Outcomes of Anterior Lamellar Keratoplasty

Introduction

Recent advances in surgical techniques have created a paradigm shift, promoting the development of lamellar keratoplasty procedures that aim to selectively replace only damaged tissue. Deep anterior lamellar keratoplasty (DALK) enables the removal and replacement of an abnormal corneal stroma with retention of the patient’s Descemet membrane (DM) and endothelium. By sparing healthy tissue, DALK offers many potential advantages for the treatment of keratoconus, stromal dystrophies, and corneal scars.

Theoretically, DALK should result in minimal trauma to the endothelium and eliminate endothelial rejection as a cause for graft failure. DALK also prevents the rare but potentially severe complications that can occur during the open-sky portion of a penetrating keratoplasty (PK); however, DALK creates a potential space and an interface between the recipient bed and the donor tissue, which has been argued to affect visual outcomes. Excellent outcomes are achieved by PK and therefore serve as the standard against which DALK results should be measured. This chapter reviews the current literature comparing DALK and PK with regard to visual and refractive outcomes, graft survival, and other peri- and postoperative factors.

Visual Outcomes

When visual outcomes are compared, it is important to consider final visual acuity and the length of time it takes to achieve a stable and optimal level of vision. A growing number of studies have compared the visual outcomes of DALK and PK for the treatment of keratoconus. Krumeich et al. found that although DALK led to better vision at 3 months (median PK = 0.5 logMAR, DALK = 0.35 logMAR, P = .001), there was no statistical difference in best corrected visual acuity (BCVA) at any point in time ranging from 6 months to 5 years postoperatively. In this retrospective study, the median BCVA at 1 year was logMAR 0.6 and 0.68 in the DALK and PK groups, respectively. After 5 years of follow-up, the median BCVA was logMAR 0.65 and 0.7, respectively.

Several other retrospective studies comparing DALK and PK have similarly concluded that there was no statistically significant difference in BCVA or contrast sensitivity ( Table 123.1 ). Comparable outcomes have also been demonstrated by randomized, prospective studies analyzing DALK and PK for the treatment of stromal opacities and keratoconus and for macular corneal dystrophy. In a Cochrane review comparing DALK and PK for the treatment of keratoconus, the authors did not find any evidence to support a difference in outcomes with regard to BCVA. However, only two studies with a total of 111 participants met their inclusion criteria and were analyzed in the review. A second review comprising the findings of 11 comparative studies of 481 DALK eyes and 501 PK eyes similarly concluded there was no significant difference in BCVA.

To the contrary, some studies have reported more favorable outcomes following PK. Coster et al. recently reported the findings of the Australian Corneal Graft Registry’s data from 1996 to 2013. Of the 3051 PKs performed for keratoconus, 67% achieved BCVA of 20/40 or better, compared with only 49% of the 317 DALKs performed for the same indication ( P <.001). To better understand these varying results and somewhat contradictory findings, it is important to consider that the term DALK has historically been used to refer to a variety of surgical techniques. The term is often used to describe both techniques that bare the DM as well as techniques that may leave residual recipient stromal tissue, such as manual dissection. Comparative case series suggest that these techniques may not all yield equivalent results; some have reported that the BCVA is inferior with manual dissection compared with PK and big-bubble DALK but that there is no difference between the big-bubble DALK and PK. ^,

Ardjomand et al. reported that eyes with a PK showed significantly better median BCVA than eyes with a DALK (0.023 logMAR vs. 0.196 logMAR, P = .035). However, in a subgroup analysis of eyes that underwent DALK, eyes with less than 20 μm of residual recipient stroma achieved a BCVA that was statistically similar to that of eyes that underwent a PK. They also reported a trend of decreasing BCVA with increasing residual stromal bed, but it did not reach statistical significance. DALK eyes with a recipient stromal thickness greater than 80 μm had a mean BCVA of 0.39 logMAR, which was significantly worse than that of the PK group ( P = .003).

The effect of the residual posterior stromal bed has been further demonstrated by Alio. Following DALK, Alio performed confocal microscopy on four eyes with suboptimal vision (mean BCVA 20/80). The posterior stromal thickness was found to range between 57 and 83 μm. The donor graft was lifted, a big-bubble technique was used to bare the DM, and the corneal graft was replaced and sutured into place. Postoperative vision improved in all eyes, reaching a mean BCVA of 20/25.

Finally, a review by Reinhart et al. concluded that there was a tendency for lower visual acuity in DALK eyes in which the DM was not bared, resulting in a residual stromal bed of greater than 10% of the total stromal thickness. These findings illustrate an important relationship between dissection depth and visual outcomes following DALK and suggest that the residual stromal bed retained by the recipient may significantly influence the final visual outcome.