The autonomic mechanisms involved in neurogenic paroxysmal hypertension are not understood. We present the first demonstration of the precise haemodynamic and autonomic changes during a complex partial seizure.

A 50 year old headmaster was investigated for an 8 year history of recurrent absence attacks, stereotyped in nature and of sudden onset, each lasting about half a minute. He became pale, sweaty, and mentally withdrawn but did not fall down. Recovery was rapid and associated with transient headache. Previous neurological investigations, including repeated EEG and MRI, were negative. Electrocardiographic Holter monitoring disclosed only sinus bradycardia so he underwent head up tilt testing to exclude vasovagal syncope. Intra-arterial blood pressure and ECG were recorded continuously. Microneurography needles were positioned in the peroneal nerve of the right leg for recording efferent postganglionic MNSA.1 This technique allows beat to beat monitoring and quantification of MNSA (bursts/min) which controls vascular tone in skeletal muscle. MNSA in turn is modulated by changes in blood pressure via the baroreflexes. Blood pressure is normally maintained during head up tilt by increased MNSA and vasocontriction.2

The patient showed normal blood pressure, heart rate, and MNSA responses to tilt initially, but after 10 minutes, he suddenly became pale, sweaty, and withdrawn for about 30 seconds. No loss of muscle tone was seen and he later confirmed that this was a typical absence attack. Coinciding with the onset of his symptoms, MNSA increased briefly for 3 seconds associated with a sudden increase in blood pressure from 138/95 to 222/150 mm Hg over 10 seconds. Heart rate simultaneously increased from 65 to 98 bpm (fig 1). Over the next 20 seconds,blood pressure and heart rate decreased and there was a major burst of MNSA followed by reciprocal oscillation of blood pressure with MNSA (0.1Hz) as blood pressure reached normal levels. During recovery, he complained of his usual transitory headache. Venous noradrenaline (norepinephrine) concentrations were 1650 pmol/l and 5250 pmol/l before tilt and during recovery respectively. Normal values in our laboratory before and after 10 minutes of tilt are 456 (SD 50) and 705 (SD 74) pmol/l.3 His absence symptoms could not be reproduced by rapidly increasing blood pressure to similar values (250/120 mm Hg for 30 seconds) with an intravenous bolus of epinephrine (100 μg). One week later, an EEG during a similar absence attack showed sharp waves arising from the left frontoparietal area (fig 2). Subsequent continuous EEG and blood pressure monitoring confirmed that focal seizure activity was simultaneous with paroxysmal hypertension. Studies with MRI showed hippocampal atrophy consistent with the diagnosis of complex partial seizure disorder. His absences were abolished with 400 mg carbamazepine daily and he has remained free of symptoms for 6 months.

• Download figure
• Open in new tab
• Download powerpoint

Figure 1

A 2 minute recording of blood pressure (BP), muscle nerve sympathetic activity (MNSA), and heart rate (ECG) during an absence attack after 10 minutes of head up tilt. At 30 seconds there was sudden mental withdrawal and a rapid increase in MNSA* followed by a severe and paroxysmal increase in blood pressure and heart rate. As blood decreased, MNSA increased and when blood pressure normalised there was a marked baseline shift in MNSA‡. During recovery, blood pressure and MNSA oscillated reciprocally (0.1 Hz)†.

• Download figure
• Open in new tab
• Download powerpoint

Figure 2

EEG recording obtained during light sleep showing sharp waves arising maximally in the left frontotemporal area. The montage consists of four sets of channels running anterior to posterior recorded from the right parasagittal, left parasagittal, right temporal, and left temporal areas respectively.

This is the first demonstration of paroxysmal neurogenic hypertension triggered by a seizure in a patient with complex partial seizure. The diagnosis of complex partial seizure was supported by the following: focal EEG changes during a subsequent absence seizure; no reproduction of absence symptoms during drug induced paroxysmal hypertension; characteristic hippocampal atrophy on MRI4; and complete response to anticonvulsant drugs. Other possible diagnoses including brain stem tumour, phaeochromocytoma, and renal artery stenosis were excluded by appropriate imaging and neurohormonal analysis. Pseudoseizures were excluded on the basis of the EEG findings and the rapid response to treatment. Although rapid increases in MNSA and heart rate have been found during panic attacks, paroxysmal hypertension and loss of consciousness are not consistent features.5

The paroxysm consisted of simultaneous hypertension and tachycardia associated with sweating and facial pallor during the absence attack. We suggest that this is secondary to a generalised increase in sympathetic activity causing vasoconstriction and increased cardiac output. This is supported by (a) increased MNSA and heart rate despite progressive rise in blood pressure; (b) symptomatic blood pressure overshoot; (c) noradrenaline increased to over seven times the normal tilt levels; (d) prominent low frequency (0.1Hz) oscillations in blood pressure and MNSA during recovery.2 These low frequency oscillations (0.1 Hz) are thought to be secondary to changes in brain stem sympathetic activity separate from the effects of respiration, which are generally of a higher frequency (0.2 Hz). We emphasise that the initial increase in MNSA occurred when blood pressure was increasing and so was not baroreflex mediated as would be expected for respiratory or normal brain stem low frequency oscillations.

We hypothesise that this generalised increase in sympathetic activity is permitted by a transient interruption of baroreflex feedback inhibition during the seizure. We think that this is a unique recording of transient baroreflex failure characterised by a rapid and generalised increase in sympathetic activity, overriding the baroreflex afferents in the brain stem. It has long been suspected that paroxysmal hypertension occurs in complex partial seizures but to date, ambulatory monitoring has only demonstrated changes in heart rate.6Ambulatory beat to beat blood pressure monitoring would allow closer study of this phenomenon and its possible relation to sudden cardiac death in epileptic patients. Finally, this a good example of an episodic medical condition which may be very difficult to diagnose. Occasionally, when an episode is seen fortuitously in the laboratory, we may identify pathophysiology previously suspected but not actually seen.