High-resolution biomechanical mapping of SMILE and SMILE with CXL using Brillouin microscopy: Insights into localized...

To evaluate the biomechanical effects of small incision lenticule extraction (SMILE) alone and with corneal collagen cross-linking (SMILECXL) on corneal stiffness, using the Brillouin Optical Scanning System (BOSS) to assess localized changes in Brillouin modulus (BM).

This retrospective study analyzed an initial cohort of 358 eyes (321 SMILE, 37 SMILECXL) from 180 patients, with propensity score matching (PSM) yielding matched cohorts of 17 eyes each, including unilateral and bilateral cases. PSM balanced baseline characteristics using 10 variables: age, gender, lenticular thickness, residual stromal thickness, spherical power, cylinder, spherical equivalent, flat keratometry, steep keratometry, and mean keratometry. To address the violation of independence in the between-group comparison of pre- and postoperative ∆BM due to the inclusion of bilateral eyes from some participants, generalized estimating equations (GEE) with an exchangeable correlation structure were used to account for within-person correlation in clustered data from bilateral eyes. BOSS (Intelon Optics, Woburn, MA, USA) measured BM via Brillouin frequency shifts (gigapascals, GPa) at 10 predefined corneal points (central, superior, inferior, nasal, temporal; coordinates -2.4 mm to +2.4 mm from apex) preoperatively and at 1 and 3 months postoperatively. Post-PSM, BM values and changes (ΔBM) were compared, including point-specific, mean, minimum, and maximum analyses. ΔBM patterns (≥ 0 for preservation/increase, p < 0.05 for inter-group significance) were categorized for regional insights.

Pre-PSM differences in age, lenticular thickness, residual stromal thickness, and refractive parameters (e.g., spherical equivalent) necessitated matching; post-PSM, all were balanced (p > 0.05). Preoperative BM values were comparable. Post-SMILE, stiffness decreased (mean ΔBM -0.021 GPa at 1 month, -0.009 GPa at 3 months), with partial recovery but 80% points below baseline; BM4/BM9 spontaneously recovered (ΔBM ≥ 0 at 3 months). SMILECXL preserved/enhanced stiffness across all points (mean ΔBM +0.015 GPa at 1 month, +0.011 GPa at 3 months), with significant differences at minimum ΔBM (p < 0.05 at 1 month) and points BM1, BM4, BM5 (p < 0.05 at 1 month), BM1 (p < 0.05 at 3 months). Four patterns emerged: (1) Resistant but potentially vulnerable (BM7/BM10: near-zero initial loss but worsening, inferior progressive loss); (2) Spontaneously and fully recovered (BM4/BM9: initial loss but recovery ≥ 0 at 3 months without CXL, superior resilience); (3) Moderately effective for CXL (BM2/BM3/BM6/BM8: loss with non-significant preservation, peripheral benefits); (4) Highly effective for CXL (BM1/BM5: pronounced loss but significant preservation [p < 0.05], nasal-temporal central protection).

BOSS's high-resolution spatiotemporal analysis reveals that CXL counters SMILE-induced stiffness loss, achieving universal preservation or enhancement across corneal zones (ΔBM ≥ 0 at all points and timepoints) and significant protection in vulnerable nasal-temporal central areas (p < 0.05). These patterns-inherent superior resilience, inferior susceptibility, zone-specific CXL efficacy-suggest BOSS's potential utility in mapping localized biomechanical changes, offering insights for monitoring postoperative stiffness in refractive procedures.