This cross-sectional study was approved by the Ethics Committee on Biomedical Research, West China Hospital of Sichuan University, and was conducted in accordance with the Declaration of Helsinki. All the participants were recruited in this study from June 2022 to December 2023. Written informed consent was obtained from all the enrolled participants.

Participants with PCA and age- and sex-matched healthy control participants were recruited from West China Hospital (the details are shown in Supplementary Methods). All participants with PCA completed a comprehensive neurological evaluation, including a clinical history interview, physical examination, and neuropsychological assessment. The clinical diagnosis of PCA was confirmed in accordance with previously proposed clinical diagnostic criteria [7]. Baseline information, including age, sex, and educational level, was recorded for all the participants, and the participants underwent a complete ophthalmic evaluation, including slit-lamp biomicroscopy, fundus examination, SS-OCT, and SS-OCTA. AO-SLO with a 2.4 × 2.4 field of view was performed in some patients (Supplementary Fig. 1). All evaluations and examinations were conducted over 1 month (Supplementary Methods).

A detailed medical history was obtained in all participants with PCA, and they underwent extensive neurological evaluation, including the following: Montreal Cognitive Assessment (MoCA), and Mini-Mental State Examination (MMSE), and Clinical Dementia Rating (CDR)assessment for general cognitive function, and APOE genotyping. 18 F-AV45 PET/MRI was performed using the GE 3T scanner (SIGNA PET/MR; GE Healthcare, Chicago, IL, USA) to identify the Aβ deposition in the brain (Fig. 1).

Representative sagittal 18 F-AV45 PET/MRI images. (A) A participant with PCA with significant cortical amyloid-β (Aβ) deposition (white arrows) and occipital-parietal atrophy of right hemisphere in brain MRI (yellow circle). (B) An age and sex matched healthy control participant with typical negative 18 F AV45-PET scan and MRI scan indicates no significant occipital-parietal atrophy

A visual assessment of PET images was performed by two experienced nuclear medicine physicians in a collaborative session, during which they were required to provide a binary diagnosis (positive or negative). To mitigate potential bias, the order of image readings was randomized, and all participant information was concealed. The criteria for identifying positive 18 F-AV45 PET images were based on visual rating guidelines for interpreting amyloid PET [31].

All participants underwent SS-OCT (VG200, SVision Imaging, Ltd., Luoyang, China) with a central wavelength of 1,050 nm, scanning speed of 200,000 A-scans/s, scanning depth of 2.7 mm in tissue, and sampling spacing of 12 mm [17]. The scanning models included (1) 6 mm × 6mm2,512 × 512 pixels, R4, centered on the fovea, (2) 33 scan lines, R16, centered on the fovea. Both SS-OCT and SS-OCTA images with signal strengths > 7/10 without severe motion artefacts were obtained and included in this study.

The SS-OCT structure parameters were automatically measured in a 6 mm × 6mm2 diameter region centered on the fovea, including the outer retinal thickness (ORT), choroidal thickness (CRT), retinal nerve fiber layer (RNFL) and ganglion cell inner plexiform layer (GCIPL). The SS-OCTA flow parameters were automatically assessed in a scanning mode of 6 mm × 6 mm,512 × 512 pixels, R4, and centered on the fovea, which included the choriocapillaris vessel density (VD), choroidal vascular volume (CVV) and choroidal vascular index (CVI). Both the SS-OCT structural parameters and SS-OCTA flow parameters were automatically measured using the Early Treatment Diabetic Retinopathy Study (ETDRS) circle, including 0–1 mm, 1–3 mm, 3–6 mm and total (0–6 mm) circles.

Two participants with PCA and two healthy control participants underwent AO-SLO scanning (Mona II, Robotrak Technologies, Nanjing, China), the field of 2.4 × 2.4° (approximately 700 × 700 μm) view on the macula. Cone segmentation is automatically generated and cone morphology properties, including cell density, cell dispersion and cell regularity, were automatically analyzed.

The Cone cell density is defined as the average cone cell number per square millimeter. The cone cell regularity refers to the uniformity of cone distribution, measured by the ratio of cells with a precise number of closest cells within a certain range. The cone cell dispersion is defined as the extent of cell dispersed or clustered, quantified by the proportion of the average distance between cone cells [32] (Supplementary Fig. 1).

The outer retinal and choroidal alterations, and thickness of RNFL and GCIPL were compared between the enrolled eyes of PCA and age- and sex-matched healthy control participants. Accordingly, subgroup comparisons between ε4 carriers and ε4 non-carriers were performed to detect the effects of APOE ε4 on the retina and choroid in participants with PCA. To investigate the potential association between the ocular alterations and the PCA features: (1) the PCA neuroimaging features, including cortical volume and cortical thickness of the occipital, parietal, and temporal lobes; (2) the PCA clinical features, including cognitive function, assessed by MMSE, MoCA and CDR examinations were assessed.

Demographic and clinical variables were analyzed using an independent t-test for continuous variables and χ 2 test for categorical variables. Generalized estimating equation (GEE) models were used to compare the differences in the SS-OCT structure and SS-OCTA flow parameters between participants with PCA and healthy control participants while adjusting for inter-eye dependencies, age, sex and signal strengths. Bonferroni method was applied for multiple comparisons. Partial correlation analyses were used to detect correlations between SS-OCT structure SS-OCTA flow parameters and PCA pathologic characteristics. All P values were performed on 2-tailed tests, and a P-value < 0.05 was considered significant. All statistical analyses were performed using IBM SPSS Statistics version 23.