mise à jour le
29 janvier 2006
Headache 1985;25:394-395
Lithium-triggered chronic cluster headache
Brainin M, Elsenstädter A
Department of Neurology, Niederösterreichisches Landeskrankenhaus
für Neurologie und Psychiatrie Klostemeuburg, Austria


Lithium carbonate, usually helpful in the treatment of cluster headache, showed a paradoxical effect in a patient now being seen for over one year. Chronic cluster headache developed after she had been taking lithium for more than eight years for psychiatric indications. Only by total discontinuation of lithium therapy could a complete remission of the headaches be brought about. This effect was observed repeatedly. It is suggested that a chronic effect of lithium at the receptor sites of the central trigeminovascular pathways plays a role.
It has been clinically established that lithium carbonate can rapidly improve chronic or episodic cluster headaches. The dose is the same as for psychiatric disorders. Though much has changed in the pathophysiological concepts of cluster headache in recent years, its mechanism has remained obscure.
A female patient is described in whom lithium therapy, started for depressive episodes eight years previously, repeatedly triggered chronic cluster headaches. To our knowledge, this is the first case in which lithium showed such a paradoxical effect
A 45-year-old female patient was refered by a neurologist for chronic cluster headache occurring during lithium carbonate therapy. Only discontinuation of oral lithium brought about symptom-free periods but lithium had to be prescribed again for recurring severe depressions.
On admission the otherwise healthy woman complained of severe facial pain starting rather abruptly with a painful burning sensation in the inner right nose and at the corner of the right eye spreading perorbitally followed by slight swelling and reddening of the upper eyelid, ciliary injection, increased lacrimation and occasionally increased rhinorrhea on the right side. Such attacks occurred either spontaneously or by sneezing, yawning or eating spiced food, especially with vinegar. Within the past two years these painful sensations had been experienced only occasionally, but in the preceding four weeks had reached an unbearable intensity. They now occurred in closely packed groups, especially in the morning.
For eight years the patient had been regularly treated with oral lithium carbonate for endogenous depressions. Since the beginning of the therapy no severe depressive episodes had been experienced and no hospitalization had been necessary. Before that, the patient, having had many depressive phases since her 18th year, had been frequently admitted to psychiatric institutions.
No neurological abnormality and no signs of lithium intoxication were noted clinically. Lithium blood cells were 0.8 meq/l. All routine laboratory tests were normal. ACT of the head was unremarkable. Carotid arteriography showed no intracranial pathology, nor did the neck arteries show any irregularities. Rhinological and ophthalmological examinations were normal. Thyroid function was normal. The EEG showed some diffuse dysrhythmic activity without paroxysmal activation. Electroneurographic studies of the peroneal and median nerves showed no slowing of motor or sensory nerve conduction velocity. Transcutaneus ultrasonic doppler flow examination of the supratrochlear artery were registered in the symptom-free interval and during a provoked attack. A marked flow acceleration on the affected side was recorded in the attack phase thus impying increased flow in the ophtalmic artery.
Within the following weeks no therapeutic response was seen with pizotifen, dihydroergotamine, indomethacine, or pindolol. Inhalation of pure oxygen gave no lasting or definite relief. Also amitriptyline, diazepam or placebo showed no effect. Finally, lithium had to be discontinued and this brought about a completely symptom-free state. Provocation by sneezing or yawning had no effect. Also small amounts of ethanol could not evoke an attack. Because of recurring severe depressive episodes with suicidal ideas, lithium was started again, resulting in new clusters of painful headaches within two days. The daily lithium dose was then reduced to correspond to a blood level of 0.4 meq.IL and carbamazepine was added in a daily dose of 600 mg. With this medication the patient stated that the cluster headaches reoccurred but less intensely. In the follow-ups no significant change was reported by her within one year.
Lithium carbonate is known to bring about almost immediate relief from cluster headache. It is often superior to other drugs but, because of its narrow therapeutic range, has been recommended mainly in cases resistant to conventional prophylactic agents. The therapeutic benefit can be long-lasting, but Ekbom noted "a tendency to diminishing effectiveness of the drug after a few months."
The present case met the diagnostic criteria for chronic cluster headache. We do not assume the cluster headache in this patient to be an equivalent symptom of the depression, as the effect of lithium has been shown to be independent of its antidepressant action. In addition, ultrasonic Doppler flow probes in our patient showed an increased flow in the ophthalmic artery of the affected side during an attack. Though conflicting evidence has been reported on the diagnostic reliability and pathophysiological significance of such findings, identical ultrasonic changes have in one series been shown to be of value in the differential diagnosis between migraine patients and patients with cluster headache. Moreover, recent work has shown that pathological changes of blood flow in cluster headache are secondary events due to primary neuronal processes within the central or peripheral trigeminal pathways.
Though we are unable to give a full explanation of this seemingly contradictory effect of lithium, any interpretation must take into account that the patient received lithium many years prior to her headache symptoms. Possibly chronic lithium intake evoked changes on a molecular level at the local trigemino-vascular receptors that led not only to a wearing-off, but to a paradoxical effect. Of the known long-term effects of lithium at synaptic levels, the most relevant phenomenon may be that chronic lithium intake is known to reduce synaptic tryptophan-hydroxylase activity, thereby causing a disequilibrium of tryptophane and serotonin. This effect is not seen in patients with newly started lithium carbonate therapy.
It is evident that additional factors must be considered: first, because the symptoms are strictly unilateral; and, second, because in a long-term follow-up of lithium patients undergoing careful clinical, laboratory, and neurophysiological testing, no symptoms similar to those in this case have been found.
1. Kudrow L: Lithium prophylaxis for chronic cluster headache. Headache 17:15-18, 1977.
2. Ekbom K: Lithium in the treatment of chronic cluster headache. Headache 17:39-40, 1977.
3. Mathew NT: Clinical subtypes of cluster headache and response to lithium therapy. Headache18:26-30,1978.
4. Ad Hoc Committee on the Classification of Headache. JAMA 179:717-718,1962. 5. Kudrow L: Thermographic and Doppler flow asymmetry in cluster headache. Headache 19:204-208, 1979.
6. Drummond PD, Lance JW: Thermographic, changes in cluster headache. Neurology (Cleveland) 34:1292-1298, 1984:
7. Okayasu H, Meyer JS, Mathew NT, Aman OT, Hardenberg J: Lithium carbonate has no measurable effect on cerebral hemodynamics in cluster headaches. Headache 24:1-4, 1984,
8. Schroth G, Gerber WD, Langohr HD: Ultrasonic Doppler flow in migraine and cluster headache. Headache 23:284-288,1983.
9. Moskowitz MA: The neurobiology of vascular head pain. Ann Neurol 16.,157-168,1984.
10. Mühlbauer HD, Gräf KJ, Müller-Oerlinghausen H. Neuro-endokrinologische Aspekte der Lithium-Prophylaxe. In: Beckmann H, ed. Biologische Psychiatrie, Stuttgart New York: Thieme, pp 211-217 1982