activation and excessive vagotonic reactions. In the patients with baroreflex failure described in the literature, uncontrolled increases of the parasympathetic activation of the heart were not encountered.3 4 5 In fact, the relatively high resting HR in most reported patients with baroreflex failure may suggest parasympathetic denervation of the heart. Thus, the clinical entity of baroreflex failure as described in the literature appears to entail lesions involving both the afferent and the efferent arcs of the arterial baroreflex. We describe the unusual case of a patient who presented with baroreflex failure and life-threatening episodes of bradycardia and hypotension.
A 54-year-old woman was admitted to the Elliot V. Newman Clinical Research Center at Vanderbilt University with the presumptive diagnosis of baroreflex failure.
This previously normotensive woman underwent anterior cervical disc excision and fusion at the level of C4 to C5 and C5 to C6. After a motor vehicle accident, she experienced increasing symptoms of a cervical radiculopathy and subsequently underwent anterior fusion of C6 and C7. Then, wide variations of her BP and HR were noted. She had episodes of hypertension with arterial BP values as high as 270/140 mm Hg and tachycardia of about 120 bpm, which were often triggered by emotional factors, sexual activity, or physical exertion. Hypertensive episodes with SBP >200 mm Hg were first associated with flushing of the upper trunk and shoulder girdle. It appeared that with greater increases in systolic pressure, the flushing spread to include the upper extremities but spared the hands.
She also had prolonged episodes of severe hypotension with SBP <50 mm Hg. These episodes were associated with symptoms such as fatigue and dizziness, which occasionally progressed to frank syncope. The most severe hypotensive episodes tended to occur early in the day, and her husband reported that he was sometimes unable to awaken her for several hours. In the early morning, her husband also observed weakness and slowing of the radial pulse during these episodes. Holter monitoring documented HR as low as 20 bpm. Twice (once after a neurosurgical procedure and once after a continuous intravenous infusion of sodium nitroprusside) she required resuscitation after periods of asystole lasting more than 1 minute. There was also an episode of severe bradycardia and hypotension after the administration of sublingual nitroglycerin for the treatment of hypertension. She did not report other symptoms involving the autonomic nervous system.
The development of chronic cough and slight hoarseness after spine surgery at the level of C4,5 and C5,6 is consistent with the possible involvement of cranial nerves or the brain stem. She reported that cough paroxysms would often be associated with hypertensive episodes.
The patient was placed on a 150-mEq sodium and 70-mEq potassium diet free of substances that could interfere with catecholamine measurements. Vasoactive medications and fludrocortisone were discontinued at least five half-lives before testing.
The patient underwent a battery of autonomic reflex tests in the supine position to assess sympathetic and parasympathetic control of the cardiovascular system. HR was determined by continuous electrocardiography, and BP changes were measured beat-to-beat by photoplethysmography (Finapres). The response of SBP to rapid (approximately 100 breaths per minute) shallow breathing for 60 seconds was determined. The patient was unable to sustain an isometric handgrip. The SBP response to pain was measured by the cold pressor test that was performed by immersing one hand in ice water for 1 minute. The SBP and HR responses to the Valsalva maneuver (40 mm Hg for 15 seconds) were also determined.
Plasma catecholamines were determined after the patient remained in the supine position overnight and after 30 minutes in the upright position before and during guanadrel therapy (10 mg BID). Blood samples were drawn from a heparin lock placed at least 30 minutes before the first blood draw. All plasma catecholamines were analyzed by high-pressure liquid chromatography as previously described.10
All drugs that were given were administered via a heparin lock in a large antecubital vein. Incremental bolus doses of sodium nitroprusside, which began with 0.1 μg/kg, were given intravenously up to a dose of nitroprusside that decreased the SBP by 25 mm Hg. This procedure was repeated, with incremental intravenous bolus doses of nitroglycerin, which started at 1 μg. Similarly, incremental doses of phenylephrine, begun at 12.5 μg, were administered to increase the SBP by 25 mm Hg. To account for spontaneous BP and HR changes, pharmacological testing was repeated on a different occasion after 1 month, and all bolus doses were given at least three times on each occasion. The contribution of the parasympathetic and sympathetic nervous system to resting HR was assessed by autonomic blockade with atropine (2.4 mg in four divided doses) and propranolol (12.2 mg in four divided doses), respectively. Pharmacological testing with sodium nitroprusside was repeated after autonomic blockade.
Power Spectral Analysis
Power spectral analysis of RR interval and systolic arterial BP was performed as described elsewhere.11 The power spectral density was estimated on data segments of 256 to 512 beats. The power spectra, obtained from RR interval and SBP variabilities, are characterized by two main oscillatory components,12 an LF component and an HF component. The HF component, centered at a frequency of ≈0.25 Hz, has been related to the vagal efferent activity that modulates the sinoatrial node. The HF component of systolic arterial pressure is thought to reflect the mechanical effects of respiration on the cardiovascular system. The LF (0.10 Hz) of the SBP variability corresponds to Mayer waves and has been proposed to be an indicator of sympathetic activity. At the end of the recording, atropine (2.4 mg in four divided doses) was given to confirm that HF RR was indeed related to parasympathetic control of the HR.
Muscle Sympathetic Nerve Activity
Sympathetic nerve activity was measured as previously described13 in the right peroneal nerve at the level of the fibular head with a tungsten needle electrode (shaft diameter, 200 μm; tip diameter, 1 to 5 μm). Recorded signals were fed to a preamplifier (gain, ×1000) and were filtered using a bandwidth between 700 and 2000 Hz. The filtered signal was rectified, amplified (gain, ×100), and integrated in a resistance-capacitance network using a time constant of 0.1 (Nerve Traffic Analysis System 662C-3). The final signal was monitored using a storage oscilloscope (S111A; Tektronics) and recorded after fourfold amplification (TA-2000 recorder, Gould Inc).
Criteria for an adequate muscle sympathetic nerve activity recording were as follows: (1) electrical stimulation produced muscle twitches but no paresthesias; (2) stretch of the tendons in the foot evoked proprioceptive afferent signals, whereas cutaneous stimulation by slight stroking of the skin did not; and (3) typical morphology of the neurogram.
Muscle sympathetic nerve activity was recorded during a 45° tilt-table test and a Valsalva maneuver. Because of technical difficulties and an inability to obtain a reliable recording, it was not possible to obtain a muscle sympathetic nerve recording during pharmacological testing.
Blood Volume and Dynamic Volume Changes
With the patient supine in an overnight fasting state, a large-gauge antecubital heparin lock (Flash-Cath, Baxter Healthcare Corp) was positioned. After 30 minutes of rest, blood volume was determined by Evans blue dye.14 Dynamic plasma volume shifts caused by standing were determined as previously described.14 Briefly, the patient stood motionless at the bedside, and blood was drawn without hemostasis from the antecubital heplock at 2.5, 5, 7.5, 10, 15, and 20 minutes for the determination of hematocrit and plasma catecholamines. Hematocrit was determined in quadruplicate with microcapillary tubes (International Equipment Co, model MB), which were centrifuged at 11 500 rpm for 10 minutes. Relative plasma volume changes were determined based on hematocrit changes from the supine value.
Fig 1⇓ illustrates the wide spontaneous changes in BP and HR. BP and HR changes occur in the same direction (ie, decreases in BP are associated with decreases in HR, and increases in BP are associated with increases in HR, which is typical for patients with baroreflex failure).5 9 Considerable decreases of BP and HR occur within several seconds. Such variability in HR could raise the consideration of sick sinus syndrome. However, BP and HR varied so similarly that the diagnosis of sick sinus syndrome would be unlikely.
Power spectral analysis of the BP showed LF oscillations indicating sympathetic control of the BP. HF oscillations were shown by power spectral analysis of the HR. Power spectral analysis demonstrated a loss of these HF RR vagal oscillations after atropine.
During phase II of the Valsalva maneuver, there was a marked decrease in the SBP from 150 to 90 mm Hg (normal, <20 mm Hg) without an adequate increase in HR or, as shown during a separate Valsalva maneuver, muscle sympathetic nerve activity. There was excessive, though delayed, overshoot of the SBP reaching a maximum of 210 mm Hg. Hyperventilation decreased the SBP by 25 mm Hg (normal, <10 mm Hg), and the cold pressor test elicited a significant increase of both HR (75 to 90 bpm) and BP (154 to 200 mm Hg).
The tilt-table test at 45° (Fig 2⇓) showed, after a stable phase over the first minutes, a continuous decrease of the SBP by 80 mm Hg (from 175 to 95 mm Hg) after 7 minutes, which was associated with symptoms of lightheadedness and blurred vision. There was no significant increase in HR. Muscle sympathetic nerve activity increased slightly initially during the tilt test but then actually began to decrease, despite the continued decrease in BP.
The dose of phenylephrine that increased the SBP by 25 mm Hg was 12.5 μg (normal controls, 225±15 μg). The dose of sodium nitroprusside that decreased the SBP by 25 mm Hg was 0.15 μg/kg (normal controls, 1.2±0.2 μg/kg). There was no change of the RR interval after phenylephrine administration (Fig 3⇓), and the decrease in BP caused by sodium nitroprusside failed to elicit an increase in HR. In fact, sodium nitroprusside caused a paradoxical decrease in HR (Fig 4⇓, top). A bolus of 4 μg nitroglycerin decreased BP by 30 mm Hg and HR from 65 to 50 bpm, at which point the cardiac pacemaker began pacing. (A cardiac pacemaker was placed after the initial evaluation [pharmacological testing and tilt study] and before the final evaluation [pharmacological test and dynamic posture study].) These responses to pharmacological testing confirmed profound dysfunction of the arterial baroreceptors. Propanolol decreased HR from 62 to 58 bpm. After β-blockade, atropine increased the HR from 62 to 95 bpm. The paradoxical bradycardic effect of nitroprusside was completely abolished after autonomic blockade (Fig 4⇓, bottom), and the dose of nitroprusside that decreased BP by 25 mm Hg was four times higher (0.6 μg/kg) than before autonomic blockade.
The plasma volume, as measured by Evans blue dye, was 8% lower than expected.14 During the dynamic posture study (ie, changing from recumbent to standing position), there was a transient increase of the SBP from 102 mm Hg to a maximum of 161 mm Hg after 2.5 minutes (Fig 5⇓, top). This rapid increase of SBP was followed by a slow decline to a minimum of 57 mm Hg after 20 minutes, associated with presyncopal symptoms. The increase of the HR upon standing was ultimately inadequate, considering the pronounced decrease in BP. There was a rapid decrease of the plasma volume with standing, which reached 18% below baseline after ≈20 minutes (Fig 5⇓, bottom). There was an increase of the plasma norepinephrine from 1.77 nmol/L (300 pg/mL) in the supine position to a maximum of 3.78 nmol/L (639 pg/mL) after 20 minutes of standing (Fig 5b⇓).
Fig 6⇓ shows supine and upright plasma norepinephrine levels before and during guanadrel treatment. Supine values of plasma catecholamines before treatment were plasma norepinephrine 0.90 nmol/L (153 pg/mL), plasma epinephrine 0.20 nmol/L (37 pg/mL), and dopamine 52.2 pmol/L (8 pg/mL). During a hypertensive episode (SBP >200 mm Hg), with flushing of the upper trunk, the plasma norepinephrine was 2.68 nmol/L (453 pg/mL), plasma epinephrine was 0.31 nmol/L (56 pg/mL), and plasma dopamine was undetectable. Urine methylhistamine obtained after one of these episodes was not elevated. Therefore, histamine or dopamine release does not appear to explain the observed flushing.
This is the first report of a patient with selective baroreflex failure presenting with alternating phases of severe hypertension and life-threatening bradycardia and hypotension.
The diagnosis of baroreflex failure was confirmed by a battery of autonomic function testing and pharmacological testing as previously described in humans5 and animals.1 There was no compensatory baroreflex-mediated HR change after administration of vasodilator or vasopressor agents or during phase II or phase IV of Valsalva’s maneuver. Her hypersensitivity to the hypotensive effects of nitroprusside and the hypertensive effects of phenylephrine were likely due to the debuffering of the baroreflex and/or receptor hypersensitivity.
The integrity of the efferent sympathetic nervous system is shown by the supranormal increase in arterial BP in response to painful stimuli (cold pressor test) and arousal. Vagal control of the heart was demonstrated by power spectral analysis of the HR12 and confirmed by atropine administration. Because parasympathetic control of the HR is exerted by vagal efferent nerve fibers, the vagus nerves must be intact. Most patients with baroreflex failure reported in the literature had relatively high resting HR and a small increase of the HR after atropine administration (10 bpm; range, 7 to 15 bpm),5 consistent with substantial parasympathetic denervation of the heart.
It has been shown in numerous animal studies and in humans that complete denervation of the carotid and aortic baroreceptors is necessary to cause baroreflex failure.2 5 9 The signals generated in the carotid baroreceptors and the aortic baroreceptors are transmitted to the brain stem via the glossopharyngeal15 and vagus16 nerves, respectively, and converge on the same brain stem nuclei.16 Thus, to elicit baroreflex failure, four separate cranial nerves or the brain stem nuclei (where these nerves project) must be damaged, explaining the rarity of this condition in clinical practice.
The ratio of afferent to efferent nerve fibers in the vagus nerves has been estimated to be 4:1.17 Therefore, the chance of having efferent vagal innervation of the heart, in the face of bilateral damage to afferent vagal nerve fibers innervating the aortic baroreceptors (selective baroreflex failure), seems to be small, which possibly explains the impairment of parasympathetic innervation of the heart in previously described patients (nonselective baroreflex failure) (Fig 7⇓). Near the brain stem, the efferent and afferent nerve fibers of the vagus nerves separate.18 We suspect that in our patient, the damage to the afferent part of the baroreflex may be, at least in part, located near the brain stem. Another less likely explanation for selective baroreflex failure in humans is a very selective damage of vagal afferent fibers in the peripheral nerve.
Patients with baroreflex failure typically present with episodes of severe hypertension, tachycardia, and elevated plasma catecholamines.3 4 5 6 7 Acute administration of clonidine decreases BP, HR, and plasma catecholamine levels in such patients5 and is therefore useful as a diagnostic test to distinguish baroreflex failure from pheochromocytoma.5 19 20 21 There can also be diagnostic confusion with orthostatic intolerance,22 because this entity is sometimes also associated with raised catecholamines and volatile BP.
Major BP changes occur in most patients with nonselective baroreflex failure over periods of at least several minutes.5 A distinct feature of selective afferent baroreceptor failure that is similar to the vasovagal syncopes is the rapidity with which profound BP and HR changes occur, especially the prolonged vagotonic phases with bradycardia and severe hypotension. One possible explanation for the brisk BP changes would be a rapid parasympathetic-mediated decrease in the HR without adequate sympathetic counterregulation. Another possible explanation is a direct cholinergic-mediated suppression of cardiac contractility.23
There was a paradoxical bradycardic effect to systemic administration of nitric oxide donors (sodium nitroprusside, nitroglycerin) that was completely abolished after autonomic blockade. Therefore, a direct bradycardic effect of nitroprusside can be excluded. The hypotensive response to sodium nitroprusside was also diminished after autonomic blockade; a fourfold higher dose was required to achieve a similar hypotensive effect. Based on animal studies, it has been suggested that nitric oxide, acting as a second messenger and/or neurotransmitter, decreases the sympathetic discharge from the central nervous system.24 25 Therefore, the bradycardic and, in part, the hypotensive effect of nitroprusside observed in our patient could be related to a central nervous system effect. A similar paradoxical bradycardia after nitroprusside administration was reported in a cardiac transplant patient.26 In healthy subjects, the central nervous system effects of nitric oxide may be masked by baroreflex-mediated changes in sympathetic and parasympathetic tone.
It is often thought that excessive BP decreases on assumption of the upright posture are prevented by baroreflex-mediated activation of the sympathetic nervous system and parasympathetic withdrawal.27 Severe orthostatic hypotension immediately after assumption of the upright posture is uncommon in patients with baroreflex failure, and some patients actually have orthostatic hypertension. Mechanisms other than the baroreflex (eg, visual, vestibular, cerebellar, and cortical activation28 ) increase sympathetic discharge upon standing. The increase of the sympathetic discharge by these mechanisms may be buffered by the baroreflex to prevent excessive orthostatic BP increases. After more prolonged standing, while the stimulation of sympathetic outflow by other mechanisms decreases and plasma volume decreases,14 29 30 baroreflex-mediated increases of sympathetic discharge may become more important for BP control. The mild baseline hypovolemia in our patient could be related to alterations of renin and vasopressin regulation as described in sinoaortic denervated rats.1
Management of this patient proved to be complex. Treatment had to be tailored to attenuate extremes of both HR and BP (Table 1⇓). Because hypotensive episodes were often associated with severe bradycardia, and there was also a history of asystole, a cardiac pacemaker was implanted before antihypertensive therapy was begun. She was also treated with fludrocortisone (0.1 mg at 8 am and noon) and a high salt diet (>4 g/d) to prevent or blunt hypotensive episodes.31 Excessive BP increases were then prevented by therapy with guanadrel (10 mg BID), which was begun to block the peripheral release of norepinephrine from sympathetic nerve endings.32 A peripherally acting sympatholytic medication was chosen, because the patient had excessive side effects (sedation) with centrally acting agents (eg, clonidine). This paradoxical combination of therapeutic strategies was associated with great symptomatic improvement, and the extremes of the BP were blunted.
We conclude that selective baroreflex failure occurs after complete loss of the afferent arc of the baroreflex with preservation of efferent sympathetic and parasympathetic nerves. Selective baroreflex failure is characterized by phases of severe hypertension and tachycardia alternating with prolonged hypotension and bradycardia. Patients with selective baroreflex failure and malignant vagotonia require a unique therapeutic strategy for control of disease manifestations, including medications that decrease sympathetic nerve traffic, medications that increase BP (eg, fludrocortisone), and, in some cases, implantation of a cardiac pacemaker.
Selected Abbreviations and Acronyms
|SBP||=||systolic blood pressure|
This study was supported in part by National Institutes of Health grants RR-00095 and NS-33460 and NASA grants NAG9-563 and NAGW-3873. Jens Jordan is a scholar of the Deutsche Forschungsgemeinschaft.
Reprint requests to David Robertson, MD, Autonomic Dysfunction Center, AA3228 MCN, Vanderbilt University, Nashville, TN 37232-2195.
- Received March 19, 1997.
- Revision received April 3, 1997.
- Accepted April 11, 1997.
Schreihofer AM, Sved AF. Use of sinoaortic denervation to study the role of baroreceptors in cardiovascular regulation. Am J Physiol. 1994;266:R1705-R1710.
Cowley AW, Liard JF, Guyton AC. Role of the baroreceptor reflex in daily control of arterial blood pressure and other variables in dogs. Circ Res. 1973;32:564-576.
Aksamit TR, Floras JS, Victor RG, Aylward PE. Paroxysmal hypertension due to sinoaortic baroreceptor denervation in humans. Hypertension. 1987;9:309-314.
Fagius J, Wallin BG, Sunlof G, Nerhed C, Englesson J. Sympathetic outflow in man after anesthesia of the glossopharyngeal and vagus nerves. Brain. 1985;108(pt 2):423-438.
Ford FR. Fatal hypertensive crisis following denervation of the carotid sinus for relief of repeated attacks of syncope. Johns Hopkins Med J. 1957;100:14-16.
Biaggioni I, Whetsell WO, Jobe J, Nadeau JH. Baroreflex failure in a patient with central nervous system lesions involving the nucleus tractus solitarii. Hypertension. 1994;23:491-495.
Robertson D. Disorders of autonomic cardiovascular regulation: baroreflex failure, autonomic failure, and orthostatic intolerance syndromes. In: Laragh JH, Brenner BM, eds. Hypertension: Pathology, Diagnosis and Management. New York, NY: Raven Press; 1995:941-962.
Pagani M, Lombardi F, Guzzeti S, Rimoldi O, Furlan R, Pizzinelli P, Turiel M, Baselli G, Cerutti S, Malliani A. Power spectral analysis of heart rate and arterial pressure variabilities as a marker of sympatho-vagal interaction in man and conscious dog. Circ Res. 1986;59:178-193.
Biaggioni I, Killian TJ, Mosqueda-Garcia R, Robertson RM, Robertson D. Adenosine increases sympathetic nerve traffic in humans. Circulation. 1991;83:1668-1675.
Jacob G, Ertl AC, Shannon JR, Furlan R, Robertson RM, Robertson D. Effect of standing on neurohumoral responses and plasma volume in healthy subjects. J Appl Physiol. In press.
Brodal A. Hjernenervene. Copenhagen, Denmark: Munksgard; 1957:1-31.
Magnus O, Koster M, Van der Drift JHA. Cerebral mechanisms and neurogenic hypertension in man, with special reference to baroreceptor control. In: De Jong W, Provoost AP, Shapiro AP, eds. Hypertension and Brain Mechanisms. Vol. 47 of Progress in Brain Research. Amsterdam, Netherlands: Elsevier North-Holland Biomedical Press; 1977:199-218.
Lefkowitz RJ, Hoffman BB, Taylor P. Neurotransmission: the autonomic and somatic motor nervous systems. In: Hardman JG, Gilman AG, Limbird LE, eds. Goodman and Gilman’s Pharmacological Basis of Therapeutics. 9th ed. New York, NY: McGraw-Hill Book Co; 1996:105-139.
Mosenthal WT. Cranial nerve X. In: Mosenthal WT, ed. A Textbook of Neuroanatomy. New York, NY: Parthenon Publishing Group; 1994:289-292.
Manger WM, Gifford RW. Differential diagnosis. In: Manger WM, Gifford RW, eds. Clinical and Experimental Pheochromocytoma. 2nd ed. Cambridge, Mass: Blackwell Science; 1996:152-204.
Streeten DHP. Orthostatic Disorders of the Circulation: Mechanisms, Manifestations, and Treatment. New York, NY: Plenum Press; 1987:1-272.
Landsberg JS, Parker JD, Gauthier DF, Colucci WS. Effects of intracoronary acetylcholine and atropine on basal and dobutamine-stimulated left ventricular contractility. Circulation. 1994;89:164-168.
Togashi H, Sakuma I, Yoshioka M, Kobayashi T, Yasuda H, Kitabatake A, Saito H, Gross SS, Levi R. A central nervous system action of nitric oxide in blood pressure regulation. J Pharmacol Exp Ther. 1992;262:343-347.
Guyton AC. Nervous regulation of the circulation, and rapid control of arterial pressure. In: Guyton AC, ed. Textbook of Medical Physiology. Philadelphia, Pa: WB Saunders Co; 1991:194-204.
Goldstein D. Stress and science. In: Goldstein D, ed. Stress, Catecholamines, and Cardiovascular Disease. New York, NY: Oxford University Press; 1995:3-55.
Youmans JB, Akeroyd JH, Frank H. Changes in the blood and circulation with changes in posture. The effect of exercise and vasodilation. J Clin Invest. 1935;14:739-753.
Youmans JB, Wells HS, Donley D, Miller DG. The effect of posture (standing) on the serum protein concentration and colloid osmotic pressure of blood. J Clin Invest. 1934;13:447-459.
Hogikyan RV, Supiano MA. Homologous upregulation of human arterial alpha-adrenergic responses by guanadrel. J Clin Invest. 1993;91:1429-1435.