Cytokine and Chemokine Profile Changes in Patients with Neovascular Age-Related Macular Degeneration After Intravitreal Ranibizumab Injection for Choroidal Neovascularization

Small immunological clocks identified by deep learning and gradient boosting

Background

The aging process affects all systems of the human body, and the observed increase in inflammatory components affecting the immune system in old age can lead to the development of age-associated diseases and systemic inflammation.

Results

We propose a small clock model SImAge based on a limited number of immunological biomarkers. To regress the chronological age from cytokine data, we first use a baseline Elastic Net model, gradient-boosted decision trees models, and several deep neural network architectures. For the full dataset of 46 immunological parameters, DANet, SAINT, FT-Transformer and TabNet models showed the best results for the test dataset. Dimensionality reduction of these models with SHAP values revealed the 10 most age-associated immunological parameters, taken to construct the SImAge small immunological clock. The best result of the SImAge model shown by the FT-Transformer deep neural network model has mean absolute error of 6.94 years and Pearson ρ = 0.939 on the independent test dataset. Explainable artificial intelligence methods allow for explaining the model solution for each individual participant.

Conclusions

We developed an approach to construct a model of immunological age based on just 10 immunological parameters, coined SImAge, for which the FT-Transformer deep neural network model had proved to be the best choice. The model shows competitive results compared to the published studies on immunological profiles, and takes a smaller number of features as an input. Neural network architectures outperformed gradient-boosted decision trees, and can be recommended in the further analysis of immunological profiles.

Ocular Vascular Diseases: From Retinal Immune Privilege to Inflammation

The eye is an immune privileged tissue that insulates the visual system from local and systemic immune provocation to preserve homeostatic functions of highly specialized retinal neural cells. If immune privilege is breached, immune stimuli will invade the eye and subsequently trigger acute inflammatory responses. Local resident microglia become active and release numerous immunological factors to protect the integrity of retinal neural cells. Although acute inflammatory responses are necessary to control and eradicate insults to the eye, chronic inflammation can cause retinal tissue damage and cell dysfunction, leading to ocular disease and vision loss. In this review, we summarized features of immune privilege in the retina and the key inflammatory responses, factors, and intracellular pathways activated when retinal immune privilege fails, as well as a highlight of the recent clinical and research advances in ocular immunity and ocular vascular diseases including retinopathy of prematurity, age-related macular degeneration, and diabetic retinopathy.

Biomarkers as Predictive Factors of Anti-VEGF Response

Age-related macular degeneration is the main cause of irreversible vision in developed countries, and intravitreal anti-vascular endothelial growth factor (anti-VEGF) injections are the current gold standard treatment today. Although anti-VEGF treatment results in important improvements in the course of this disease, there is a considerable number of patients not responding to the standardized protocols. The knowledge of how a patient will respond or how frequently retreatment might be required would be vital in planning treatment schedules, saving both resource utilization and financial costs, but today, there is not an ideal biomarker to use as a predictive response to ranibizumab therapy. Whole blood and blood mononuclear cells are the samples most studied; however, few reports are available on other important biofluid samples for studying this disease, such as aqueous humor. Moreover, the great majority of studies carried out to date were focused on the search for SNPs in genes related to AMD risk factors, but miRNAs, proteomic and metabolomics studies have rarely been conducted in anti-VEGF-treated samples. Here, we propose that genomic, proteomic and/or metabolomic markers could be used not alone but in combination with other methods, such as specific clinic characteristics, to identify patients with a poor response to anti-VEGF treatment to establish patient-specific treatment plans.