Glaucoma: Practicing Evidence-Based Medicine
Data from recent landmark clinical trials make it easier to determine the most effective treatment.
BY ROBERT NOECKER, M.D.

One of the things that makes glaucoma a dangerous disease is that it's asymptomatic. Patients usually don't know that they have it until they come in for an exam. Then we have to do an intervention, which can cause side-effects and make their eye red or cause it to sting, leaving some patients quite unhappy.

As a result, when a patient's pressure isn't too high, it's easy to treat conservatively. In fact, we may settle for therapies that are inadequate just because we don't want to upset the patient.

Today, data from five landmark studies provide solid information about the effectiveness of conservative vs. aggressive treatment. The overriding theme of their results is that patients who are treated aggressively do better, no matter what type of glaucoma they have, or what stage they're at.

Reviewing the Data

Here are a few highlights of the data produced by these five landmark trials:

The Ocular Hypertension Treatment Study (OHTS). OHTS followed more than 1,600 patients with high IOP but no definitive visual field defect or optic nerve changes. One subgroup had their pressure lowered medically by 20%; a second subgroup was observed but not treated.

At 5 years, the risk of developing a visual field defect or optic nerve change dropped from 9.5% to 4.4% among the treated patients. (OHTS also showed that patients who started with higher pressures tended to do worse over time.)

The Advanced Glaucoma Intervention Study (AGIS). AGIS followed more than 750 patients with advancing glaucoma to determine whether ALT or trabeculectomy was the better option when patients were not responding sufficiently to medical intervention. Blacks did better if ALT was used as the next resort; Caucasians did better with trabeculectomy as the next option.

Perhaps even more important, a second analysis compared patients who had IOPs below 18 mmHg on every follow-up visit with those whose pressure was below 18 mmHg at fewer than half of their visits. The first group had close to zero mean change in visual field defect score; the second group's visual field defect score worsened by an estimated 0.63 units.

This suggests that it may be important to not only keep IOP low, but to maintain it consistently.

The Normal Tension Glaucoma Study (NTGS). NTGS was designed to find out whether lowering IOP in patients who've never had high IOPs makes a difference. (The highest IOP in the study was 24 mmHg.) The data showed that a 30% pressure reduction, regardless of method used to achieve it, reduced the average risk of progression from 36% to 12% in the 5-year follow-up period.

The Collaborative Initial Glaucoma Treatment Study (CIGTS). CIGTS is an ongoing study comparing medical IOP lowering with surgical IOP lowering as initial therapy. On average, the patients in CIGTS have had their IOPs lowered more than in any other study -- from 38% to 45%. So far the data show that surgery produces a greater lowering of pressure than treatment with drugs, but both groups have done very well; few subjects have had disease progression.

The Early Manifest Glaucoma Trial (EMGT). This study looked at patients with early visual field defects; the researchers either treated patients, lowering IOP about 25%, or just observed them. During the 5-year follow-up period 45% of the treated group progressed; 62% of the control group progressed. (Progression occurred significantly later in treated patients.)

Equally important, another analysis of the data found that every millimeter of IOP lowering was accompanied by a 10% decrease in progression during the follow-up period.

Again, the message is clear: Getting IOP lower is a good thing to do -- even in patients in the early stage of disease.

How Low Do You Go?

Given this data, I choose a target pressure based on the patient's status:

► For an ocular hypertensive -- someone with an elevated IOP (in the vicinity of 30 mmHg) but no visual field defect and no definitive optic nerve damage -- the OHTS study suggests that a 20% reduction from baseline, with a final pressure lower than 24 mmHg, should reduce the patient's risk of progression by about half. I may adjust this number downward closer to 20 mmHg, depending on other risk factors.

► If a patient has an early visual field defect, a little nasal step or a thin spot on their nerve fiber layer or optic nerve, I typically accept an IOP in the upper teens, as long as there's been at least a 30% reduction in pressure.

► If the patient has a moderate visual field defect, or noticeable optic nerve thinning in one area, my goal is
15 mmHg (i.e., the mid-teens).

► If the patient is very damaged, with extensive visual field loss, I aim for an IOP of 12 mmHg (i.e., the low teens). In this situation you can't afford to be wrong and risk the patient's disease progressing, so I drive the pressure down as low as I can.

In general, I think that lowering IOP by 30% to 35% for starters is a reasonable thing to do.

Choosing the Right Option

Once you've decided how low you need to go, you have to decide how you're going to get there. To lower pressure, you have three options: medical, laser and surgical.

Medication. A generally accepted practice in the United States is that if a patient with glaucoma presents with a pressure in the vicinity of 30 mmHg, the first choice for treatment should be medication. Prostaglandins or pros-tamides such as bimataprost, travaprost or latanoprost are my first choice; they lower pressure about 30% on average, and they're very safe systemically. Beta blockers lower pressure by about 25% when used as monotherapy; alpha2 agonists lower IOP by a similar amount. The topical carbonic anhydrase inhibitors (CAIs) -- dorzolamide and brinzolamide -- typically only lower pressure 15% to 20%. They haven't proven effective as primary therapy.

If a prostaglandin doesn't lower IOP sufficiently by itself, you have to decide what medication to add:

► Studies have shown that adding beta blockers to prostaglandins only lowers IOP an additional 10% to 15%.

► Alpha agonists have shown good additivity with latanoprost, on the order of 4 mmHg -- better than what beta blockers often do. (Lee DA et al., J Glaucoma. 2001; 10:220-226, and Walters TR., Abstract presented at Ameri-can Academy of Ophthalmology Annual Meeting, Novem-ber 2001, New Orleans.) Recent experience has also shown them to be effective when added to bimataprost (Netland PA et al., Adv Ther. 2003; in press) and travaprost.

► Adding a combination drug is another option. At the moment, Cosopt, which combines dorzolamide and timolol, is the only one available in the United States. Its primary advantage is that it consolidates the number of drops the patient is using.

Laser treatment. If medications don't lower pressure sufficiently, or the patient appears to be progressing, I'd consider performing laser trabeculoplasty, which typically lowers IOP by about 20% -- a 3- to 6-mm reduction. Typically, the higher the pressure is, the more lowering you get.

These days, I only perform SLT (selective laser trabeculoplasty), not ALT (argon laser trabeculoplasty). Even though SLT doesn't work for about 20% of patients, studies have shown that it doesn't cause structural damage. ALT burns away some of the trabecular meshwork, which could actually make some patients worse.

SLT might be considered a reasonable first-line therapy for an ocular hypertensive patient, where the studies indicate a 20% reduction could be acceptable. It can also be an add-on option if a patient is unable to be compliant with more than a single drug therapy.

Surgery. Patients who have serious damage probably need surgery, although you have to weigh the risks and benefits. (Sometimes you get lucky; drug treatment may be sufficient to achieve the low IOP these patients need.) If the patient has a long time to live, surgery may be the way to get his pressure down into the low teens for the longest period of time.

A trabeculectomy with antimetabolites is still the primary surgical option; this may get a patient down to single digits if he doesn't have a lot of scarring from previous surgery. Endoscopic cyclophotocoagulation (ECP) is another alternative, but it's more variable. You don't always get the very low pressures that you would with a trabeculectomy.

Reach Your Target Pressure

The bottom line? If you're going to treat glaucoma, treat it. Make the patient understand that he has a serious problem -- but also inform him that people who are treated aggressively are unlikely to get worse. Then set a target IOP and achieve it, using whatever therapy makes the most sense.

Let the studies guide you; practice evidence-based medicine rather than just treating things anecdotally. Be realistic about what you expect to achieve with each therapy, and use it at the right place in the algorithm. And be sure that the patient has appropriate expectations regarding whatever treatment you choose.

Most important, don't stop until you reach the target pressure.

Dr. Noecker is associate professor of ophthalmology at the University of Pittsburgh. He is also vice chairman and director of the glaucoma service.