Non-invasive Assessment of Neurovascular Coupling After Aneurysmal Subarachnoid Hemorrhage: A Prospective Observational Trial Using Retinal Vessel Analysis

Abstract

Objective: Delayed cerebral ischemia (DCI) is a common complication after aneurysmal subarachnoid hemorrhage (aSAH) and can lead to infarction and poor clinical outcome. The underlying mechanisms are still incompletely understood, but animal models indicate that vasoactive metabolites and inflammatory cytokines produced within the subarachnoid space may progressively impair and partially invert neurovascular coupling (NVC) in the brain. Because cerebral and retinal microvasculature are governed by comparable regulatory mechanisms and may be connected by perivascular pathways, retinal vascular changes are increasingly recognized as a potential surrogate for altered NVC in the brain. Here, we used non-invasive retinal vessel analysis (RVA) to assess microvascular function in aSAH patients at different times after the ictus. Methods: Static and dynamic RVA were performed using a Retinal Vessel Analyzer (IMEDOS Systems GmbH, Jena) in 70 aSAH patients during the early (d0-4), critical (d5-15), late (d16-23) phase, and at follow-up (f/u > 6 weeks) after the ictus. For comparison, an age-matched cohort of 42 healthy subjects was also included in the study. Vessel diameters were quantified in terms of the central retinal arterial and venous equivalent (CRAE, CRVE) and the retinal arterio-venous-ratio (AVR). Vessel responses to flicker light excitation (FLE) were quantified by recording the maximum arterial and venous dilation (MAD, MVD), the time to 30% and 100% of maximum dilation (tMAD30, tMVD30; tMAD, tMVD, resp.), and the arterial and venous area under the curve (AUCart, AUCven) during the FLE. For subgroup analyses, patients were stratified according to the development of DCI and clinical outcomes after 12 months. Results: Vessel diameter (CRAE, CRVE) was significantly smaller in aSAH patients and showed little change throughout the whole observation period (p < 0.0001 vs. control for all time periods examined). In addition, aSAH patients exhibited impaired arterial but not venous responses to FLE, as reflected in a significantly lower MAD [2.2 (1.0-3.2)% vs. 3.6 (2.6-5.6)% in control subjects, p = 0.0016] and AUCart [21.5 (9.4-35.8)%*s vs. 51.4 (32.5-69.7)%*s in control subjects, p = 0.0001] on d0-4. However, gradual recovery was observed during the first 3 weeks, with close to normal levels at follow-up, when MAD and AUCart amounted to 3.0 [2.0-5.0]% (p = 0.141 vs. control, p = 0.0321 vs. d5-15) and 44.5 [23.2-61.1]%*s (p = 0.138 vs. control, p < 0.01 vs. d0-4 & d5-15). Finally, patients with clinical deterioration (DCI) showed opposite changes in the kinetics of arterial responses during early and late phase, as reflected in a significantly lower tMAD30 on d0-4 [4.0 (3.0-6.8) s vs. 7.0 (5.0-8.0) s in patients without DCI, p = 0.022) and a significantly higher tMAD on d16-23 (24.0 (21.0-29.3) s vs. 18.0 (14.0-21.0) s in patients without DCI, p = 0.017]. Conclusion: Our findings confirm and extend previous observations that aSAH results in sustained impairments of NVC in the retina. DCI may be associated with characteristic changes in the kinetics of retinal arterial responses. However, further studies will be required to determine their clinical implications and to assess if they can be used to identify patients at risk of developing DCI. Trial Registration: ClinicalTrials.gov Identifier: NCT04094155.

Keywords: aneurysmal subarachnoid hemorrhage; cerebral infarction; delayed cerebral ischemia; microvascular function; neurovascular coupling; retinal vessel analysis.

Conflict of interest statement

IMEDOS Systems GmbH provided the Retinal Vessel Analyzer for research purposes only. The company did not have any additional role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Copyright © 2021 Albanna, Conzen, Weiss, Seyfried, Kotliar, Schmidt, Kuerten, Hescheler, Bruecken, Schmidt-Trucksäss, Neumaier, Wiesmann, Clusmann and Schubert.

Figures

Figure 1

Static and dynamic retinal vessel analysis (RVA). (A) Average retinal vessel response to flicker light excitation (FLE) in healthy subjects with the different parameters determined for dynamic RVA. The stimulation period is indicated above and by light gray shading. Parameters used for quantification of the response comprised maximum arterial or venous dilation (MAD/MVD), time to maximum arterial or venous dilation (tMAD/tMVD) relative to flicker initiation, time to 30% of the maximum arterial or venous dilation (tMAD30/tMVD30) relative to flicker initiation, and arterial or venous area under the curve (AUCart/AUCven) during the FLE. (B) Example of a monochromatic fundus image used for RVA. Arterial (A) and venous (V) segments are indicated in red and blue, respectively.

Figure 2

Results from static RVA in aSAH patients and control subjects. Comparison of (A) central retinal arterial equivalent (CRAE) as a measure of arterial diameter, (B) central retinal venous equivalent (CRVE) as a measure of venous diameter and (C) retinal arterio-venous-ratio (AVR) in control subjects and aSAH patients during the early (aSAHd0−4), critical (aSAHd5−15) and late (aSAHd16−23) phase and at follow-up more than 6 weeks after the ictus (aSAHf/u). ***p < 0.001 vs. control; **p < 0.01 vs. control; *p < 0.05 vs. control.

Figure 3

Results from dynamic RVA in aSAH patients and control subjects. (A) Average relative retinal arterial responses to flicker light stimulation measured in control subjects and aSAH patients during the early (aSAHd0−4), critical (aSAHd5−15), and late (aSAHd16−23) phase and at follow-up more than 6 weeks after the ictus (aSAHf/u). Stimulation period is indicated by light gray shading. (B) Maximum arterial dilation (MAD) quantified from data shown in A. (C) Arterial area under the curve (AUCart ) quantified from the data shown in A. (D) Average retinal venous responses to flicker light stimulation measured in control subjects and aSAH patients at the same times as in A. (E) Maximum venous dilation (MVD) quantified from the data shown in D. (F) Venous area under the curve (AUCven) quantified from the data shown in A. ***p < 0.001 vs. control; **p < 0.01 vs. control; *p = 0.05 vs. control; ##p < 0.01 vs. aSAHf/u; #p < 0.05 vs. aSAHf/u.

Figure 4

Comparison of retinal arterial responses in aSAH patients with and without DCI. (A) Average relative retinal arterial responses to flicker light stimulation measured during the early phase (days 0–4) in aSAH patients with or without delayed cerebral ischemia (DCI). Stimulation period is indicated by light gray shading. Inset shows the rising phase on an expanded time-scale and after scaling the responses to the same amplitude to highlight kinetic changes. (B) Time to 30% of the maximum arterial dilation (tMAD30) and (C) time to maximum arterial dilation (tMAD) determined during the early phase in patients with and without DCI. (D) Time-course of changes in median [q1–q3] tMAD30 (bottom) and tMAD (top) observed in patients with or without DCI. (E) tMAD30 and (F) tMAD determined during the late phase (days 16–23) in patients with and without DCI. For a summary of all parameters and p-values see Supplementary Table 2.

Figure 5

Dependence of retinal vessel properties in aSAH patients on nimodipine treatment. (A) Average relative arterial responses to flicker light stimulation measured with or without nimodipine (nimo) treatment. Inset shows the same data but scaled to the same amplitude for comparison of response kinetics. Stimulation period is indicated by light gray shading. (B) Average relative venous responses to flicker light stimulation measured with or without nimodipine (nimo) treatment. Inset shows the same data but scaled to the same amplitude for comparison of response kinetics. Stimulation period is indicated by light gray shading. (C–H) Comparison of (C) central retinal arterial equivalent (CRAE), (D) central retinal venous equivalent (CRVE), (E) maximum arterial dilation (MAD), (F) maximum venous dilation (MVD), (G) time to 30 and 100% of maximum arterial dilation (tMAD30 & tMAD), and (H) time to 30 and 100% of maximum venous dilation (tMVD30 & tMVD) in aSAH patients with or without nimodipine treatment during the early (aSAHd0−4), critical (aSAHd5−15), and late (aSAHd16−23) phase.