Retina thickness measurements provide a sensitive and specific diagnostic tool in Parkinson’s disease

Purpose To test the diagnostic ability of optical coherence tomography (OCT) for the detection of Parkinson’s disease (PD). Retinal thickness depends in part on the retinal pigment epithelium, one of the principal tissues producing levodopa. Methods PD patients (n=111) and healthy subjects (n=200) were enrolled. The Axonal application of the Spectralis OCT was used to obtain the circumpapillary RNFL thickness and retinal measurements in each participant. Two linear discriminant functions (LDF) were developed, one using RNFL parameters and another LDF using retinal thickness. A validating set (120 eyes from healthy individuals and 66 eyes from PD patients) was used to test the performance of both LDFs in an independent population. Receiver operating characteristic (ROC) curves were plotted and compared between both LDFs and with the standard parameters provided by OCT. Sensitivity and specificity were used to evaluate diagnostic performance. Results The Retinal LDF combines only retinal parameters and provided the best performance: 31.1 + 0.02 x Temporal Outer - 0.2 x Superior Outer + 0.16 x Nasal Outer - 0.2 x Inferior Outer - 0.06 x Superior Inner + 0.05 x Foveal thickness. The largest areas under ROC curve were 0.902 for Retinal LDF and 0.845 for RNFL LDF. At 95% fixed specificity, the Retinal LDF yielded the highest sensitivity values. Conclusion Measurements of retinal thickness obtained with Spectralis OCT had a very good ability to differentiate between healthy and PD individuals. Based on the area under the ROC curve, the Retinal LDF performed better than any single parameter or diagnostic test used for PD.