Microvascular blood flow velocities measured with a retinal function imager: inter-eye correlations in healthy controls and an exploration in multiple sclerosis

Liang Wang, Ohemaa Kwakyi, James Nguyen, Esther Ogbuokiri, Olwen C. Murphy, Natalia Gonzalez Caldito, Laura J. Balcer, Elliot M. Frohman, Teresa C. Frohman, Peter A. Calabresi, Shiv SaidhaEye 2018 · DOI

TL;DR

This study investigates the consistency of blood flow velocity (BFV) in the retinal vasculature between a person's left and right eyes, finding strong correlations in healthy individuals. Crucially, it also reveals that individuals with Multiple Sclerosis (MS) exhibit reduced retinal BFV in both eyes, regardless of optic neuritis history, suggesting a potential global microvascular issue in MS.

What Problem Does This Paper Address?

The retina, being an extension of the brain, offers a unique window into the health of the brain's microvasculature. Measuring retinal blood flow velocity (BFV) using techniques like the Retinal Function Imager (RFI) is a promising, non-invasive way to assess this. However, before using RFI-derived BFV in clinical studies, especially those comparing conditions like Multiple Sclerosis (MS) to healthy controls, it's essential to understand the inherent variability. Specifically, this paper addresses two key questions: 1) How consistent are BFV measurements between the left and right eyes of healthy individuals? This is critical for determining if one eye can be used as a reliable control for the other, or if eye selection needs careful consideration. 2) Do individuals with MS show differences in retinal BFV compared to healthy controls, and if so, does this differ based on a history of optic neuritis (ON)? Understanding these differences could shed light on the role of vascular dysfunction in MS pathology.

Methods

The researchers utilized a Retinal Function Imager (RFI), an FDA-approved device that measures blood flow velocity in retinal vessels by tracking red blood cells. The RFI captures images of the retina with a 35-degree field of view. For analysis, secondary and tertiary vessel segments were manually outlined, and proprietary software automatically calculated BFV.

  • Healthy Controls (HCs): Both the right (OD) and left (OS) eyes of 20 HCs were imaged.
  • Multiple Sclerosis (MS) Patients: One eye from each of 38 MS patients was imaged. These patients were further categorized into those with a history of optic neuritis (MSON) and those without (MSNON).

Statistical analyses included paired t-tests and Pearson correlations to assess inter-eye consistency in HCs, and mixed-effects linear regression to compare BFV between groups (HCs vs. MS, MSON vs. MSNON), adjusting for potential confounding factors like age and sex.

Datasets

The study used data collected from:

  • 20 healthy control participants.
  • 38 patients diagnosed with Multiple Sclerosis, further divided into:
  • 13 patients with a history of optic neuritis (MSON).
  • 25 patients without a history of optic neuritis (MSNON).

No new datasets or benchmarks were introduced; the study utilized existing RFI technology and patient cohorts.

Key Insights and Ideas

  • Remarkable Inter-Eye Consistency in Healthy Individuals: The study found that blood flow velocities in the retinal arterioles and venules are highly similar and strongly correlated between the left and right eyes of healthy individuals. This is a crucial finding for future research, suggesting that either eye can be reliably used for comparative studies.
  • Global Reduction in Retinal BFV in MS: A significant finding is that individuals with MS, regardless of whether they have experienced optic neuritis, show reduced arteriolar and venular blood flow velocities compared to healthy controls. This points towards a potential widespread issue with microvascular function in MS, not just localized to areas of optic nerve damage.
  • Vascular Health as a Potential MS Biomarker: The consistent reduction in BFV across MS patients suggests that retinal microvascular measurements could serve as a valuable, non-invasive biomarker for assessing disease processes or treatment efficacy in MS.

Key Results

  • Healthy Controls:
    • BFVs in arterioles and venules were not significantly different between OD and OS (p=0.10 for arterioles, p=0.06 for venules).
  • Very strong inter-eye correlations were observed: r=0.84 (p < 0.001) for arterioles and r=0.87 (p < 0.001) for venules.
  • Multiple Sclerosis Patients:
    • Arteriolar BFV in MSNON eyes (3.48 ± 0.88 mm/s) and MSON eyes (3.59 ± 0.69 mm/s) were significantly lower than in HC eyes (p=0.009 and p=0.046, respectively).
  • Venular BFV in MSNON eyes (2.75 ± 0.53 mm/s) and MSON eyes (2.80 ± 0.45 mm/s) were significantly lower than in HC eyes (p=0.005 and p=0.048, respectively).
  • There was no significant difference in BFV between MSON and MSNON eyes (p=0.42 for arterioles, p=0.48 for venules).

Limitations and Future Work

The authors acknowledge several limitations:

  • Small Sample Size for MSON Cohort: The limited number of patients with a history of optic neuritis means that definitive conclusions about differences between MSON and MSNON eyes are difficult to draw and may be underpowered.
  • Single Eye Imaging in MS Patients: For practical reasons (pupil dilation, time), only one eye of MS patients was imaged. While previous studies suggest this is sufficient, it limits direct inter-eye comparisons within the MS group.
  • Manual Vessel Tracing: The RFI measurements were not fully automated, involving manual tracing of vessels, which can introduce some operator variability.
  • FOV Differences: This study used a 35-degree FOV, while some previous studies used a 20-degree FOV, making direct comparisons of correlation strengths challenging.

Future work should focus on:

  • Larger, Longitudinal Studies: To more definitively assess differences between MSON and MSNON eyes and to track changes over time.
  • Exploring Global Microvascular Alterations: Further investigation is needed to confirm if the observed BFV reductions in MS are indeed a global phenomenon.
  • Relating BFV to Other Biomarkers: Investigating the relationship between retinal BFV and established retinal biomarkers of neurodegeneration could provide a more comprehensive understanding of MS pathology.

Summary updated

Abstract

Inter-eye arteriolar and venular BFVs do not differ significantly in HCs and are strongly correlated. Our findings support prior observations that arteriolar and venular BFVs may be reduced in MS eyes. Moreover, this seems to be the case in both MS eyes with and without a history of ON, raising the possibility of global blood flow alterations in MS. Future larger studies are needed to assess differences in BFVs between MSON and MSNON eyes.
BibTeX
@inproceedings{wang2018microvascular,
  author = {Liang Wang and Ohemaa Kwakyi and James Nguyen and Esther Ogbuokiri and Olwen C. Murphy and Natalia Gonzalez Caldito and Laura J. Balcer and Elliot M. Frohman and Teresa C. Frohman and Peter A. Calabresi and Shiv Saidha},
  title = {{Microvascular blood flow velocities measured with a retinal function imager: inter-eye correlations in healthy controls and an exploration in multiple sclerosis}},
  booktitle = {Eye},
  year = {2018},
  doi = {10.1186/s40662-018-0123-0},
}
Figure