No evidence for amyloid in the retina of Alzheimer's disease patients

Background The retina is increasingly recognized as a promising target to identify early changes associated with Alzheimer’s disease (AD). Previous studies showed presence of retinal amyloid in AD patients in‐vivo as well as in post‐mortem retinal tissue of AD patients. The objective of this study is to non‐invasively discriminate AD patients from controls with fluorescent retinal imaging, using curcumin as labeling fluorophore. Method 26 AD patients (age 66(+9), MMSE≥18) and 14 controls (age 71(+12)) were enrolled from a tertiary memory clinic. Subjects were confirmed amyloid positive or negative based on CSF analysis and/or amyloid PET. We used three curcumin formulations: Longvida (10 days 4000 mg), Theracurmin (5 days 180 mg) and Novasol (4 days 300 mg). Plasma levels of curcumin and its derivatives were determined. Baseline and follow up scans of 2‐6 retinal regions were performed with blue autofluorescence imaging (λ = 486nm). Retinal images were visually assessed in a multidisciplinary setting. In addition a selection of retinal scans of participants receiving Longvida and Novasol were quantitatively analyzed. Result Plasma analysis yielded mean curcumin levels of 198.7 (±143.4) with Longvida, 576.6 nM, (±211.5) with Theracurmin and 1605.8 nM (±524.6) with Novasol. Visual analysis of baseline images showed no focal hyperfluorescence or other changes in AD‐patients compared to controls. Also, no differences were found when comparing pre‐and post‐curcumin images within AD patients and controls. Quantitative analyses confirmed a similar amount of focal hyperfluorescent spots in AD patients and controls. In addition no difference in increase of focal hyperfluorescence after curcumin intake could discriminate between AD patients and controls. Conclusion We found no focal retinal hyperfluorescence in AD patients or controls pre‐and post curcumin, even when using curcumin formulations that yielded higher plasma levels than Longvida, a formulation previously successfully used for this purpose. Implication: More and more initiatives are taken to assess retinal amyloid in AD patients and currently even commercial scans are brought to the market. We could not confirm that retinal changes described in previous studies represent retinal amyloid, and we were unable to replicate previous work discriminating AD patients from controls based on retinal amyloid visualization.