Article Text


Treatment of paroxysmal sympathetic storm with labetalol
  1. D DO,
  2. V L SHEEN,
  1. Departments of Neurology, Brigham and Women's Hospital, 75 Francis Street, Boston, MA 02115, USA and Massachusetts General Hospital, Harvard Medical School, Boston, MA, USA
  1. Dr E Bromfield ebromfield{at}

Statistics from

First described by Penfield in 1929, paroxysmal sympathetic storm is characterised by episodic hyperhidrosis, hypertension, hyperthermia, tachypnoea, tachycardia, and posturing. It has commonly been associated with closed head traumatic brain injury, agenesis of the corpus callosum, hydrocephalus, and suprasellar or diencephalic tumours.1 2 Penfield hypothesised that these sympathetic spells were caused by epileptiform discharges in thalamic nuclei irritated by increased intracranial pressure, thereby leading him to name this entity “diencephalic autonomic seizures”.2Electroencephalograms obtained on patients during these autonomic attacks, however, have not shown epileptic activity, and anticonvulsant therapies have not proved useful in their treatment.2 3Bromocriptine and morphine have been the standard treatments for paroxysmal sympathetic storm, and propranolol has been shown to reduce the hyperpyrexia seen during autonomic spells.3 4 In this case report, we describe a patient treated successfully with labetalol, but not metoprolol, suggesting that β1 antagonism alone is not sufficient to suppress paroxysmal sympathetic storm.

A 21 year old white man was an unrestrained passenger in a motor vehicle accident and developed a closed head shear injury. He was admitted to a hospital where a head CT showed hydrocephalus necessitating a ventriculoperitoneal shunt placement. A head MRI showed abnormal T2 signal in the corpus callosum and the dorsal midbrain consistent with shear injury. Although initially comatose, he improved to near baseline over the next few months. He was admitted to our hospital 15 months later with a shunt infection, necessitating treatment with vancomycin, shunt externalisation, and, eventually, replacement. During and after resolution of the shunt complications, he developed episodes of sympathetic hyperactivity while under continuous monitoring in an intensive care unit. These attacks were characterised by (1) diaphoresis throughout the entire body, (2) tachycardia (heart rate 140–160 bpm) measured by automated pulse oximetry or by ECG, (3) hypertension (blood pressure 170–180/100 mm Hg) measured by arterial line pressure transducers or by sphygmomanometer, (4) fevers to 39.1 °C orally, and (5) flexor posturing. He was alert during these episodes, and responded to questions appropriately with denial of any acute onset of discomfort or pain. Furthermore, these attacks were not correlated with periods of bladder distension (Foley catheter in place) or impaction (radiograph not suggestive of retained stool). Individual episodes lasted 5–10 minutes and recurred at 5–10 minute intervals. Clusters of these spells would last 1 to 2 hours with more than three clusters a day.

Multiple CSF and blood cultures were negative. Serial head CT showed marked reduction of hydrocephalus and no brain stem abnormalities after shunt correction. No other intracranial pathology was noted. Plain films and MRI of the spine showed no myelopathic findings suggesting autonomic dysreflexia. Abdominal and pelvic CT did not show any hidden masses or lesions. Toxicology screen at onset of symptoms was negative. Electroencephalograms obtained during these episodes of dysautonomia disclosed theta and delta slowing with some sharply contoured waves, but no definite ictal or interictal epileptiform activity (figure).

EEG obtained during episodes of paroxysmal sympathetic storm (tachycardia with heart rate of 120 bpm) shows predominant delta and theta waves (greater on the right than on the left) with no clear epileptiform activity, indicating that these attacks are not of seizure origin.

Although initially treated successfully with bromocriptine (5 mg twice daily) and morphine (15 mg every 6 hours), he was withdrawn from morphine with a methadone taper at the request of his parents secondary to concerns over addiction. He was then started on metoprolol (25 mg thrice daily) with little effect on the frequency or severity of the hyperautonomic episodes. Replacement with 100 mg labetalol twice daily led to reduction in the frequency of events to about one a day. Subsequent increase of the medication to 200 mg twice daily resulted in a marked decrease to less than one paroxysmal sympathetic storm over several days. At the time of discharge, the patient had returned to his preadmission baseline.

The current observations lend support to the prevailing view that paroxysmal sympathetic storm may represent disruption of autonomic function in the diencephalon and brainstem. Bullard has proposed that the clinical syndrome may be the result of a release phenomenon within the brainstem and/or diencephalon from loss of overriding cortical or subcortical inhibition.3 More recent case studies suggest localisation to the central sympathoexcitatory regions including the paraventricular hypothalamic nucleus, lateral periaqueductal grey matter, lateral parabrachial nucleus, or rostral ventrolateral medulla.3 Compromised autonomic neuronal integrity centrally is not surprising in the setting of infection after traumatic brain injury.

Various medications can potentially be used in managing central sympathetic storm. Imidazoline agonists and specific α2 adrenoceptor antagonists, such as clonidine and methyldopa,5 have recently been shown to have sympathoinhibitory actions centrally within the rostral ventrolateral medulla. These agents have so far been used in the treatment of essential hypertension, tetanus, or autonomic dysreflexia. β Blockers such as propranolol however, have long been the mainstay of treatment of the hypertension, tachycardia, and hyperpyrexia associated with paroxysmal sympathetic storm.6 7 This non-selective β adrenergic antagonist acts through inhibition of peripheral catecholamine activity, and being highly lipophilic, may also exert central effects through membrane stabilisation or receptor blockade. Moreover, propranolol may reduce sustained muscle contraction.7 Taken together, these findings suggest that non-selective β receptor antagonism is sufficient to inhibit the clinical manifestations of diencephalic seizures.

The present case suggests that β1 receptor antagonism alone is not sufficient to treat hyperautonomia during paroxysmal sympathetic storm. This patient was initially placed on starting doses of metoprolol, a selective β1 antagonist, with little clinical effect in controlling the frequency of the autonomic attacks; however, labetalol, an α1 and β1-β2 adrenergic receptor antagonist did lead to an observable decline in symptoms. Both sympatholytic agents were given at doses typically used in initiating treatment of systemic hypertension, suggesting that the observed response seen with labetalol could not be explained solely by a dosage phenomenon. Prior studies also demonstrate that small amounts of propranolol (20 mg four times a day) can achieve similar responses to those seen with labetalol,6 7further arguing against a dose dependent effect. Thus, at a minimum, either α1 or β2 receptor blockade, likely in addition to β1 blockade, is necessary in the treatment of paroxysmal sympathetic storm.

The discrepancy in response between metoprolol and labetalol could result from their different effects on the cardiovascular system or CNS. The β1-β2 adrenergic receptor blockade by labetalol decreases blood pressure and heart rate through negative inotropic and chronotropic effects, and by inhibiting renin release. In addition, labetalol has vasodilator properties resulting from α1 blockade and partial β2 agonism. These reduce peripheral vascular resistance, blood pressure, and coronary vascular resistance, a potential advantage over other β blockers. Alternatively, differences in central activity may explain the increased efficacy of labetalol over metoprolol. As both agents are lipophilic, their central access should not differ significantly; rather, differences in receptor antagonism (β1 versus α1, β1, β2) would more likely explain the therapeutic discrepancy. As proposed with propranolol,4 inhibition of β2 receptors by labetalol may exert a stabilising effect within the CNS through indirect inhibition of sympathetic nerve activity.

In the present case, we report the use of labetalol as an alternative agent in the treatment of paroxysmal sympathetic storm. It likely exerts both a central and peripheral blockade of α1 and β adrenergic receptors to produce inhibition of autonomic dysregulation. The clinical ineffectiveness of metoprolol further suggests a necessary role for β2 and/or α1 receptors in the clinical presentation of paroxysmal sympathetic storm. Labetalol may prove an alternative equal to or better than morphine in the treatment of these spells, especially when addiction and dependency are of concern.


View Abstract

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.