Monitoring for the Conversion of Dry to Wet Age-Related Macular Degeneration
Nalin J. Mehta, MD,
Colorado Retina Center, Denver, CO
WHAT IF I could show you a way to improve the visual prognosis of patients with exudative age-related macular degeneration (AMD) by 2-3 lines WITHOUT utilizing expensive pharmaceuticals or laser treatments? More importantly, what if this improvement was additive to present treatment modalities, effectively doubling the improvement in visual outcomes of
current treatments when compared to the natural history of this devastating disease?
Interested? Read on…
For the past several years, efforts have concentrated on pharmacologically decreasing the rate of vision loss in patients with exudative AMD. We have evolved from goals of damage control with treatments such as thermal laser, to expectations of visual stability and even improvement with intravitreal anti-VEGF pharmacotherapy. Our efforts, however, continue to be limited by the size, location and severity of the exudative process upon diagnosis. Studies show that with routine care, the average size of a subretinal neovascular membrane (SRNVM) lesion upon diagnosis is approximately 3300um1.
Physicians intuitively believe that treating disease in its earliest stages leads to better outcomes, and careful sub-group analysis of data derived from several of the more-recent AMD pharmacologic studies would support this observation. In the TAP and VIP studies, increased initial lesion size at the time treatment was initiated was inversely correlated with two-year visual acuity outcomes2. Sub-group analyses of both the ANCHOR and MARINA 1-year data also show an inverse relationship between initial lesion size and mean change in vision, with erosion of any visual improvement noted more so at 0.5 mg ranibizumab dosing than at 0.3mg3,4.
Subgroup analysis of data derived from the pegaptanib studies5 demonstrated prognostic trends similar to those noted in the verteporfrin studies. SRNVM lesions grew by at least 25% after 54 weeks of treatment in all dose groups (0.3 mg-3.0mg). This growth pattern, not surprisingly, correlated with a decrease in vision of approximately one line or more. Interestingly, the sham group showed similar trends (though the visual outcome was, of course, superior in the pegaptanib treatment groups).
Contrast the above results with visual outcome data from the ANCHOR and MARINA studies. Over the 12-24 months of treatment, visual acuities essentially stabilized after the first three or four months. Patients treated with ranibizumab had minimal observable lesion growth.
The Complications of Age-Related Macular Degeneration Trial (CAPT) may have failed in altering the natural course of non-neovascular AMD6, but the intense patient monitoring required during this study demonstrated the benefits of frequent monitoring of high-risk non-neovascular AMD. The percentage of patients with subfoveal SRNVM lesions upon presentation was 57%, with 66% of patients having a visual acuity of 20/40 or better in the eye with SRNVM (as opposed to 80% and 20%, respectively, with usual care7). It is, therefore, reasonable to conclude that earlier detection of SRNVM should result in lesions which are smaller upon initiation of treatment, and should result in an improved long-term visual prognosis with treatment.
Until recently, the best tool for the detection of conversion from dry to wet AMD between office visits has been the Amsler grid; however, our brains adapt to subtle metamorphopsia through phenomena such as cortical completion and crowding. Further, patients may have difficulty maintaining fixation, assuming they comply with Amsler grid usage. Color and pattern variations of the Amsler grid may help to increase its sensitivity, but are still susceptible to the brain’s adaptive abilities. As a result, a patient noticing symptoms or Amsler grid changes has often already suffered irreversible vision loss. The Amsler grid is, therefore, a rather crude and archaic tool, giving a false sense of security to both the doctor and the patient.
Snellen acuity (visual acuity) is also a poor indicator of conversion to exudative AMD because the best candidates for treatment would be early extrafoveal lesions that may not affect visual acuity until the lesion has significantly progressed toward the fovea.
Hyperacuity (Vernier acuity) is the ability to perceive minute differences in the relative spatial localization of two objects; it is the ability to detect misalignment. For example, Snellen acuity of 20/15 resolution subtends approximately 1 degree of arc, while Vernier resolution is as fine as two seconds of arc: this translates into the ability to perceive misalignment equal to the width of a pencil at 300 meters! The measurement of hyperacuity, furthermore, is not significantly affected by cataract, media opacity or moderate loss of visual acuity.
The Amsler grid is only a crude detector of hyperacuity. New technologies, however, such as the Preferential Hyperacuity Perimeter (PHP) can help to detect these lesions when they are less than 1,000 microns in diameter, giving doctors a head start in initiating meaningful treatment of exudative AMD. The PHP achieves this by projecting individual lines in an automated fashion which may or may not contain hyperacuity disturbance. The patient then preferentially marks either the computer-generated distortion, or the patient’s own distortion, depending on which is more prominent. The instrument then quantifies the intensity of any true, patient-generated distortion and maps this along with relative intensity and statistical significance by comparing the same to an extensive normative database. With 82% sensitivity and 88% specificity, the PHP compares favorably with automated visual field testing8. It is important to note, however, that significant geographic atrophy and RPE detachment can cause false positive PHP results9.
By projecting individual target lines each for a finite period of time, and redirecting fixation after each stimulus, the PHP is able to bypass the brain’s corrective abilities, allowing detection of hyperacuity disturbance from exudative AMD lesions less than 1000 microns in diameter, or 70% smaller than detection with the conventional standard of care.
If we extrapolate from verteporfrin and 0.5 mg ranibizumab data, this would translate into a roughly 2-3 line improvement of visual acuity after 1-2 years of treatment. More importantly, this improvement would be in addition to that derived from the pharmacologic treatment itself! Based on earlier discussion, moreover, this benefit would be further enhanced by the increased likelihood that the CNV lesion would be extrafoveal and associated with relatively good visual acuity at the time of treatment, therefore increasing the likelihood of favorable long-term visual prognosis.
By increasing the frequency of monitoring patients, and utilizing newer technology such as the Preferential Hyperacuity Perimetry, significant improvements in the visual outcomes of our AMD patients can be effected by relatively simple means; the logistics of patient management and referral, however, are equally important. Ideally, PHP testing would be of greatest benefit if performed quarterly by the general eye care practitioner. Rather than using PHP as a blanket screening tool, which risks over-utilization, PHP is best suited for patients with high risk of progressing to exudative AMD; the best category of patients would coincide with the AREDS category of moderate non-neovascular AMD, those with numerous medium-sized drusen, or one or more large drusen, in one or both eyes10, or those patients with exudative AMD in one eye (i.e., testing the opposite eye, which is at higher risk for conversion with each passing year). Although ideally suited to primary eye care providers, PHP testing could be utilized by the retinal community as well to monitor patients between clinical examinations; indeed, usage can be tailored to individual practice and referral patterns.
1 Olsen TW, Feng X, Kasper TJ, et al. Fluorescein angiographic lesion
type frequency in neovascular age-related macular degeneration.
Opthalmology 2004; 111:250-5.
TAP and VIP Report 1, American Journal of Ophthalmology, September 2003.
ANCHOR Study Group, Subgroup Analysis, Ranibizumab in Neovascular Age-Related Macular Degeneration.
MARINA Study Group, Subgroup analysis of the MARINA study of ranibizumab in neovascular age-related macular degeneration, Ophthalmology 2007; 114:246-252.
VEGF Inhibition Study in Ocular Neovascularization Clinical Trial Group. Pegaptanib for neovascular age-related macular degeneration. New England Journal of Medicine. 351 (27): 2805, Table 4. December 2004.
6 Complications of age-related macular degeneration prevention trial (CAPT). Ophthalmology 113:1974-1986. 2006.
Alexander, J. Presented at ARVO. May 2004.
8 Alster, Y, et al. Preferential hyperacuity perimeter (Preview PHP™) for detecting choroidal neovascularization study
Mehta, NJ, Alster, Y. Cross-sensitivity analysis among a population of AMD patients monitored with the PreView PHP™. Presented at ARVO 2006.
The AREDS system for classifying age-related macular degeneration from stereoscopic fundus photographs. AREDS Report No. 6. Archives of Ophthalmology, 132: 668-681. October 2001.
Ingrid U. Scott,
MD, MPH, Editor
Professor of Ophthalmology and
Public Health Sciences,
Penn State College of Medicine