Interface haze has been described to occur frequently after lamellar keratoplasty. Opacification at the donor-to-recipient interface may be one of the major drawbacks of a lamellar procedure, since it may result in reduced best corrected visual acuity and contrast sensitivity. To my knowledge, no histological reports are available to define the cause(s) of interface haze after lamellar keratoplasty.

In laser in-situ keratomileusis (LASIK) the lamellar interface usually shows minimal interface opacification, if any. Also, in automated lamellar keratoplasty (ALK), haze formation is most often minimal at the interface between the donor and recipient tissues. These observations suggest that the presence of a lamellar interface by itself may not cause interface opacification, but that the method of dissection is decisive. Apparently, microkeratome dissection produces far less amounts of interface opacification in comparison to manual stromal dissection, probably because a smoother cut is obtained by the oscillating blade of a microkeratome than by manual dissection.

Textbooks on lamellar keratoplasty advocate to dissect the recipient bed by lifting the anterior corneal lamella upward, thereby stretching the interlamellar adhesions and cutting the adhesions with a various dissection blades. Blunt spatula dissection is most often advocated for preparation of the donor tissue. To evaluate the smoothness of the stromal surfaces with these dissection techniques, blunt corneal dissection and sharp dissection while lifting the anterior lamella was performed in a human eye bank eye model. The histologic appearance of the dissections was compared to that of a microkeratome and to sharp corneal dissection in a lamellar plane (without lifting the anterior lamella).

With light microscopy, blunt dissections showed abundant torn stromal lamellae, with distortion of the parallel orientation of the stromal lamellae. Disrupted lamellae and an irregular stromal surface were also seen following sharp dissection while lifting the anterior corneal lamella. Sharp dissection in a lamellar plane better approached the smoothness of a microkeratome dissection.

By using dissection techniques commonly advocated in textbooks, a relatively irregular stromal surface may be produced in both the donor and recipient tissues. As a result, apposition of these stromal surfaces may distort the layered structure of the stroma, and subsequent scarring at the interface may add to the opacification seen clinically.

Absence of clinically significant interface opacification, i.e. smoother interfaces in deep lamellar keratoplasty may be obtained by a combination of three factors. First, a custom made spatula was designed with ultrasharp edges, to create smooth interface in the recipient. Second, an intended dissection depth in the recipient at 95% of the stromal thickness may give a smoother interface. Third, an intended dissection depth in the donor at 95% of the stromal thickness may give a relatively smooth interface.