Bridge treatments for cluster headaches

Bridge treatments (sometimes called transitional treatments) are short-term therapies you take at the very start of a cluster cycle to suppress attacks quickly, while a longer-acting preventive ramps up. For example, verapamil (the most common preventive) can take two to five weeks to reach its effective dose. A bridge fills that gap.

A bridge isn't something you take long term. The bridge and the preventive are often started on the same day (though most preventives can and should ideally be started before the beginning of a cycle, for episodic patients), and the bridge is tapered off over two to three weeks as the preventive starts to work.

There is one exception. If you're episodic and your bouts reliably last only two to four weeks, a bridge on its own may be enough, because its window of effect covers the entire cycle. In that case you may not need a long-term preventive at all. It's worth revisiting this with your doctor if a future bout starts running longer, since cycles can lengthen over the years.

Each treatment section below follows the same structure: a brief description, then Protocol (how it's given), Evidence (what the data show), and Side effects and considerations (what to watch for, plus practical issues).

---

Summary of bridge treatments for cluster headaches

Treatment	Speed	Duration of benefit	Evidence	Main barrier
Oral prednisone taper	1-3 days	2-3 weeks (covers taper)	RCT (strong)	Steroid side effects; can't repeat often
Greater occipital nerve block (GON block)	1-3 days	2-4 weeks	RCT (strong)	Needs a trained clinician
IV DHE protocol	1-2 days	Until next bout	Open-label	Requires hospital admission
IV methylprednisolone	Days	Weeks	Limited	Hospital only; same risks as oral steroids
Long-acting oral triptan	A few days	Length of a short bout	Small case series	Triptan contraindications
IV ketamine course	Hours to days	Weeks to months	Small case series	Specialist clinic; abuse potential
Sphenopalatine ganglion (SPG) block	Minutes to hours	Days to weeks	Case series only	Needs a specialist; weak evidence
SPG pulsed radiofrequency (PRF)	Days	Months	Small case series	CT guidance; specialist only

---

Oral prednisone taper

A short, high-dose course of an oral steroid (usually prednisone, or prednisolone in some countries) is the most powerful bridge available. Most patients feel better within one to three days, and many become attack-free in the first week.

Protocol

A common schedule is 100 mg per day for five days, then dropping by 20 mg every three days, finishing in just under three weeks. Some doctors prefer a gentler start at 60 to 80 mg with a slower taper. Take the pill in the morning, with food: steroids can disrupt sleep if taken late in the day, and they can irritate the stomach lining if taken on an empty one.

The taper itself isn't optional. Don't stop prednisone abruptly. Tapering down slowly gives your adrenal glands time to wake back up. If you've been on prednisone for more than about two weeks, ask your doctor about a simple blood test (sometimes called an ACTH stimulation test) before fully stopping, to confirm your adrenals are working again on their own.

Evidence

The strongest evidence comes from the PREDCH trial, which randomized 116 patients with episodic cluster headache to prednisone or placebo. Both groups also took verapamil. The trial counted attacks in the first week: 7.1 on prednisone versus 9.5 on placebo, and about a third of prednisone patients had at least a 50% reduction by day 7, compared to 13% on placebo.^[1] In real-world practice the effect is usually larger than the trial's strict endpoint suggests: roughly 60% to 75% of patients see meaningful attack reduction during the taper, with many becoming attack-free in the first week. Every major guideline endorses a short oral steroid course as a first-line bridge.^[2][3][4]

Side effects and considerations

A two- to three-week course is generally safe. Common short-term effects include trouble sleeping, increased appetite, mild mood changes (a "wired" feeling or irritability), raised blood sugar, mild fluid retention, and a flushed face. Taking it in the morning with food keeps most of these manageable.

The reason it's short-term only is the cumulative damage from repeated long-term steroid exposure: weight gain, weakened bones, high blood pressure, stomach ulcers, suppressed immunity, cataracts, and at high cumulative doses, damage to the hip joint. If you've been using steroids at the start of every cycle, ask your doctor about greater occipital nerve blocks instead (see below), since they deliver the steroid locally and limit how much reaches the rest of your body.

Tell your doctor before starting if you have diabetes, uncontrolled high blood pressure, glaucoma, osteoporosis, an active infection, or a history of stomach ulcers.

---

Greater occipital nerve block (GON block)

A greater occipital nerve block (GON block) is a small injection of numbing medication plus a long-acting steroid placed near the greater occipital nerve, a nerve that runs up the back of your skull on each side. The numbing medication quiets the nerve immediately, and the steroid keeps it calmed for weeks. That's enough to suppress most cluster attacks during the window the steroid is active. Because the injection delivers the steroid locally to the area whose nerves drive cluster pain, very little reaches the rest of your body, which is what makes GON blocks safer than an oral prednisone taper.

Most patients feel relief within 24 to 72 hours, and the benefit usually lasts two to four weeks, sometimes longer. The GON block is the only bridge given the highest evidence rating by the American Headache Society.

Protocol

The injection is given at the back of your skull, on the side where you get your attacks. The clinician finds the right spot by touch (no imaging or scan is needed), cleans the skin, and injects a small volume just under the scalp. The injection itself takes a few seconds. You can usually go home immediately afterwards. The full appointment takes 10 to 15 minutes.

Some clinicians give a single injection, while others give three injections two to three days apart, which can extend the duration of benefit. The numbing medication is typically lidocaine or bupivacaine. The steroid is typically betamethasone, methylprednisolone, triamcinolone, or cortivazol. The exact choice varies by country and clinician.

To limit how much steroid your body absorbs over the years, most guidelines suggest no more than four GON block injections per year.

Evidence

In the landmark double-blind trial, a single injection made 11 of 13 patients (85%) attack-free at one week, compared to 0 of 10 on placebo.^[5] Eight of those patients were still attack-free at four weeks. A later trial used three injections of a longer-acting steroid (cortivazol) and gave 20 of 21 patients near-complete relief during the first 15 days.^[6] A 2024 trial confirmed the same effect with methylprednisolone and lidocaine.^[7]

The GON block is the only preventive given a Level A recommendation by the AHS, and EAN 2023 strongly recommends it.

Side effects and considerations

Side effects are minimal and include brief soreness at the injection site and occasional small bruising. With repeated injections over time, there's a small risk of hair thinning or skin changes (a small dent in the fat layer) at the injection spot. Allergic reactions are rare but possible. There are no systemic interactions with other cluster headache medications.

Because the steroid stays mostly localized, a GON block is the safer bridge for anyone with diabetes, uncontrolled high blood pressure, glaucoma, osteoporosis, or a history of stomach ulcers, all of which make oral steroids riskier.

The main practical hurdle is access. A GON block requires a clinician trained to perform it. Neurologists, pain specialists, and some headache nurses do them routinely but primary care doctors usually don't. If your doctor doesn't offer GON blocks, ask for a referral to a headache clinic.

---

Other bridges

A few other treatments are used less commonly and only in specific situations. These are listed roughly from strongest to weakest evidence for use as a bridge.

• IV dihydroergotamine (DHE). This is reserved for the most stubborn bouts that haven't responded to anything else. It is given as an in-hospital infusion, typically 0.5 to 1 mg every eight hours for three to five days, together with metoclopramide (an anti-nausea drug). About 84% of patients become attack-free during the admission. DHE can't be used within 24 hours of a triptan, and shouldn't be used if you have heart disease, peripheral vascular disease, or kidney problems. See the abortives chapter for more on DHE.
• IV methylprednisolone. This is used at some European centers when oral steroids aren't an option. The dosage is typically 250 to 500 mg per day for three to five days. It carries the same risks as oral steroids, and it is only delivered in hospital.
• Long-acting oral triptans. These are typically frovatriptan (2.5 to 5 mg per day) or naratriptan (2.5 mg twice daily), given for the length of a short bout (under about eight weeks). Evidence is limited to small case series. The usual triptan cautions apply (no heart disease, no ergot drugs in the previous 24 hours, not in pregnancy).
• IV ketamine. A clinic-based infusion (typically 0.5 mg/kg over 30 to 60 minutes, sometimes combined with magnesium) that can reduce attack frequency for weeks to as long as 18 months in small case series, even though the drug itself clears the body in hours. This bridge-like behavior is the reason ketamine is included here as well as in the abortive chapter. Evidence remains limited to open-label series; two randomized trials are underway. See the abortive chapter for the full protocol and side-effect profile.
• Sphenopalatine ganglion (SPG) block. A small amount of local anesthetic, sometimes mixed with a steroid, delivered through the nose to a nerve cluster behind the nasal cavity. It is usually performed by a neurologist or pain specialist using a thin catheter. It is used in refractory chronic cluster headache, but the evidence is weak: there's no randomized trial in cluster headache, and the available case series are small. Patients often confuse this with the Pulsante SPG stimulator, a small implanted device that's a different procedure entirely and isn't currently available for new implants.
• SPG pulsed radiofrequency (PRF). A more durable version of the SPG block. Instead of bathing the nerve cluster in anesthetic, a thin needle is guided to the SPG under CT imaging and delivers short bursts of high-frequency electrical current, which interrupt nerve signaling for months at a time. A case series of 16 patients with refractory cluster headache reported significant attack reduction lasting 12 to 30 months.^[8] It is only available at a few specialist pain clinics, and evidence is still limited to small uncontrolled studies.

---

Choosing between bridge treatments

The choice between oral prednisone and GON block depends on a few practical factors:

• How fast you need relief. Both work within a few days. Oral prednisone is slightly faster on average (the GON block usually catches up by day 3).
• Your medical history. A GON block is safer if you have diabetes, high blood pressure, glaucoma, osteoporosis, or a history of ulcers.
• Whether you've used a bridge before. If you've already had steroids at the start of recent cycles, a GON block is preferred next time to limit your total steroid exposure.
• Access. A GON block requires a trained clinician whereas oral prednisone only requires a prescription.

In practice, most clinicians give a GON block at the first appointment, start titrating verapamil the same day, and add an oral prednisone taper only if attacks keep breaking through. The bridge fades out over two to three weeks as the verapamil reaches its effective dose.

---

References

1. ↩ Obermann M, Nägel S, Ose C, et al. (2021). Safety and efficacy of prednisone versus placebo in short-term prevention of episodic cluster headache: a multicentre, double-blind, randomised controlled trial. Lancet Neurology, 20(1), 29–37. Link
2. ↩ May A, Evers S, Goadsby PJ, Leone M, Manzoni GC, Pascual J, et al. (2023). European Academy of Neurology guidelines on the treatment of cluster headache. European Journal of Neurology, 30(10), 2955–2979. doi:10.1111/ene.15956
3. ↩ Robbins MS, Starling AJ, Pringsheim TM, Becker WJ, Schwedt TJ (2016). Treatment of Cluster Headache: The American Headache Society Evidence-Based Guidelines. Headache, 56(7), 1093–1106. doi:10.1111/head.12866
4. ↩ National Institute for Health and Care Excellence (2021). Headaches in over 12s: diagnosis and management (CG150). NICE Clinical Guideline. Link
5. ↩ Ambrosini A, Vandenheede M, Rossi P, et al. (2005). Suboccipital injection with a mixture of rapid- and long-acting steroids in cluster headache: a double-blind placebo-controlled study. Pain, 118(1-2), 92–96. Link
6. ↩ Leroux E, Valade D, Taifas I, et al. (2011). Suboccipital steroid injections for transitional treatment of patients with more than two cluster headache attacks per day: a randomised, double-blind, placebo-controlled trial. Lancet Neurology, 10(10), 891–897. Link
7. ↩ Chowdhury D, Kordcal SR, Nagane R, Duggal A (2024). ANODYNE study: a double-blind randomized trial of greater occipital nerve block of methylprednisolone and lignocaine versus placebo as a transitional preventive treatment for episodic cluster headache. Cephalalgia, 44(10), 3331024241291597. Link
8. ↩ Fang L, Jingjing L, Ying S, Lan M, Tao W, Nan J (2016). Computerized tomography-guided sphenopalatine ganglion pulsed radiofrequency treatment in 16 patients with refractory cluster headaches: twelve- to 30-month follow-up evaluations. Cephalalgia, 36(2), 106–112. Link