Consensus of the Brazilian Headache Society on the treatment of chronic migraine

Chronic migraine poses a significant personal, social and economic burden and is characterized by headache present on 15 or more days per month for at least three months, with at least eight days of migrainous headache per month. It is frequently associated with analgesic or acute migraine medication overuse and this should not be overlooked. The present consensus was elaborated upon by a group of members of the Brazilian Headache Society in order to describe current evidence and to provide recommendations related to chronic migraine pharmacological and nonpharmacological treatment. Withdrawal strategies in medication overuse headache are also described, as well as treatment risks during pregnancy and breastfeeding. Oral topiramate and onabotulinum toxin A injections are the only treatments granted Class A recommendation, while valproate, gabapentin, and tizanidine received Class B recommendation, along with acupuncture, biofeedback, and mindfulness. The anti-CGRP or anti-CGRPr monoclonal antibodies, still unavailable in Brazil, are promising new drugs already approved elsewhere for migraine prophylactic treatment, the efficacy of which in chronic migraine is still to be definitively proven.

Chronic migraine is a widespread disorder that poses a significant personal, social and economic burden. Although less prevalent than episodic migraine, chronic migraine patients present with more comorbidities and are more refractory to treatment 1,2,3 . After the publication of the third edition of the International Classification of Headaches Disorders (ICHD-3) 4 and considering new evidence published since the Latin American Consensus on Guidelines for the Treatment of Chronic Migraine was reached in 2012 5,6 , the Brazilian Headache Society (Sociedade Brasileira de Cefaleia) recognized the need to elaborate on a new consensus on this subject.

METHODS
A group of 17 members of the Brazilian Headache Society met on September 29-30, 2018, after a process of literature searches made by five subgroup coordinators. The first day was dedicated to discussion of the relevant literature and to the drafting, by five independent subgroups, of the text sections. On the second day, the full text was read, discussed and modified in a plenary session. The text was then reviewed by one of the members (FK) and submitted for translation into English, which was approved by all members before being submitted for publication. Classes of evidence and levels of recommendation were described according to the American Academy of Neurology Classification for the Rating of a Therapeutic Study and Classification of Recommendations whenever possible 7,8,9 . According to the latter, studies vary from class I (high quality well-designed randomized double-or triple-blind controlled trials) to class IV (uncontrolled studies, consensus and expert opinions). Levels of recommendation are described in Table 1. The Brazilian Headache Society received a grant from Allergan Specialty Care designated specifically to the logistics and editorial expenses.
The authors did not receive any fee. Allergan Specialty Care staff did not interfere in the authors' selection process, nor took part in any step of the manuscript elaboration.

CONCEPTUALIZATION
The ICHD-3 classifies headaches into three groups. The first group comprises the primary headaches, conditions whose mechanisms are eminently neurochemical -in other words, the tendency to develop headache attacks is the disorder itself. The secondary headaches, which compose the second group, are caused by some underlying disorder, e.g. intracranial structural lesions, whereas the third group includes the painful cranial neuropathies, other facial pains, and unspecified or nonspecific headache 4 .

DIAGNOSTIC CRITERIA, EPIDEMIOLOGY AND IMPACT
According to ICHD-3, chronic migraine belongs to the primary headache group and is defined by the diagnostic criteria listed in Table 2.
It is estimated that up to 5% of patients with migraine meet the criteria for the chronic form of the disease 1 . In systematic reviews of population studies, the prevalence of chronic migraine ranges from 0.9% to 5.1% 1,10 , while in the Brazilian infant population it is around 1.7% 11 . Since the publication of the last report by the World Health Organization's Global Burden of Disease, migraine has come to be considered the second major cause of disability among all diseases, behind only mental disorders, led by depression 12 . Over the period of a year, 25% of the patients with chronic migraine return to the episodic form of the disease and another 40% oscillate between the episodic and chronic forms 13 .
Nearly 50% of individuals with headache self-medicate 5 . Overuse of symptomatic drugs is believed to cause headache in individuals previously affected by a primary headache. This headache is classified in the ICHD-3 among secondary headaches, as item 8.2 Medication-overuse headache. From 55% to 70% of the patients seen at specialized centers in Latin America receive this diagnosis 14 . The term chronic migraine was first mentioned in ICHD-2, in 2004 15 , and its diagnostic criteria were reviewed in 2006 16 . In previously-proposed classifications, these diagnoses were grouped, and the terms transformed migraine and chronic daily headache were used 17 . Currently, a patient who meets the criteria for chronic migraine (Table 2) and for medication-overuse headache (triptans, ergot drugs, common analgesics alone or in combination and/or nonsteroidal anti-inflammatory drugs) should initially receive both diagnoses. It should be emphasized that patients who present with headache on 15 or more days per month may have primary headaches other than chronic migraine, such as chronic tension-type headache, new daily persistent headache and hemicrania continua 15,18 .
Pathophysiological mechanisms probably involved in migraine include cortical hyperexcitability (or deficient inhibition), dysmodulation in areas of the brainstem and hypothalamus, and central and peripheral trigeminal sensitization, with release of vasoactive/neuroactive peptides, such as calcitonin gene-related peptide (CGRP). Episodic trigeminovascular system activation, faulty filtering of sensorial stimuli and heredity are characteristics that point to the neural basis of this disorder 19,20 .

ADDITIONAL INVESTIGATION
Additional investigation, such as neuroimaging, cerebrospinal fluid analysis and blood tests, should be employed judiciously and parsimoniously in chronic headache patients, considering the inherent risks and costs. Patients who meet the ICHD-3 criteria for chronic migraine and have a normal neurological examination do not require, in most cases, additional tests 21 . On the other hand, these tests may become necessary to rule out alternative diagnoses in the presence of various secondary headache red flags, listed in Table 3.
In the presence of tinnitus, visual changes, headache aggravated by Valsalva-like maneuver and/or morning headache with improvement throughout the day, lumbar puncture with cerebrospinal fluid manometry (plus cerebral venous angioresonance, when appropriate) is recommended even in the absence of papilledema, to rule out idiopathic intracranial hypertension (pseudotumor cerebri) without papilledema (class of evidence IV, level of recommendation U). On the other hand, lumbar puncture is not necessary when the clinical picture is typical and these symptoms are absent 21 . Although epidemiological data do not correlate chronic migraine with obstructive sleep apnea headache 22 , this differential diagnosis may be difficult, especially in patients with headache attributed to medication overuse.
In the presence of strictly morning headaches, polysomnography may be indicated 21 .

MANAGEMENT OF HEADACHE ATTRIBUTED TO MEDICATION OVERUSE ASSOCIATED WITH CHRONIC MIGRAINE
The concept of medication-overuse headache has evolved since the first edition of the ICHD, e.g. the inclusion of triptan-overuse headache, elimination of the need for a specific headache pattern and the establishment of different thresholds for each type of analgesic/antimigrainous agents. Medication-overuse headache, present in ICHD-3 as item 8.2, is defined as a secondary headache that occurs in ≥ 15 days per month, in a patient with a pre-existing primary headache, that develops as a consequence of the regular overuse of symptomatic medication for the treatment of headache (ergotamine, triptans, opioid analgesics or combined analgesics: ≥ 10 days/month; common analgesics or NSAIDS: ≥ 15 days/month) 4 . Smoking, sedentary lifestyle, metabolic syndrome, obesity, musculoskeletal diseases, psychiatric comorbidities and excessive use of caffeine increase the risk of developing medication-overuse headache 23,24,25,26 .
Treatment of patients with medication-overuse headache is often difficult, specific guidelines are lacking and the available evidence does not establish the superiority of a strategy. The treatment modalities proposed in the literature are summarized in Table 4. The main poor prognosis risk factors are the amounts of overused drugs, the association of drugs from multiple classes, use of opioids, recurrence of a previous episode and psychiatric comorbidities 27,28,29 .

General aspects of the pharmacological management of chronic migraine
Drugs included in this section were tested in randomized controlled trials (RCTs) with well-defined inclusion criteria, double-blind or open, involving chronic migraine Table 3. Warning signs for secondary headache ("red flags").
1. Headache of sudden onset (thunderclap*) 2. New headache or progressive worsening of a pre-existing headache 3. Change in the pattern of a previously stable headache 4. Headache that is aggravated or alleviated by postural changes 5. Headache that awakens the patient 6. Headache precipitated by physical activity and/or Valsalva maneuver 7. Onset ≥ 50 years of age 8. Symptoms and/or focal neurologic signs, altered level of consciousness, seizures 9. History of cranial and/or cervical trauma 10. Neck stiffness and/or history of fever 11. Neoplasia, HIV, immunosuppression, active infection *reaching maximum intensity in < 1 minute.
patients. Uncontrolled or poor-quality trials, with small samples and/or published in periodicals considered to be unreliable ("predatory journals") were discarded. Papers published prior to the International Headache Society's adoption of diagnostic criteria for chronic migraine 15,16 were included, provided that they had enrolled participants presenting with "chronic daily headache" and that the methods section made clear that the majority of the study population suffered from chronic migraine if the current criteria were applied 4 . Table 5 lists oral drugs used in the treatment of chronic migraine in adults, along with levels of evidence and recommendation, doses used in daily practice and risk categories regarding use during pregnancy or breastfeeding 60,61 .
Topiramate is the only oral drug with class I evidence and level A recommendation for preventive treatment in chronic migraine. Its efficacy was initially demonstrated in a small single-center double-blind RCT published in 2003 62 , involving patients with the diagnosis of "chronic daily headache" and past migraine without aura (classified as chronic migraine), overuse of acute medication and at least four previous preventive treatment failures. Topiramate (50 mg/day) caused a significant reduction in headache frequency and was well tolerated. This was followed by another positive single-center double-blind RCT 63 and two larger multicenter double-blind placebo-controlled trials 64,65 . These two to three-month duration trials showed significant reduction in monthly migraine days with topiramate (slow titration, mean final dose around 100 mg/day). The most important adverse events were paresthesias, fatigue, difficulties with memory, concentration and/or attention, taste perversion. When analyzing the results of both studies together, the authors concluded that topiramate was safe and effective in chronic migraine patients with or without concomitant medication overuse 66 .
As of January 2019, three subcutaneous-administered monoclonal antibodies against CGRP 67 or its receptor 68 , were not yet available in Brazil but have already been approved by the Food and Drug Administration (FDA) for the preventive treatment of migraine (without specification of the type or frequency of attacks): erenumab, fremanezumab and galcanezumab (eptinezumab, an intravenous-administered CGRP antagonist, is awaiting approval). Positive results from phase III ( fremanezumab, galcanezumab) and phase IIb (erenumab) RCTs, specifically addressing chronic migraine patients, were recently published 67-69 . Practical aspects of the pharmacological management of chronic migraine The choice of prophylactic drugs should favor those with greater evidence of efficacy, considering the individual patient preferences, associated clinical conditions and absolute and relative contraindications.
Drugs acting on the central nervous system and/or with marked systemic effects should be started with low initial doses followed by gradual increments. The final therapeutic dose should be the lowest effective dose, considering the maximum dose used in each clinical trial or, eventually, at the discretion and under the responsibility of the prescriber, the maximum tolerated dose. A drug should not be considered ineffective before two to three months after the recommended dose or maximum tolerated dose has been achieved. Laboratory tests must be ordered in a predetermined manner or on demand, looking for potential hydroelectrolytic disturbances, hematological, metabolic and/or intoxication imbalances, according to the prescribed drug or drug combinations.
The strategy of associating different prophylactic drugs is not supported by high level evidence. A class II trial did not show superiority of the association of topiramate with longacting propranolol versus topiramate alone 81 . However, the so-called rational polytherapy-the association of effective drugs with different mechanisms-can be used in monotherapy refractory patients 82 . Regarding comparative efficacy, one single-center double-blind RCT showed equivalence between onabotulinum toxin A (100 units at fixed points plus 100 units at "follow the pain" points) and topiramate (maximum dose of 200 mg), with better tolerability and adherence in the onabotulinum toxin A group 83 , while one single-center open-label study showed comparable efficacy between amitriptyline (25-50 mg/day) and onabotulinum toxin A (250 U/15 sites), also with better tolerability and compliance in the group treated with onabotulinum toxin A 84 . On the other hand, and acupuncture performed better than topiramate in two small single-center open-label RCTs 77,85 . Although there is no evidence of response consistency in long term, it is usually recommended that individuals who have undergone reversal of chronic migraine for episodic migraine should be maintained on prophylactic treatment for at least one to two years 86 (class of evidence IV, level of recommendation U).

Pharmacological management of chronic migraine during pregnancy and breastfeeding
Pharmacological treatment of pregnant and breastfeeding women should be avoided, whenever possible, in favor of nonpharmacological methods. Considering the lack of RCTs specifically dedicated to the treatment of chronic migraine in pregnant women, the following recommendations (Table 5) are based on evidence from trials involving nonpregnant women in the light of current knowledge about potential consequences to the fetus. Atenolol and flunarizine may be considered as first-choice migraine prophylactic drugs in pregnancy, if atenolol is discontinued in the third trimester to prevent bradycardia and intrauterine growth retardation. Amitriptyline (despite its classification as category C) and gabapentin, which must be discontinued early due to possible interference with fetal bone development, may also be used 87,88 . On the other hand, topiramate and valproate should be avoided during pregnancy. Some authors consider that all prophylactic migraine drugs are poorly excreted in breast milk, but others contraindicate the use of atenolol and flunarizine in high doses. There is a consensus that topiramate may be poorly metabolized by premature infants, therefore, in these cases, discontinuation or close monitoring of the infant is necessary 89,90 .

Pharmacological management of chronic migraine in the elderly
In the pharmacological management of chronic migraine in the elderly, the presence of other clinical conditions, potential adverse effects of medications and possible drug interactions should be considered. Some drugs, such as beta-blockers, amitriptyline and flunarizine require special attention in this age group 91,92,93 .

Pharmacological management of chronic migraine in children
The only RCT involving drug treatment of chronic migraine in children and adolescents points towards the efficacy of the combination of amitriptyline and cognitive-behavioral therapy in the treatment of this condition 94 . Despite the lack of trials, drugs with proven efficacy in episodic migraine, such as flunarizine, have been used in clinical practice.

Onabotulinum toxin A (class I evidence, level A recommendation)
Onabotulinum toxin A is a purified neurotoxin derived from the Clostridium botulinum bacteria. It has been approved since 2011 by Anvisa (Agência Nacional de Vigilância Sanitária: the Brazilian government agency equivalent to the FDA) for use in the prophylactic treatment of chronic migraine in adults. This approval was based on a pooled analysis of two large double-blind clinical trials (PREEMPT 1 and 2), which demonstrated significant reduction in headache days and improvement in the patients' quality of life during the trial period, effects also observed in those patients with medication overuse (class of evidence I, level of recommendation A) [95][96][97] . In a 56-week treatment period, 1,384 patients were assigned into two groups: active and control treatment (onabotulinum toxin A and placebo, respectively). In the first, double-blind, 24-weeks phase of the trial, the patients received 31 to 39 injections of onabotulinum toxin A (active treatment) or saline (control) at predetermined points of the head and shoulders, every 12 weeks. In the second phase (open), all participants received onabotulinum toxin A injections, according to the same protocol, every 12 weeks, for 32 weeks. As each injection of 0.1 ml contained 5 units of the toxin and the injections were performed on 31 fixed sites and up to eight random ("follow-the-pain") sites, each participant received 155 U to 195 U per session (Table 6).
Open-label prospective long term and observational studies conducted in different centers, according to the same injection paradigm, have corroborated the efficacy, tolerability and safety of onabotulinum toxin A 98,99,100,101,102,103,104,105 . The Chronic Migraine Onabotulinum Toxin A Prolonged Efficacy Open Label trial, for example, evidenced that onabotulinum toxin A had its efficacy sustained for as long as 108 weeks, without further safety or tolerability problems 101 . Increase in productivity and a reduction of about 60% in expenses with medical visits and hospitalizations were suggested by pharmacoeconomics studies 106,107 .
There is no consensus in the literature regarding the number of onabotulinum toxin A cycles required for the management of chronic migraine. Some trials suggest an increasing efficacy with regular cycle repetition for more than one year, including in patients that overuse symptomatic drugs (three class II trials, level B recommendation) 104,108,109 . To date, no clinical features predicting responses to onabotulinum toxin A (recommendation level B) have been identified 100,101 .
An onabotulinum toxin A trial including children 110 showed a significant reduction in pain days, but without improvement in quality of life or reduction of pain intensity during the attacks. It should be noted that onabotulinum toxin A has its use in chronic migraine approved in Brazil for patients ≥ 18 years old.
The adverse effects caused by this treatment are rare, transient and mild. The most frequently reported were neck and shoulder muscle weakness, post-application Table 5. Oral drugs evaluated in the treatment of chronic migraine in adults: evidence of efficacy, class of recommendation, recommended doses and risk categories regarding use during pregnancy and lactation.

Peripheral nerve blocks
Not many trials have been conducted to investigate the efficacy of cephalic segment peripheral nerve blocks in the management of chronic migraine. The heterogeneity of study cohorts, designs and techniques do not allow consistent conclusions 111,112,113,114 . However, it is worth mentioning a doubleblind RCT in which 44 participants were submitted to weekly bilateral greater occipital nerve blockades for one month, with 0.5% bupivacaine or saline solution (1.5 mL). The active group showed a significant reduction in headache days, which lasted for three weeks after the end of treatment 111 (evidence class I, recommendation level B.)

NONPHARMACOLOGICAL TREATMENT AND COMPLEMENTARY THERAPIES
The number of trials contemplating nonpharmacological treatments in patients with chronic migraine is limited, but some of these showed promising results, especially when these methods were used in association with conventional treatment with prophylactic drugs.

Psychological and behavioral therapies
Positive results were seen in some trials addressing cognitive behavioral therapy, biofeedback and meditation techniques such as mindfulness. In one RCT, cognitive behavioral therapy, combined with amitriptyline, was more effective than amitriptyline alone in reducing days of headache in children and adolescents with chronic migraine 115 . Two RCTs showed biofeedback efficacy in adult patients with chronic migraine and medication-overuse headache when combined with pharmacological treatment 116,117 , with higher reduction of analgesic use and number of headache days in the long run. These results are supported by an RCT that demonstrated a reduction of oxidative stress indexes in women with chronic migraine who were submitted to biofeedback 118 . The mindfulness meditation technique has been tested in prophylactic management of chronic migraine and medication-overuse headache in a one-year follow-up RCT comparing prophylactic medication versus six weeks of mindfulness training. There was a comparable reduction in headache frequency and in the use of analgesic medication in both groups 119 . Two other trials with similar design demonstrated reduction of inflammation biomarkers 120 and catecholamine production 121 in both groups.

Physical activity
One RCT comparing isolated amitriptyline versus amitriptyline associated with aerobic exercise three times a week in the management of chronic migraine showed a greater reduction in the frequency of headache days in the combined approach group 122 .

Physiotherapeutic measures
One RCT evaluated patients with chronic migraine using prophylactic drugs (amitriptyline or topiramate). They were allocated into two groups, one of which was submitted to physiotherapy sessions (diaphragmatic breathing exercises, cervical traction and mobilization, massage therapy, myofascial release, manual compression of muscle trigger points and passive stretching of cervical muscles). A greater reduction of days of headache was observed in patients with chronic migraine and associated neck pain with the combined approach in comparison with drug treatment alone 123 .

Acupuncture
A multicenter RCT involving 205 chronic migraine patients showed acupuncture's superiority (one session per week for 12 weeks) compared with sham acupuncture, regarding the reduction in number of headache days 124 . Two subsequent class II trials comparing acupuncture with onabotulinum toxin A and topiramate demonstrated a similar or superior efficacy of acupuncture 85,125 .

Nutraceuticals
Individuals with chronic migraine using amitriptyline were enrolled in a double-blind RCT that demonstrated a greater reduction in headache days in the group that received the polyunsaturated long-chain fatty acid eicosapentaenoic acid 400mg/day and the polyunsaturated very long-chain fatty acid docosahexaenoic acid 350mg/day 126 .

Transcutaneous supraorbital stimulation
Although daily transcutaneous supraorbital stimulation using the Cefaly ® device is approved both by ANVISA and the FDA as a prophylactic treatment for episodic migraine, a single preliminary open prospective trial was performed in patients with chronic migraine, in which the goal of 50% reduction on headache days was achieved by only 30% of participants 127 .

Transcutaneous electrical vagus nerve stimulation
In a double-blind RCT including patients with chronic migraine receiving a 1 Hz or 25 Hz stimulation in the auricular branch of the vagus nerve, only patients who underwent 1 Hz stimulation achieved a reduction greater than or equal to 50% on headache days 128 . Another multicenter doubleblind RCT did not show a reduction in the number of attacks compared with a sham procedure control group, except in the open phase, after eight months of treatment 129 .

Invasive electrical stimulation of the occipital nerves
The results obtained with invasive occipital electrical stimulation of patients with chronic migraine considered refractory are controversial. A multicenter RCT with 110 participants showed that only 39% of them achieved a reduction of 50% or more in the number of headache attacks per month 130 . Another RCT compared effective stimulation with sham stimulation and no stimulation. In this trial, participants who underwent effective stimulation showed a reduction of pain severity 131 . On the other hand, a third double-blind multicenter RCT found no difference between the active stimulation and sham stimulation groups in relation to the primary endpoint of 50% or more reduction of headache days after twelve weeks 132 .

Transcranial magnetic stimulation
Two transcranial magnetic stimulation trials have shown opposite results in the management of chronic migraine. A first double-blind RCT showed a reduction of more than 50% in the frequency and intensity of attacks in the active group compared with the placebo group 133 . However, a more recent RCT demonstrated a significantly greater reduction in the frequency of headache days in the sham group in relation to the active group 134 . Table 7 summarizes level of recommendation and class of evidence for the nonpharmacological modalities in the treatment of chronic migraine. It should be remembered that the evaluation of the efficacy of these methods is compromised by the limited number and heterogeneity of the trials, as well as by the variability of methods and conflicting results.

CONCLUSION
Chronic migraine carries a substantial burden and is recognized as a difficult-to-treat condition, particularly when associated with acute medication overuse. Despite this, the use of some oral or non-oral pharmacological approaches are supported by good quality evidence. Additionally, nonpharmacological techniques are currently being subjected to a more profound scrutiny and some of them seem to be probably helpful. Novel biological therapies are currently being tested for efficacy in chronic migraine and new evidence is likely to be available soon.  Should be: