LAUSR

Considerations in the Understanding of Venous Outflow in the Retinal Capillary Plexus For many decades, fluorescein angiography has been the gold standard imaging modality for the assessment of retinal perfusion and the evaluation of retinal vascular disorders.1 Fluorescein angiography, however, does not provide sufficient detail of the deep retinal capillary plexus.1–3 More advanced retinal imaging technology, including optical coherence tomography angiography, can now provide depthresolved visualization of the complex anatomy of the retinal vasculature4 and identification of three distinct plexuses in the central macula including the intermediate retinal capillary plexus (ICP),5–7 in addition to the superficial capillary plexus (SCP) and deep capillary plexus (DCP). The ability to image all three plexuses may have important clinical implications for retinal ischemic disorders such as paracentral acute middle maculopathy (PAMM) and acute macular neuroretinopathy, which may be due to capillary ischemia at the ICP and/or DCP levels.8,9 Furthermore, imaging the three plexuses may allow for earlier detection of microvascular changes in disorders such as diabetic retinopathy and macular telangiectasia Type 2.4,5,10 The interrelationships of the retinal capillary plexuses are incompletely understood. Based on their observations using a novel projectionresolved optical coherence tomography angiography algorithm, Campbell and associates recently proposed a model of the three trilaminar microvascular plexuses in which the SCP, ICP, and DCP are arranged in a hammock-like configuration, each communicating through vertically oriented major arteries and veins that are shown to connect with all three plexuses.7 Others have proposed different and/or more complex models.11–15 In this editorial, we wish to offer our interpretations of certain structural features of the retinal microvasculature, particularly as pertains to venous drainage at the level of the DCP. Studies in humans and in animal models have suggested that major arterioles and especially major venules may independently connect to the DCP without first communicating with the SCP, and that the DCP may serve a primarily venous role.5,12,16–19 In a 1996 study, Foreman and associates presented stereoscopic confocal scanning laser microscopic images of the retinal capillary system in human eyes prepared from depth-resolved maximum-intensity immunofluorescence projections.18 Careful inspection of the en face stereopair photographs from this article demonstrates a larger superficial arteriole or venule communicating independently with the DCP, without any connection with the SCP. As Shimizu described in his landmark book entitled Structure of Ocular Vessels: “It even happens, rather frequently, that an (outer) retinal capillary directly drains into a major retinal vein at its posterior aspect.”16 Shimizu’s images of the retinal microvascular casts of macaque monkeys demonstrated this pattern. In a study of neural–vascular relationships in the central retina of macaque monkeys, Snodderly and associates noted that “the deep capillary converged onto tributaries which drained directly into large venous trunks.”17 Studies in mice,12 rats,19 and pigs20 have also suggested that the DCP drains independently into venules, with fewer direct connections between the SCP (and likely the ICP) and the major venular system. We would also like to highlight several other critical aspects regarding the morphology of the DCP. It is generally accepted that the SCP (and perhaps ICP) is arranged as a series of “hammocks” situated between a major artery and vein.14 However, reports have consistently demonstrated that the DCP does not have this “hammock” morphology suggested in the model by Campbell and associates,7 but is instead organized as radial capillaries that converge into a central “vortex” venule that then drains directly into a major venule.5,12,14 In addition, capillary anastomoses directly connecting the SCP, ICP, and DCP have been demonstrated,5,7,12,15,21 the presence of which would imply D. Sarraf receives research grants from Allergan, Genentech, Heidelberg, Regeneron, and Optovue and is a consultant for Amgen, Bayer, Genentech, Novartis, and Optovue. K. B. Freund is a consultant for Genentech, Optovue, Optos, Bayer Healthcare, and Heidelberg Engineering. The other authors have no financial/ conflicting interests to disclose.