In vivo biomechanical mapping of normal and keratoconic corneas

INTRODUCTION

Corneal mechanical strength is critical to withstanding intraocular pressure and maintaining normal shape^1,2. In keratoconus, the mechanical stability is compromised³, and this may lead to progressive morphological changes. Therefore, a noninvasive technique capable of accurately measuring the mechanical properties of the cornea may help understand the mechanism of keratoconus development and improve detection and intervention in keratoconus. We have previously developed Brillouin microscopy based on light scattering from inherent acoustic waves in tissues⁴ and shown that this technique can provide quantitative estimates of local longitudinal modulus⁵, which correlate to the Young's/shear moduli of the cornea^2,6. Using a clinically viable instrument, for the first time to our knowledge, here we mapped the elastic modulus of normal and keratoconus patients in vivo. We found distinctive biomechanical features that differentiate normal and keratoconic corneas and, therefore, have potential to serve as diagnostic metrics for keratoconus.

METHODS

The study recruited six healthy volunteers (Age=37±15) and five patients with advanced keratoconus (Age=43±7). All participants signed an informed consent form approved by the Partners Human Research Committee (IRB), in accordance with the principles embodied in the Declaration of Helsinki. We constructed a laser-scanning confocal Brillouin microscope (wavelength: 780 nm, power: 1.5 mW, lateral/axial resolution: 5 μm/30 μm, sensitivity ~10 MHz). The instrument was equipped with wide field-of-view imaging to allow real-time pupil detection and beam positioning (lateral accuracy <0.5 mm). For healthy subjects, about 5 mm by 5 mm areas in the central region of the cornea were scanned. For keratoconus patients, similar regions, but including the center of the cone, were scanned as confirmed by their tomography images (Pentacam, Oculus). To construct Brillouin maps, axial scans were taken at various transverse locations; the anterior mean Brillouin shift was computed from each axial scan by averaging the measured Brillouin shift values over the anterior portion of the corneal stroma; a color-coded elasticity map was obtained by two-dimensional interpolation of mean Brillouin shift in the anterior portion.

RESULTS

Normal corneas were found to have relatively uniform anterior Brillouin shifts in the central region (Fig. 1a). By contrast, keratoconic corneas presented strong spatial variations of Brillouin shifts (Fig. 1b). Figure 2 shows the average anterior Brillouin shifts of normal corneas (N=7) and of keratoconus corneas (N=6) in the cone region (<1 mm from thinnest point) and outside the cone region (> 3 mm away from thinnest point). A highly statistically significant decrease (unpaired t-test, p<0.005) was found in the keratoconic cone region with respect to normal corneas. Also, a highly statistically significant difference (unpaired t-test, p<0.005) was observed between the cone and outside-cone regions. The regions outside the cone showed no statistically significant difference compared to the normal corneas.

Fig. 1. Brillouin elasticity maps.

Representative maps of the mean anterior Brillouin shift for (a) a 53-year-old healthy subject and (b) a 40-year-old patient with advanced keratoconus. Insets are the respective curvature and pachymetry maps with outlined Brillouin scanned areas.

Fig. 2. Focal weakening in keratoconus.

The mean Brillouin shifts of the keratoconic corneas (N=6) in the cone region versus outside-the-cone region, in comparison to average normal cornea values (N=7). ***, P <0.005; n.s., not statistically significant difference.

DISCUSSION

We have reported the distribution of elastic modulus within keratoconic and normal corneas in vivo. The elasticity maps show remarkable spatial variations around the cone. The reduction of 100 MHz in the keratoconic cone region (Fig. 2) corresponds to a ~3% decrease in longitudinal modulus and about ~70% reduction in shear modulus⁵. The regions away from the cone in the keratoconus corneas have similar Brillouin shifts as normal corneas, which is consistent with our ex vivo data⁵. This finding supports the long-standing hypothesis that keratoconus involves a spatially-localized mechanical alteration in the cornea¹. Importantly, it emphasizes the need for spatially resolved measurements for accurate analysis of the biomechanical anomalies in keratoconus. Future research is warranted to understand the relationship between the focal or heterogeneous mechanical weakening and morphological changes (i.e. thinning and steepening) and to develop biomechanics-based metrics for improved diagnosis and prognosis of keratoconus, screening of at-risk subjects for post-LASIK ectasia, and monitoring the effects of corneal collagen crosslinking.