American Academy of Ophthalmology Web Site: www.aao.org
The Glaucoma Suspect Puzzle
Suspicious signs of glaucoma often precede frank disease, and many physicians offer some of these patients early treatment. But which patients,and how early?
If It’s Not Broken, Don’t Fix It. Or Do You?
That’s the question now revolving around the treatment of glaucoma suspects, those patients who have elevated intraocular pressure but no clinical signs of disease, or, alternatively, patients whose pressures are within range but who show other signs of concern.
One glaucoma specialist is said to pose the dilemma in automotive terms. “Are you better off,” he asks, “changing the oil in your car every 3,000 miles? Or could you wait until you were a quart low and then change it?” There seems to be no consensus—not among car owners, and not among ophthalmologists. In fact, almost everyone, from those who strongly favor a preventive approach to the most ardent “watchful waiter,” is likely to say: It depends.
The factors that come under consideration include everything from a patient’s fear of going blind to the physician’s desire to do no harm. They also include one’s attitude toward the expenditure of scarce economic resources on a potentially unnecessary treatment.
High pressures ahead.Ocular hypertension without detectable visual field loss and/or optic nerve damage loss may not be the most urgent condition confronting ophthalmologists, but it is one of the most common. An estimated 4.5 to 9.4 percent of the population above age 40 has the condition, with ocular hypertension defined as IOP greater than 21 mmHg. “You’re talking about millions of people,” said Michael A. Kass, MD, chairman of ophthalmology and visual sciences at Washington University in St. Louis and principal investigator for the Ocular Hypertension Treatment Study (OHTS). In part, the OHTS sparked this debate with the finding that the cumulative probability of developing primary open-angle glaucoma without treatment was 9.5 percent over five years, compared with only 4.4 percent in medicated patients. But OHTS also found that for every 20 patients given medication, only one patient benefited.
It’s the kind of finding that prompts glaucoma experts like Dale K. Heuer, MD, to ask: Is there a penalty for waiting to treat someone until they start to show early glaucoma? “That’s really the $64,000 question,” he said. Dr. Heuer is chairman of ophthalmology at the Medical College of Wisconsin in Milwaukee.
Suspects: the obvious and not-so-obvious. Some cases are easier to sort out than others. Consider, said Dr. Kass, a 95-year-old patient with a 30 percent chance of developing glaucoma in five years. “Why would you treat somebody to prevent the disease, if they’re not expected to live long enough to develop visual impairment?” On the other hand, a 16-year-old who could be exposed to elevated pressure for 70 years is more likely to be offered treatment.
The challenge for clinicians lies in the vast middle ground. Robert D. Fechtner, MD, director of the glaucoma division at New Jersey Medical School in Newark, posed a more challenging scenario in which a patient is diagnosed a suspect. “That means probably their pressure was high or their nerve fiber or optic nerve looked suspicious.” After a full evaluation, including measuring central corneal thickness, observing the optic nerve and nerve fiber layer and recording these structures with an imaging device such as HRT or GDx, and after testing visual field with white-on-white perimetry, and perhaps blue-on-yellow or frequency doubling testing, you’re stuck. “We’ve gotten to the end of that process and made the determination that this does not yet meet our threshold for glaucoma,” Dr. Fechtner said. “Whom are we going to treat? That’s where you’re going to get lots of opinions, because you’re not going to get lots of fact.”
Ultimately, said Dr. Kass, the decision is individualized for the patient. “It’s such a cliché. But it’s the best approach.”
Enter the Risk Calculators
Now there are tools to help individualize each treatment decision. Prior to the OHTS, the overall annual risk for conversion to glaucoma was estimated to be 0.5 percent to 1 percent. But the OHTS found that certain ocular hypertensives had a 5 to 7 percent annual risk of conversion to glaucoma. The case for selective treatment of ocular hypertension—for thinking of patients in terms of high, medium and low risk—was made. From there, the OHTS data were spun into a calculator that predicts the odds a patient will develop glaucoma.
In addition to making decisions based on IOP, physicians now can plug in a patient’s other risk variables, such as age, central corneal thickness, vertical cup-disc ratio and visual field pattern standard deviation, to derive that patient’s odds for converting to glaucoma.
“Clinicians need this tool,” said Steve L. Mansberger, MD, MPH, creator of an earlier risk calculator, also based on OHTS, that is posted on the Devers Eye Institute Web site. “Without it, they may overtreat or undertreat their patients.” Dr. Mansberger is an assistant scientist at Devers.
Robert N. Weinreb, MD, who with Felipe A. Medeiros, MD, developed the first validated risk calculator, noted that these predictive models will evolve over time. And the estimates of risk should not replace the judgment of a clinician. Dr. Weinreb is director of the Hamilton Glaucoma Center in San Diego and professor of ophthalmology at the University of California, San Diego .
A patient is not a population. Dr. Kass, who unveiled the newest risk calculator at the Academy’s Joint Meeting in November, agrees. While Dr. Kass called the calculator, “the best evidence-based medicine we have right now,” he said that the calculations were derived for large populations, not the single patient. “You will do pretty well if you have 1,000 people. But is that prediction right for an individual patient? That’s hard to know,” he said. “They’re not as accurate for individual patients. But they’re the best we have.” The goal in creating a calculator, he said, was “trying to point to which patients are at higher risk and need closer follow-up and, perhaps, preventive treatment. Conversely, other patients may be at low risk and could be followed less frequently without treatment.”
Still, it’s a useful tool for thinking of patients in terms of high, medium and low risk. “Considering the gaps in our knowledge, it makes good sense to me to reduce risk in the higher-risk population, if I can do so conveniently, effectively and without causing patients intolerable side effects,” Dr. Fechtner said.
The Risk is Calculated. Now What?
So what is an acceptable level of risk? “Everybody will come up with a different number,” said Dr. Fechtner, who sets his own high-risk threshold at 15 to 20 percent or more over five years if the patient has a reasonably long life expectancy. “This is where it becomes abstract.” One group of glaucoma experts agreed that treatment strategies should focus on moderate- and higher-risk patients, and set the treatment threshold at greater than 15 percent risk over five years.1
Ten percent is the treatment threshold for Steven M. Kymes, PhD, assistant professor of ophthalmology and visual sciences at Washington University in St. Louis and member of the OHTS study group. Dr. Kymes found that it is cost-effective on a population basis to treat people with IOP greater than 24 mmHg who have a 10 percent risk of developing glaucoma over five years. In fact, Dr. Kymes and colleagues took issue with the small, but vocal, “watchful waiting” group, calling that approach “unnecessarily conservative.”2
At the same time, they found that it is not cost-effective to treat everyone with ocular hypertension.
Interestingly, the newest calculator, which combines data from the OHTS and the European Glaucoma Prevention Study, does not suggest a risk level at which to initiate treatment. “We specifically tried to avoid telling people when to treat because there are so many exceptions,” Dr. Kass said
Physician, use thy judgment. Use the risk calculator, said Dr. Kass, but remember that other variables enter the equation, including age, health status and patient preference.
“It’s really the patient’s choice,” said Dr. Heuer. “Part of the decision-making process is doing an assessment of the patient’s treatment willingness and risk aversion. Some don’t want to use any medications. Other people may have a cousin by marriage who went blind from glaucoma and you tell them they have a two percent risk and they go on treatment. The best we can do is help patients make a decision.”
Dr. Mansberger regards a 20 percent risk over five years as a useful treatment threshold. At the same time, he rarely recommends treatment when the risk falls below 10 percent. “Between 10 percent and 20 percent, it’s a discussion between the patient and me,” he said. And what does he tell the patient? “If we don’t treat you, we’ll be able to pick up visual field loss and changes to your optic nerve before you notice it.” But like Dr. Heuer, he’ll treat the patient who says: “I’m scared of going blind, my mother went blind.”
Finally, consider the objective. “My goal is not to prevent the development of glaucoma,” said Dr. Fechtner, “but rather, to prevent disability from glaucoma.”
The Case Against Treatment
Alan L. Robin, MD, argues against presumptive treatment of glaucoma suspects. Dr. Robin, who is an associate professor of ophthalmology at Johns Hopkins University and who frequently brings a public health perspective to clinical questions, explained his position in a letter cosigned by four colleagues in Archives of Ophthalmology.3 “Before expensive and quality of life–limiting therapy is begun,” he writes, “we must ensure that a disease actually exists and that a glaucomatous process has begun.”
At issue is the assumption that by treating early, you’re preventing a devastating blinding condition. “The assumption,” said Dr. Robin, “is that glaucoma is like cancer. Pick it up early and prevent it from metastasizing and spreading and becoming overly aggressive. But glaucoma isn’t like cancer.”
Cancer, if caught at an early stage, he elaborated, is easier and cheaper to treat and more likely to yield a cure than if treated at a later stage, when the disease can be fatal. But glaucoma is different. “What are you really preventing? You’re not preventing anything.” His advice to clinicians: Follow your patients more carefully. Rather than treating moderate- to high-risk individuals, he suggests annual or semiannual examinations, with high-quality baseline documentation of the optic nerve and visual field. And be sure to schedule follow-up evaluations.
To wait is OK. Watchful waiting is “a reasonable approach,” Dr. Kass said. “We have never said treating ocular hypertension is the only way to manage these patients. We have said it’s a reasonable way to manage these patients.”
Dr. Robin noted that, ultimately, it’s the patient’s choice. After explaining the odds to the patient, the physician should ask: “Would you like me to treat or not?”
And if the patient refuses treatment? Dr. Mansberger would be comfortable with not treating, since most patients with glaucoma don’t notice visual field loss until they’ve lost a lot of visual field. “In most patients, the damage occurs over many years,” he said. “The thought is, perhaps you don’t need to start treating at the beginning. Maybe you can wait until they develop the earliest functional loss, then start treating them.”
Future Answers to Today’s Questions
Over time, more data and newer generations of risk calculators, incorporating everything from genetics to functional testing, could settle some stubborn questions. Does the rate of perimetric loss, for example, diminish if treated early? Does the disease progress more rapidly if therapy is begun after perimetric damage occurs? The OHTS II study now under way could answer the question of whether there is a penalty for delaying treatment, because the original observation participants are now taking pressure-lowering medication. Investigators will be watching to see if those who convert to glaucoma will progress at a faster rate than those who were on medication from the start. “That’s what most of us are interested in,” Dr. Heuer said. “What is the threshold of risk?”