Sruthi Arepalli1,2, Charles C Wykoff3, Joseph R Abraham1,2, Leina Lunasco1, Hannah Yu3, Ming Hu1,4, Sunil K Srivastava1,2, Jamie L Reese1, David Brown3, Justis P Ehlers5,6. 1. The Tony and Leona Campane Center for Excellence in Image-Guided Surgery and Advanced Imaging Research, Cole Eye Institute, Cleveland Clinic, Cleveland, OH, USA. 2. Vitreoretinal Service, Cole Eye Institute, Cleveland Clinic, Cleveland, OH, USA. 3. Retinal Consultants of Texas, Retina Consultants of America, Houston, TX, USA. 4. Quantitative Health Sciences, Cleveland Clinic, Cleveland, OH, USA. 5. The Tony and Leona Campane Center for Excellence in Image-Guided Surgery and Advanced Imaging Research, Cole Eye Institute, Cleveland Clinic, Cleveland, OH, USA. ehlersj@ccf.org. 6. Vitreoretinal Service, Cole Eye Institute, Cleveland Clinic, Cleveland, OH, USA. ehlersj@ccf.org.
Abstract
BACKGROUND/ OBJECTIVES: Retinal vein occlusion (RVO) is the second most common retinal vascular disorder. Despite promising advances with anti-VEGF therapy, select patients are unresponsive to therapy. A precision medicine-based approach for therapeutic decision-making based on underlying biomarkers may facilitate treatment based on the underlying pathway. This study aims to identify the baseline and longitudinal cytokine profiles of RVO-related macular oedema and correlating these expression profiles with higher order OCT features using a novel retinal segmentation and feature extraction platform. SUBJECTS/ METHODS: The IMAGINE study is a post-hoc assessment of aqueous humour cytokines with correlation to higher level analysis of imaging studies. OCT scans underwent machine learning enhanced segmentation of the internal limiting membrane (ILM), ellipsoid zone (EZ) and retinal pigment epithelium (RPE), as well as evaluating volumetric fluid metrics. Samples of aqueous humour were obtained at baseline, as well as months 4 and 9 prior to treatment. These samples were analysed for the expression of multiple cytokines. Patients were divided into Responders and Non-Responders based on OCT profiles. Additionally, patients were categorised as a Rebounder if their CST increased by 50% after initial improvement. RESULTS: Twenty-six eyes were included. The OCT-based response schema identified 21 Responders (81%) and 5 Non-Responders (19%). VEGF levels directly correlated with intraretinal fluid volume and angiogenin was inversely correlated with fluid indices. Multiple cytokines, including ANGPTL4, were directly correlated with ellipsoid zone disruption. The baseline VEGF levels were significantly higher in all responders compared to Non-Responders (p = 0.02). Rebounders tended to have significantly decreased levels of angiogenin and TIMP-1 (p = 0.019, p = 0.015). CONCLUSIONS: Cytokine expression was linked to specific OCT features and treatment response in RVO. Identification of an imaging phenotype that could serve as a surrogate for underlying active disease pathways could enhance treatment decision-making and precision medicine.
BACKGROUND/ OBJECTIVES: Retinal vein occlusion (RVO) is the second most common retinal vascular disorder. Despite promising advances with anti-VEGF therapy, select patients are unresponsive to therapy. A precision medicine-based approach for therapeutic decision-making based on underlying biomarkers may facilitate treatment based on the underlying pathway. This study aims to identify the baseline and longitudinal cytokine profiles of RVO-related macular oedema and correlating these expression profiles with higher order OCT features using a novel retinal segmentation and feature extraction platform. SUBJECTS/ METHODS: The IMAGINE study is a post-hoc assessment of aqueous humour cytokines with correlation to higher level analysis of imaging studies. OCT scans underwent machine learning enhanced segmentation of the internal limiting membrane (ILM), ellipsoid zone (EZ) and retinal pigment epithelium (RPE), as well as evaluating volumetric fluid metrics. Samples of aqueous humour were obtained at baseline, as well as months 4 and 9 prior to treatment. These samples were analysed for the expression of multiple cytokines. Patients were divided into Responders and Non-Responders based on OCT profiles. Additionally, patients were categorised as a Rebounder if their CST increased by 50% after initial improvement. RESULTS: Twenty-six eyes were included. The OCT-based response schema identified 21 Responders (81%) and 5 Non-Responders (19%). VEGF levels directly correlated with intraretinal fluid volume and angiogenin was inversely correlated with fluid indices. Multiple cytokines, including ANGPTL4, were directly correlated with ellipsoid zone disruption. The baseline VEGF levels were significantly higher in all responders compared to Non-Responders (p = 0.02). Rebounders tended to have significantly decreased levels of angiogenin and TIMP-1 (p = 0.019, p = 0.015). CONCLUSIONS: Cytokine expression was linked to specific OCT features and treatment response in RVO. Identification of an imaging phenotype that could serve as a surrogate for underlying active disease pathways could enhance treatment decision-making and precision medicine.
Authors: Jeffrey S Heier; Peter A Campochiaro; Linda Yau; Zhengrong Li; Namrata Saroj; Roman G Rubio; Phillip Lai Journal: Ophthalmology Date: 2012-02-01 Impact factor: 12.079
Authors: Rohit Varma; Neil M Bressler; Ivan Suñer; Paul Lee; Chantal M Dolan; James Ward; Shoshana Colman; Roman G Rubio Journal: Ophthalmology Date: 2012-07-18 Impact factor: 12.079
Authors: Peter A Campochiaro; Jeffrey S Heier; Leonard Feiner; Sarah Gray; Namrata Saroj; Amy Chen Rundle; Wendy Yee Murahashi; Roman G Rubio Journal: Ophthalmology Date: 2010-04-15 Impact factor: 12.079