Second eye at risk: predicting the development of advanced age-related macular degeneration

Abstract

Hypothesis 1: In people who develop a choroidal neovascular membrane (CNVM) secondary to Age-Related Macular Degeneration (AMD), presenting visual acuity varies. Eyes with worse visual acuities at presentation have worse visual acuity outcomes compared to eyes that present with better visual acuities despite treatment.

Aim 1: To determine and compare the visual acuity of first and second eyes diagnosed with CNVM secondary to AMD at presentation and through two years of follow-up to determine the long-term visual consequences of poor presenting visual acuity.

Methods 1: Retrospective review of clinical notes of patients presenting with newly diagnosed CNVM secondary to AMD in their first eye from the Royal Victorian Eye and Ear Hospital from January 2012 to April 2014. Visual acuity was recorded up to two years. Visual acuity in the second eye developing CNVM was recorded at diagnosis and for a further two years from the time of diagnosis.

Results 1: First eyes presented with a visual acuity of 6/52 and second eyes presented with a visual acuity of 6/21. Although this did not reach statistical significance, first eyes tended to have worse visual acuity during the 2 years of follow up. This suggests that there is room to identify and begin treatment earlier once CNVM has developed.

Hypothesis 2: Reduced microperimetric macular sensitivity, or a decrease in microperimetric macular sensitivity over time, is associated with an increased risk of developing CNVM in an eye with large drusen whose fellow eye has CNVM secondary to AMD.

Aim 2: To determine if reduced macular function, or a decrease in macular function over time, as determined by microperimetry, can predict the development of CNVM in eyes with large drusen whose fellow eye has CNVM secondary to AMD.

Methods 2: Participants with unilateral CNVM secondary to AMD in one eye and large drusen in the other eye (study eye, group 1) were recruited prospectively from the Royal Victorian Eye and Ear Hospital, East Melbourne, Victoria Australia between February and May 2014. All participants underwent microperimetry testing at baseline and 12 months study eye developed CNVM or GA.

Results 2: Reduced microperimetric sensitivity is associated with the development of CNVM or GA collectively (P=0.045). All eyes that developed CNVM or GA had a mean macular microperimetric sensitivity less than 25dB at baseline. This confirms our hypothesis, and provides further evidence that macular function, as tested by microperimetry, may be a useful biomarker for predicting disease progression.

Hypothesis 3: Microstructural changes, such as drusen, EZ disruption, HF, RPD and nGA, are independently associated with decreased retinal function, which partially explain the reduced microperimetric retinal sensitivity seen in eyes that progress to late AMD from Aim 2.

Aim 3: To determine if macular structure-function relationships explain the reduced microperimetric macular sensitivity in fellow eyes with large drusen that ultimately go on to develop late AMD of people with unilateral CNVM secondary to AMD in their other eye.

Methods 3: Group 1 from Aim 2 was used in Aim 3. Controls with bilateral intermediate AMD in both eyes (group 2) were recruited as a comparison group. All participants from groups 1 and 2 underwent multimodal imaging, including spectral domain optical coherence tomography, fundus autoflourescence, and fundus infrared reflectance.

Results 3: Mean microperimetric macular sensitivity is lower in cases compared to controls (P<0.001) at baseline and is independently associated with the development of CNVM or GA collectively (P=0.039), and GA alone (P=0.005) when taking into account microstructural changes.

Conclusion: People with CNVM secondary to AMD are presenting with poor vision in an Australian population, which limits their long-term visual outcome despite treatment. Within our study population, all eyes developing late AMD had a baseline mean microperimetric macular sensitivity less than 25dB. Furthermore, microperimetric macular sensitivity is associated with the development of late AMD whose fellow eye already has established CNVM independent of microstructural changes. This provides substantial evidence that microperimetric macular sensitivity, in conjunction with the quantification of structural changes, may prove to be a useful biomarker that could potentially allow for greater risk stratification of those likely to develop late AMD. As such, macular microperimetry has the potential to become a useful clinical tool with practical implications in terms of monitoring.