[Treatment of exudative age-related macular degeneration].

I PROPHYLACTIC TREATMENT

We followed 75 eyes contralateral to eyes with exudative age-related macular degeneration (AMD), using indocyanine green angiography (IA), for more than one year. Hyperfluorescent areas in the late phase of IA were seen in 19 eyes at the initial examination, and in 25 eyes during follow-up. Exudative AMD developed in 9 of the 25 eyes. Using timetable analysis, we estimated that 11% of these 27 eyes developed AMD within one year and 55% within three years. The hyperfluorescent areas seen on IA appeared to be latent choroidal neovascularization (CNV) under the retinal pigment epithelium. We propose that photocoagulation aimed at hyperfluorescent areas should be considered in such cases. We performed prophylactic laser photocoagulation in 21 eyes, which were then followed up for at least six months. These eyes all had 10 or more serous drusen within 1,500 microns of the fovea and did not show hyperfluorescence, suggesting latent CNV in the late phase of IA. The majority or a small fraction of the serous drusen disappeared in 48% and 18% of the 21 eyes, respectively. CNV appeared adjacent to the laser scar in one eye (5%). Judging from these results, it is important to establish a method of definitively abolishing drusen and preventing the development of CNV. II TREATMENT OF CNV: Of 229 eyes which showed occult CNV in fluorescein angiography (FA), 124 eyes (54%) showed classic CNV outside the fovea on IA. One hundred and two of the 124 eyes (45%) underwent laser photocoagulation. We evaluated indocyanine green guided laser photocoagulation of extrafoveal CNV in 139 eyes. The success rate was 81% at 3 months after laser photocoagulation. This was estimated using timetable analyses to have decreased to 78% at one year and 71% at three years. Eighty percent of successfully treated eyes showed maintained or improved visual acuity. These results did not differ significantly from those obtained with laser photocoagulation based on FA findings. When classic CNV is not detected on FA, IA is indicated and if classic CNV is detected outside the fovea, photocoagulation should be performed aggressively. We compared the histological findings of removed subfoveal CNV with the IA findings prior to removal. IA findings of CNV before removal were divided into four types. Types I and II showed numerous vascular lumina in von Willebrand factor staining. Prominent proliferation in types I and II was demonstrated with Ki-67. Vascular endothelial growth factor (VEGF) showed strong staining in types I, II and III. Therefore, CNV in eyes with types I, II and III, and especially type I, requires urgent treatment. We compared visual outcomes one year after treatment between 52 eyes that underwent removal and 56 eyes that underwent laser photocoagulation of subfoveal CNV. When CNV size was one disc diameter or less, mean visual acuity, the percentage of eyes with improved vision and with a visual acuity of 0.1 or more, was significantly greater in operated eyes than in coagulated eyes. Removal is therefore superior to photocoagulation in subfoveal CNV of one disc diameter or less in size. We evaluated preoperative factors influencing the best corrected visual acuity at least six months after removal of subfoveal CNV in 67 eyes. Factors influencing the best corrected visual acuity were IA findings, feeder vessel ingrowth sites and CNV sizes. Good indications for removal included types I, II and III IA findings, extra foveal location feeder vessel ingrowth sites, and small CNV. III LOW VISION CARE: We performed reading tests using the MN read J chart for 90 eyes with the scar stage of CNV. Reading speed profiles of the 90 eyes were classified into two groups: 48 eyes showing a plateau with a relatively constant reading rate at the maximum level; and 42 eyes showing steadily increasing reading rate without a plateau. Eyes without a plateau had worse visual acuity, worse reading acuity, and slower reading speed. We evaluated newspaper reading performance by comparing conventional versus new methods based on the critical print size of the MN read J chart. Magnification power based on the critical print size were similar to those finally chosen as the best aid by the patients. On the other hand, magnification power was lower when conventional methods were used. The critical print size of the MN Read J chart appears to be valuable for determining low vision aids which facilitate reading by AMD patients.