The First-Line Glaucoma Drop That Flatters Its Own Pressure Reading

The Academy 2026 meeting opened in Anaheim on September 30. There, Daniel Epshtein, an associate clinical professor at SUNY College of Optometry, told a room of eye doctors that the drug they reach for first in glaucoma may be flattering the one number they use to judge whether it works. His evidence came from patients at the Albuquerque VA Medical Center, and as Healio reported from the meeting, it pointed somewhere nobody prescribing these drops should be comfortable with.

The drug is a prostaglandin analog, latanoprost and its relatives. European and Asia-Pacific glaucoma guidelines recommend the class as first-line treatment, according to a 2023 review in the Indian Journal of Ophthalmology. It goes in once a day at bedtime, it is cheap, and it brings down intraocular pressure, which is the one glaucoma risk factor doctors can actually do something about. For decades the arrangement has been simple: the patient puts in the drop, the number on the chart falls, and the falling number is taken as proof the optic nerve is being spared.

The trouble starts with how that number is obtained. In a routine exam nobody measures eye pressure directly. A clinician presses a small probe against the front of the eye and reads how much force it takes to flatten a fixed patch of cornea. The method, Goldmann applanation, carries a built-in assumption about how thick and how stiff that cornea is, so a thinner, more yielding cornea gives way more easily and the gauge reports a pressure lower than the one actually bearing down on the nerve. Prostaglandin analogs thin the cornea.

Epshtein’s group drew patients with glaucoma, ocular hypertension or suspected glaucoma from a longitudinal study at the Albuquerque VA and measured their corneas with the same ultrasound pachymeter at baseline and again about three years later. Of the 128 eyes, 67 were on a prostaglandin and 61 were not. Over those three years the treated corneas thinned by an average of 11.7 microns, against 3.4 microns in the untreated eyes, and one patient’s cornea lost 35 microns. The treated eyes also scored significantly lower on corneal hysteresis and corneal resistance factor, two measures of how the tissue absorbs and resists force. They were turning into exactly the kind of cornea a tonometer misreads.

To size the error, the team set the ordinary Goldmann-correlated reading beside a corneal-compensated measurement designed to see past the cornea’s distortion. In treated eyes that lost at least 20 microns of thickness, a group Epshtein called “super-thinners,” the two numbers diverged by an average of 3.3 mm Hg; in untreated eyes past the same cutoff, the gap was 1.4 mm Hg. He gave the room a bedside version: a chart that reads 15 when the compensated pressure is 18 means “you’re underestimating eye pressure by about 3 mm Hg.” Three points is not a rounding error in a disease managed to targets a few points wide.

The next question is whether the flattering reading costs anyone their sight, and the Albuquerque numbers lean the wrong way. Eyes whose treatment eventually had to be stepped up had thinned more than eyes that held steady, yet their average Goldmann pressure was 15.8 mm Hg, against 16 in the rest, which is if anything a shade better. The readings matched. The corneas did not. The super-thinners also lost visual field at 0.32 decibels a year, against 0.18 for eyes that thinned less, nearly twice the rate of decline. Over the long run, Epshtein warned, the underestimation “may lead to a misleading perception of adequate IOP control.”

The study has the limits you would expect from one VA clinic. It is observational, it was presented at a meeting rather than published in a journal, and the super-thinners are a subset of only 128 eyes. It cannot show that the drops caused the faster vision loss, since a patient on aggressive therapy may simply have aggressive disease. What it does have is company. A 2022 systematic review in the Journal of Clinical Medicine pooled 22 prospective studies covering 3,880 participants in 12 countries and found prolonged prostaglandin use associated with “a small but statistically significant reduction” in corneal thickness. Severe thinning, 25 microns or more, was uncommon, but in the two studies that reported it, it hit 3.8 to 14.8 percent of participants. A side effect that is modest on average and severe in a few eyes is easy to miss. It is easier still to miss when it nudges the instrument used to judge the drug toward a better-looking number.

Nobody is arguing that the drops fail to lower pressure. The problem is the routine built around them, in which corneal thickness gets measured once at the start of care and filed away as though it were a fixed trait, like eye color. Epshtein’s prescription was blunt: “We should be rechecking central corneal thickness in our patients using prostaglandin analogs.” He followed it with the admission that “this is not something that is routinely done in clinic practice.” His own disclosures run long: consulting and speaking for Zeiss and Tarsus Pharmaceuticals, advisory-board seats with Alcon, Apellis, Harrow, Iveric Bio/Astellas and Regeneron, and presentation support from Annexon. Those are the ordinary entanglements of almost anyone at a podium in this field, and the measurements have to stand on their own. The fix he is asking for, though, needs a device most glaucoma clinics already own. The pachymeter is sitting in the room. Nobody has told anyone to pick it back up.

Which leaves the patient in the exam chair, whose pressure has read a steady 15 for years while the gray edge of the visual-field printout creeps inward, trusting a number that the bottle in the medicine cabinet has been teaching, a few microns at a time, to lie.

Sources

1. Healio: Prostaglandin analogs may lead to IOP underestimation in glaucoma
2. Kim and Yim, Effect of topical prostaglandin analogue therapy on central corneal thickness: a systematic review, Journal of Clinical Medicine (2022), research summary
3. Prostaglandin analogs in ophthalmology, Indian Journal of Ophthalmology (2023)