Determining the Efficacy of Corneal Cross-Linking Protocols using Brillouin Microscopy

ABSTRACT: Keratoconus and related corneal ectatic diseases cause significantly decreased quality of life, are the leading cause for full thickness corneal transplant in the US, and are significantly more prevalent than previously thought. Corneal cross-linking (CXL) has emerged as a clinical technique to halt keratoconus progression by stiffening the corneal stroma. Despite being used clinically for more than a decade, it is currently impossible to assess CXL protocol efficacy or predict the long-term stability and because we lack quantitative biomechanical measures to inform predictive models. Currently available metrics to characterize CXL responses are morphologic and have not proven predictive of clinical outcomes. The major gap is the lack of measurement techniques that can accurately and non-perturbatively characterize corneal mechanics with three-dimensional (3-D) resolution in vivo. To address this need, in the past decade we have pioneered Brillouin microscopy, a high-resolution optical technology that can measure corneal longitudinal modulus in situ in 3-D without contacting or perturbing the eye. Brillouin microscopy has provided the first and only direct mechanical evidence of decreased modulus in keratoconus corneas in vivo and the only 3-D maps of CXL-induced corneal stiffening. The overall goal of this research program is to combine 3-D Brillouin corneal maps and finite element (FE) modeling to quantitatively predict corneal shape outcomes after CXL protocols. In strong preliminary data, we demonstrated that, by accounting for tissue hydration, we can establish the quantitative relationship between Brillouin-measured longitudinal modulus and Young’s modulus. Thus, our central hypothesis is that spatial maps of corneal Young’s modulus derived from quantitative Brillouin microscopy will enable accurate prediction of corneal shape behavior via FE modeling. The development of this noninvasive measure of corneal stiffness also enables us to use a rabbit model to evaluate, for the first time, both morphologic and mechanical evolution in longitudinal studies in vivo, validated by direct mechanical analysis using experimental protocols that cannot be performed in human subjects. We will test our central hypothesis through the three specific aims: 1) Validate in vivo Brillouin mechanical measurements after CXL; 2) Quantify the in vivo mechanical outcomes of novel CXL protocols; and 3) Link CXL biomechanical impact to morphologic outcome with Brillouin imaging and FE modeling. This research is significant because accurate nondestructive, nonperturbative elasticity-based metrics will drive a paradigm shift in how CXL protocols are evaluated, developed, and performed clinically as well as ultimately allow us to develop individualized CXL treatment protocols.