A theoretical study of vascular configurations of retinal capillary plexi based on OCTA data

Abstract

Purpose : Optical coherence tomography angiography (OCTA) has made available a huge amount of imaging data regarding the superficial (SCP), intermediate (ICP) and deep (DCP) capillary plexi in the retina. However, their interpretation is not unanimous: the question of whether capillary plexi are connected in series (S) or in parallel (P) remains a matter of debate (Campbell et al 2016; Fouquet et al 2017). To address this question, we utilize a mathematical model to quantify the impact on hemodynamics and oxygenation associated with S and P vascular configurations (denoted by Sconf and Pconf) (FIG 1)

Methods : The retinal vasculature is modeled via an hydraulic-electric analogy, where the two configurations are tested (FIG 1). Model parameters are based on OCTA data. The capillary flow computed with the model provides an oxygen (O₂) source within the tissue, where, given the metabolic consumption, we calculate the O₂ levels. The model simulates how retinal hemodynamics and oxygenation would change in the cases of central retinal vein occlusion (CRVO) (FIG 2A) and IOP elevation (IOP-el) (FIG 2B), depending on whether Sconf or Pconf are considered

Results : CRVO is simulated by reducing the CRV radius in FIG 1. The results reported in FIG 2A show that:(i) in the Sconf, the plexi exhibit a differential response, with DCP and ICP experiencing larger pressure drops than SCP;(ii) in the Pconf, the blood flow redistributes uniformly across them. The Sconf results are coherent with studies on paracentral acute middle maculopathy (Garrity et al 2017; Sarraf et al 2013), where, secondary to CRVO, DCP and ICP are affected by ischemia. In the Sconf, the O₂ profiles of the outer nuclear layer (ONL), outer plexiform layer (OPL) and inner nuclear layer (INL) are those most affected by CRVO, whereas INL and glanglion cell layer (GCL) in the P-conf. The simulated O₂ levels are positive until reaching a CRVO of 80%. Similar results are obtained upon elevating IOP (FIG 2B-top): Sconf results are coherent with clinical studies on IOP-el (Bui et al 2018)

Conclusions : Mathematical modeling allowed us to simulate two possible vascular configurations of the retinal capillary plexi, thereby aiding the interpretation of OCTA data. Results suggest that S-conf better describes the physiology of the retinal vasculature

This abstract was presented at the 2019 ARVO Annual Meeting, held in Vancouver, Canada, April 28 - May 2, 2019.

 

Schematic of retinal vascular model

 

Simulated hemodynamic and oxygenation changes due to CRVO and IOP-el in Sconf and Pconf