Abstract
Short-lasting unilateral neuralgiform headache with conjunctival injection and tearing (SUNCT) is a primary headache syndrome that has been reported to be resistant to treatment with intravenous lidocaine. We report four cases of SUNCT in whom intravenous lidocaine (1.3-3.3 mg kg−1 h−1) completely suppressed the headaches for the duration of the infusion. The headache returned after cessation of treatment. Two patients went on to have their symptoms controlled on topiramate (50-300 mg daily). One patient had typical migrainous aura in association with some of the attacks of pain but never migrainous headaches. These cases suggest that treatment with lidocaine can be considered when acute intervention is required to suppress a severe exacerbation of SUNCT, and further broaden the therapeutic and clinical background of this syndrome.
Introduction
Short-lasting unilateral neuralgiform headache with conjunctival injection and tearing (SUNCT) syndrome is a rare form of strictly unilateral headache that occurs in association with prominent cranial autonomic features (1). The pain is abrupt in onset, of short duration, stabbing or burning in quality, and accompanied by prominent ipsilateral cranial autonomic features, particularly lacrimation and conjunctival injection. The natural history of the condition is poorly understood. Generally, a clustering pattern is reported, with active phases erratically alternating with remissions. During symptomatic periods, the frequency of attacks varies from less than one per day to more than 30 per hour. Functional magnetic resonance imaging (fMRI) in spontaneous attacks has demonstrated activation of the ipsilateral hypothalamic grey (2).
Diagnostic criteria for SUNCT syndrome have been suggested (3) that require at least 30 attacks of unilateral, moderately severe, orbital or temporal, stabbing or throbbing pain, lasting for 15–120 s and associated with at least one of the following cranial autonomic features: conjunctival injection, lacrimation, nasal congestion, rhinorrhoea, ptosis and eyelid oedema. The attacks have to occur at a frequency of 3–100 daily during the active phase. SUNCT was not included in the first edition of the International Headache Society classification (4) but will appear in the second (5). Until recently, SUNCT was thought to be highly refractory to treatment; several categories of drugs used in the treatment of headache syndromes, including intravenous lidocaine, have been reported to be ineffectual (6). We report four cases of SUNCT syndrome highly responsive to intravenous lidocaine.
Case material
Case 1
A 60-year-old, right-handed female presented with a 5-year history of headaches in September 2000. There were no precipitants at onset. A typical attack was strictly unilateral on the right. The pain was either centred on the orbit and forehead with radiation to the upper lip and gum or centred on the upper lip and gum with radiation to the orbit and cheek. The pain was severe and had an electric shock-like quality.
The usual duration of the pain was 10–30 s, although it could range from 5 s to 2 min. The pain came on rapidly, was maintained at a plateau phase and then resolved rapidly. She denied any superimposed spikes or variations in the severity of the pain. At presentation, she had 5–30 attacks per hour when awake but was pain free during sleep. The attacks were associated with prominent ipsilateral conjunctival injection, lacrimation, rhinorrhoea and eyelid oedema. She denied other cranial autonomic features, and had no nausea, vomiting, photophobia, phonophobia or osmophobia. She felt restless during the attack. She denied any aura symptoms. The attacks could be triggered by talking, chewing, touching the right upper lip, brushing her hair on the right side of the head, blowing her nose, sneezing, wind blowing on the face or sudden jerking movements. There was no refractory period to triggering after an attack.
Between the onset of the headaches in 1995 and 1997, she had bouts of 5–10 episodes daily for a few weeks alternating with remission periods lasting a few weeks. In addition, the headache showed a slight preponderance of being centred on the right upper lip and gum rather than the right orbit and forehead, and was shorter in duration, lasting 5–15 s. She was started on carbamazepine 300 mg daily from which there was a moderate benefit; the carbamazepine dose was increased to 700 mg daily and phenytoin 200 mg daily was added without any further improvement, but the patient became moderately ataxic. In April 1997, right trigeminal radio-frequency thermocoagulation was performed, after which she was pain free. Post-operatively, she had marked hypo-anaesthesia over the right maxillary and mandibular trigeminal distributions; trigeminal motor function and reflexes were preserved. The sensory changes gradually improved over 2 years.
The headaches recurred in May 2000 in the absence of any obvious precipitants. She had daily attacks without any remission periods. Prior to the presentation, she had trials of gabapentin 900 mg daily, fluoxetine 20 mg daily, indomethacin 50 mg three times daily, mefenamic acid 500 mg three times daily, ibuprofen 400 mg three times daily and paracetamol/codeine combined analgesics, all of which were ineffective. At presentation she was taking carbamazepine 800 mg daily, which had moderately reduced the frequency of the attacks though the severity had remained unchanged. Previous medical history included hypertension treated with lisinopril 5 mg daily. There is no family history of headaches. She is a non-smoker and rarely drinks alcohol.
General and neurological examination was normal except for a small patch of mildly diminished sensation below the right eye in the distribution of the maxillary division of the trigeminal nerve. She weighed 95 kg. Routine haematological and biochemical screening was normal. An MRI scan of the brain was normal including the posterior fossa. A modified Indotest (double-blind, placebo-controlled intramuscular indomethacin 100 mg) was negative. A diagnosis of SUNCT syndrome was made. The patient was admitted to hospital because the pain had become unbearable and she was unable to eat or talk. An infusion of intravenous lidocaine was started at a rate of 1 mg/min. Three hours after initiation of the lidocaine infusion, the frequency of the headaches reduced to 1–5 attacks per hour. In addition, the patient reported that the severity of the attacks was markedly reduced. However, she developed marked nausea and vomiting which necessitated treatment with domperidone 30 mg four times daily. The intravenous lidocaine infusion rate was increased to 2 mg/min (1.3 mg kg−1 h−1), which completely suppressed the headaches. Unfortunately this led to the recurrence of nausea and vomiting which were inadequately controlled despite optimum antiemetic treatment; for this reason, the infusion rate was reduced to 1 mg/min. The intravenous lidocaine infusion was continued for the next 7 days, during which period she had only 0–5 attacks per hour. The infusion was stopped on three occasions and on each occasion the headache frequency increased dramatically 30–45 min later. When the infusion was restarted, the headache frequency reduced within 15–30 min. While the patient was receiving the lidocaine infusion, she reported that the pain was bearable and she was able to eat and talk normally.
While the patient was receiving the lidocaine infusion, the carbamazepine dose was reduced to 200 mg twice daily and she was started on topiramate 25 mg once daily; the topiramate dose was increased over 4 days to 75 mg twice daily, before the lidocaine infusion was stopped. On this combination of carbamazepine 400 mg daily and topiramate 150 mg daily, she had less than five attacks daily. Over the next year, the carbamazepine dose was gradually reduced before being stopped completely and the topiramate dose gradually increased to 150 mg twice daily. Topiramate 300 mg daily has rendered the patient pain free for the last 13 months. The only side-effect she reports is mild hypersomnolence. She has attempted to reduce the topiramate dose on two occasions but the headaches recurred within 1–2 days.
Case 2
A 60-year-old, right-handed, male taxi driver presented with a 12-year history of headaches in November 2001. There were no precipitants at onset. Until 1997, he had 2–4 headaches per week; they usually occurred at night, waking him from sleep, and were of moderate intensity. In 1997, there was a marked increase in the frequency and severity of the headaches over a few weeks; there was no obvious trigger for this deterioration. From 1997 up to his presentation, there was a further gradual worsening of the headaches. Over this period, he experienced only one spontaneous remission of 2 weeks’ duration. A typical attack was strictly unilateral on the left, centred on the retro-orbital and orbital regions with radiation to the temple, nose and left upper teeth. The pain was very severe and had a stabbing or burning quality. The usual duration of the pain was 5–20 s, although it could last for up to 1 min. The pain came on rapidly, was maintained at a plateau phase and then resolved rapidly. He denied any superimposed spikes or variations in the severity of the pain. He had 200–300 attacks daily at presentation. The attacks were evenly distributed over the 24 h. The attack frequency was not altered by sleep and, in fact, the patient was woken up at regular intervals by the pain, which resulted in poor sleep quality. The attacks were associated with prominent ipsilateral conjunctival injection, lacrimation and ptosis. He denied all other cranial autonomic features, nausea, vomiting, photophobia, phonophobia or osmophobia. Movement had no effect on the pain and he did not feel restless during the attack. He denied any aura symptoms. Infrequently and inconsistently, stroking his hair on the left side of the head and chewing could trigger the attacks. There was no refractory period after an attack. There was no past history of headaches.
The patient had previously tried carbamazepine 400 mg daily, gabapentin 2400 mg daily, lamotrigine 50 mg daily, indomethacin 100 mg daily, amitriptyline 75 mg daily, clonazepam 1 mg once daily, baclofen 15 mg daily, tramadol 400 mg daily, buprenorphine 600–1200 µg daily and dihydrocodeine 240 mg daily. At presentation he was taking sodium valproate 2 g daily and imipramine 150 mg daily (for depression). There had been no improvement with any of these medications. In March 2000, a left trigeminal microvascular decompression was performed following which the patient's symptoms worsened. He had also previously tried a transcutaneous electrical nerve stimulation (TENS) machine, acupuncture and a maxillary bite appliance, all without any benefit. In the past medical history he had had depression for 10 years, which was well controlled on tricyclic antidepressants. His mother had a 8-year history of trigeminal neuralgia, which was well controlled on carbamazepine. He is a teetotaller and smokes 30 cigarettes daily.
A detailed general and neurological examination was normal. He weighed 93 kg. Routine haematological and biochemical screening was normal. A MRI scan of the brain (with gadolinium), including dedicated trigeminal cuts, did not reveal any abnormalities. A modified Indotest, as described above, was negative. A diagnosis of SUNCT syndrome was made. An infusion of intravenous lidocaine was initiated at a rate of 1 mg/min and the rate increased after 4 h to 2 mg/min (1.3 mg kg−1 h−1). Two hours after initiation of the lidocaine infusion, the headaches reduced in severity and frequency by approximately 50%. The headaches were suppressed completely about 15 min after the infusion rate was increased to 2 mg/min. The intravenous lidocaine infusion was continued at the rate of 2 mg/min for 48 h; during this period the patient did not have any headaches. The headaches recurred 15–20 min after the infusion was stopped. Over the next 3 days, the intravenous lidocaine infusion was administered at a rate of 2 mg/min for 10 h overnight; on each occasion the headaches were completely suppressed within 15–30 min of initiation of the treatment and recurred 15–30 min after the treatment was stopped. The patient denied having any side-effects.
Sodium valproate was stopped. Topiramate was started at a dose of 12.5 mg once daily and the dose increased to 25 mg twice daily over 1 week. On topiramate 50 mg daily the SUNCT attacks were completely suppressed. The patient has now been pain free on topiramate for 1 year. He denies any side-effects with this agent. He has attempted to stop topiramate on three occasions but the headaches recurred within 2–4 days.
Case 3
A 49-year-old, right-handed male presented with a 21-month history of headaches in December 2002. There were no precipitants at onset. Initially the headaches occurred infrequently but over a few months become more frequent and daily. A typical attack was strictly unilateral although the side could vary, being equally frequent on both sides. The pain was centred on the retro-orbital region and the temple. It was very severe and had a boring quality. The usual duration of the pain was 30 s to 1 min. The pain came on rapidly, was maintained at a plateau phase, and then resolved rapidly. He denied any superimposed spikes or variations in the severity of the pain. The attacks would occur between six and 30 times daily, and could wake him from sleep. There was associated conjunctival injection, lacrimation, ptosis, eyelid oedema and nasal congestion; although there was prominent conjunctival injection, the lacrimation was not prominent. He felt restless during an attack. He denied other cranial autonomic features, and had no nausea, vomiting, photophobia, phonophobia, osmophobia or any aura symptoms. The attacks could be triggered by touching the face. There was no refractory period and attacks could occur repetitively. There was a constant mild to moderate background pain bilaterally across the forehead. This constant pain was completely featureless.
He has tried paracetamol/codeine combinations, aspirin, diclofenac, indomethacin 50 mg daily, celecoxib 200 mg daily, lofepramine 210 mg daily, sodium valproate 600 mg daily, pizotifen 1.5 mg daily and prednisolone 60 mg daily, all of which were ineffective. A trial of high-dose, high-flow-rate oxygen inhalation (100% at 12 l/min) and subcutaneous sumatriptan 6 mg had no effect on the short-lasting attacks or the constant background pain.
In the past he had supraventricular tachycardias for which he was treated with propranolol 160 mg for 5 years. He had chemical meningitis 18 years previously after exposure to a bacterioside, BTZ-RX-31. He had also undergone a right inguinal hernia repair 2 years previously, and surgery for a calcaneal spur in 1996. He did not smoke, drank very little alcohol, and worked since the age of 18 in the paper processing industry. There was no family history of headaches. On examination he weighed 89 kg. His heart rate was in sinus rhythm. A detailed neurological examination was unremarkable. Prior to presentation at our clinic, a gadolinium-enhanced MRI of the brain revealed scattered white matter lesions with no enhancement and he was found to have a reduced creatinine clearance (72 and 87 ml/min on two separate occasions). Subsequently, he had been investigated for an inflammatory disorder: blood tests including erythrocyte sedimentation rate, immunoglobulins, serum electrophoresis, autoantibodies including extractable nuclear antigen and double-stranded DNA, anti-thrombin III and lupus anticoagulant were all normal or negative; cerebrospinal fluid examination for protein, blood, glucose and oligoclonal bands was normal; skin biopsy for CADASIL was negative; and cerebral angiography, renal biopsy and a right frontal cerebral biopsy were all normal.
We made a diagnosis of SUNCT syndrome. The patient was admitted to hospital and commenced on an intravenous lidocaine infusion at a rate of 1 mg/min (0.67 mg kg−1 h−1); on this dose the patient reported a reduction in the duration of the headaches after 5 h, but there was no change in attack frequency. In addition, he developed mild nausea. The infusion rate was increased to 2 mg/min (1.33 mg kg−1 h−1) with which both the frequency and severity of attacks were reduced. The infusion rate was further increased to 3 mg/min (2 mg kg−1 h−1) with which both the short-lasting attacks and the constant background headache were completely suppressed. However, he complained of persistent nausea which required treatment with domperidone. Unfortunately, the patient developed marked depressive thoughts 6 h after the infusion rate was increased. Cessation of the lidocaine infusion led to the recurrence of the headaches at the usual frequency and severity, and resolution of depressive thoughts and nausea, within 12 h. The patient is currently undergoing a trial of lamotrigine.
Case 4
A 53-year-old, right-handed, female nurse presented in November 2002 with a 5-month history of headaches. There were no precipitants at onset. The first episode began with the sudden onset of right-sided facial tingling involving the frontal and maxillary region. The sensory symptoms lasted about 5 min. After 15 min, she suddenly developed an excruciating, sharp pain in the right eye that lasted 10–15 min. Two days later, she had another episode that was identical except that the right-sided facial paraesthesia were followed by a numb sensation. Three days later, she had a further episode that was identical but she was left with a continuous background ache and numbness in the same region. Since that time she has had 1–10 attacks daily. A typical episode starts with right facial tingling in the frontal and maxillary region that lasts 5–10 min. After a further 5–15 min she develops the pain that is strictly unilateral on the right, centred on the retro-orbital and orbital regions. The pain is excruciating and has a sharp and burning quality. The usual duration of the pain is 8–20 min. The pain comes on rapidly, is either maintained at a plateau phase or has a fluctuating, saw-tooth pattern, before resolving rapidly. The attacks are associated with prominent ipsilateral conjunctival injection, lacrimation, nasal blockage and ptosis. She also gets phonophobia and right-sided (homolateral) photophobia in association with the headaches. She denied all other cranial autonomic features, nausea, vomiting or osmophobia. She feels restless during the pain. The headaches usually occur when she is awake, although infrequently she is woken from sleep by an attack. There is no periodicity to the attacks. The attacks can be triggered by eating, wind blowing on the face or having a shower. Alcohol does not trigger a headache. There is no refractory period to triggering after an attack. There was no past history of headaches.
The patient had previously tried carbamazepine 900 mg daily, indomethacin 25 mg three times daily, amitriptyline 25 mg daily and morphine sulphate 7.5 mg four times daily, all of which were ineffective except carbamazepine, which led to a slight reduction in the severity of the pain. In the past medical history, she had Herpes Zoster in the right maxillary distribution of the trigeminal nerve in August 2001, which was treated with a course of acyclovir; the numbness and burning sensation resolve completely within a few weeks. She had post viral fatigue 5 years ago. Her son has migraine. She is a non-smoker and drinks alcohol occasionally. General and neurological examination was normal except for diminished sensation in the distribution of the maxillary and mandibular divisions of the trigeminal nerve. Trigeminal motor function and reflexes were normal. She weighed 54 kg. Routine haematological and biochemical screening was normal. An MRI scan of the brain with gadolinium contrast did not reveal any abnormalities. A lumbar puncture revealed normal opening pressure and constituents. A modified Indotest, as described above, was negative. She had trials of subcutaneous sumatriptan 6 mg, inhaled oxygen (100% at 12 l/min for 15 min) and intranasal lidocaine (two drops of 4% solution in each nostril), all of which were ineffective. Sublingual nitroglycerine 1.2 mg did not trigger a headache. A diagnosis of SUNCT syndrome was made.
An infusion of intravenous lidocaine was started at a rate of 1 mg/min and increased to 2.7 mg/min (3.0 mg kg−1 h−1) over 24 h. She was continued on this infusion rate for 4 days during which period the headache frequency reduced to one to two attacks daily and the severity of the attacks was moderately reduced; in addition, the background headache was completely suppressed. However, she developed marked nausea that necessitated treatment with intravenous granisetron 1 mg three times daily. The lidocaine infusion rate was then increased to 3 mg/min (3.3 mg kg−1 h−1), which completely suppressed the headaches. Unfortunately, the patient developed paranoid ideation 23 h after the infusion rate was increased. Cessation of the lidocaine infusion led to the recurrence of the headaches at the usual frequency and severity, and resolution of paranoia and nausea, within 24 h. The patient is currently undergoing a trial of lamotrigine.
Discussion
We describe four cases of SUNCT that respond very well to intravenous lidocaine. Each had rapid improvement in attack frequency, or severity, or both, and had a resumption of attacks after the infusion was ceased. We find intravenous lidocaine very useful in the hospital management of SUNCT provided care in administration is observed in terms of cardiac monitoring, control of nausea and awareness of the psychiatric manifestations.
The phenotype of the headache for cases 1, 2 and 3 satisfies current suggested diagnostic criteria for SUNCT, though the strictly unilateral but side-variable attacks manifested by case 3 are atypical; strictly unilateral, side-altering SUNCT syndrome has been described (7, 8). In case 4, the patient described strictly unilateral, excruciating headaches centred on the ophthalmic trigeminal distribution and associated with prominent ipsilateral cranial autonomic features, on a background of constant pain. The differential diagnosis for these episodes includes cluster headache, paroxysmal hemicrania, hemicrania continua and SUNCT syndrome. Paroxysmal hemicrania and hemicrania continua are unlikely in view of a lack of response to indomethacin (4, 9). Cluster headache is equally unlikely in view of lack of: circadian periodicity; triggering by alcohol and nitroglycerin; and lack of response to subcutaneous sumatriptan, oxygen and intranasal lidocaine (10). Though the headache is relatively long-lasting, several features of the clinical phenotype are consistent with a diagnosis of SUNCT syndrome including: the site, intensity and character of pain; frequency of attacks; associated prominent conjunctival injection and lacrimation; presence of typical triggers; lack of refractory period; and the partial response to carbamazepine (11). In addition, there are four reported cases of SUNCT syndrome in the literature in which prolonged attacks, lasting 1–2 h, are described (2, 12, 13). A persistent dull interictal ache has been described in association with SUNCT syndrome (12).
Intravenous lidocaine has been demonstrated to provide effective analgesia in a variety of acute and chronic pain states (14, 15). It has been reported to be effective in several headache syndromes including trigeminal neuralgia (16), chronic migraine (17) and cluster headache (18). It has been proposed as treatment for status migrainosis (19) and chronic daily headache (20), although a brief infusion in migraine did not prove effective in one trial (21). Pareja and colleagues have reported that intravenous lidocaine 100 mg bolus followed by 4 mg/min over 48 h was ineffective when administered to two patients with SUNCT syndrome (6). In contrast to their observations, we found that intravenous lidocaine was highly effective at completely suppressing the headaches in our four patients with SUNCT syndrome. Remarkably, we have not encountered a convincing case of SUNCT that has not responded to intravenous lidocaine when we have had the opportunity to use it. Indeed, we have seen other patients who probably have SUNCT, where the phenotype may be argued, who also have responded to intravenous lidocaine. It is not clear why the outcomes are so completely different.
Occasionally, patients are encountered with SUNCT syndrome experiencing severe exacerbations with frequent, easily triggered, high-intensity pain attacks (22, 23). In this situation acute interventions are needed because the patients are severely affected and may not be able to eat or drink because these actions trigger attacks. Our report draws attention to the possibility of utilizing intravenous lidocaine in this situation to ameliorate the attacks temporarily while conventional therapy is being optimized. Table 1 outlines the protocol we employ when administering intravenous lidocaine. Treatment with intravenous lidocaine is associated with significant side-effects, especially when high doses are utilized (24–26). Lidocaine mediates its effect through blockade of sodium channels. It blocks sodium channels in the heart, brain and peripheral nerve, causing adverse effects at each site, including pro-arrhythmic effects, cognitive impairment, dizziness, nausea, and diarrhoea. For this reason, it is imperative to monitor the patient carefully for adverse effects.
Protocol for use of intravenous lidocaine
Cases 1 and 2 reported a complete response to topiramate and withdrawal of the agent rapidly led to recurrence of the attacks. Open-label use of topiramate has previously been reported to be effective in one case report (13). Recently, lamotrigine (8, 27–29) and gabapentin (30–32), used in an open-label fashion in nine and three patients, respectively, have been reported to be effective in SUNCT syndrome. Although the ultimate confirmation of the utility of these drugs in the treatment of this debilitating syndrome should come from randomized, double-blind, placebo-controlled clinical trials, for now lamotrigine is the treatment of choice whereas gabapentin and topiramate are reasonable second-line agents in patients who fail a trial of lamotrigine. It should be remembered that a number of patients with SUNCT have a partial and useful response to carbamazepine and its use should be considered routinely. Cases 1 and 2 also serve to draw attention to the role of surgery in SUNCT syndrome. Case 1 showed an apparent benefit following trigeminal radiofrequency thermocoagulation that lasted for 3 years. On the other hand, trigeminal microvascular decompression led to a deterioration of symptoms in case 2. There are six case reports of apparently successful treatment of SUNCT syndrome with surgical procedures, including the Jannetta procedure, retrogasserian glycerol rhizolysis, percutaneous trigeminal ganglion compression and trigeminal nerve balloon compression (33–35). However, follow-up in some of these patients was limited to less than 18 months, which makes it difficult to assess the actual effectiveness of the procedures given the episodic nature of the syndrome. Moreover, there can be considerable diagnostic difficulties when considering the SUNCT vs. ophthalmic division trigeminal neuralgia (36–38). Furthermore, Black and Dodick (39) reported on two SUNCT cases refractory to various surgical procedures. The first patient underwent a glycerol rhizotomy, gamma-knife radiosurgery and microvascular decompression of the trigeminal nerve while the second patient underwent gamma-knife radiosurgery of the trigeminal root exit zone and two microvascular decompressions of the trigeminal nerve. Hannerz and Linderoth (35) made a brief reference to a patient who had not benefited from trigeminal vascular decompression and two gamma-knife radiosurgeries of the trigeminal root. Given the uncertain efficacy of trigeminal procedures together with the well-known potential for complications, surgery should be considered only as a last resort and only when the pharmacological options have been exploited to the fullest.
An interesting facet of the presentation of case 4 was the presence of a migrainous sensory aura. An aura has not, hitherto, been described in association with SUNCT syndrome. The occurrence of aura in SUNCT raises the possibility of coexistence of migraine and SUNCT. However, our patient denied ever having suffered from migraine or other headaches. Aura symptoms have been described in association with cluster headache (40, 41), chronic paroxysmal hemicrania (42) and hemicrania continua (43). Hence, migrainous aura may be seen with the range of trigeminal autonomic cephalgias and may represent the presence of aura-susceptibility gene expression rather than typical migraine biology (44).
In summary, four cases of SUNCT are presented, each with very severe and frequent headache attacks that responded rapidly and well to intravenous lidocaine. Our cases suggest that intravenous lidocaine is a useful treatment for the acute management of SUNCT in patients severely distressed by the frequency and severity of their attacks, while appropriate preventative strategies are being implemented or simply to provide some respite for the patient from their suffering.
Footnotes
Acknowledgements
A.S.C. is a Wellcome Trust Fellow. P.J.G. is a Wellcome Trust
Senior Research Fellow.