A Case of Symptomatic Hypnic Headache Syndrome

The hypnic headache syndrome is a rare, benign, nocturnal, non-familial headache disorder that occurs only while asleep. Since the first description by Raskin (1) in 1998, the syndrome has been characterized by a diffuse headache that awakens patients at a consistent time of the night, sometimes during a dream, and lasts for 30–180 min. This disorder mainly affects elderly subjects and both sexes equally (2).

The pathophysiological mechanism of hypnic headache is not known, but its circadian periodicity and responsiveness to lithium carbonate suggest that perturbation of chronobiological rhythms causes this headache (1). We describe a patient with hypnic headache, which had developed after pontine infarction. To our knowledge, this is the first report of symptomatic hypnic headache that developed after stroke and that suggests direct evidence of a potential pathophysiological mechanism and relationship of hypnic headache to the sleep cycle.

Case report

A 71-year-old man visited our headache clinic with complaint of noctural headache. He had a history of hypertension and diabetes mellitus. Two years previously he developed right hemiparesis, dysartheria following vertigo and dysequilibrium of sudden onset. Antiplatelet agents were given to him after the event. Two weeks later, the headache that awakened him from sleep developed. He described the headache as bilateral, dull pressure-like and moderately severe. He was awakened at a consistent time, usually between 02.00 h and 03.00 h, due to headache. The attacks began 2–3 h after he fell asleep and resolved within 1–2 h. He could not tell relationship with dreams exactly. He denied vomiting, nausea, photophobia, phonophobia, or any autonomic symptoms such as lacrimation, nasal congestion, rhinorrhoea or redness of eye. During the attack, the patient tended to isolate himself, usually sitting on a chair, and await the resolution of pain. The pain failed to respond to simple analgesics, ergotamine, or amitriptylline. A lighter headache of dull nature sometimes awakened the patient in the early morning, but headache never occurred during the day. He had previously had neither migraine nor any other kind of headache. The patient had a history of snoring and sleep apnoea, which developed in middle age. He took an antihypertensive agent, antiplatelet agent and sleep medications such as zolpidem.

Blood pressure was 147/93 mmHg and other vital signs were normal. Physical examination was unremarkable. There were ataxia and increased tone in right limbs with extensor planter reflex on neurological examination. The results of basic blood tests and urine analysis were normal. We found a small lesion on magnetic resonance imaging that was performed after stroke. The lesion was located in the ventrolateral portion of the midrostral upper pons (Fig. 1). The lesion did not enhance with gadolinium. Their topographic localization corresponded to the pontine reticular formation, where the neural network generating rapid eye movement (REM) sleep is presumed to be located.

Figure 1

Flair-weighted magnetic resonance image shows a hyperintense lesion without mass effect in central portion of the upper pons.

A polysomnographic study revealed heavy snoring and moderate obstructive sleep apnoea syndrome with mean oxygen saturation (pulse oximetry) of 87%. The arousal index was 18.4, most of which was due to disturbed breathing. The patient awoke with a severe headache 70 min after the onset of sleep, and this coincided with the first REM period.

The morning headaches of dull nature resolved immediately after treatment with continuous positive airway pressure but nightly headache remained. We did not prescribe lithium due to his hypertension. Instead, we recommended caffeine (a cup of coffee before bedtime), but there was no benefit. Indomethacin was then tried (100 mg orally) for a month. This treatment slightly reduced the pain intensity, but showed no benefit on the development or frequency of headache.

Discussion

An infrequent hypnic headache syndrome only develops during night sleep on a nearly daily basis, and it usually occurs at the same time (1–4). The neurological and neuradiological findings are normal. The pathomechanism of hypnic headache is still unknown. Raskin et al. suggested that these headaches could be associated with REM sleep and be caused by some alteration of the biological pacemaker that is modulated by the serotonergic system (1). This idea was suggested on the clinical basis of the timing of the headaches, the cyclic recurrence of the headaches through the night, and the awakening of some patients with a headache during a vivid dream(1–3). Polysomnography is helpful in defining the relationship between the sleep cycle and this headache and in elucidating the sleep stage during which these headaches consistently occur. In another report, it was captured in a patient arising out of REM sleep at a time of severe oxygen desaturation (5).

In our patient, stereotypical nocturnal headaches occurred nightly for 2 years after a stroke. The age at onset, the frequency, intensity, duration, location and type of pain, and the lack of associated autonomic features are all typical of hypnic headache. While undergoing polysomography, the patient awoke with a typical headache during the onset of the first REM period, thus reinforcing the likelihood that hypnic headache syndrome is likely to be a REM-related phenomenon. Surprisingly, the ischaemic lesion corresponded to the pontine reticular formation, where the neural network generating REM sleep is located.

REM sleep is mediated by a more discrete system in the pons and includes the dorsolateral pontine reticular formation, particularly the region just ventral and lateral to the locus ceruleus. The dorsolateral pontine neurons, interacting with other neurons in the brainstem and forebrain, are involved with reciprocal activation and deactivation of the neuronal groups that mediate REM sleep. The suprachiasmatic nuclei of the hypothalamus, the ‘biological clock’, project to and receive afferent from the brainstem periaqueductal grey, so that a functional link to the pain-modulating system is feasible. Furthermore, serotonergic-containing terminals arising from the dorsal raphe nuclei distribute in a dense plexus in the suprachiasmatic nuclei. The dorsal raphe nuclei have been implicated recently as a headache-generating locus (6).

There is substantial evidence that lithium carbonate stabilizes and enhances serotonergic neurotransmission in the central nervous system, and suppresses REM sleep (2). The theory that a biological clock is a clinically relevant site of lithium's action is supported by the effectiveness of this agent in phasic disorders such as cluster headache, hypnic headache syndrome, cyclical migraine and bipolar illness (6).

We posited that the lesions within the pontine reticular formation would alter the relationships between REM sleep generator and chronobiological centre and cause the hypnic headache syndrome in our patients. The headache might be related to regeneration rather than destruction of the lesion, because headache developed 2 weeks later. There is a striking consistency in clinical characteristics, polysomnographic results, and location of imaged area.

There are no previous reports of symptomatic hypnic headache with brainstem lesions that suggest direct evidence of a potential pathophysiological mechanism and a relationship of hypnic headache to the sleep cycle. Finally, our experiences help clarify the relationship between this type of headache and REM sleep, which may shed light generally on the biological association between sleep and headache.