Abstract
There are detailed reports of patients with otherwise typical cluster headache (CH), who never develop autonomic symptoms during the attacks (1). Recently, three cases of cluster-like attacks sine headache (2, 3) have also been reported, suggesting that these two core manifestations of CH may fully dissociate. We describe a patient with CH which evolved from cluster attacks without autonomic symptoms to recurring attacks of cranial autonomic manifestations without pain.
Case report
A 47-year-old businessman was examined because of recurring ocular symptoms. He had a past history of episodic CH beginning 24 years earlier. The cluster periods occurred once every year, or every 2 years, usually by Christmas time, and lasted between 3 and 5 weeks. These bouts were heralded by a series of stabbing-like pain, felt in the left frontal region. Typical cluster attacks developed a couple of days later, consisting of an excruciating pain, recurring every 24–48 h, mostly during the night. The pain was felt in the left fronto-orbital region, radiating through the temple to the left side of the neck. It built up to its maximum intensity in 20 min, and lasted around 60 min in the first attacks of each bout. As the cluster period progressed, the attacks became longer in duration (60–120 min), in such a predictable way that a prolonged attack announced the end of the active period. He never experienced autonomic symptoms. The pain was relieved with tablets of ergotamine tartrate (0.75 mg). During the attack, the patient felt restless and pressed the painful site with ice, or against cold surfaces (mirrors, glasses, etc.), attempting to relieve pain.
His last cluster period began earlier than usual, at the beginning of October 2001. He was prescribed subcutaneous sumatriptan (6 mg) that produced a complete pain relief in less than 10 min. Yet, because of marked drowsiness, he decided not to repeat the injection. Oral sumatriptan (50 mg) had no effect on pain. By the end of October, the attacks became longer, as if the bout was coming to an end. Then, for the first time, he noticed a left-sided ptosis and lacrimation during the attacks. The pain subsided in a few days. However, he began experiencing recurring episodes of left-sided autonomic symptoms without pain that have persisted for more than 2 years.
By the time he was first examined by the authors (March 2002) he was experiencing two or three such attacks daily. The attacks occurred in the morning, at dinnertime or late in the evening, and were obvious to anyone around him, but not always to himself. They consisted of a left-sided conjunctival injection, ptosis and miosis, lasting between 1 and 2 h. There was no pain, but he occasionally felt a cutaneous hyperaesthesia on the left side of the scalp. Friends and colleagues would call his attention to the fact that there was some problem with his eye. He began using dark sunglasses all day long, to avoid this embarrassment. He had been prescribed several oral drugs (naproxen 500 mg b.i.d., indomethacin 75 mg b.i.d. and verapamil 40 mg t.i.d.) with no improvement. Sumatriptan was not tried at this point.
He is a smoker (seven cigarettes and three cigars a day) and drinks three glasses of scotch (150 g alcohol) and about 20 cups of coffee a day. He had suffered from renal calculi in the past, but was otherwise healthy.
He was examined during one attack. He was calm, showing no signs of anxiety or agitation. Physical examination revealed normal blood pressure (110/60 mmHg), marked left-sided ptosis, miosis and conjunctival injection that lasted 1 h. There was no sensory impairment in the distribution of the trigeminal nerve and no abnormalities were noticed in physical or neurological observation. Routine blood and serum chemistries were normal. Brain magnetic resonance imaging (MRI) showed a tortuous left vertebral artery in contact with the anterior pons, but no encephalic abnormalities.
He was advised to abstain from alcohol and was prescribed an initial dosage of lithium carbonate (400 mg daily) with a very transient improvement. Topiramate (25 mg b.i.d.) reduced the attack frequency to three per week. He never increased the dosage. Two years later (February 2004) he continues to experience attacks (at present, once a week, lasting 10–15 min), with no fixed timetable. There is no evident association with alcohol and the scalp hyperaesthesia has disappeared. He decided to stop prophylactic medication 18 months ago. He has not experienced cluster pain since the onset of the autonomic episodes.
Discussion
Pain and cranial autonomic symptoms (CAS) may dissociate in CH. We describe a patient who expressed, sequentially, two possible dissociations: pain alone and chronic recurring attacks of cranial autonomic features sine pain. He experienced typical CH attacks only for a short period of time, but he always kept the duration and recurrent pattern characteristic of CH.
Previous reports have shown that both dissociations are possible: CH without CAS, which is well recognized and described in large series (1, 4, 5) (including cases with an autosomal dominant transmission) (3), and CAS without pain (including proband 1 of the same family) (3). It has also been documented that these atypical syndromes can evolve to or from typical CH (for example, proband 2 of the same family) (2, 3). In addition, there is the report of one patient in whom pain and CAS were simultaneous but developed on opposite sides (6).
CH without autonomic symptoms corresponds to 3–7% of all cluster patients (4, 5, 7). Although it is acknowledged that patients may not be aware of autonomic symptoms (4, 7), it has also been proposed that it represents a milder form of CH (8). Thus, CAS are no longer compulsory for the diagnosis of CH (9), whenever the attack is associated with agitation or a feeling of restlessness, another typical aspect of this syndrome (7, 10, 11).
Our patient presented both dissociations at different times, thus a double dissociation, showing that both symptoms may alternate and not be a direct consequence of each other. It is also noteworthy that agitation and restlessness disappeared with pain, suggesting that they may somehow be linked. The effect of subcutaneous sumatriptan in this acephalgic form was not explored since the patient was unwilling to repeat the injection. Although a tortuous left vertebral artery was seen in the MRI, it is unlikely that the new presentation of symptoms resulted from a neurovascular contact, since it involved only the parasympathetic outflow of the VII nerve, sparing its motor component.
CH is a trigemino-autonomic cephalalgia (TAC) (9, 12), a group of headaches phenotypically related to the activation of a trigemino-autonomic reflex, by which painful stimuli reaching the trigemino-cervical complex activate the superior salivary nucleus (SSN) of the facial nerve and its parasympathethic outflow through the pterygopalatyne ganglion. This will cause lacrimation, rhinorrhoea, nasal obstruction and a release of vasoactive intestinal peptide and nitric oxide, producing vasodilation and increased blood flow in the meningeal and extracranial arteries (13–16), and vasodilation of the internal carotid artery. The sympathetic pericarotid plexus may then become dysfunctional, resulting in ptosis and miosis (17). Symptoms of this nociceptive reflex have been elicited in other circumstances of intense craniofacial pain (13, 18, 19), suggesting that the autonomic component is due to the trigeminal activation (8). Reports of CH without pain show that different components of the syndrome may be activated separately, favouring the modular theory of headache (20).
The modular theory conceptualizes symptoms of primary headaches as being linked to neuronal networks (modules) with a specific anatomo-physiological identity. One may recognize at least three independent modules in CH: (i) a generator module, responsible for the periodicity of the syndrome and initiation of the attacks, located in the homolateral posterior inferior hypothalamic grey matter (21); (ii) the trigeminocervical complex (with a contralateral inhibition module), responsible for the unilateral pain; and (iii) the parasympathetic SSN producing the autonomic manifestations. The first module is preferentially linked to the trigeminal nucleus. From there, the SSN is activated, as occurs in other TACs (CH, chronic paroxysmal hemicranias, SUNCT) and also in trigeminal neuralgia of the ophthalmic branch. Activation of the trigeminal nucleus, however, may not spread to the SSN, in which case one would see cluster attacks without CAS. In addition, on rare occasions (anatomical/genetic variants, different types of sensitization, inhibitory processes), activation of the hypothalamic module spreads directly to the autonomic module, bypassing the locus of pain. This is observed in cases of CAS without pain (2, 3) and after trigeminal root section for the treatment of CH (14).
According to Young et al. (20), a bidirectional process of reinforcement may govern the interaction between two modules. In the case reported, CAS did not add up to a previous syndrome, but actually suppressed the previous pattern. Two years after the onset of this new pattern of symptoms, the patient had no more annual cluster pain episodes. This change somehow modulated the generator module leading to a chronic, instead of a seasonal, pattern.
Reports of dissociations in CH and other trigemino-autonomic cephalalgias may help to individualize syndromes that may not gain the attention of neurologists (interpreted as pure ophthalmological disorders) and may reveal how modules can operate independently. It may also help in understanding the mode of action of certain pharmacological agents.