Abstract
Introduction
Hemicrania continua (HC) is an indomethacin-responsive primary headache disorder. Response to indomethacin is an essential criterion for the diagnosis of HC. The response to indomethacin is usually prompt and enduring (1). However, HC-like headache unresponsive or partially responsive to indomethacin is also described in the literature (2,3). HC patients usually respond dramatically to a therapeutic dose of indomethacin, usually within 24 h. A reduction of the dose may be successful after some time (4). Here we report three patients with HC in whom complete response to indomethacin was noted a few months after the initiation of indomethacin therapy.
Case reports
Case 1
A 52-year-old man had a 10-year history of right-sided continuous mild to moderate headache with superimposed exacerbations of stabbing pain and cranial autonomic features. He had never had periods of remission since the onset of headache. The exacerbations, in bouts of a few hours to a few days (up to 2 days), had a frequency of about two to three attacks per day to two to three attacks per week. The pain was maximal in the frontal area with radiation to the orbit and temporo-parietal areas. The exacerbations were associated with nausea, conjunctival injection, lacrimation and rhinorrhoea in about one-quarter of the exacerbations. He did not identify any triggering factors for the exacerbations. There were occasional nocturnal exacerbations. He was non-alcoholic and a non-smoker. There was no other significant past medical history. There was no family history of headache. On direct questioning the patient admitted having mild anxiety and depressive symptoms (restlessness, irritability, disturbed sleep and feelings of tiredness).
He had been extensively investigated during these 10 years. He had undergone magnetic resonance imaging (MRI) of brain and cervical spine on many occasions, which did not reveal any abnormality. Routine haematological and biochemical screening done on many occasions in the past had been normal. Physical examination during exacerbations showed conjunctival injection. Prior treatment with various drugs had produced minimal or no effect. The patient was not on any analgesic at the time of consultation in our out-patient clinic.
A diagnosis of HC was made, and indomethacin was started at a dose of 25 mg t.d.s. The patient did not feel any improvement on this dose. The dose was gradually escalated to 75 mg t.i.d. There was minimal improvement in back ground headache with this dose. However, he had no exacerbation with this dose during the next 3 weeks. The dose was further increased (up to 125 mg t.i.d.). However, a mild background headache did not subside even with this dose in the next 3 weeks. Therefore, we planned to reduce the dose of indomethacin and dose reduction was successful up to 75 mg t.i.d. As the patient was satisfied with his symptoms (mild headache with no exacerbation), we asked him to continue indomethacin at this dose. Eight weeks later, he reported complete suppression of the background headache. The suppression of background headache was gradual. In the next 1-year follow-up, he experienced no similar type of headache. However, skipping doses of indomethacin always resulted in headache recurrence.
Case 2
A 46-year-old man presented with a 14-year history of continuous right-sided headaches. The pain was felt on the whole right side of the head. It was described as a constant ache of mild to moderate intensity with superimposed exacerbations of severe to excruciating throbbing pain lasting 60 min to 12 h. The exacerbations used to extend even to the orbital area. The exacerbations occurred two to five times per week, and were associated with ipsilateral conjunctival injection, lacrimation, ptosis and rhinorrhoea in about 50% of the attacks. The patient never developed aura, nausea, vomiting, photophobia, phonophobia or restlessness (pacing) during the exacerbations. Nocturnal attacks were not reported. There were no precipitating or aggravating factors. The patient had no history of head trauma. Family history for headache was negative. On direct questioning the patient admitted having mild anxiety and depressive symptoms (irritability, disturbed sleep, feelings tense and feelings of tiredness). Physical examinations and investigations (including brain MRI) were normal. Prior treatment with various drugs had not produced any response on background headache and exacerbations, and he was not on any drug at the time of consultation in our out-patient clinic.
A diagnosis of HC was made, and indomethacin was started at a dose of 25 mg t.i.d. There was no response on this dose. At a dose of 50 mg t.i.d., there was no exacerbation in the next 2 weeks. However, background headache persisted at the same intensity. The dose was increased up to 100 mg t.i.d. The background headache gradually reduced in intensity after about 6 weeks (at times the background headache also disappeared). A feeling of anxiety used to increase the awareness of the background headache. Complete suppression of the background headache was noted in the next 8 weeks. The patient did not feel any headache in the further 4 months' follow-up. The patient did not skip any dose of indomethacin in these 4 months and wanted to continue the same dose.
Case 3
A 32-year-old woman presented with a 4-year history of continuous right-sided headache with superimposed exacerbations. The background pain was described as a steady pressure of moderate severity in the right temporo-parietal areas. Exacerbations, one to three times daily, lasted 30 min to 60 min and were associated with ipsilateral conjunctival injection and lacrimation. There were no nocturnal attacks. She was non-alcoholic and a non-smoker. There was no other significant past medical history. There was no family history of headache. On direct questioning the patient admitted having mild anxiety and depressive symptoms (disturbed sleep and irritability). Prior treatments with various drugs had been without benefit. There was no history suggestive of analgesic overuse. Physical examinations and investigations (including brain MRI) were normal. Treatment with indomethacin 100 mg t.i.d. (titrated over 2 weeks) markedly reduced the background headache and abolished the exacerbations. The patient claimed that thinking about the headache increased the awareness of the background headache. The background headache subsided gradually with the same dose after 8 weeks. Tapering of indomethacin done after 4 months was successful up to 50 mg t.i.d. Further tapering led to recurrence of the headache.
Discussion
More than 25 years have passed since the first description of HC by Medina and Diamond (5). However, its natural history is still being determined. Information regarding HC is rapidly changing. The diagnostic criteria of HC have been revised on many occasions in the past. Even the present International Headache Society (IHS) criteria have been criticized by a few authors (3). One of the controversial issues regarding HC is the use of indomethacin. Initial diagnostic criteria of HC considered indomethacin responsiveness as a confirmatory trait, instead of an obligatory diagnostic criterion (6). However, complete response to therapeutic doses of indomethacin is an essential feature in the present IHS diagnostic criteria for HC (1). The range of therapeutic doses has not been described in the criteria. The effective dose of indomethacin in patients with HC is highly variable (25 mg every 2 days to 300 mg daily) (4). The response to treatment as an essential feature is well acknowledged by IHS criteria in a few other headache disorders. The therapeutic response of drugs in all these headache disorders should appear within a defined time frame (within 72 h for Tolosa–Hunt syndrome, 2 weeks for retropharyngeal tendonitis, and 48 h for giant cell arteritis) (1). However, time frame has not been defined for indomethacin-responsive headache, including HC. HC patients usually respond within 24 h. However, it may take up to 10 days to show the response to indomethacin (7).
All of our cases fulfilled the International Classification of Headache Disorders, 2nd edn criteria of HC, as they showed complete response to indomethacin. The possibility of paroxysmal hemicrania (PH) with interictal pain does not exist, as all three had ‘moderate’ background headache, and the frequency and duration of exacerbation were much lower than the defined IHS criteria for PH. The duration and frequency of exacerbations of case 3 may have a differential diagnosis of indomethacin-responsive cluster headache (CH) with interictal pain. However, the presence of ‘moderate’ background headache, absence of nocturnal attacks and absence of restlessness during the exacerbations were against the diagnosis of CH. Complete response to indomethacin was not immediate in all three cases. Periodic exacerbations subsided within 2 weeks in all three cases. The background headache took > 2 months to achieve complete remission. All of our cases received at least 300 mg of indomethacin daily. Therefore, the possibility of suboptimal management does not exist in these patients. The recurrence of headache on skipping or tapering of indomethacin in cases 1 and 3 rules out the possibility of placebo effect. Skipping and tapering were not done in case 2. Therefore, we can not rule out the possibility of placebo response in this patient. However, this possibility is less likely, as he had a very long history (14 years) of HC.
Hemicrania continua-like headache unresponsive to indomethacin is well reported in the literature. However, to the best of our knowledge, such a type of delayed response of indomethacin in patients with HC has not been reported previously. On review of the literature, we noted many patients with HC in whom response to indomethacin was incomplete. Marmura et al., in the largest case series of HC, reported at least four patients in whom the response to indomethacin, although dramatic, was associated with persistence of baseline headache (2). There are many other similar case reports (i.e. marked response with persistence of baseline headache with indomethacin) in the literature (3,8–10). Besides these, there are many other cases where response to indomethacin has been described as marked or dramatic. It is difficult to draw any conclusion as to whether these patients had complete improvement or not (2,11). Long-term follow-up is lacking in all the above-mentioned case reports/series. Therefore the possibility of delayed response by indomethacin even in these patients remains open.
It is difficult to speculate on the mechanisms of delayed response of indomethacin in all of our three cases, as current knowledge of the pathophysiology of HC is limited. The different responses to indomethacin (immediate and complete, marked with persistent background headache, and delayed response) suggest that various mechanisms may be involved in the pathogenesis of HC.
There may be a number of possible reasons why these patients showed complete response after > 2 months' treatment. These possible mechanisms may be the pain memory hypothesis, hallucination, delusion, other psychological factors, etc.
The pain memory may be the principal mechanism for the delayed response to indomethacin in patients with HC. This mechanism may be responsible even for those patients who showed marked but incomplete response to indomethacin. According to the pain memory hypothesis, information about chronic or severe pain is somehow stored in the brain and later on re-experienced (continuously or intermittently) (12,13). Apkarian et al. write that ‘chronic pain is a persistence of the memory of pain and/or inability to extinguish the memory of pain … ’ (14,15).
The pain memory hypothesis is one well-known mechanism for phantom sensation or phantom pain. Phantom sensation or phantom pain is usually described in relation to surgical removal of the body parts such as limb, breast, rectum, penis, testicle, tongue, teeth, etc. However, a lesion of the peripheral nerves or the central nervous system can also cause phantom pain. Phantom sensation or pain can also be induced even in normal individuals (12). ‘Phantom nose’ illusion has been produced by a few authors (16). Ramchandran et al. were able to project tactile sensation even onto inanimate objects (such as shoe and table) by a specific repeated procedure on the patients (16). As a phantom sensation can be induced even in normal persons by repeated procedure, the possibility of pain memory in patients with HC can not be ruled out.
Information about the body and its sensation depends on the neuromatrix, which is a network of neurons including thalamus, primary somatosensory cortex (SI), secondary somatosensory cortex (SII), reticular formation, limbic system, etc. (13,17). Amputation creates an abnormal input into the neuromatrix owing to a lack of normal or overactivity related to the abnormal firing pattern of damaged nerves. This input leads to an altered neurosignature and the experience of a phantom (13,17). It has been suggested that long-lasting noxious stimuli may lead to long-term changes in the neuromatrix (especially at the cortical level). These changes have been demonstrated in patients with chronic low back pain (CLB) as increased representation zone of the back in the SI cortex (12). No such observation has been made in patients with HC. However, Matharu et al. have reported an increase in regional cerebral blood flow in the areas known to be a part of the neuromatrix (including primary and secondary cortices) in patients with HC (18). Therefore, we speculate that there are cortical changes similar to CLB (or following amputation of body parts) because of the chronic, continuous nociceptive input in patients with a history of longstanding HC. These changes may be responsible for the formation of a pain memory (and delayed or incomplete response of indomethacin) in patients with HC. PH is another indomethacin-responsive primary headache disorder. However, delayed or incomplete response is not reported in patients with PH. This further strengthens the pain memory hypothesis in patients with HC, as long-lasting/continuous noxious stimuli (as in CLB) are more likely to produce a change in the neuromatrix.
The pain memories established prior to amputation are powerful elicitors of phantom limb pain. Phantom limb pain is usually similar to the pain that existed (if any) in the limb prior to amputation (13,14,17). The persistence of background pain (unchanged quality) with indomethacin therapy in our patients (for few months) and in the literature again suggests the possibility of pain memories in patients with HC.
There is a suggestion that persistent pain may be just a special type of hallucination, according to which there is no current disturbance causing a person to feel pain, but that there was a disturbance that caused the pain at some time in the past. However, direct evidence for this hypothesis is lacking in the literature (19).
Apkarian et al. have suggested that chronic pain is a state of continuous learning in which emotional associations are continuously made simply due to persistent pain. This emotional association itself may cause some suffering of chronic pain (14). Thinking about the pain (of HC) and feelings of anxiety (restlessness, irritability) led to a feeling of background headache in two of our patients. This suggests that emotional associations (as suggested by Apkarian et al.) may be another mechanism for the persistence of pain in our cases. Emotion is linked with the limbic system, which itself is part of the neuromatrix (hence, with pain memory system). Therefore emotions may trigger pain memories. Giummarra et al. have described eight cases who reported that their phantom pain was triggered by observing, thinking about the pain, or inferring that another person was in pain (13).
Psychiatric comorbid states have not been studied in patients with HC. However, all of our cases on direct questioning admitted having features suggestive of depression and/or anxiety. Presence of psychiatric comorbid states suggests a poor prognosis in patients with headache, especially with chronic daily headache (20). Mechanisms of this poor prognosis have not been well described in the literature. Therefore, the presence of psychiatric comorbid features may be one of the mechanisms for delayed response or presence of mild background headache. However, this may not be the main mechanism in our patients, as psychiatric comorbid states were very mild, and there was complete remission of exacerbations of HC (only baseline headache persisted).
Besides these mechanisms, we can not rule out the possibility of other mechanisms (such as some pharmacological–biological interaction or some other organic mechanisms) for the delayed response of indomethacin in our patients.
In conclusion, our case reports and review of the literature suggest that delayed response and marked response with persistence of background headache (mild) probably exist in patients with HC. Pathophysiological studies of such patients, especially concerning the neuromatrix, are required to confirm this.
Footnotes
Competing interests
None to declare.