Flitcroft Ian, Lingham Gareth, Kerin Eoin, Nkansah Ernest Kyei, Mackey David A, Lee Samantha Sze-Yee, Kobia-Acquah Emmanuel, Loughman James
From the Centre for Eye Research Ireland, Environmental Sustainability and Health Institute, Technological University Dublin (I.F., G.L., E.K., E.K.N., E.K.-A., J.L.) Dublin, Ireland; Department of Ophthalmology, Mater Misericordiae Hospital (I.F.), Dublin, Ireland.
From the Centre for Eye Research Ireland, Environmental Sustainability and Health Institute, Technological University Dublin (I.F., G.L., E.K., E.K.N., E.K.-A., J.L.) Dublin, Ireland; Centre for Ophthalmology and Visual Science (incorporating Lions Eye Institute), University of Western Australia (G.L., D.A.M., S.S.-Y.L.), Perth, Australia; Centre for Eye Research Australia (G.L.), Melbourne, Australia.
Am J Ophthalmol. 2025 Jun 28;278:346-355. doi: 10.1016/j.ajo.2025.06.041.
A thin choroid is a recognized risk factor for myopia-associated complications and visual impairment in later life. This study aims to develop a clinical tool to identify individuals whose choroidal thickness varies from that expected for their age, sex and refraction, and who might therefore be at higher or lower risk of future myopic complications.
Post-hoc patient-level meta-analysis SUBJECTS: Participants aged 6-30 years from four clinical studies: Myopia Outcome Study of Myopia in Children (MOSAIC), Treatment Optimization of Atropine Study (TOAST), Western Australia Atropine for the Treatment of Myopia study (WAATOM), Kidskin-Young Adult Myopia Study.
Spherical equivalent refraction (SER) was measured by cycloplegic autorefraction and axial length (AXL) by partial coherence interferometry. Choroidal thickness (ChT) was measured by Swept Source-OCT (Triton Plus, Topcon) or Spectral Domain-OCT (Spectralis, Heidelberg). Multiple linear regression and Machine Learning approaches were applied to create prediction models for ChT as a function of age, sex, SER and AXL.
Deviation of subfoveal ChT from the expected value a function of age, sex, SER and AXL.
Ordinary least square (OLS) regression with restricted cubic splines and a linear mixed model with non-linear spline terms for age and AXL both estimated ChT well, explaining over 44% of the variance. Of the remaining variance, approximately 50% was due to inter-individual differences in ChT (likely reflecting genetic, environmental or lifestyle factors), with 44% of participants having a ChT that was >50 µm thicker or thinner than expected. A clinical nomogram was generated from the OLS model to facilitate comparison of the observed ChT with that expected on the basis of SER, AXL and demographic factors.
ChT is a known myopia biomarker that is not fully utilised to inform clinical practice. The resultant clinical ChT nomogram is a simple, visual clinical tool that provides an objective method to gauge and potentially track a myopic child's risk of myopia-related complications, such as myopic maculopathy, based on the mismatch between their measured and expected choroidal thickness.
脉络膜变薄是近视相关并发症及晚年视力损害的一个公认风险因素。本研究旨在开发一种临床工具,以识别脉络膜厚度与其年龄、性别和屈光度预期值不同的个体,这些个体可能因此面临更高或更低的未来近视并发症风险。
事后患者水平的荟萃分析
来自四项临床研究的6至30岁参与者:儿童近视结局研究(MOSAIC)、阿托品治疗优化研究(TOAST)、西澳大利亚阿托品治疗近视研究(WAATOM)、儿童-青年成人近视研究。
通过睫状肌麻痹自动验光测量等效球镜度(SER),通过部分相干干涉测量法测量眼轴长度(AXL)。通过扫频源光学相干断层扫描(Triton Plus,拓普康)或谱域光学相干断层扫描(Spectralis,海德堡)测量脉络膜厚度(ChT)。应用多元线性回归和机器学习方法创建ChT作为年龄、性别、SER和AXL函数的预测模型。
中心凹下ChT与年龄、性别、SER和AXL预期值的偏差。
采用受限立方样条的普通最小二乘(OLS)回归以及对年龄和AXL采用非线性样条项的线性混合模型均能很好地估计ChT,解释了超过44%的方差。在其余方差中,约50%归因于ChT的个体间差异(可能反映遗传、环境或生活方式因素),44%的参与者其ChT比预期厚或薄超过50 µm。从OLS模型生成了临床列线图,以便于比较观察到的ChT与基于SER、AXL和人口统计学因素预期的ChT。
ChT是一种已知的近视生物标志物,但尚未充分用于指导临床实践。由此产生的临床ChT列线图是一种简单的可视化临床工具,它提供了一种客观方法,可根据近视儿童测量的脉络膜厚度与预期厚度之间的差异,来评估并潜在地跟踪其发生近视相关并发症(如近视性黄斑病变)的风险。