Power and clinical utility of mesopic microperimetry analysis strategies in age-related macular degeneration

Purpose: 

This study evaluates whether mesopic microperimetry (MMP) provides a more robust measure of retinal function compared to visual acuity (VA) in age-related macular degeneration (AMD) clinical trials, with a focus on optimal analysis strategies. 

Method: 

Fellow-eyes of unilateral neovascular AMD were prospectively studied. Presenting VA was measured. The Macular Integrity Assessment Microperimeter (MAIA) was used with a 4 to 2 staircase strategy with a 10° diameter grid containing 37 loci. Three analysis strategies were calculated: the mean of 37 sensitivity thresholds (MS), the per cent reduced threshold (PRT), and the log-transformed candela mean (MS cd log). Sample size requirements were calculated for 12- and 24-month follow-ups using a paired one-sided T-test (α = 0.05, power = 0.80).

Results:

N = 123 were analysed (82 (65.5%) females; mean age 74.2 (7.8) years). Ranked high to low, the required sample size at 12 months was: MS (n = 51), MS cd log (n = 52), PRT (n = 139), and VA (n = 203). Similar trends were seen at 24 months, with MS requiring the smallest sample size (n = 85) and VA the largest (n = 1673). 

Conclusion: 

All MMP analysis strategies outperformed VA, and MS required the least number of patients to show significant changes. This trend was consistent for both 12 and 24 months. These findings provide strong statistical arguments for the use of MMP in longitudinal within-subjects clinical trials and suggest that averaging decibels is optimal.