Glaucoma expert looks back to future

He began with the year 1973, in which the magnitude of the problem was unknown; there were no criteria for glaucoma screening, diagnosis, or progression; and IOP was the only known risk factor. Since then, a number of developments in diagnostic techniques and treatment options have occurred, but Dr. Sommer focused on conceptual issues.

"Now, IOP is recognized as a risk continuum, at least by some," said Dr. Sommer, professor of ophthalmology and epidemiology, Johns Hopkins University, Baltimore. "We understand there is redundancy built into the optic nerve axons, and we understand more about risk factors and calculating risk.

In reviewing the changes that have occurred in glaucoma, Dr. Sommer took a look back to the future, referring to a classic paper written by Grant and Burke in the early 1980s (Ophthalmology. 1982;89:991-998). Grant and Burke undertook a chart review-based, case-control study to identify why some patients lose vision from glaucoma. Although they were limited by their study methods and did not perform rigorous statistical analyses of their data, they made a number of important observations on risk factors, including IOP, race, age, and existing glaucomatous damage.

IOP issues

One of the key findings in the paper by Grant and Burke was that IOP was related to the risk and severity of subsequent primary open-angle glaucoma (POAG)-related optic nerve damage.

Dr. Sommer noted, however, there has been some persistence of erroneous thought that there is something "magical" about an IOP of 21 mm Hg for defining "normal" and risk of glaucomatous progression. He referred to this issue as "persistent perseveration of the pressure principle."

The concept of "normal" IOP derives from population-based studies undertaken in the United Kingdom and northern European countries to enhance the efficiency of glaucoma screening by identifying an IOP cut-off warranting further evaluation, he explained.

In calculating the mean IOP and standard deviation, those studies identified that persons with an IOP above 21 mm Hg represented the upper 2% to 3% of the population. Although the investigators recognized their analyses were meant to identify a subset of the population that was more likely to have glaucoma because of higher IOP, at some point, the data were translated into a definition of normal IOP.

Application of what is really an arbitrary IOP level as a threshold for screening also results in a population of patients with diagnosed glaucoma being overly enriched by those with an IOP at or above this level. In contrast, in the Baltimore Eye Survey and the Early Manifest Glaucoma trial, about half of the population with POAG had an IOP below 21 mm Hg.

"The 'normal range' is simply what is common, and the inference is that what is common is normal," Dr. Sommer said. "However, we realize that for most biologic phenomena that have a continuous measurement, there is no dichotomization of normal.

"In 1989, I wrote in an editorial that IOP and glaucoma are a continuum and it is wrong to dichotomize them and make up a disease based on a magic number," he continued. "That is the bottom line. IOP helps identify people at higher and lower risk of developing glaucomatous optic nerve atrophy, but it does not define the disease."