Drusen-specific dark adaptation profiles in intermediate age-related macular degeneration

Aims: To investigate the influence of macular drusen phenotypes on dark adaptation (DA) in intermediate age-related macular degeneration (iAMD). Methods: This cross-sectional, multicentric study enrolled 57 eyes of 43 iAMD patients. Drusen were subclassified as cuticular, soft, reticular pseudodrusen (RPD), or combined soft + RPD based on multimodal imaging. Dark adaptometry (AdaptDx; 20-minute test-time) assessed DA function through rod intercept time (RIT), last measured log sensitivity (LMLS; primary outcome) and area under the DA curve (AUDAC; secondary outcome). OCT volumes (30°x20° field) were analyzed using an AI-enhanced algorithm providing a quantification of chorioretinal layers and pigment epithelium detachments (PED) volumes. Multivariable tobit and linear regression analyzed associations between drusen phenotypes and DA outcomes. Results: Drusen phenotype distribution was: cuticular 13 eyes (22.8%), soft 13 eyes (22.8%), isolated RPD 21 eyes (36.8%), and combined soft + RPD 10 eyes (17.5%). Nearly all eyes reaching RIT within the 20-minute test ceiling (10 of 11 eyes) had cuticular drusen, which showed preserved rod function (LMLS: 2.9 ± 0.2 log-units; AUDAC: 10.3 ± 3.2 log-units·min; RIT: 14.9 ± 4.5 min) despite presenting the highest PED volumes (301 ± 152 nL). Conversely, isolated RPD demonstrated most severe dysfunction (LMLS: 1.7 ± 0.3; AUDAC: 23.7 ± 5.39). Soft drusen showed intermediate impairment (LMLS: 2.2 ± 0.2; AUDAC: 18.4 ± 2.4). Drusen phenotype was the most significant predictor of DA outcomes in both regression models (p < 0.001). Conclusion: Dark adaptometry reveals distinct patterns of DA impairment across drusen phenotypes, necessitating precise drusen classification when dark adaptometry serves as a functional endpoint in iAMD clinical trials. Future studies should employ extended test protocols (40-minute ceiling) to confirm these findings. Clinical trial number: Not applicable.