Abstract
Introduction:
Till date, there is no prospective study in patients with hemicrania continua (HC).
Methods:
Patients fulfilling the international classification of headache disorders criteria for HC were evaluated prospectively. All patients were subjected to a detailed clinical interview, based on a structured questionnaire. Before starting indomethacin, all patients were instructed to fill a headache diary for at least 5 days. Gradual tapering of indomethacin was done at regular intervals.
Results:
We enrolled 41 patients over 4.5 years, 31 of whom met the criteria after confirming the indomethacin response. The mean age was 41 years, and 55% were female. The mean duration of headache was 43.6 months. All patients had continuous strictly unilateral pain with episodic exacerbations. At least one cranial autonomic feature was noted in 81% of patients. Twenty-five patients (81%) felt a sense of restlessness during exacerbations. The mean follow-up was 2.5 years. Three-fourths of patients noted a reduction in indomethacin dose after an average 2.5 of years follow-up. The mean reduction of the dose in the follow-up was statistically significant (172 mg vs. 110 mg, p < 0.001). All patients missed the drug for various reasons over the observation period. The headache reappeared within 48 h in 97% of patients.
Conclusion:
Misdiagnosis of HC is still very common. Patients may not volunteer about the background pain and will focus only on the exacerbations. HC rarely remits, but indomethacin requirements may decrease over time. Skipping of the effective drug leads to the immediate reappearance of pain.
Background
Hemicrania continua (HC) is a primary chronic headache disorder, which is classified under the heading of trigeminal autonomic cephalalgias (TACs). 1 More than 1000 cases of HC have been reported in the literature. It constitutes 1.3–2.3% of all headache patients in the clinical settings. 2 After cluster headache (CH), it may be the second most common TACs. 3 Various large case series have been reported in recent years. 2 Despite this, there remains a conundrum about HC, specifically regarding clinical features, diagnostic criteria, natural course, and therapeutic interventions. Diagnostic criteria of HC have been modified several times over the years. 2 A case of HC may be missed even by headache specialists and neurologists. 4 The objective of this prospective study was to find out the clinical features, natural course, and therapeutic effects of drugs in patients with HC. We also aimed to find out the reasons for misdiagnosis of HC.
Materials and methods
This prospective study was conducted in a neurology outpatient clinic of a large tertiary hospital between February 2014 and July 2018. The Institutional Ethics Committee approved the study (SVRC/ON/MEDZ/FP/15009). All patients provided written informed consent before enrollment. The diagnosis of HC was made by at least two headache specialists or neurologists together in the outpatient clinic. The inclusion criteria were as follows: (i) age >18 years, (ii) patients fulfilling the International Classification of Headache Disorders (ICHD-3β) 5 criteria of HC, and (iii) a minimum 12 months of follow-up. The exclusion criteria included (i) age <18 years, (ii) a possibility of secondary HC, (iii) a follow-up of <12 months, (iv) incomplete or no response to indomethacin, and (iv) patients not subjected to neuroimaging.
To get the maximum number of cases of HC, we encouraged primary physicians in the surrounding areas to refer the side-locked headaches to our department.
Data collection
All patients with a putative diagnosis of HC underwent a detailed clinical interview based on a structured questionnaire. The questionnaire included the following items: demographic data, age at onset, duration of illness, pattern of headache, details of continuous pain, details of exacerbations, side and site of headache, details of cranial autonomic symptoms (CAS), a sense of restlessness or agitation, migrainous features, auras, other accompanying symptoms, and precipitating factors. The corresponding author interviewed each patient and fulfilled the questionnaire. All patients underwent a detailed neurological examination. We also reviewed each individual’s previous medical case sheets or records to get other details, including previous diagnoses, investigations, and therapeutic interventions. Magnetic resonance imaging (MRI) of the brain was done in all patients to rule out underlying secondary pathologies. All patients were subjected to routine biochemical parameters, including renal and hepatic function.
Before starting indomethacin, all patients with a putative diagnosis of HC received a headache diary. We asked them to note details of both continuous background headaches and episodic exacerbations prospectively. We asked each patient to grade pain intensity prospectively according to visual analog scale (VAS) for both background pain and exacerbations. They were asked to note CAS, restlessness, and other associated symptoms prospectively. We instructed each patient to start indomethacin only when they fulfilled the headache diary for at least 5 days. Oral indomethacin was started at a dose of 25 mg three times daily. The dose of indomethacin was gradually escalated (increment of 25 mg three times daily at every 3–5 days intervals) up to 100 mg three times daily or until the patient had a complete pain relief. Injectable indomethacin was not given due to unavailability. A diagnosis of HC was considered only if there was a complete response to indomethacin. Each patient received a gastric protective agent.
We followed all patients for at least 12 months. Patients were asked to keep a headache diary and to note if they felt any type of recurrence. We instructed them to note down if they got symptoms related to indomethacin toxicity. Although patients were advised not to skip doses of indomethacin, we asked them to record if they skipped or missed the drug for any reason. We also asked each patient to note if they felt a recurrence of pain on skipping the drug.
We had planned to taper indomethacin at every 4–6 months intervals. Gradual dose reduction (25 mg at every 5–7 days intervals) was done till pain reappeared or the patients got completely off indomethacin. In case of reappearance of the symptoms, the patients were advised to increase the dose by 25 mg/day at every 2–3 days intervals till they got complete remission. Follow-up visits were planned at every 4- to 6-month intervals. We contacted a few patients by phone to complete the follow-up details.
We used Microsoft Excel and Graph Pad for statistical analysis. Data are presented as a percentage or as an arithmetic mean with standard deviation (SD). Student’s t-test was used to compare the continuous data. The χ 2 test with Yates’s correction was done for categorical data. All p-values were two-tailed, and a p value <0.05 was considered as statistically significant.
Results
We noted 44 patients with a putative diagnosis of HC over the study period. Thirteen patients were excluded for the following reasons: (i) no or incomplete response to indomethacin: six patients, (ii) a follow-up of <12 months: three patients, (iii) not subjected to neuroimaging: one patient, (iv) age <18 years: one patient, (v) secondary HC: one patient (pituitary macroadenoma), and (vi) not willing to participate in the study: one patient. Finally, 31 patients were included in the analyses (Figure 1). Routine biochemical parameters, including renal and hepatic function, were normal in all patients.
Flow chart of patient enrollment.
Epidemiology and clinical features
The epidemiological data and various aspects of clinical features are summarized in Table 1. The mean age of the patients at the time of presentation was 41 years, 55% were female. The mean duration of headache was 43.6 months (±42.6 months). Sixteen (52%) patients had side-locked pain on the right side. Twenty-nine patients (96%) had an unremitting form of headache. The pain was localized in the orbital/retro-orbital area in 28 (90%) patients. Frontal and temporal were the two other common sites for perceiving pain.
Epidemiological profiles and clinical features of 31 patients with hemicrania continua.
The VAS for the continuous background pain varied between 2 and 7 with a mean of 4.0 ± 1.3. The background pain did not hamper any physical activity in any patients. The severity of exacerbations varied from 6 to 10 (mean 8.1 ± 0.8) and hampered routine activities in all 31 patients during exacerbations.
Table 2 provides the details of the exacerbations. The duration of exacerbations varied from a few minutes to 1 week. The frequency of the exacerbations varied from more than five attacks in a day to one attack in 2-week duration. About 65% of patients had at least one attack daily. Initially, about 55% of patients reported at least one CAS during exacerbations. However, this rose to 81% when we asked to note down CAS prospectively in the headache diary. Conjunctival injection (71%), tearing (65%), and nasal stuffiness (55%) were common CAS. Feeling of foreign body sensation in the eye was noted in 42% of patients. Twenty-five (81%) patients noted a sense of restlessness during the exacerbations. Rubbing or pressing or holding the aching part was the most common form of restlessness (65%). Only 16% of patients were able to stay comfortably in the bed during the attacks. At least one migrainous feature was noted in 71% patients during the exacerbations. Nausea was the most common migrainous symptom (58%).
Details of exacerbations in HC (n: 31 Patients).
HC: hemicrania continua.
Response to indomethacin
Tables 3 and 4 show the various aspects of the response to indomethacin. Ninety-seven percent of patients responded to a dose between 75 mg and 225 mg. The mean daily dose was 172 mg/day (±55 mg). All patients showed at least some response within 24 h. However, only three patients showed a complete response within 24 h. Others responded only after escalation of the dose of indomethacin. A total of 81% of patients noted a complete response within a week. Nineteen percent of patients responded in the second and third week.
Details of indomethacin dose.
Clinical features noted during follow-up.
Follow-up observations
The mean follow-up was 30 months (±10.6) (Table 4). The dose of indomethacin had been reduced at every 4- to 6-month intervals. One patient successfully tapered off indomethacin on four different occasions (with 2- to 5-month symptom-free period). Two other patients were successfully weaned off of indomethacin for 2–4 months duration. Readministration of indomethacin again led to the complete cessation of pain in all three patients.
Changes in the dose were noted in a total of 77% of patients. Table 3 compares the initial effective dose of indomethacin to the effective dose noted in the most recent follow-up. There was a significant reduction in the daily mean dose of indomethacin (172 mg vs. 110 mg, p < 0.001). The initial effective dose of indomethacin was >150 mg/day in 11 (36%) patients. However, none of the patients required >150 mg/day of indomethacin on the follow-up. Indomethacin was well-tolerated in most patients. Mild-to-moderate gastrointestinal side effects (occasional abdominal pain, nausea, and abdominal fullness) were noted in 42% of patients. Patients managed these mild symptoms with the help of their family physicians (that included stopping of indomethacin for a few days). Three patients had severe gastrointestinal disturbances (continuous abdominal pain, vomiting, diarrhea, etc.). This leads to a trial of other drugs for them. One patient showed a complete response to topiramate, and he continued it till the most recent follow-up. Two other patients restarted indomethacin.
We divided all patients into two groups: (i) HC with a history of <12 months of pain and (ii) HC with a history of >12 months (Table 5). No epidemiological or clinical differences were noted between the two groups. However, the initial effective dose of indomethacin was significantly lower in HC with <1-year history (136 mg vs. 191 mg; p < 0.006 ). There was a reduction in mean effective dose in the follow-up in both groups (96 mg vs. 118 mg). The change in the mean dose was more in patients with HC of longer history (41 mg vs. 74 mg, p < 0.001).
A comparative study between HC of <12 months to HC of >12 months duration.
HC: hemicrania continua.
All patients missed the drug for 1–7 days period for various reasons. All patients noted recurrence of symptoms. Pain reappeared within 24 h in 84% of patients. Another 13% of patients had symptoms recurrence between 24 h and 48 h. One patient had been switched to topiramate (because of severe gastrointestinal symptoms). Skipping of even topiramate led to reappearance of symptoms within 24 h on two occasions.
Evaluation of previous records of patients
None of the patients received the correct diagnosis before visiting to our neurology clinic. The mean time to diagnosis was 43.6 ± 42.6 months (range 4 months to 10 years). Table 6 summarizes the various aspects of diagnostic and therapeutic interventions patients received before being correctly diagnosed.
Diagnostic and therapeutic features noted from previous medical case sheet or records.
Sixty-one percent patients were initially seen by neurophysicians prior to correct diagnosis. Migraine was the most common misdiagnosed condition, and it was a diagnosis in 71% of patients. Other misdiagnoses were CH, trigeminal neuralgia, sinus-related problems, eye-related problems, atypical facial pain, dental pain, cervicogenic headache, temporomandibular joint (TMJ) dysfunctions, and temporal arteritis.
We evaluated all the records or related documents of all patients. All enrolled patients had continuous background pain. However, medical records/prescriptions of at least 20 patients (65%) described the headaches as an “episodic pain.” On inquiry, patients admitted having continuous background headache as the beginning. As exacerbations were more disturbing than continuous pain, they focused only on the exacerbations part. They considered continuous headaches as “normal” or “not significant” and thus did not volunteer about background continuous pain during history taking. Mentioning of cranial autonomic features were noted in medical records of only 26% of patients. Restlessness or agitation was mentioned in only 10% of cases.
None of the patients received indomethacin prior to enrollment. Nine (29%) patients never showed any responses to any drugs or any drug combinations. Fourteen (45%) patients had noted a complete improvement in the past by certain drugs, usually a combination of two or more drugs. However, all these patients noted immediate reappearance of headaches on stopping the drugs. Topiramate was the most common drug in those combinations. Twenty-nine percent of patients were subjected to some sort of surgical interventions, without a positive response.
Discussion
This is the first prospective study with a long-term follow-up on patients with HC. We evaluated the clinical features prospectively. In parallel, we tried to look for the natural course and drug requirement during follow-up. We also evaluated the records of all patients to find out the diagnostic and therapeutic issues.
Continuous side-locked pain
The central feature of HC is a continuous background side-locked headache. 2 A diagnosis of HC can be missed if you do not get a history of background pain. The mean VAS of the continuous background pain usually varies from 3.3 to 5.2 in various studies, and it may be less disturbing than the severe exacerbations. 2 Therefore, a large number of patients may not volunteer about the background pain. It is incumbent on the physician to ask specifically about the background pain in all patients having strictly unilateral headaches.
Exacerbations
The intensity of exacerbations usually varies from 6 to 10 in VAS score, and it is comparable to the intensity of CH and paroxysmal hemicrania (PH). 2 The frequency and duration attacks are highly predictable and well defined in ICHD-3 criteria for CH and PH. The duration of attacks is 2–30 min for PH and 15–180 min for CH. 1 However, the exacerbations in HC are highly variable in most of the studies, ranging from a few minutes to a few days (may be up to a few weeks in some cases). The duration of exacerbation varied between a few minutes to >1 week in our study. In Cittadini and Goadsby study, the exacerbations were between 30 min and a week. 6 Therefore, the exacerbations of HC may match with other TACs and migraine. So, missing a history of the background pain will lead to misdiagnosis as a migraine or other TACs. 2,7 Therefore, asking for background pain is of utmost importance in all side-locked headaches. Chronic CH and chronic paroxysmal headache may also have interictal pain. However, interparoxysmal pain is usually not present throughout the day. 8
Cranial autonomic features
The prevalence of cranial autonomic features varies between 60% and 100%. 4,6,9 It was noted in 81% in our observations. About 95% of patients reported at least one cranial autonomic feature in the study by Cittadini and Goadsby. 6 Conjunctival injection and tearing are the two most common cranial autonomic features in most of the studies. There were 10 different cranial autonomic features (including aural fullness and facial flushing) in ICHD-3β. 5 However, aural fullness and facial edema were removed from recent ICHD-3 criteria. The prevalence of facial flushing and aural fullness was noted in 19% and 13%, respectively, in our case series. Our observations suggested that a subset of patients may have subtle cranial autonomic signs and patients may not be aware of it. The prevalence of CAS increased when we asked the patients to look for CAS prospectively. Therefore, an objective assessment of CAS is important before labeling it as “not present.”
Restlessness or agitation
The prevalence of restlessness during exacerbations ranges from 30% to 69%. 4,6 However, the pattern of restlessness during exacerbations has never been explored in patients with HC. We evaluated behavioral abnormalities during headache exacerbations through a structured questionnaire. To the best our literature search, none of the studies have evaluated the details of restlessness. More than 80% of patients reported at least one symptom suggestive of restlessness. The pattern of restlessness noted in our patients was comparable to that of CH. 10 Rubbing or pressing or holding the aching part (mainly eye) was the most common form of agitation.
Migrainous features
Up to 90% of patients may have at least one migrainous symptom during exacerbations. 2 In our cohort, 71% of patients noted at least one migrainous symptom. As a large number of patients have exacerbations of >4 h, a misdiagnosis as migraine is likely if you miss the history of background pain. Up to 70% of patients may meet the criteria for migraine in the exacerbation period. 7 Topiramate is one of the commonly used medications for chronic migraine. A few patients of HC may also show a response to topiramate. 2 Response to topiramate may further reinforce a (mis)diagnosis as chronic migraine.
Response to indomethacin
The mean daily dose (171 mg/day) was comparable to other large case series. It was 176 mg/day in Cittadini et al. observation of 32 patients. 6 The response to indomethacin may be immediate. 11 However, a recent review suggested that a large number of patients may take >1 week to show a complete response (depending upon the titration of the effective dose). 2 About 81% of patients noted a complete response within a week. It has been suggested that patients with more chronic headache may take more time to show a complete response.
Indomethacin was largely well tolerated. Only 10% of patients reported persistent gastrointestinal symptoms, requiring switching to another drug. The incidence of indomethacin related side effects in HC patients in various case series varied between 23% and 50%. 6,12,13 However, serious side effects, requiring hospitalization or major interventions are relatively rare in patients with HC. It is said that patients with CH tolerate various drugs in a higher dose much better than other patients. 14 Probably, just like CH, HC patients tolerate indomethacin well.
HC may respond to a number of other drugs. There are about 100 patients in the literature who responded to drugs other than indomethacin. 2 Topiramate is probably the second most effective drug in HC. Three patients of our series got a trial of topiramate. It was effective in one patient.
Follow-up observations
This is probably the first study on HC where patients were prospectively observed for drugs requirement, drug-related side effects and recurrence of the symptoms. More than three-fourths of patients had a reduction in indomethacin dose during the follow-up. The effective dose may be significantly less if you catch the patients in the early stage. HC patients with a long history of pain may require a higher dose of indomethacin to achieve a complete response. So, correct diagnosis at an early stage may be important. Only 10% were completely weaned off from indomethacin for a limited period. So, none of the patients were completely off the drug in 2.5 years of follow-up, suggesting a possibility of HC being as a lifelong disorder.
Pareja et al. noted a change in the dose of indomethacin in 47% of the patients with HC in a retrospective observation. 13 They suggested several explanations for the reduction of doses in indomethacin: (i) indomethacin metabolism may slow down with time, (ii) the underlying disorders weakness over time, and (iii) progressive tolerance in pain. The indomethacin dose in our study was escalated in a block of 75 mg (25 mg three times a day) (i.e. after 75 mg next dose was 150 mg, then it was 225 mg, etc.). However, tapering was planned at the lower dose (reduction of 25 mg indomethacin at every 5–7 days interval. Therefore, it is possible that a few patients received a higher dose than required dose since the beginning. So, we suggest a gradual reduction in the dose of indomethacin in each patient at every 3- to 6-month intervals. In this way, we can find out the remission phase of HC or the lowest effective dose for a patient.
Reappearance of pain on skipping or missing indomethacin is an important issue in HC. The reappearance of pain on skipping indomethacin is almost immediate. Headaches reappeared within 48 h in 97% of patients. It had been within 24 h in 86% of patients. Headaches reappeared between 6 h and 48 h in all 22 patients who skipped the drug in Prakash and Golwala series. 15 Headache reappeared immediately in several patients of Newman et al. case series. 12 In one patient, skipping a dose of indomethacin resulted recurrence of pain within 2 h. Various authors have suggested that the reappearance of the headaches is a strong character of HC. 16,17 Antonaci and Sjaastad write “The fact that the pain returns upon indomethacin discontinuation is a strong testimony, as regards HC, perhaps even stronger than that of the Indotest itself.” 17 They further write that “If pain does not recur upon indomethacin discontinuation, this indicates that either: the pain has disappeared spontaneously, or the diagnosis is wrong.” In fact, the reappearance of headache on skipping the drug is not limited with indomethacin only. Discontinuation of other effective drugs may have a similar effect. In our cohort, 14 (45%) patients had a complete improvement in the past by certain drugs. However, all these patients categorically stated that “response continued as long as they received the drugs.” Immediate reappearance of headaches had been noted on stopping these effective drugs. So, our observations reaffirmed the fact of “immediate reappearance of headaches on discontinuation of indomethacin or other effective drugs.” This may be a diagnostic clue in patients with a side-locked headache. 16 Patients having a history of complete response to any drugs as long as they continued the drugs should raise a possibility of HC. Such patients should be enquired for continuous pain, autonomic features, and restlessness. A trial of indomethacin is warranted in such patients to confirm the diagnosis.
Limitation
Our observations cannot be generalized as our sample of patients may not be a true representation of HC due to selection and referral biases. Moreover, it is a single center-based study that was done in an adult tertiary neurology outpatient clinic. We excluded patients below <18 years of age. We could not perform “indotest” due to the unavailability of injectable indomethacin.
Conclusions
Misdiagnosis of HC is still very common. Patients may not mention about the background pain and will focus only on the exacerbations part, unless specifically asked. Frequency and duration of exacerbations are highly variables. Response to indomethacin is “sine qua non” for HC. However, the reappearance of the headaches on skipping indomethacin or other effective drugs is a strong character of HC. In 2.5 years of follow-up, none of the patients had remission for a long period.
Footnotes
Declaration of conflicting interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.
Funding
The author(s) received no financial support for the research, authorship, and/or publication of this article.
