Abstract
Introduction
The hypnic headache syndrome (HHS) or ‘alarm clock’ headache was first described by Raskin in 1988 (1). Less than 100 cases have been reported in the international literature, while the estimated incidence of the syndrome is 0.07%. The female/male ratio is now considered to be 1.7/1 (2–4). HHS was included in the classification of the International Headache Society (IHS) in 2004, coded under the category ‘4. Other Primary Headaches’ (2, 5).
The HHS is a benign syndrome characterized by recurrent episodes of headache occurring always during sleep. The episodes characteristically result in the awakening of the patient and they are not accompanied by any autonomic symptoms. The onset of the syndrome occurs usually after the age of 50 years, while lithium is the most effective treatment in the case of HHS (2, 3, 6). The pathogenetic mechanism of HHS remains unidentified. No other case of HHS occurring after the withdrawal of long-term lithium treatment has been reported in the literature to date.
Case presentation
A 54-year-old female patient, suffering from bipolar disorder, received lithium treatment for 6 continuous years. Her disorder included mainly depression and less hypomanic episodes. It was the first time she underwent lithium treatment. During the first 2 years of treatment she received 900 mg of lithium divided into three daily dosages. In the following 4 years she received 600 mg of lithium/24 h as prophylactic therapy. While she was on the medication the patient was under close monitoring, having frequent measurements of lithium plasma concentrations, which remained at therapeutic levels (0.8–1.5 mEq/l). The treatment stopped due to side-effects regarding mostly mild renal dysfunction and electrocardiographic alterations. Furthermore, the patient was stabilized regarding her mood disorder presentations. Lithium withdrawal took place slowly (gradual discontinuation over 30 days). One month after the withdrawal of the medication she developed episodic headache, which occurred always during nocturnal sleep, approximately 4 h after its onset. The episodes woke her from sleep yet she did not report the coincident dreams before the headache attacks. The pain was described as dull, of moderate intensity and unilateral frontoparietal location. The duration of each episode was 2–3 h. The patient was relieved from the pain without the use of analgesics. The frequency of the episodes was one every night, while the whole disorder lasted for about 1 month.
The patient did not report the coincidence of other symptoms such as nausea, vomiting, photophobia, phonophobia or osmophobia. There were no signs of autonomic nervous system (ANS) disorder. The headache attacks' onset was not associated with any exogenous factors. The episodes remitted spontaneously without the administration of medical treatment.
The patient did not suffer from hypertension, lung disease, migraine or tension headache, obstructive sleep apnoeas or other sleep disorders, except for episodes of insomnia while she was on lithium administration. The episodes remitted spontaneously. She did not report insomnia when the headache attacks occurred. She did not receive drugs that might have affected her sleep architecture yet she did not undergo polysomnography during or after lithium treatment or during the period of time when she developed the hypnic headache attacks. Polysomnography was not performed due to the limited duration and transient character of the hypnic headache syndrome. The family history was free regarding migraine and other forms of headache. A thorough neurological examination was normal. The systolic blood pressure ranged from 110 to 130 mmHg. The conducted blood tests including full haematological and biochemical screening, ESR, vitamin B12 and thyroid hormone levels were within normal limits. The EEG and the brain MRI scan were normal.
The patient did not report any similar or different headache attacks during the next 2 years. She is currently receiving the following treatment: clorazepate dipotassium 60 mg/24 h, venlafaxine—HCl (XR) 300 mg/24 h and valproate sodium 1500 mg/24 h. The new treatment started after the headache episodes remitted, approximately 2 months after the withdrawal of lithium, due to relapse of the bipolar disorder. Clorazepate was administrated in order to control concomitant anxiety symptoms while valproate was used as mood stabilizer.
Discussion
Several primary headaches, including migraine, chronic paroxysmal hemicrania and cluster headache, occur frequently during sleep and mostly in the early morning hours. As far as cluster headache is concerned, 50% of the attacks occur exclusively during sleep, while migraine attacks occur mostly between 06.00 and 08.00 (7). Hypnic headache develops exclusively during sleep and the episode always awakens the patient. In the presented case the patient did not mention any symptoms accompanying the headache while the clinical features fulfill the IHS diagnostic criteria for HHS (5).
The interesting point in the presented case, which differentiates it from the other cases of HHS, is the time of onset of the headache attacks, which was shortly after the withdrawal of the lithium medication. No similar case of transient HHS has been reported in the literature to date.
The pathogenetic mechanism of hypnic headache is not yet fully recognized. It is probably a chronobiological disorder because the episodes happen exactly at the same time and only during sleep (3, 8). The hypothalamus is considered to play a part in the development of HHS. It projects directly to the locus coeruleus, the periaqueductal grey matter, the dorsal raphe nuclei (pain control and sleep regulation centers) and certain brain stem structures that are involved in the ANS control (6). Melatonin is also considered to play significant part in the pathophysiology of HHS.
The hypothalamic suprachiasmatic nucleus, through its projection to the pineal gland, is essential for the melatonin secretion. It is also the main pacemaker of the endogenous circadian rhythm and the regulator of the sleep-alertness circle. During the ageing process the suprachiasmatic nucleus cells undergo gradual degeneration. The result of this degeneration is the reduction of the function of the hypothalamus-pineal gland circuit and the consequent disorder of melatonin secretion. This process probably explains the main incidence of HHS in the elderly even though there is a report of a 9-year-old patient suffering from the syndrome (7–9). Another hypothesis suggests that the age-related functional impairment of the hypothalamic suprachiasmatic nucleus could activate disnociceptive procedures, leading to awakening and pain due to reduction of the dorsal raphe nuclei function (4).
Tryptophan and serotonin are precursor substances for the synthesis of melatonin. Melatonin has a circadian secretion rhythm and plays an important role in the pathophysiology and suppression of headache through variant mechanisms, which are the following: (i) anti-inflammatory and anti-oxidant action; (ii) inhibition of NO synthesis; (iii) inhibition of dopamine secretion; (iv) stabilization of the cell membrane; (v) opioid analgesic properties; (vi) inhibition of glutamic acid and its neurotoxic effect; (vii) neurovascular regulation; (viii) modification the total GABAergic activity; (ix) regulation of the serotoninergic activity; and (x) chemical structure similar to that of indomethacin (10).
Polysomnographic studies have shown that hypnic headache attacks usually occur during the REM phase of sleep. A possible explanation for this observation is the lack of activity in the locus coeruleus and in dorsal raphe nuclei during the REM phase of sleep. These structures, along with the periaqueductal grey matter, are essential centres of the human analgesic system and they are in constant interaction with the hypothalamus. We should add though that there are studies presenting cases of hypnic headache attacks that occur during the NREM phase of sleep. There is also evidence that migraine, as well as cluster headache attacks, is also related to the REM phase of sleep. These observations suggest that the relationship between HHS and the REM phase of sleep is not a specific one (7, 11–13).
Lithium is the most effective treatment in the case of HHS. The preferable dosage in terms of prevention is 400 to 600 mg once a day, administrated usually at night. Lithium has been used for the therapeutic management of various chronobiological disorders such as cluster headache and bipolar disorders. It accumulates selectively to the hypothalamus and one of its actions is the increased tryptophan absorption and the consequent enhancement of serotonin synthesis. Furthermore, lithium is involved in the serotonin storage, release and receptor uptake, leading to increased serotoninergic neurotransmission in the CNS (8).
We believe that in the presented case the withdrawal of lithium, which was administrated for a long period of time, modified the regulation of serotonin and melatonin levels, resulting in the emergence of hypnic headache. This hypothesis may support the implication of the hypothalamus and melatonin secretion in the pathophysiology of HHS.