MORPHOMETRIC HISTOPATHOLOGY OF THE PHYSIOLOGICALLY FIXED KERATOCONIC CORNEA

Purpose: Clinically, keratoconus is a condition associated with corneal ectasia, thinning, steepening and scarring, and sometimes necessitates transplant surgery. Exactly where this corneal pathology starts, what structures are involved and what is lost in the thinning process is poorly defined in literature. The purpose of this study was to investigate the etiology, pathophysiology and the structural collapse leading to ectasia by utilizing a histopathological approach.

Methods: A total of 15 surgically removed keratoconic corneas and 7 eye bank control corneas were used. The corneas were fixed in 2% glutaraldehyde in 80mM sodium cacodylate (pH 7.4, 320 to 340 mOsm/kg) and prepared for transverse sectioning with light (Olympus BX51 digital) and transmission electron microscopy (Jeol 100C, initially, and the Tecnai G2 12 twin). Serial electron micrographs were taken of the geometric center of normal corneas and the central cone region of keratoconic corneas. Montages were created for full stromal thickness lamellar counts and for assessment of the anterior limiting lamina (ALL)/stromal interface. Lamellar counts adhered to a specific set of criteria, while a morphometric approach was utilized to define the features of the ALL/stromal interface. The central cone and mid-peripheral regions of the keratoconic cornea were evaluated using light microscopy (LM) and larger anatomical and pathological structures were quantified with the aid of NIH image software.

Results: The normal cornea contained 242+4 lamellae, while the keratoconic cornea averaged 319+105 but, if an extremely thinned keratoconic cornea was excluded, the average was 360+22. In the keratoconic cornea, the middle and, to a degree, posterior lamellae had broken up into smaller units. In the normal cornea, the ALL and stromal layers showed some shallow overlapping of <1µm in either direction. The anterior stromal lamellae were delicate and interweaved extensively. In central and peripheral LM assessments of the keratoconic cornea, the ALL was thinned or lost over 60% of the area examined and the thin anterior lamellae were absent. Anterior lamellae in central control cornea appeared, at intervals, to terminate surrounded by dense staining particles. This unreported stromal feature was termed electron dense formation and was not observed in the keratoconic cornea. The epithelial thickness varied greatly across the keratoconic cornea (14-92µm), while the epithelium in normal maintained a uniform thickness. Keratoconic specimens contained densely stained, distorted epithelial cells and in 58%, a grossly thickened basement membrane. Stroma in the normal cornea contained predominantly one cell type, the keratocyte, along with the occasional Schwann cell and neuron, but in keratoconus, other cells were present in areas where apparent stromal and ALL disassembling occurred.

Conclusions: The keratoconic stroma had paradoxically 50% more lamellae than the normal and this was explained by the fragmentation of lamellae – not an actual increase in lamellae, but rather the breaking apart of lamellae, which has not been described elsewhere. The complete loss of anterior interweaving lamellae in the keratoconic corneas examined likely contributes to the thinning of the keratoconic corneas seen clinically. The terminating lamellae in the normal central cornea suggests that stromal lamellae do not span the full width of the cornea, which is in contradiction to existing literature. The present study reported a clear and well-defined epithelial contribution to pathological changes in keratoconus, but this work does not establish which layer - the epithelium or stroma – is affected first. The extensive destruction of ALL reported here is in contrast to previous statements in literature. It appears that the removal of ALL and anterior lamellae is accomplished by stromal cells that may have been recruited from outside the cornea.

This morphometric, ultrastructural study has demonstrated the anterior focus of keratoconus and the epithelial involvement in the disease, all of which is of importance to contact lens practitioners and surgeons. The new knowledge on the normal human cornea, e.g. number of lamellae and anterior structural integration, generated by this work has multiple surgical and clinical applications. The loss and disintegration of lamellar structure in keratoconus appears to be an important factor in provoking ectasia and its accompanying devastating effect on vision.