Corneal Confocal Microscopy and Brain MRI: Surrogate Markers of Neuronal Pathology in Schizophrenia

Background: Both neurodevelopmental and neurodegenerative processes have been proposed to contribute to the development of schizophrenia. However, there is no validated surrogate marker of neuronal pathology for schizophrenia. To quantify neuronal pathology in subjects with schizophrenia using corneal confocal microscopy (CCM) and brain volumetric MRI. Methods: In this cross-sectional study, subjects with DSM-5 based criteria for schizophrenia enrolled from the Psychiatry Hospital in Qatar and healthy controls were recruited between August 26, 2019 and March 11, 2020. Subjects with schizophrenia were assessed for positive and negative symptoms, and with controls underwent assessment of cognitive function, CCM and brain volumetric MRI. Neuronal pathology was assessed by quantifying corneal nerve fibre density (CNFD), length (CNFL), branch density (CNBD) and CNBD:CNFD ratio, and MRI volume of whole brain, cortical gray matter, ventricle, cingulate gyrus, hippocampus, amygdala, thalamus and entorhinal cortex. Findings: Subjects with schizophrenia (n=36) and controls (n=26) with a comparable age and gender were studied. After adjusting for confounders, subjects with schizophrenia had lower cognitive function (P≤0.01), CNFD, CNFL, CNBD and CNBD:CNFD ratio (P<0.0001) compared to controls. Quantitative MRI demonstrated a significant reduction in the cingulate gyrus volume (P<0.05) with no difference in whole brain (P=0.61), cortical gray matter (P=0.99), ventricle (P=0.47), hippocampus (P=0.10), amygdala (P=0.68), thalamus (P=0.20) and entorhinal cortex (P=0.26) volumes compared to controls. Corneal nerve measures and cingulate gyrus volume did not significantly correlate with symptom severity (P=0.35) or cognitive function (P=0.051). Corneal nerve measures had no significant association with metabolic syndrome (P=0.61) or diabetes (P=0.057). The area under the ROC curve to distinguish subjects with schizophrenia from controls was 88% for CNFL (P<0.0001), 84% for CNBD and CNBD:CNFD ratio (P<0.0001), 79% for CNFD (P<0.0001), 73% for the cingulate gyrus volume (P<0.05) and 70% for cognitive function (P<0.01). Interpretation: This study demonstrates a reduction in corneal nerve fibres and cingulate gyrus volume with a high diagnostic accuracy of CCM for schizophrenia. Funding Statement: Qatar National Research Fund (BMRP-5726113101) and Hamad Medical Corporation (IRGC-04-SI-17-166). Declaration of Interests: The authors declare no competing interest. Ethics Approval Statement: The study was approved by the IRB of Weill Cornell Medicine in Qatar and Hamad Medical Corporation. All participants gave informed consent to take part in the study. The research adhered to the tenets of the declaration of Helsinki.