Doheny Eye Institute, Los Angeles, CA, United States of America.
Department of Ophthalmology, David Geffen School of Medicine at UCLA, Los Angeles, CA, United States of America.
PLoS One. 2020 May 29;15(5):e0232785. doi: 10.1371/journal.pone.0232785. eCollection 2020.
Alzheimer's disease (AD) pathology precedes symptoms and its detection can identify at-risk individuals who may benefit from early treatment. Since the retinal nerve fiber layer (RNFL) is depleted in established AD, we tested whether its thickness can predict whether cognitively healthy (CH) individuals have a normal or pathological cerebrospinal fluid (CSF) Aß42 (A) and tau (T) ratio.
As part of an ongoing longitudinal study, we enrolled CH individuals, excluding those with cognitive impairment and significant ocular pathology. We classified the CH group into two sub-groups, normal (CH-NAT, n = 16) or pathological (CH-PAT, n = 27), using a logistic regression model from the CSF AT ratio that identified >85% of patients with a clinically probable AD diagnosis. Spectral-domain optical coherence tomography (OCT) was acquired for RNFL, ganglion cell-inner plexiform layer (GC-IPL), and macular thickness. Group differences were tested using mixed model repeated measures and a classification model derived using multiple logistic regression.
Mean age (± standard deviation) in the CH-PAT group (n = 27; 75.2 ± 8.4 years) was similar (p = 0.50) to the CH-NAT group (n = 16; 74.1 ± 7.9 years). Mean RNFL (standard error) was thinner in the CH-PAT group by 9.8 (2.7) μm; p < 0.001. RNFL thickness classified CH-NAT vs. CH-PAT with 87% sensitivity and 56.3% specificity.
Our retinal data predict which individuals have CSF biomarkers of AD pathology before cognitive deficits are detectable with 87% sensitivity. Such results from easy-to-acquire, objective and non-invasive measurements of the RNFL merit further study of OCT technology to monitor or screen for early AD pathology.
阿尔茨海默病（AD）的病理学先于症状出现，其检测可以识别出处于高危状态的个体，这些个体可能受益于早期治疗。由于在已确诊的 AD 中视网膜神经纤维层（RNFL）已经耗竭，我们测试了其厚度是否可以预测认知健康（CH）个体的脑脊液（CSF）中 Aβ42（A）和 tau（T）比值是否正常或存在病理学改变。
作为一项正在进行的纵向研究的一部分，我们招募了认知健康的个体，排除了有认知障碍和明显眼部疾病的个体。我们使用 CSF AT 比值的逻辑回归模型将 CH 组分为两组，正常组（CH-NAT，n = 16）或病理组（CH-PAT，n = 27），该模型可以识别出 85%以上有临床确诊 AD 诊断的患者。对 RNFL、神经节细胞内丛状层（GC-IPL）和黄斑厚度进行了光谱域光相干断层扫描（OCT）检查。使用混合模型重复测量和基于多变量逻辑回归的分类模型测试组间差异。
病理组（n = 27；75.2 ± 8.4 岁）的平均年龄与正常组（n = 16；74.1 ± 7.9 岁）相似（p = 0.50）。病理组的平均 RNFL（标准误差）薄 9.8（2.7）μm；p < 0.001。RNFL 厚度可将 CH-NAT 与 CH-PAT 分类，其敏感性为 87%，特异性为 56.3%。
我们的视网膜数据预测了哪些个体具有 CSF 生物标志物的 AD 病理学改变，而这些改变在认知缺陷可检测之前就已经出现，其敏感性为 87%。这些来自易于获取、客观和非侵入性的 RNFL 测量结果值得进一步研究 OCT 技术，以监测或筛查早期 AD 病理学改变。
