Customize your JAMA Network experience by selecting one or more topics from the list below.
Latanoprost is a phenyl-substituted isopropyl ester of prostaglandin
F2α that enhances uveoscleral outflow and lowers intraocular
pressure. Its potency, once-daily dosing schedule, and low incidence of adverse
effects led to widespread use immediately after its release in the United
States. Since then, adverse effects such as anterior uveitis and cystoid macular
edema1 have been reported. This report describes
3 patients who experienced new onset of migraine headache after taking latanoprost;
2 patients had to discontinue the drug. This adverse effect previously has
not been reported in association with latanoprost use. The trigeminovascular
theory of migraine pathogenesis could explain this association.
Report of Cases
A 65-year-old man with primary open-angle glaucoma that was previously
well controlled with topical therapy became allergic to dipivefrin hydrochloride
and dorzolamide hydrochloride. He was intolerant of β-blockers because
of a cardiac arrhythmia. He was prescribed latanoprost for both eyes at bedtime.
Although the patient had no prior history of migraine, in the ensuing months
he began to have migraine headaches. The frequency and severity increased
until the headaches were occurring daily. The pain was throbbing, bifrontal,
and associated with lightheadedness and photophobia. It was not relieved by
over-the-counter or narcotic analgesics, and the patient was virtually incapacitated.
He consulted his neurologist, but because he had been examined 1 year previously
for vertigo, no new tests were performed. Eight months later he informed his
ophthalmologist, and latanoprost treatment was discontinued. He had almost
immediate relief, with only 1 migraine the following week. He was headache-free
for the following 10 months.
The patient agreed to a rechallenge with latanoprost. After the second
night of therapy, he woke up with another migraine headache. He continued
the latanoprost for 2 more nights but again had to stop the drug because of
incapacitating pain. Forty-eight hours later, his headache resolved. It has
not recurred within 4 months of follow-up.
A 67-year-old white man with primary open-angle glaucoma taking levobunolol
hydrochloride at a dose of 0.5% was prescribed a nighttime dose of latanoprost
for both eyes to treat a mild rise in intraocular pressure. The next morning
he awoke with a severe bifrontal headache. It was throbbing in nature and
associated with photophobia and slight blurring of vision. He did not normally
suffer from headaches and described it as the worst headache he had ever had.
It was not relieved with an over-the-counter analgesic. It intensified, and
4 days later he went to the emergency room. A computed tomagraphic scan and
neurologic consultation were performed and found to be normal. On the sixth
day, he realized the temporal association with latanoprost and called the
eye clinic, whereupon he was instructed to discontinue the drug. His headache
disappeared within 24 hours and has not recurred with 1 year of follow-up.
A 54-year-old woman with primary open-angle glaucoma taking betaxolol
hydrochloride at a dose of 0.5% for both eyes had mild progression of visual
field loss in her left eye. Latanoprost was added nightly to her left eye.
At approximately 3 AM on the night after her first dose, she was awakened
by a severe unilateral pounding headache extending from the left eye and brow
to the left cranium. There were no associated neurologic symptoms. The patient
does not normally suffer from headaches. The headache was not relieved by
acetaminophen or codeine phosphate but spontaneously resolved the next day.
It recurred for 3 more nights after instillation of latanoprost. On the fourth
night, the headache did not occur. The patient continued to take the drug,
and the headache has not returned.
Migraine headache is a significant cause of morbidity and lost work
time in the United States. About 6% of men and 18% of women experience regular
migraines and have at least 1 attack per year.2
Peak incidence of migraine onset is in the teens; onset after age 50 years
is rare. Patients with a history of migraine are not usually examined unless
"alarm symptoms" are present. These include sudden changes in attack duration,
frequency, or severity; abnormal neurological examination results; or onset
after age 50 years.3 All of the patients
described in this report were older than 50 years and had no prior history
of migraine headache.
Many theories of migraine pathogenesis have been proposed, but the exact
mechanism is still not understood. People prone to migraine have underlying
cortical hyperexcitability, and the attack is started by 1 or more triggers.
Humoral mediators have been extensively investigated, but the exact role of
each remains in debate.3 The visual aura
is caused by a wave of depolarization called cortical-spreading depression.
In animal experiments, this is accompanied by changes in cerebral blood flow
similar to those seen in migraine, suggesting that the phases of vasoconstriction
and vasodilation are only epiphenomena. A new theory of migraine pathogenesis
suggests that the trigeminovascular system plays a key role via activation
of pain fibers from the large cerebral vessels and dura, which lead centrally
to the thalamus and cortical pain centers.4
Ocular abnormalities can cause headaches in several ways. Ciliary muscle
spasm can cause pain referred to the brow but did not appear to be present
in these patients because they did not display acute myopia. In addition,
latanoprost causes relaxation rather than contraction of the ciliary muscle,5 which may be how it increases uveoscleral outflow.
Ciliary body irritation due to anterior uveitis was also considered, but this
was not present on examination either.
Latanoprost may have caused migraine via activation of the trigeminovascular
system. The parent compound, prostaglandin F2α, is a vasoactive
agent. Clinical trials with the tromethamine salt of prostaglandin F2α revealed a 50% incidence of conjunctival hyperemia, ocular
pain, and headache.6 The ocular hyperemia
induced by prostaglandin F2α isopropyl ester is partially
mediated by nitric oxide release, although the mechanism by which it causes
headache has not been explained.7 Latanoprost
stimulates the production of endogenous prostaglandins D2, E2, and F2α in the ciliary body, all of which are vasoactive
agents.8 Thus, a small concentration of
the drug could lead to significant levels of vasoactive prostaglandin in the
eye. In addition, if migraine can indeed be triggered by direct stimulation
of the trigeminovascular system, the small amounts of latanoprost entering
the eye could be sufficient to do so. The half-life of latanoprost in the
aqueous humor and ciliary body is 3 hours,9
which could circumvent the advantage of low systemic absorption.
Migraine headache associated with latanoprost has not been previously
described. However, it is a major adverse effect of a similar prostaglandin
analog investigated for use in glaucoma.6
Because of the atypical initial symptoms, 2 of the patients had neurologic
consultations, 1 of whom was investigated to rule out a serious neurologic
problem. These 2 patients had to discontinue the drug for relief. New prostaglandin
analogs are in clinical trials for use in glaucoma; there is reason to be
concerned that they will cause headache as well. Ophthalmologists should first
discontinue latanoprost in a patient developing headache, because they may
spare their patient the inconvenience and risk of an investigation.
This study was supported by grant EY 12962 from the National Institutes
of Health, Bethesda, Md.
Corresponding author and reprints: Bonnie C. Weston, MD, FRCSC, Department
of Ophthalmology, University of South Alabama, HSB Room 2500, 307 University
Blvd, Mobile, AL 36688.
Weston BC. Migraine Headache Associated With Latanoprost. Arch Ophthalmol. 2001;119(2):300–301. doi:
Coronavirus Resource Center
Create a personal account or sign in to: