Changes in the posterior corneal surface after LASIK

In this era of increasing concern over post laser in situ keratomileusis (LASIK) keratectasia, we read with interest the article by Lee et al.¹ on apparent changes in the posterior corneal surface after LASIK. With corneal residual bed thickness generally implicated in the etiology of iatrogenic keratectasia, the authors' use of intraoperative ultrasound for direct measurement of residual bed thickness is to be highly commended, as is their novel investigation into the relationship between posterior corneal elevation and the ablation ratio per total corneal thickness. However, we have some concerns about the results and conclusions of their study. Although Orbscan^® (Bausch & Lomb) is in widespread clinical use and has been shown to be highly accurate for anterior surface elevation measurement,² its accuracy and reproducibility in assessing posterior surface elevation has yet to be scientifically validated. Indeed, Orbscan pachymetry measurements are known to be significantly higher than ultrasound preoperatively and significantly lower (with greater variability) post LASIK³; therefore, it is not unreasonable to conjecture that postoperative measurement of posterior elevation is similarly altered.

As in similar studies,^4,5 Lee et al. used the peripheral 7.0 to 10.0 mm fit zone for the preoperative and postoperative best-fit spheres (BFS) on the logical assumption that this removes central bias by restricting the BFS fit zone to that area presumed to be unchanged by the ablation. However, difference maps produced by Orbscan II are identical regardless of the fit zone used (full corneal fit—0 to 10.0 mm; peripheral fit—7.0 to 10.0 mm; or center fit—0 to 7.0 mm), the alignment used (floating, apex, axial, or pinned), or the BFS calculation option chosen (recalculated using defaults, last user selected sphere, or same calculated sphere).

Although simple regression analysis showed that the change in posterior corneal elevation, as a function of change in residual bed thickness, was statistically significant, the R² value was low at 0.293, indicating that residual bed thickness accounts for only 29.3% of the variation in outcome and 77.7% of the variation in the outcome is therefore accounted for by unknown predictors. Similarly, the ablation ratio accounts for only 22.3% of the variation in outcome. Application of multiple linear regression techniques may reveal other significant predictors.

Although Lee et al. demonstrated a significant change in “forward shift” of the posterior corneal surface relative to the residual bed thickness within Group 1 (<250 μm), Group 2 (250 to 300 μm), and Group 3 (300 to 350 μm), they did not report significant differences between these groups. Similarly, no significant changes were reported between ablation ratio groups. It may therefore be erroneous to conclude that increased forward shift of the posterior corneal surface correlates with residual corneal thickness. Finally, Figure 1 provides a scale from 300 to 600 μm for the residual bed thickness, which would seem a highly unlikely range post LASIK.

While we welcome such detailed analysis of changes in posterior corneal elevation post LASIK, we would certainly advise caution in drawing firm conclusions based on these Orbscan data.

Susan Ormonde FRCOPHTH

Christine Waterman MSC

Charles Mcghee PHD, FRCOPHTH

^aManchester, United Kingdom

^bAuckland, New Zealand