Safety and predictability
The safety analysis showed 32% of SMILE eyes lost 1 line from preoperative BCVA. This finding was expected because of the minification effect of the surgery and similar to the result in the LASIK group where 25% of eyes had a 1-line loss of BCVA, Dr. Reinstein said.
The predictability analysis showed that SEQ was within ±0.5 D of attempted in 65% of SMILE eyes and in 53% of LASIK eyes. In the regression analysis, the r2 value was higher for SMILE than for LASIK, 0.59 versus 0.48, he noted.
“This is interesting, because we would expect to see tighter correlation in an analysis including more eyes,” Dr. Reinstein said. “Therefore, based on this limited comparison, it seems there is less scatter of results with SMILE treatment for higher hyperopia.”
Stability of the SMILE correction was good during the short follow-up. Mean change in SEQ from 1 to 3 months was about 0.1 D in both the SMILE and LASIK groups.
Other analyses compared optical zone centration, topographic optical zone diameter, and spherical aberration induction in a cohort of 60 consecutive SMILE eyes and two LASIK control groups treated at 6.5- and 7-mm optical zones, respectively, and matched 1:1 by treated SEQ [Reinstein DZ, et al. J Refract Surg. 2017;33:150-156. Reinstein DZ, et al. J Refract Surg. 2017;33:370-376].
All eyes had BCVA of 20/40 or worse and the mean optical zone for the SMILE group was 6.37 mm.
The results showed optical zone centration was similar for SMILE and LASIK. Mean topographic optical zone diameter was significantly larger for SMILE compared with both LASIK groups, and mean induced spherical aberration in the SMILE eyes was similar to the mean change in the 7-mm LASIK group and significantly less than after 6.5-mm LASIK.
“The explanation for the optical zone finding is that with the excimer laser procedure, there are projection errors in the periphery of the cornea that are never fully compensated,” Dr. Reinstein said. “SMILE cuts precisely at the location geometrically calculated by the software and gives a truer-sized transition zone. Knowing that transition zone size is key to regression stability and treatment accuracy might also be the explanation for why we saw less scatter in the 3-month predictability outcomes with SMILE.”