How much glaucoma damage is pressure-dependent?

Paul Palmberg

Bascom Palmer Eye Institute, University of Miami School of Medicine, Miami, FL, USA

The dose-response relationship of intraocular pressure and visual field progression

How effective is pressure lowering in preventing further visual field loss in openangle glaucoma? How low does the pressure have to be to achieve the maximum benefit? Finally, definitive data have arrived and the answers to these questions will have a major impact on our theoretical understanding of the disease and treatment decisions.

The investigators in the Advanced Glaucoma Intervention Study (AGIS) have reported what they found in analyzing the relationship between intraocular pressure (IOP) and visual field loss in 789 eyes followed for six to 11 years.1 The patients in that National Eye Institute (NEI) sponsored clinical trial had, on average, moderate visual field loss, and had failed medical therapy. They were randomized to either initial argon laser trabeculoplasty or initial trabeculectomy without an antimetabolite. Subsequent management was dictated by a detailed protocol. The initial IOP averaged 25 mmHg, and was reduced at one year to about 16 in the surgery group and 17 in the laser group. At seven years’ follow-up, the investigators reported that, in white patients, 34% treated initially with laser had suffered field progression versus 20% treated initially with surgery, while, in black patients, 25% of laser and 24% of surgery patients progressed.2 But those results did not tell the whole story.

Now the investigators have provided an additional analysis to delve into the dose-response relationship between IOP and visual field progression in the total group of patients. Of several complex analyses presented in the paper, one is fundamental. In it, the patients were divided, for analysis purposes, into four groups, based upon the percentage of time they achieved an IOP < 18 mmHg. The four groups achieved the goal in either 100%, 75-99%, 50-74%, or less than 50% of visits, respectively. The corresponding mean IOP values of the groups were 12.3, 14.7, 16.9 and 20.2 mmHg, respectively.

As can be seen in Figure 1, the lowest pressure group had a flat curve throughout follow-up, meaning that, on average, glaucoma damage was halted. Each suc-

Address for correspondence: Paul Palmberg, MD, PhD, Bascom Palmer Eye Institute, University of Miami School of Medicine, P.O. Box 016880, Miami, FL 33101, USA

Glaucoma in the New Millennium, pp. 21–27

Proceedings of the 50th Annual Symposium of the New Orleans Academy of Ophthalmology, New Orleans, LA, USA, April 6-8, 2001

edited by Jonathan Nussdorf

© 2003 Kugler Publications, The Hague, The Netherlands

Fig. 1. The relationship between IOP and visual field loss in the Advanced Glaucoma Intervention Study.

cessively higher pressure group had a correspondingly greater mean progression of field loss. The visual field defect scores were measured in a scoring system particular to the study, AGIS units.3 The mean baseline field defect was 8.4 AGIS units, which corresponded to a mean deviation of -10.5 dB in the more generally understood Humphrey visual field scoring.

The comparison of mean change in visual field defect scores for groups of patients in this analysis is far more powerful statistically than would be a comparison of the percentage of patients found to progress by any specific criteria. Any ‘noise’ (variability not due to true progression or true improvement) is cancelled. The AGIS investigators reported that, in the lowest pressure group, 15% of patients would have met their formal criterion for progression, but also, 15% would have moved in the direction of improvement by the same degree. How much was due to noise and how much to a true getting worse or getting better at these pressures? Perhaps a trend analysis, such as that with the Progressor Program developed at Moorfields Eye Hospital, would identify progression in individual cases more reliably, or a random walk through the data in the low pressure group could estimate inherent noise in the testing procedure. The data are currently undergoing further analysis in order to clarify the signal/noise components.

What are the implications and limitations of these data? Firstly, they are specific for patients with primary open-angle glaucoma with moderate damage, and do not apply to persons with ocular hypertension only, or to persons with normaltension glaucoma (NTG).

Fig. 2. The relationship between IOP and the risk of progressive visual field loss at five years.

Patients with ocular hypertension clearly have less sensitivity to IOP, and their dose-response relationship between IOP and risk of visual field loss must be shifted far to the right. We should learn about that relationship when the results of the Ocular Hypertension Treatment Trial are eventually reported.

On the other hand, persons with NTG have a greater sensitivity to IOP than is seen in the AGIS data. In the Collaborative Normal Tension Glaucoma Trial, patients with documented field progression or with split fixation were randomized to observation or to treatment aimed to reduce the IOP by 30%. In those in whom a 30% IOP reduction was actually achieved, and after correcting for the effects of surgically induced cataracts upon the fields, the investigators reported that only 20% of treated patients suffered field progression at five years, at an average IOP of 11 mmHg, while 60% of controls had progressed at five years, at an average IOP of 16 mmHg.4 Those results correspond to a shift to the left from the AGIS data.

The AGIS data correspond well with the dose-response relationship between IOP and risk of visual field loss in our meta-analysis of other surgery outcome studies. That analysis was first distributed at an American Academy of Ophthalmology regional update meeting in 1988,5 and was then included in the first Preferred Practice Pattern for Primary Open-Angle Glaucoma issued by the American Academy of Ophthalmology in 1989.6 In that guide, I introduced the term ‘target pressure’7 as a way of thinking about glaucoma management in the light of our emerging understanding of the dose-response relationship between IOP and risk of further field loss. The meta-analysis was updated this year8 (Fig. 2) to include our studies of patients undergoing primary filtering surgery with either 5-FU (Table

1)or mitomycin C (Table 2), the NTG study, and earlier results of the AGIS study. In our primary 5-FU and MMC filtering procedure study, we took a group quite

similar to the AGIS patients (failure of medical therapy, mean IOP 26 mmHg, baseline Humphrey MD of –14 dB) and reduced the IOP throughout the followup to 10-11 mmHg, with resultant flat curves of both the mean of the mean deviation and the mean of the pattern standard deviation over time, in agreement with the result for the best-controlled IOP group in AGIS. In our data, 17% of patients ‘improved’ and 15% ‘worsened’ from a regrettably single baseline field by 3 dB

IOP: mean intraocular pressure; MD: mean of the Humphrey visual field mean deviation; CPSD: corrected pattern standard deviation; n: sample size

Table 2. IOP and visual field results of our study of primary filtering surgery with mitomycin C

IOP: mean intraocular pressure; MD: mean of the Humphrey visual field mean deviation; CPSD: corrected pattern standard deviation; n: sample size

in mean deviation and/or pattern standard deviation, a result seen at one year of follow-up and which remained unchanged for six years. But inspection of sequential fields indicated that only about 5% had an improvement or worsening that exceeded variability. Thus, we conclude that 95% of glaucoma progression in primary open-angle glaucoma is pressure-dependent and avoidable. The one caveat is that, in both AGIS and in our study, only patients who could be followed with Humphrey fields were included, thereby omitting patients with very advanced visual field damage.

Patients with milder glaucoma damage, as might be seen at diagnosis, may not be as sensitive to pressure as either the AGIS patients or ours. Such patients were randomized in the Comparison of Initial Glaucoma Treatments Study (CIGTS, NEI) to initial medical treatment or initial surgery, and are currently being studied.9 We should await the results of that study before abandoning the usual practice of trying medical therapy and/or laser trabeculoplasty before using surgery, as the dose-response curve may be shifted to the right for such patients in comparison to the AGIS cohort.

Implications of the data for our understanding of the mechanism of damage in primary open-angle glaucoma

It is surprising that as much as 95% of glaucoma damage has turned out to be pressure-dependent. There was much evidence to indicate that, in glaucoma, IOP might only play a small role. In epidemiological investigations, such as the Baltimore Eye Survey,10 a receiver-operator curve using pressure alone to separate those in the population with glaucoma damage from those without, would indicate that only about 35% of glaucoma damage was due to pressure. Even worse, careful retrospective studies of patients undergoing medical therapy for glaucoma have often found virtually no correlation between IOP and the risk of visual field progression.11,12

How can we reconcile these disparate results? Firstly, we have to distinguish between that part of glaucoma damage which is due to elevated IOP (35%) and that part which is due to abnormal sensitivity to pressure, for which we can apparently compensate by achieving low-normal pressures. Further analysis of our patients operated upon with either 5-FU or mitomycin C indicates that there was no detectable dose-response relationship in the 5-11 mmHg range. Thus, getting down to about 10 mmHg gives the maximum benefit for primary open-angle glaucoma patients. We propose that this result is consistent with the hypothesis of Anderson13 that glaucoma damage is often due to faulty vascular autoregulation in the optic nerve head, occurring either chronically or episodically. Once the IOP is reduced to the level present in the ocular veins, about 10 mmHg, further lowering would not be expected to improve ocular perfusion. However, it would not have been possible to predict the degree to which IOP lowering could compensate for deficient autoregulation.

The failure to show a strong correlation of IOP to visual field progression in retrospective studies of medical therapy has been disconcerting, and has led some to doubt that IOP control could play much of a role in slowing field loss in patients with IOPs in the normal range. Now that the AGIS results have shown a strong correlation of IOP to risk of field progression, we can wonder what was wrong with the retrospective study approach to the problem. A likely possibility is that the clinicians who were guiding the therapy of the patients in these studies were treating patients more aggressively who had had greater or more rapid damage in the past, resulting in a levelling of risk.

That would be analogous to a handicap golf tournament, in which players are assigned an adjustment in their score based upon past performance, with the intent of equalizing the likelihood of anyone winning. If we were to look at the results of such handicap tournaments, we would conclude that neither the ability of the players, as reflected in the raw scores, nor the handicap assigned, would correlate with the adjusted scores. Similarly, in the retrospective studies, the mean IOP did not correlate with field progression. In addition, in the retrospective studies, patients with advanced damage and elevated pressures were operated upon, thus eliminating the top of a potential dose-response curve between IOP and risk of progression.

26 P. Palmberg

Implications for practice

In patients with at least moderate visual field loss, in whom medical therapy has failed, I believe a comparison of the AGIS results for surgery without antimetabolites to results for surgery with either 5-FU or MMC favors antimetabolite use. Low normal pressures can be achieved with the use of antimetabolites, and this is not likely to occur without them. The consequence is a marked reduction in the risk of visual field progression from 20% to 5%, and in blindness from 7% to 2%. However, precautions should be taken to reduce the potential complications of antimetabolite use.14

When treating patients medically, the AGIS results suggest that, in patients with moderate to severe glaucoma damage, we should strive to achieve pressures in the low-normal range. This is likely to require the use of multiple medications, and to bring the use of combination therapies into favor. The era in which it was reasonable to place a glaucoma patient with moderate glaucoma damage on a beta-blocker alone, and to observe the course at a pressure in the upper teens (just 25% lower than the initial pressure), should now be over. Except in emergencies, one medication should be added at a time, with one-eye trials being carried out to prove the effectiveness in pressure reduction, but medications should continue to be added until a low normal pressure has been achieved, if feasible, as well as at least a 3050% lowering of IOP. Supplemental laser trabeculoplasty will often be needed to achieve such results.

These recommendations are based on results for groups of patients, since we cannot currently determine the true target pressure needed to prevent damage in a particular optic nerve. It would be highly desirable to have an ‘axon screamometer’ that could detect distress in an optic nerve, allowing us to titrate the pressure down just far enough, thus enabling us to maximize the risk/benefit of therapy.

However, our enthusiasm for low normal target pressures should be balanced by knowledge of the dose-response curve relating IOP to risk of future damage, and especially by the past history of the patient at hand. The risk of progression in patients with primary open-angle glaucoma and at least moderate damage is about 5% at an IOP of 11 mmHg, 20% at 15 mmHg, 35% at 17 mmHg, and 50% at 20 mmHg. If the next step is surgery, which has risks of cataract progression and infection, the surgeon may reasonably choose to observe carefully at a pressure that is not optimal. If a patient’s visual fields have been stable for five or more years at the current level of IOP, there is even better reason to observe.

Just as was shown in the Diabetes Control and Complications Trial for insulindependent diabetes,15 it appears that attempted ‘tight control’ is beneficial in glaucoma as well. However, we must always weigh the benefits and risks of medical and surgical therapy when choosing how to treat an individual patient.

References

1.The AGIS Investigators. The Advanced Glaucoma Intervention Study (AGIS). 7: The relationship between control of intraocular pressure and visual field deterioration. Am J Ophthalmol 130:429-440, 2000

2.The AGIS Investigators. The Advanced Glaucoma Intervention Study (AGIS). 4: Comparison of treatment outcomes within race: seven year results. Ophthalmology 105:1046-1064, 1998

3.The AGIS Investigators. The Advanced Glaucoma Intervention Study (AGIS). 2: Visual field test scoring and reliability. Ophthalmology 101:1445-1455, 1994

4.Collaborative Normal-Tension Glaucoma Study Group: Comparison of glaucomatous progression between untreated patients with normal-tension glaucoma and patients with therapeutically reduced intraocular pressures. Am J Ophthalmol 126:487-497, 1998

5.Palmberg P: The rationale and effectiveness of glaucoma therapy. Distributed at the American Academy of Ophthalmology, Regional Update Meeting, Miami, FL, December 1998

6.Preferred Practice Pattern for Primary Open-Angle Glaucoma, 1989 Edn. San Francisco, CA: American Academy of Ophthalmology 1989

7.Palmberg P: Clinical controversies: target pressures—what are they? In: Leader BJ, Calkwood JC (eds) Peril to the Nerve: Glaucoma and Clinical Neuro-Ophthalmology, Proceedings of the New Orleans Academy of Ophthalmology 1996, Vol 45, pp 87-95. The Hague: Kugler Publ 1998

8.Palmberg P: Target pressure. In: Alm A (ed) The Gullstrand Foundation Meeting. April 1, 2001. CD-ROM. Uppsala: Uppsala University 2001

9.Sommer A, Teilsch JM, Katz J, Quigley HA, Gottsch JD, Javitt J, Singh K: Relationship between intraocular pressure and primary open angle glaucoma among white and black Americans: the Baltimore Eye Survey. Arch Ophthalmol 109:1090-1095, 1991

10.Chauhan BC, Drance SM: The relationship between intraocular pressure and visual field progression in glaucoma. Graefe’s Arch Clin Exp Ophthalmol 230:521-526, 1992

11.Martinez-Bello CM, Chauhan BC, Nicolela MT, McCormick TA, LeBlanc RP: Intraocular pressure and progression of glaucomatous visual field loss. Am J Ophthalmol 129:302-308, 2000

12.Musch DC, Lichter PR, Guire KE, Standardi CL, the CIGTS Study Group: The comparison of initial glaucoma treatments study: study design, methods and baseline characteristics of enrolled patients. Ophthalmology 106:653-662, 1999

13.Anderson DA: Introductory comments on blood flow autoregulation in the optic nerve head and vascular risk factors in glaucoma. Surv Ophthalmol 43(Suppl 1):S5-9, 1999

14.Palmberg P: Surgery for complications of filtering surgery. In: Albert D (ed) Ophthalmic Surgery: Principles and Techniques, pp 476-478. Molde, MA: Blackwell Scientific 1999

15.The Diabetes Control and Complications Trial Research Group: Progression of retinopathy with intensive versus conventional treatment in the Diabetes Control and Complications Trial. Ophthalmology 102:647-661, 1995