gms | German Medical Science

Purpose: Subjects with intermediate AMD (iAMD) have been shown to have difficulty in low luminance and low contrast conditions. To provide further evidence for visual function deficits in iAMD and to evaluate how different aspects of visual function characterized by a wide variety of clinical outcome measures changes longitudinally with disease progression or conversion to late AMD, we here report functional data, obtained in the MACUSTAR clinical study during ≥3 years.

Method: The MACUSTAR clinical study is a multi-center study at 20 European sites across seven countries, funded by the European Innovate Medicine Initiative. The study included a total of 718 individuals, mostly female (66%), 585 with iAMD, 34 early AMD, 43 late AMD and 56 with no AMD. The mean (SD) age at baseline was 71.9±7.0 years. 46 (8%) participants with iAMD had late AMD in the non-study eye. Visual function assessments performed in the study were: best-corrected visual acuity (BCVA); low luminance visual acuity (LLVA); Moorfields acuity chart (MA); Pelli-Robson contrast sensitivity (CS); International Reading Speed Test (IReST); mesopic and scotopic fundus-controlled perimetry (mAT and sAT; Macular Integrity Assessment, iCare, Finland) and dark adaptometry (AdaptDx, Lumithera, USA). Quality of the visual function data was evaluated centrally every 6 months till the very recent data cut at 3.5-years median follow-up. Baseline analyses comparing visual function outcomes across different disease stages were performed by linear regression models adjusted for age, sex, and phakic status with multiple testing correction. To evaluate the proportion of participants with abnormal visual function for each assessment, one-sided normal reference limits (95th or 5th percentile), based on the No AMD group, were used to establish thresholds and dichotomize the iAMD group.

Results: At baseline, all VF measures except IReST were significantly worse in those with iAMD compared to those with no AMD, and in general showed more prominent impairment, when assessed in subjects with late AMD. However, absolute differences between iAMD and No AMD were smaller than the limits of repeatability and therefore cannot be deemed clinically significant. Using the established thresholds for separating normal from abnormal function, we discovered distinct functionally abnormal iAMD subgroups. Their visual function breaches the normal reference limits for the different assessments with proportion relative to all iAMD participants that range from around 13% for IReST to over 30% for CS and Rod Intercept Time (RIT). Overall, about 71% of iAMD subjects showed at least one functional deficit outside the normal reference limits. The prognostic value of such functional baseline deficiencies for conversion to late stages of AMD will be shared. Preliminary longitudinal analyses relative to baseline for the whole iAMD group revealed changes for functional tests, including RIT, LLVA and MP over a period of 3.5 years.

Conclusion: The value of the observed heterogeneous baseline functional deficits in over 70% of iAMD patients as inclusion criteria for enrichment of trial populations and for prognosis of conversion to late AMD as well as the qualification path towards a clinical trial endpoint for functional assessments that describe functional decline within iAMD will be discussed with regulators in the near future.

Conflict of interest statement: UFOL is employee of F. Hoffmann La Roche ltd.