Migraine

Migraine treatments: what can be done today

This is what I tell everyone who asks me what can be done about migraine today. It is general information: it does not replace a consultation and it is not medical advice for your particular case. If something here hits close to home, we review it at a consultation before changing anything. I start with the urgent part, because it is the only thing here that cannot wait.

By Dr. Mauricio Casarsa · 16 min read · Last reviewed: Aug 15, 2026

Migraine is treated on two separate fronts: stopping the attack once it has started, and preventing it so that it happens less often. On each front there are drug options — anti-inflammatories, aspirin, paracetamol (acetaminophen), triptans, ditans and gepants for the attack; classic oral preventives and CGRP-targeting preventives for prevention — and interventional options, such as the greater occipital nerve block. Botulinum toxin has a single approved indication, chronic migraine (15 or more headache days per month), and migraine is controlled, not cured: a good result is lowering frequency and intensity and getting the predictability of your days back.

First, the urgent part

There are headaches that do not wait for an appointment. Before anything else, this:

Go to an emergency department today if:
  • The pain reached its peak in less than a minute, or it is the worst of your life.
  • It comes with fever and a stiff neck, with confusion or unusual drowsiness, with loss of strength, with difficulty speaking, with a change in your vision or with projectile vomiting.
  • It started after a blow to the head, or you are taking blood thinners (anticoagulants).
  • It is set off by coughing, by straining or by sex, or it clearly changes when you lie down or stand up.
  • It is a new or different headache and you are pregnant or have just had a baby.
  • It is a new or different headache and your immune system is weakened: HIV, transplant, chemotherapy, corticosteroids for a prolonged period.
  • The attack has already gone on for more than 3 days without letting up, or you are vomiting so much that you cannot keep any medication down.
Any one of these, on its own, is enough.

If what you have is cluster headache, what follows will not help you

This is a page about migraine. There is another condition that gets confused with it and is treated differently. It is a pain around one eye or in the temple, always on the same side, brutal, which untreated lasts between 15 minutes and 3 hours and can repeat several times in the same day. On that side your eye waters, your nose blocks up or your eyelid droops. And instead of lying down in the dark you cannot keep still: you end up pacing back and forth. It usually comes in bouts of weeks or months and then disappears for months or years. Within the bout the attacks tend to repeat at the same time of day and to wake you at night, and alcohol sets them off.

It is treated differently: the rescue treatment is not a tablet, because by the time an oral tablet starts to work that attack is already over, and the preventive is not the same one either. And it is not a "men's" headache: it also occurs in women, and considerably more often than was believed. If you recognize yourself in this description, message me and we will look at it separately. A first bout is investigated with imaging, because there are other causes that mimic it.

It is not "a bad headache"

If you have a headache on more days than not, you have already tried several medications and you feel like you organize your life around the attacks, this is for you.

Migraine has its own diagnostic criteria and it is the one that causes the most disability at the population level. In population studies in the United States, with more than 100,000 people, migraine affected 12% of people. Tension-type headache is more frequent, 38%, but it is less disabling. Many people have both mixed together.

We speak of chronic migraine when the headache occurs on 15 or more days a month for more than 3 months, with migraine features on at least 8 of those days. All headache days count toward that total, not just the migraine ones. Migraine is controlled, not cured: a good result is not "never having pain again", it is lowering frequency and intensity and getting the predictability of your days back. I have written about this point separately: chronic headaches: when to see a pain specialist →

Stopping the attack

Treating early, while the pain is still mild, gave better results than waiting for it to settle in. Early in the pain, not "just in case", and without that changing the limit on days of use per month (the numbers are further down).

There are the anti-inflammatories, aspirin, paracetamol (weaker), the specific family of the triptans and two new families: ditans and gepants for rescue. Now, the fact that they exist does not mean you can get them: some of these families are still not available in Argentina or are only just arriving, and that changes often, so it gets checked at the consultation.

The numbers are always read with the placebo alongside: with oral sumatriptan, depending on the dose, around 3 in 10 people are pain-free at 2 hours, versus 1 in 10 with placebo. Which means most people are not completely pain-free with a tablet. "It did nothing for me" often means it was taken too late, or that that particular molecule is not the one that works for you. That gets settled at the consultation: triptans have formal cardiovascular contraindications, they are not an ordinary pain reliever and they are not something to pass around among friends. And if you vomit early, the oral route stops being a route and the strategy changes.

Opioids are not the rescue treatment for migraine. In the American Headache Society guideline for the acute attack in the emergency department, injectable opioids were placed among what should not be offered. They are also among the ones that push you fastest into the trap that comes further down.

Preventing

When is it right to think about a preventive? The Consenso sobre el uso de anticuerpos monoclonales en la migraña en Argentina (Consensus on the use of monoclonal antibodies in migraine in Argentina), published in 2020 by a panel of Argentine neurologists, sets these thresholds for episodic migraine:

In all three cases the requirement is that it has been sustained over the previous 3 months: one bad month is not enough, and that detail is precisely what keeps people from classifying themselves too readily.

Careful with how this is used. These are thresholds for deciding whether it is worth trying a preventive, any preventive; they do not say which one, and they are not a basis for demanding a particular one. The criteria for the monoclonal antibodies are different, more demanding, and they come further down.

A preventive is considered to have worked when it cuts the frequency at least in half, at an adequate dose and with a minimum of 3 months of treatment.

That is the explanation for almost every "it didn't work for me": the oral ones are started low, raised slowly and need time. In an analysis of thousands of patients with chronic migraine in the United States, half had stopped their first preventive before two months. That figure measures behavior, not efficacy, but those two months are exactly the ones you have to get through.

The classic families are several: topiramate, amitriptyline, propranolol, divalproex, among others. And there are the new ones, targeting CGRP, a molecule involved in triggering the attack; they come as monthly or quarterly injectables and also as tablets. Both kinds reduce, on average, 1 to 3 migraine days per month compared with placebo. In episodic migraine, in people who have already failed two to four classes of preventives, the latter maintain their benefit.

None of these families is interchangeable with another: each has its own contraindications (asthma, low blood pressure, glaucoma, arrhythmias, pregnancy) and choosing badly is not a minor detail.

Two local clarifications. The American Headache Society has recommended them since 2024 as a first-line option, without requiring previous failures. The Argentine consensus, which is four years earlier, reserves them for after two or more failed preventives: it is a criterion that the document itself links to cost and to the evidence available at that time. This is not something you decide: it is the framework we work with here, and it is discussed with whoever prescribes your preventive. And a treatment existing is one thing, its being available in the country is another, and its being covered is yet another: three different things, and none of them automatic. It is handled case by case.

Topiramate and valproate/divalproex are contraindicated in pregnancy when used to prevent migraine, and they require effective contraception while they are being taken. And be careful, because some of these drugs interfere with hormonal contraceptives: that point is reviewed with whoever prescribed it. If you are on one of them, that is a reason to consult, never to stop it on your own.

What you cannot buy at the pharmacy

Sleep, aerobic exercise, stress management and identifying your own triggers (food, hormonal, postural) are more decisive than most people assume. They do not replace treatment and they are not "the natural solution", but without that part in order, everything else works less well. It is the part that gets skipped the most.

The frequent pain reliever trap

It is a secondary headache: it happens to people who already had headaches and use rescue medication often. It is not your fault — you took the pain reliever because it hurt. More than half of the people with pain on 15 or more days a month have this component.

The thresholds are in days of use, not tablets: 15 days a month for the common pain relievers, which count as a single family, meaning that if you rotate among several they add up; and 10 days a month for triptans, opioids, ergotamine derivatives and combinations. If you mix different families, 10 days in total is enough. Always for more than 3 months.

The combinations (several drugs in a single tablet) and the ergotamine derivatives fall under the lower threshold: 10 days, not 15. Almost everyone who writes to me recognizes Migral: check the box to see what yours contains and bring it to the consultation, because they are not all the same. Those numbers come from expert consensus, not from an experiment.

Do not stop anything on your own. With some drugs the discontinuation is done with a slow, supervised taper, and withdrawal produces a temporary worsening that in one study with hospitalized, supervised patients lasted between 4 and 10 days depending on the drug. It is a hard week, and that is why it is best to go through it with someone alongside you. The good news: usually, though not always, the picture resolves once the overuse stops, and only then can you properly assess whether the preventive was working or not.

The most useful thing you can do today, before any consultation: a calendar with two columns, days with pain and days with medication. All medication, including anything someone lent you.

Interventional treatments

The greater occipital nerve block is an outpatient procedure, it takes minutes and it is done in the office. It is a procedure I perform myself, so let me tell you its real scope, which is smaller than it is usually made out to be.

In the American Headache Society guideline on managing the acute attack in the emergency department, the occipital nerve block was placed in the highest efficacy category, a tier it shares with injectable prochlorperazine, dexketoprofen and sumatriptan. In other words: it is for the attack in the emergency department, not a long-term treatment, and it is not "the best thing out there" — it sits alongside those other options.

Now, the other half of the data, so as not to undersell it: in that same guideline, when it moves from efficacy to recommendation, the occipital nerve block and injectable prochlorperazine are the only two that ended up at the highest level, the must be offered tier, for someone arriving at the emergency department with an attack; injectable dexketoprofen and sumatriptan ended up a tier below. And the same review that gives it low certainty as a preventive gives it moderate certainty for the attack: it reduces pain severity at 30 minutes. It is still a treatment for the acute moment, not a long-term one.

As a preventive in chronic migraine there is a low-certainty benefit, and with a schedule of weekly, bilateral injections. For a one-sided block, for the ones that include a corticosteroid (alone or added to the anesthetic) and for episodic migraine, the same review says the data are not enough to draw conclusions. Whether the benefit is sustained over time is not clear either.

In people with pain reliever overuse, adding the block to withdrawal improved the pain, but the rate of successful withdrawal was the same (an open-label study, with a corticosteroid added to the anesthetic): it is an adjunct, not the way out. Its goals are narrow: relieving an attack, breaking a cycle, helping someone get out of an overuse pattern.

In the trials no serious adverse effects appeared, and the mild ones are frequent and also occur with the placebo injection. But these are small studies, of a few hundred patients: not having found serious complications is not the same as saying they do not exist. There are situations that call for precautions (anticoagulation, pregnancy, allergy to anesthetics), which is why it is assessed beforehand.

The sphenopalatine ganglion block is another matter. That same guideline, the one on the migraine attack in the emergency department, left it without a recommendation for insufficient evidence, which is not the same as saying it does not work: it ended up on the same list as injectable lidocaine, ibuprofen and ketamine. Where it has the most support, according to a 2017 systematic review, is in cluster headache, and with the caveat the review itself makes: the controlled studies are small and unreplicated. In other conditions it is a different conversation, not a dismissal.

For me, getting the indication right is worth more than doing a lot of them. Cervicogenic headache, which is where cervical blocks have their most specific indication, affects around 4% of the population and accounts for about 3% of the headaches that reach an office. Pain being felt in the neck is not enough: migraine hurts in the neck a great many times without being cervicogenic, and findings on a cervical MRI are frequent in people without headache too. You can see the full range on the services and procedures page.

Botulinum toxin: who it is for and who it is not

The same caveat as above applies: I perform this one too.

The approved indication that appears in the product's prescribing information is prevention in chronic migraine: 15 or more headache days a month, with headache lasting 4 hours or more per day. That same document states that safety and efficacy have not been established in episodic migraine, that is, 14 headache days a month or fewer. And that was put to the test: a 2025 trial with 775 patients with episodic migraine found no difference against placebo and was stopped early for lack of efficacy.

In chronic migraine, the numbers: starting from almost 20 headache days a month, at 24 weeks the treated patients came down by 8.4 days per month and those who received placebo by 6.6. In a later analysis of those same trials, signed in part by the manufacturer, 45 out of 100 halved their headache days with the toxin and 34 out of 100 with placebo. The advantage attributable to the toxin is real and solid, but moderate.

It does not stop today's attack: it is given at 31 sites every 12 weeks. In the trials the primary outcome was measured after two cycles; the Argentine consensus suggests assessing the response only after three, that is, at nine months. In practice we look at it cycle by cycle, and if no benefit appears that justifies continuing, we stop. The details of this treatment are on its own page: botulinum toxin for chronic migraine →

The most frequent adverse effects were neck pain, localized muscle weakness in the injected area and temporary drooping of the eyelid. The prescribing information also warns about the effect spreading beyond the injection site: it is uncommon, but if after a treatment session you develop difficulty swallowing, breathing or speaking, double vision or generalized weakness, you need to seek care immediately.

With most health plans the toxin requires prior authorization. That is prepared from the office, but it takes time and it does not always come through.

When to seek care without waiting

Here it is again, because it is the only thing here that cannot wait. Go to an emergency department today if:

Any one of these, on its own, is enough.

Having had migraine for 20 or 30 years does not protect you. If your usual pain has changed — it hurts differently, in a different place, at a different frequency or with new symptoms — that needs to be checked. The same goes for a new headache after 50, especially with pain in the temple, pain on chewing or visual disturbance; and for any new headache if you have a history of cancer.

It does not work the other way around: these signs make consulting mandatory, but not having them rules nothing out.

How to move forward

Start the calendar. And do not change what you have been taking on your own: bring it to the consultation, with the boxes. The first step is to get to know your case in depth: history, triggers, medication, what you have tried, how and for how long. Only then can a plan be put together. In headache care I work alongside neurology, not instead of it.

I see patients at two locations: Hepta, in San Isidro (Av. Fondo de la Legua 577) and CIAREC, in Villa Urquiza (Av. Monroe 4770, CABA). I also offer virtual first consultations to point you in the right direction and organize the next steps. For appointments, message me here or call +54 9 11 5895-3260.

This is general information: it does not replace a consultation and does not constitute medical advice for your case. If you are in the middle of an attack that will not let up, or if you have been hospitalized for this, do not try to sort it out by message.

Migraine treatments — frequently asked questions

Can an attack that has already started be stopped?
Yes, and treating early — while the pain is still mild — gave better results than waiting for it to settle in. But the numbers are always read with the placebo alongside: with oral sumatriptan, depending on the dose, around 3 in 10 people are pain-free at 2 hours, versus 1 in 10 with placebo. Which means most people are not completely pain-free with a tablet. "It did nothing for me" often means it was taken too late, or that that particular molecule is not the one that works for you. Triptans have formal cardiovascular contraindications: they are not an ordinary pain reliever and they are not something to pass around among friends. And treating early does not change the limit on days of use per month.
Does botulinum toxin work for any kind of migraine?
No. The approved indication that appears in the product's prescribing information is prevention in chronic migraine: 15 or more headache days a month, with headache lasting 4 hours or more per day. That same document states that safety and efficacy have not been established in episodic migraine, that is, 14 headache days a month or fewer. And that was put to the test: a 2025 trial with 775 patients with episodic migraine found no difference against placebo and was stopped early for lack of efficacy.
Why does taking a pain reliever every day end up making the pain worse?
Because a secondary headache from overuse can develop: it happens to people who already had headaches and use rescue medication often. It is not your fault — you took the pain reliever because it hurt — and more than half of the people with pain on 15 or more days a month have this component. The thresholds are in days of use, not tablets: 15 days a month for the common pain relievers, which count as a single family, and 10 days a month for triptans, opioids, ergotamine derivatives and combinations; if you mix different families, 10 days in total is enough, always for more than 3 months. Do not stop anything on your own: with some drugs the discontinuation is done with a slow, supervised taper.
When is it right to start a preventive treatment?
The Consenso sobre el uso de anticuerpos monoclonales en la migraña en Argentina, published in 2020 by a panel of Argentine neurologists, sets three thresholds for episodic migraine: migraine without aura with more than 6 attack days a month, regardless of intensity; migraine without aura with 3 or more days with pain a month, if those episodes cause moderate or severe disability measured with a scale; and migraine with aura with 1 or more episodes a month. In all three cases it has to have been sustained over the previous 3 months. These are thresholds for deciding whether it is worth trying a preventive, any preventive: they do not say which one. And a preventive is considered to have worked when it cuts the frequency at least in half, at an adequate dose and with a minimum of 3 months of treatment.
Does the greater occipital nerve block work?
Its scope is smaller than it is usually made out to be, and it is a procedure I perform myself, so I will say it plainly. In the American Headache Society guideline on managing the acute attack in the emergency department it was placed in the highest efficacy category, a tier it shares with injectable prochlorperazine, dexketoprofen and sumatriptan: in other words, it is for the attack in the emergency department, not a long-term treatment. As a preventive in chronic migraine there is a low-certainty benefit, and with a schedule of weekly, bilateral injections; for a one-sided block, for the ones that include a corticosteroid and for episodic migraine, the same review says the data are not enough to draw conclusions. Its goals are narrow: relieving an attack, breaking a cycle, helping someone get out of an overuse pattern.

Do you organize your life around the attacks?

Start with the calendar: days with pain and days with medication, all medication. With that in hand we can see what is going on and put together a plan for your case. Message me and we can go over it at a consultation.

💬 Do you organize your life around the attacks?

References and further reading

  1. Headache Classification Committee of the International Headache Society (IHS). The International Classification of Headache Disorders, 3rd edition. Cephalalgia. 2018;38(1):1-211. PubMed ↗
  2. Robbins MS. Diagnosis and Management of Headache: A Review. JAMA. 2021;325(18):1874-1885. PubMed ↗
  3. Derry CJ, Derry S, Moore RA. Sumatriptan (all routes of administration) for acute migraine attacks in adults — overview of Cochrane reviews. Cochrane Database Syst Rev. 2014;2014(5):CD009108. PubMed ↗
  4. Robblee J, Minen MT, Friedman BW, et al. 2025 guideline update to acute treatment of migraine for adults in the emergency department: The American Headache Society evidence assessment of parenteral pharmacotherapies. Headache. 2026;66(1):53-76. PubMed ↗
  5. Doctorovich ED, Martín-Bertuzzi F, Goicochea MT, et al. Consenso sobre el uso de anticuerpos monoclonales en la migraña en Argentina. Rev Neurol. 2020;70(4):149-158. PubMed ↗
  6. Charles AC, Digre KB, Goadsby PJ, et al. Calcitonin gene-related peptide-targeting therapies are a first-line option for the prevention of migraine: An American Headache Society position statement update. Headache. 2024;64(4):333-341. PubMed ↗
  7. Atraszkiewicz D, Ünal E, Bassett P, et al. Greater occipital nerve block for the treatment of migraine: An umbrella review, systematic review, and meta-analysis. Cephalalgia. 2025;45(12):3331024251398390. PubMed ↗
  8. Arab A, Khoshbin M, Karimi E, et al. Effects of greater occipital nerve block with local anesthetic and triamcinolone for treatment of medication overuse headache: an open-label, parallel, randomized, controlled clinical trial. Neurol Sci. 2022;43(1):549-557. PubMed ↗
  9. Ho KWD, Przkora R, Kumar S. Sphenopalatine ganglion: block, radiofrequency ablation and neurostimulation — a systematic review. J Headache Pain. 2017;18(1):118. PubMed ↗
  10. Pozo-Rosich P, Blumenfeld AM, Lipton RB, et al. OnabotulinumtoxinA for the preventive treatment of episodic migraine: Results from the phase 3, multicenter randomized, double-blind, placebo-controlled phase of the PRECLUDE trial. Cephalalgia. 2025;45(10):3331024251370769. PubMed ↗
  11. Dodick DW, Turkel CC, DeGryse RE, et al. OnabotulinumtoxinA for treatment of chronic migraine: pooled results from the double-blind, randomized, placebo-controlled phases of the PREEMPT clinical program. Headache. 2010;50(6):921-936. PubMed ↗
  12. Silberstein SD, Diener HC, Dodick DW, et al. The Impact of OnabotulinumtoxinA vs. Placebo on Efficacy Outcomes in Headache Day Responder and Nonresponder Patients with Chronic Migraine. Pain Ther. 2020;9(2):695-707. PubMed ↗
Dr. Mauricio Casarsa

Dr. Mauricio Casarsa

Anesthesiologist. Postgraduate training in Pain Medicine and Interventional Pain (UBA — Fundación Dolor; UNLP — CAIDBA). Staff physician at Hospital Alemán, Buenos Aires.

License MN 137.756 · San Isidro · Villa Urquiza

Pedir turno