Ivan Gabrić, Filip Grulović, Ivan Samuel Modrić, Nataša Drača, Tomislav Kuzman
In this large single-surgeon series, pooling data from the original SmartSight workflow and the SmartSight NOVA, KLEx on the SCHWIND ATOS delivered accurate, safe, and stable 12-month outcomes with excellent uncorrected vision and only modest higher-order aberration induction. As a single-surgeon experience using an evolving offset and two treatment workflows that are temporally confounded with each other and with surgical experience, these findings describe a descriptive association between the offset strategy and favorable outcomes rather than validation of a fixed nomogram or of SmartSight NOVA specifically; the current + 0.75 D offset was not represented at one year and requires prospective, workflow-controlled validation.
INTRODUCTION: To report the visual, refractive, astigmatic, and ocular higher-order aberration (HOA) outcomes of keratorefractive lenticule extraction (KLEx) with the SCHWIND ATOS femtosecond laser platform in a large single-surgeon cohort.
METHODS: In this retrospective consecutive series, 2518 eyes of 1292 patients underwent single-surgeon KLEx (SmartSight procedure, SCHWIND ATOS platform) for myopia or myopic astigmatism. Data from two treatment workflows were pooled: the original SmartSight lenticule workflow (1886 eyes) and, from 2025, SmartSight NOVA with CenTrax-assisted centration (632 eyes); workflow-stratified outcomes are reported separately in Table 4. Refraction, uncorrected and corrected distance visual acuity ( (UDVA and CDVA, respectively), and pyramidal-wavefront aberrometry (OSIRIS-T corneal topographer and ocular aberrometer; 4 and 6 mm) were recorded for 12 months of follow-up (primary; 3 months secondary). The target was emmetropia; astigmatism used the Alpins vector method. Denominators were endpoint specific, and inference was clustered by patient.
RESULTS: Mean preoperative spherical equivalent (SEQ) was - 3.86 ± 1.93 diopters (D). At 12 months post-procedure, in 830 eyes with a complete acuity record and 930 with paired SEQ, mean SEQ was - 0.04 ± 0.31 D, with 92.9% within ± 0.50 D and 98.2% within ± 1.00 D; achieved correction tracked attempted (slope 0.98). UDVA was 20/20 or better in 94.9% of eyes, and 94% saw uncorrected at least as well as preoperatively corrected. One of 830 eyes lost exactly two CDVA lines; none lost more than two. Astigmatic outcomes are documentation dependent: among eyes with an explicitly recorded postoperative cylinder, 68.6% were within 0.50 D, whereas the whole-cohort estimate (89.9%) rests on the unvalidated convention that a blank cylinder field denotes plano, and postoperative axis was not captured, so vector indices are exploratory. Paired mean SEQ changes from month 3 through month 12 were < 0.05 D. An enhancement-as-failure composite (any enhancement before 12 months counted as failure) was met by 89.6% of eyes. In a paired wavefront subset (412 eyes), HOA induction was modest and coma-dominated: total ocular HOA rose from 0.14 to 0.18 µm at 4 mm (P < 0.001) and was not significantly changed at 6 mm (P = 0.10). Preoperative spherical aberration was positive; ocular spherical aberration decreased modestly while total corneal spherical aberration was preserved at 12 months (P = 0.59).
CONCLUSION: In this large single-surgeon series, pooling data from the original SmartSight workflow and the SmartSight NOVA, KLEx on the SCHWIND ATOS delivered accurate, safe, and stable 12-month outcomes with excellent uncorrected vision and only modest higher-order aberration induction. As a single-surgeon experience using an evolving offset and two treatment workflows that are temporally confounded with each other and with surgical experience, these findings describe a descriptive association between the offset strategy and favorable outcomes rather than validation of a fixed nomogram or of SmartSight NOVA specifically; the current + 0.75 D offset was not represented at one year and requires prospective, workflow-controlled validation.