Abstract
Symptomatic cluster-like headaches have been described with lesions of the trigeminal and parasympathetic systems. Here, we report the case of a 44-year-old woman with continuous auricular pain and a positional cluster-like headache associated with red ear syndrome. Clinical data and morphological investigations raised the hypothesis of a neurovascular compression between the C3 root and vertebral artery. Neurosurgical exploration found a fibrosis surrounding both the C3 root and the vertebral artery. The excellent outcome after microvascular cervical decompression suggests a causal relationship between the cluster-like headache and the vertebral constraint on the C3 root.
Introduction
Cluster headache is a severe retro-orbital and/or temporal pain characterised by a recurrent circadian pattern of unilateral short-lasting attacks associated with both cranial parasympathetic activation (i.e. lacrimation, rhinorrhoea) and sympathetic dysfunction (i.e. Claude-Bernard Horner sign) (1). Although most cluster headaches are defined as primary headaches, patterns mimicking cluster headache are more often related to lesions of the trigeminal nerve and parasympathetic nervous system (2–4). Clinical features have been proposed for the identification of symptomatic cluster-like headaches (CLHs), such as onset of disease >42 years, abnormal neurological examination, attacks lasting more than 180 min or extratrigeminal localisation of pain (3). Here, we report the case of a patient with positional CLH projecting on the trigeminal and extratrigeminal areas associated with red ear syndrome. Surgery found a fibrosis surrounding both the C3 root and the vertebral artery (VA) that responded dramatically to microvascular cervical decompression.
Case report
A 44-year-old woman without any history of headache was admitted to our pain centre because of chronic headache attacks that had started 7 months earlier. Initially, she described only mild paresthesiae of the right ear lobe, but, within 3 months, she developed a typical red ear syndrome concomitant with stabbing and shooting pain attacks in the ear. The pain attacks then spread to the right jaw, right shoulder and right eye. The patient reported ipsilateral lacrimation, nasal congestion, rhinorrhoea and hypersialosis, without any other neurological signs. The initial attacks lasted 90–120 min and occurred one or two days per week. The frequency of these attacks increased progressively to one to three times per day. The attacks were either spontaneous or triggered by the activities of lowering her head and inclining it towards the right or performing hard physical activities. Subcutaneous injection of 6 mg of Sumatriptan stopped the CLH attacks within 5 min but was ineffective on the auricular pain. This dramatic response remained constant with time. Introduction of Verapamil (360 mg/day) initially decreased the frequency of CLH attacks by 30%. Later, the patient developed severe continuous otitis-like pain with unpleasant dysesthesiae on the right ear and continuous cervicobrachial pain that caused her to stop working. Because of adverse effects, Verapamil titration could not be increased; therefore, we decided to add anti-epileptic treatments (gabapentine, topiramate, clonazepam) and non-steroidal anti-inflammatory drugs including indomethacin. These different treatments were ineffective.
Neurological and otolaryngological examinations were completely normal. Interestingly, we reproduced the usual attacks by applying deep pressure to the paravertebral area at the emergence of the C2 and C3 roots or by asking the patient to perform an anteflexion of the head combined with a rotation toward the right side. Laboratory tests were normal. Nociceptive-evoked potentials elicited by laser CO2 stimulation in the areas innervated by the mandibular nerve and the C3 root were normal in latency and amplitude. Brain magnetic resonance imaging (MRI) and magnetic resonance angiography were normal and eliminated any arterial dissection. On conventional angiography, during the injection of the right vertebral artery (VA), rotation of the head toward the bottom right induced a mild plicature of the right VA at C3/C4 level and reproduced the usual CLH attack, whereas a neutral position of the head or rotation of the head toward the left side did not induce any radiological or clinical signs (Figure 1). The mechanical triggers that were observed both clinically and during angiography raised the hypothesis of a neurovascular compression between the right VA and the third cervical root.
Conventional arteriography. During the injection of the right vertebral artery (VA), when the head was in the neutral position (A, coronal view; B, lateral view), the patient reported no pain, the right VA seemed normal at C3/C4 level. On turning her head towards the bottom right (C), the patient immediately reported an ipsilateral CLH, while we observed a mild plicature of the right VA (white arrow).
Because the pain remained intractable, neurovascular decompression of the right VA was performed 14 months after the onset of attacks. An anterolateral cervical approach was used. When the anterior edges of the transverse foramen of the C3 and C4 vertebrae were removed, we observed that the C3 root and the VA were surrounded by a significant fibrosis in the transverse canal. Microvascular dissection made it possible to separate the fibrosis from the C3 root and VA. No abnormality was seen at the level of the C2 root. After 24 h, the patient experienced pain relief in the right ear and was CLH-free. We observed mild hypoesthesia in the cutaneous area innervated by the right C3. During the postoperative course, a dysfunction of her right shoulder related to an incomplete lesion of the spinal nerve was observed. The outcome was excellent under ambulatory physiotherapy. Six months after surgery, the patient stopped taking Verapamil. A few weeks after, she reported one episode with a mild CLH that had been triggered by very intense physical activity. This headache did not require any treatment since it was shorter and less intense than the previous CLH. The patient progressively returned to work as an assistant nurse and has now been CLH-free for 2.5 years without any medical treatment.
Discussion
In the present case, an initial diagnosis of primary chronic cluster headache was proposed on the basis of the characteristics of the attacks, their frequency, and the associated dysautonomic disturbances. However, the presence of four symptoms, each of which alone would not be sufficient to exclude primary cluster headache, raised the hypothesis of symptomatic CLH. The first was the localization of the pain (3). Two recent studies of patients with chronic cluster headache reported localisation of the pain in the ear (25–27% of patients), shoulder (7–25%) and neck (31–35%) (5,6). Among the lower forms of cluster headache that have been previously reported (7–14), pain most often radiated from the trigeminal area to these extratrigeminal areas, whereas, in our case, the pain was initially located in the ear and extended secondarily to the face and neck. The second symptom was re-inforcement induced by head or neck movement. This has been reported by 17% of patients with chronic cluster headache (6), but is very rarely as selective as in our patient (here only one kind of movement was able to trigger CLH). The third symptom, the persistent painful auricular state observed between cluster headache attacks, was intriguing. Even though 42% of patients with chronic primary cluster headache report such a symptom (6), this interictal pain is rarely associated with dysesthesiae. The presence of red ear syndrome represented the fourth warning sign in our case. Even though this syndrome has been observed in different kinds of primary trigemino-autonomic cephalalgia (15–17), it suggested a lesion of the C3 root. The localisation of interictal pain exclusively on the C3 dermatoma and the triggering of CLH by anteflexion of the head towards the right side, but not the left side, represented two other arguments to explore the right C3 root. The dramatic response to Sumatriptan observed in the present case did not exclude a secondary form of cluster headache, since cases of symptomatic CLH responsive to triptans have been described (18). Interestingly, Sumatriptan may stop trigeminal autonomic processes involved in CLHs but this drug has probably no action on neuropathic processes involved in the C3 dermatoma.
We cannot totally discard the possibility of a mere coincidence of primary CLH process and structural cervical lesion since natural remissions of primary cluster-headaches have been described (5). However, the disappearance of CLH attacks observed immediately after neurovascular decompression, and the excellent outcome reported 2.5 years after surgery, suggest a strong causal relationship between this CLH and the cervical neurovascular compression. The recurrence of one brief painful attack triggered by intense physical activity 6 months after surgery may be explained by a residual hypersensitivity or a plasticity of the C3 root.
A wide range of lesions affecting the trigeminal nerve and cranial parasympathetic nervous system can lead to CLHs (2–4), but cervical lesions are infrequent. CLH has been previously reported in cervical meningioma (19), C3–C7 syringomyelia associated with Arnold Chiari malformation (20), cervical spinal epidural abscess (21), a prolapsed disk compressing the C3 root (22) and after stimulation or blockade of the greater occipital nerve (23,24). Vaisman and Nimgade (25) reported a case with a combination of refractory cluster headaches and cervicogenic headaches, both of which were responsive to percutaneous radiofrequency thermocoagulation of the medial branches of the primary posterior rami from C3 to C6. Jansen and colleagues (22) reported pain relief after cervical decompression in 12 out of 16 patients with refractory hemicrania. Surgical microvascular decompression has been proposed in cases of cervicobrachial neuralgia associated with VA loop formation compressing the corresponding cervical root (26), whereas other authors prefer a conservative approach (27).
From a physiopathological point of view, the convergence of dural and cervical afferents in the greater occipital nerve and in the trigeminocervical complex explains how an inappropriate neuronal activation occurring in upper cervical segments can produce activation of second order neurons in the trigeminocervical complex, resulting in trigeminal-autonomic reflex and CLH (28–30). The CLH may also have originated directly from irritation of the neurons innervating the VA, which was completely surrounded by the fibrosis in the transversal canal. Indeed, symptomatic CLH has been reported after lateral medullary and cervical cord infarctions at the C1 or C2 levels (31–35), isolated VA dissection (36,37), ectasy or aneurysm of the VA (38) or subclavian steal syndrome (39). Vascular irritation or compression of the C2 nerve root and VA fibrosis have been reported by Jansen and colleagues (22) in 16 patients with one-sided hemicranial headache attacks that were resistant to conservative treatments (two of whom had CLH attacks).
Conclusions
In practice, patients with CLH triggered by specific cervical movements and clinical signs of C3 irritation must be carefully investigated for a neurovascular compression between the VA and C3 in the transverse foramen. Surgical decompression may be proposed if the chronic pain becomes refractory to any medical treatment.