Retinal optical coherence tomography angiography as a bedside guide for perioperative resuscitation

We explored the feasibility of utilizing optical coherence tomography angiography (OCTA) device in a peri-operative setting to assess the perfusion status of the retinal microcirculation by perturbating a quantitative macrocirculatory target - mean arterial pressure (MAP), and by extension, the cerebrovascular microcirculation. Using a porcine model, we were able to induce hypercarbia, epinephrine-led resuscitation, and hemorrhage followed by autologous re-transfusion scenarios. OCTA imaging was utilized to quantify layer-specific vascular density (%) of the retinal macular region - defined as the percentage of the total scanned area occupied by blood vessels. Comparisons across stages were performed via ANOVA or Kruskal-Wallis tests with appropriate post-hoc adjustments (Tukey’s HSD or Dunn’s test). Vascular density of superficial capillary plexus (VD-SCP) showed an average increase in perfusion of approximately 4% and 1.2% from baseline for resuscitation and hypercarbia stages, respectively ( baseline : 46.65 ± 1.74% vs. hypercarbia : 47.19 ± 1.56% vs. resuscitation : 48.50 ± 2.04% vs. hemorrhage and re-transfusion : 44.83 ± 1.90%; p = 0.0007). Conversely, re-transfusion stage showed an approximate average reduction in VD-SCP by 3.9% from baseline. Vascular density of the deep capillary plexus layer (VD-DCP) showed a significant increase (6.31%) from baseline in for hypercarbia stage but remained unchanged for the rest of the stages ( baseline : 48.46 ± 0.70% vs. hypercarbia : 51.52 ± 1.58% vs. resuscitation : 50.34 ± 0.99% vs. hemorrhage and re-transfusion : 50.58 ± 1.48%; p = 0.0003). Regional variation in perfusion was more pronounced in the SCP layer. VD-SCP exhibited significant, positive association with MAP ( r = 0.43, p = 0.0096) and other physiological parameters, whereas VD-DCP did not exhibit any such associations. Regional vascular density asymmetry changes (Asym) across each stage were captured by SCP, but not DCP. OCTA biomarkers were able to detect porcine retinal microcirculatory changes when MAP varied between 38.67 ± 4.45 mm Hg to 84.83 ± 24.68 mm Hg. The porcine SCP layer exhibited responsive, spatial changes to MAP as compared to the porcine DCP layer.