To treat or not to treat? Medication underuse headache,a novel reframing

Are migraine medications disease-modifying? This question often arises for clinicians recommending preventive therapy, and for patients deciding whether to take any migraine-targeted medications. Why take medication if sleeping off an attack will help? Why use a daily medication if symptoms are intermittent or not severe?

In their study entitled ‘Medication “underuse” headache’, Rattanawong et al. explore the impact of suboptimal migraine treatment (1). They propose a new term, medication underuse headache, to suggest that the treatments we use matter and may help prevent chronification of migraine.

The paper reviews evidence around “underuse” of migraine medications, which the authors define as ineffective use of appropriate and inappropriate medication, underutilization, inappropriate timing of usage, or patient dissatisfaction with medication. Although the concept of medication underuse may feel foreign, the headache field could look to areas such as multiple sclerosis where disease modification is a well-established goal. When medications are disease-modifying, their underuse can negatively impact the course of the disease. Is this the case for migraine medications? Read on and find out, then read the fascinating article by Rattanawong et al!

The goals of therapy in migraine are shifting sands. Traditional migraine preventive trial targets of a 50% reduction in attack frequency help assess an intervention versus a comparator, but may not imply meaningful improvement in quality of life. As Rattanawong et al. point out, increasing evidence suggests that headache begets headache (1,2). This implies that the more time the brain spends in a migraine-on state, or walking the migraine path, as we explain to patients, the more easily activated that pathway can become until eventually the path becomes a highway. Letting the plants grow over the (trigeminovascular) pathway, whether by effective acute treatments that turn off attacks quickly, or preventive options that decrease attack frequency, may keep migraine burden from worsening.

Currently accepted risk factors for migraine chronification include the modifiable (e.g. smoking, obesity, caffeine intake, medication overuse, potentially modifiable comorbidities such as depression) and nonmodifiable (e.g. female sex, lower socio-economic status, adverse childhood experiences, stressful life events). Some risk factors such as suboptimal acute treatment are harder to categorize (3,4).

The paper focuses on underuse of medications, while acknowledging that non-pharmacological measures such as lifestyle modifications and yoga can be powerful tools in migraine management. Unfortunately, these types of interventions are challenging to study, and no similar work on their underuse is available. Perhaps someday healthcare professionals will advise avoiding underuse of mindfulness!

Among modifiable risk factors for migraine progression, acute medications are vilified as potential perpetrators of medication overuse headache. Many acute therapies can perpetuate headache in patients with migraine, but this messaging sometimes causes the pendulum to swing too far – patients may wait to treat acutely, decreasing likelihood of efficacy (5). Acute medications such as triptans typically work best before central sensitization/allodynia onset, which is usually within 60 minutes of attack onset (6). Importantly, the newer gepant medications are not thought to cause medication overuse headache, even when taken frequently, enabling early treatment even for patients with frequent attacks.

Rattanawong et al. propose that patients with multiple headache days per week should wait to see whether the pain will intensify ‘with unilateral throbbing and associated symptoms’ and then take medication once sure of a migraine attack. The authors also note that acute medication should be taken within the 60-minute window before onset of allodynia. These recommendations may be hard to implement since for many patients, attacks intensify over hours, during which time the acute treatment window may close. The authors also examine different strategies regarding stepped versus stratified care, comparing the American Headache Society recommendation (stratify medication choices depending on attack characteristics) with the European guideline (start with untargeted, cheaper options for a few weeks before determining whether migraine-targeted options are needed). Some strategies may restrict patients’ ability to take medications early enough for efficacy, and after enough trials of unhelpful medications, patients may give up and underuse medications. In Canada we commonly recommend the traffic light of migraine approach, in which patients categorize attacks by a color reflecting functional impact (green = I can go, yellow = I am slowed down, red = I have to stop) and move directly to migraine-targeted options or combinations of options for more functionally impactful attacks (7). Goal-setting and education may increase acute treatment success.

Despite advances in migraine-targeted therapies, some patients have not yet been helped by available options. As Rattanawong et al. note, patients with a suboptimal response to acute therapy are at higher risk of migraine progression (1,8), and one study found that patients with an inadequate response to triptans had higher headache frequency than triptan responders (9). However, we do not know whether differences in acute medication responsiveness reflect differences in disease process. Put differently, is there a biological difference in the migraine pathway between those who benefit from currently available medications and those who do not? This is an area for further study, and the ongoing identification of new targets such as PACAP and ATP-sensitive potassium channels holds promise. Pharmacogenetics/pharmacogenomics are another burgeoning area since genetic polymorphisms likely impact medication efficacy and tolerability (10).

Medication tolerability can play a large role in whether a medication is used or not. As the authors highlight, numerous studies have shown high patient dissatisfaction with efficacy and tolerability of both acute and preventive treatments. Preventive studies suggest that few patients stay on treatment longer than a few months. Newer migraine-targeted agents seem better tolerated, which may reduce this barrier to appropriate medication use. Indeed, a recent open-label study comparing erenumab with traditional oral preventives in patients with episodic migraine who had been failed by 1–2 oral agents found that only 2.9% of patients who initiated treatment on erenumab discontinued because of adverse events over one year, compared to 23.3% of patients who started treatment on oral migraine preventive medications (11). At the same time, 56.2% of patients started on erenumab achieved the primary endpoint of staying on the medication for a year and experiencing a 50% or greater reduction in monthly migraine days at 12 months, compared to only 16.8% of those started on a traditional oral preventive.

Timing matters when starting a preventive, and Rattanawong et al. highlight growing evidence that starting effective treatments before the migraine pathway has enlarged, i.e. when patients are still at a relatively low migraine attack frequency, may help prevent migraine chronification.

Studies place the threshold of migraine chronification risk at four or more migraine days per month (12). If this is the case, we should be plastering billboards and subway stations with educational material about migraine to reach patients (and educate providers) well before most ever seek care for headache, and initiate treatment before they cross the four day per month threshold. This choice can be individualized depending on patient preference, but the current study strongly suggests we reframe the way we conceptualize medications. We could take this further and consider elevating our treatment goals. Prevention is better than cure, as the adage says, so especially with newer treatments that show better efficacy and tolerability, the time to shift treatment goals towards three or fewer migraine days per month or even no headache may be upon us. Targeting 50% reduction may do patients a disservice and keep them at an attack frequency where they remain vulnerable to worsening.

We should remain cognizant of the relatively short duration of long-term safety data available for CGRP-targeted therapies, measured in years rather than decades. Studies indicate that some but not all patients worsen upon stopping CGRP monoclonal antibodies (13). A lengthier time on treatment and on a more favorable migraine path may increase the likelihood of successfully stopping therapy without migraine disease worsening, as a recent study which examined stopping CGRP monoclonal antibodies after one year versus two years suggests (14).

Rattanawong et al. acknowledge that studying a lack of something, such as their concept of medication underuse, is not straightforward. They conclude with a suggestion to focus future research on optimal treatment initiation timing, to clarify when it is better to treat than not to treat.

Educating patients and healthcare providers about the impact of migraine as a brain disease, and that newer treatments are often effective and well-tolerated, may significantly lessen the number of patients who progress to higher migraine disease burden.

Let’s let the plants grow.