Abstract
Purpose:
A priori prediction of post-refractive surgery ectasia risk remains a challenge. Current clinical screening paradigms rely on incomplete corneal shape characterizations and surrogate surgical risk factors. We present the first patient-specific computational analyses of clinically documented post-PRK and post-LASIK ectasia cases and assess differences in load-induced stress/strain in affected and unaffected eyes.
Methods:
Preoperative and postoperative tomography data from ectatic and unaffected contralateral eyes were imported into custom finite element meshing software. Epithelium, flap, wound and residual stromal bed (RSB) layers were each defined in the LASIK models. The PRK model consisted only of epithelium and RSB layers. Stress/strain distributions were obtained using an iterative method. Munnerlyn ablation algorithms were implemented in simulations. The cornea was modeled as a fiber-reinforced material with homogenous solid matrix. At each integration point within the model, splay of fibers was modeled and angularly integrated. Each fiber was represented by a 3D helical spring in order to capture crimping behavior of collagen fibrils. In addition to actual preop, actual postop and simulated postop simulations, additional simulations modeled focal reductions in fiber or matrix modulus. All simulations were performed with 15mmHg and 30 mmHg loads.
Results:
In the eye that developed post-LASIK ectasia, maximum principal strain was 10% higher and more asymmetrically distributed than the stable eye. Simulated LASIK procedures closely matched actual postop geometries and produced 10% higher von Mises stresses in the ectatic eye with a more asymmetric, eccentric distribution than the stable eye. For the PRK case, similar but smaller differences in strains and von Mises stresses were observed in the ectatic eye with shifts in von Mises stress toward the inferior-temporal cornea where the ectatic region manifested clinically. Similar shifts were observed in models based on actual postoperative geometry. In both cases, when corneal shear modulus was reduced, the cornea thinned slightly and the steep feature shifted peripherally.
Conclusions:
Structural simulations using patient-specific geometry and a microstructurally motivated fiber-reinforced model reveal potential disease-predisposing differences in case-specific mechanical behavior that may be useful for prediction of post-refractive surgery ectasia.