Introduction

Modern refractive corneal surgery has seen a multitude of next generation laser assisted lenticule extraction (LALEX) procedures in the last decade. These procedures differ in the shape of the lenticule being created in the stroma of the cornea1. The different LALEX procedures currently in the market are small incision lenticule extraction (SMILE, Carl Zeiss AG, Germany), SmartSight (Schwind eye-tech-solutions, Germany) and CLEAR (Z8, Ziemer, Switzerland)2,3,4. Currently, these procedures are used only for myopic spherical and cylindrical corrections in patients. Since LALEX only involves a small incision in the stroma, this procedure can reduce epithelial defects, postoperative pain and dryness, and possibly lead to faster wound healing and greater functional optical zones compared to other refractive procedures5,6,7,8,9.

ELITA™ is the latest new generation ophthalmic femtosecond laser surgical system developed by Johnson and Johnson Surgical Vision, Inc. (Milpitas CA, USA), which received CE Mark approval in March 2023 for Smooth Incision Lenticular Keratomileusis (SILK™) and LASIK flap creation10. Salient features of the ELITA system are low energy of the laser pulse (40–90 nJ per pulse) through a combination of small focus spot size (approximately 1 μm), ultrashort pulse duration (100–200 fs), and ultrafast pulse frequency (10 MHz). ELITA creates virtually no pulse separation (mean spot to spot is about 1 μm), provides continuous tissue resection with minimal residual tissue bridges, and therefore enables easier lenticule manipulation for tissue removal by the surgeon. ELITA uses a biconvex lenticule shape, where the anterior and posterior surfaces are matched, so that corneal micro-folding is minimized or eliminated following lenticule removal.

LALEX procedures use custom lenticule shapes to induce refractive correction. While SILK uses a biconvex shape, SMILE uses a plano convex shaped lenticule to induce refractive correction. Therefore, some differences in postoperative corneal shape and resultant ocular aberrations may be expected. In a recent study, simultaneous assessment of ocular aberrations and depth of field at different reading distances (20 ft, 60 cm and 40 cm) was performed11. The study observed that there was a slight amount of postoperative cylinder in excess of that observed in LASIK patients postoperatively though both LASIK and SMILE patients had similar visual acuity at near (40 cm) and intermediate (60 cm) distance11. A recent study reported excellent visual, safety and refractive outcomes after SILK procedure for correction of myopia with and without astigmatism12. The objective of this study was to assess the depth of field of those patients postoperatively via simultaneous measurements of ocular aberrations and visual acuity at distance, near and intermediate distances. Further, the data from these patients was compared with reported data from wavefront optimized LASIK in our earlier study11.

Methods

This was an observational and cross-sectional study approved by the ethics committee of Narayana Nethralaya Eye Hospital, Bangalore, India (Approval No: ECR/187/Inst/Kar/2013/RR-19). The study adhered to the Declaration of Helsinki. These patients were part of a clinical trial registered on Clinical Trials registry of India (CTRI/2018/12/016733)12. Written informed consent was obtained from the patients prior to SILK surgery. The study included 69 eyes that underwent SILK surgery using the first nomogram developed for SILK. Either one or both eyes of the patient were treated on the same day based on the correction needed. The inclusion criteria were age 18 years or older, myopic spherical equivalent (SE) refractive error of less than or equal to − 12.00 diopters (D) and astigmatism less than or equal to − 6.00 D, calculated residual corneal stromal thickness greater than or equal to 250 microns, uncorrected distance visual acuity (UDVA) of 20/40 or worse, corrected distance visual acuity (CDVA) of 20/20 or better and CDVA better than UDVA by at least 2 lines. Further inclusion criteria included difference between cycloplegic and manifest refraction sphere less than or equal to 0.75 D, a stable refractive error for at least 12 months, and at least 4 weeks (rigid) or 2 weeks (soft) off contact lens use prior to surgery. Exclusion criteria included a history of medical conditions or systemic medications that may impair wound healing, diabetes mellitus, implanted electronic device, pacemaker or defibrillator, prior intraocular or corneal surgery, active ophthalmic disease or abnormality including clinically significant dry eye, glaucoma, keratoconus, corneal dystrophy or degeneration, pregnancy, lactating or planning pregnancy, and known sensitivity to any of the medications used in postoperative care.

Surgical technique

All surgeries were performed by a single surgeon (RS). SILK was performed with the new ELITA femtosecond laser system. The lenticule optical zone diameter was 6 mm with a 3 mm entry incision at the superior position. The lenticule depth at the anterior apex was 110 μm. The ELITA femtosecond laser system was designed for ultralow pulse energy operation in the range of 40–90 nJ. In this study, the pulse energies used for the posterior lenticular cut, ring cut (lenticule edge cut), anterior lenticular cut, and entry incision were between 65 and 67 nJ, 64 nJ, 67–69 nJ, and 72–74 nJ, respectively. The nomogram for sphere adjustment was given by -0.25 D + 5.75% x manifest sphere and the nomogram for cylinder adjustment was zero regardless of preoperative manifest cylinder (i.e., the cylinder treated was equal to the preoperative manifest cylinder in all cases). The laser treatment time for each eye did not exceed 16 s. All SILK procedures were performed as per treatment instructions provided by the ELITA system and were targeted for emmetropia. Following the laser treatment, the biconvex shaped lenticule was removed via the 3 mm entry incision.

Topical antibiotics, corticosteroids and lubricating eye drops were prescribed after surgery. Further details of SILK surgical procedure are described in our earlier publication11. Postoperatively at 6-months, each eye underwent ocular aberrometry with iTrace aberrometer (Tracey Technologies, Houston, TX) under monocular conditions after routine clinical examinations. The iTrace has an ocular aberrometer which measures the wavefront aberrations of the eye and uses Zernike coefficients to quantify the aberrations map. The device also has a Placido topographer which measures the curvature of the anterior surface of the cornea. The topographer also calculated the corneal aberrations. Subtraction of the corneal aberrations from the ocular aberrations yielded the internal aberrations of the eye (lens and media). For each eye, 3 reading tests were performed by placing a visual acuity chart at distance (20 ft); intermediate (60 cm) and near (40 cm) in front of the aberrometer. The Placido disk of the iTrace aberrometer had a hole in the center. While the ocular aberrations were measured by the device, the patient read the visual acuity chart (mounted on a rod) through the hole at the center of the Placido disk. The chart was kept at an appropriate distance (20 ft, 60 and 40 cm). Thus, the patient’s eye was fixed on the letters of the chart and a stable aberrometry measurement was ensured at each distance.

While one eye was being tested, the other eye was covered with a patch and the internal fixation light of the device was turned off. Further details on the measurement procedure can be found in our previous work11. At all reading distances, the luminance was 100 lx approximately. These measurements were compared with the depth of field data of eyes that underwent LASIK in our clinic11. All the LASIK eyes were treated with a wavefront optimized profile (WaveLight FS200 femtosecond laser and WaveLight EX500 excimer laser, Alcon Laboratories, Inc., Fort Worth, TX). The flap had a diameter of 9 mm and thickness of 110 μm. The optical zone was 6 mm. Additional details of the LASIK procedure can be found in our previous work11.

Data extraction and statistical analyses

The spherical error (D), cylindrical error (D) and total eye aberrations (µm) measured by the device were re-calculated for further analyses. The recalculation was done so that the same pupil diameter was used for analyses at all reading distances for a specific eye11. This was done by choosing the minimum pupil size measured for the 3 reading tests: distance, intermediate and near for each eye. This ensured that the pupil diameter was not a confounder while assessing the Zernike polynomial coefficients at the 3 reading distances for a given eye. The device by default reported Zernike coefficients up to 6th order. Only measurements that met the in-built quality standards of the iTrace aberrometer were used for statistical analyses.

Normality of the data was assessed using Kolmogorov-Smirnov test. All parameters were represented as median and 95% confidence interval as some parameters were not normally distributed. The visual acuity at all distances were represented in LogMAR units. Among the ocular aberrations, the indices of interest were defocus (µm), root mean square of lower order aberrations (LORMS; µm), root mean square of coma (Coma RMS; µm), root mean square of higher order aberrations (HORMS; µm) and primary spherical aberration (SA; µm). The non-parametric Kruskal-Wallis test with post-hoc assumption of multiple comparison (with Dunn’s correction) was used to determine statistical significance between the groups. A P-value less than 0.05 was considered statistically significant. MedCalc version 20.2.15 (MedCalc Software Ltd, Ostend, Belgium) software was used for statistical analyses.

Results

A total of 214 eyes (50 emmetropes, 95 LASIK and 69 SILK eyes) were assessed in the study. The emmetropic and LASIK eyes were from an earlier study11. This sample size was adequate based on an earlier study using the same methods11. Table 1 shows the preoperative spherical and cylindrical errors between the groups. The median preoperative spherical error was statistically different between the SILK and LASIK eyes (P < 0.01). However, median preoperative cylindrical error was similar between the 2 groups (P = 0.16). Similarly, preoperative CDVA was also comparable between the groups (P = 0.14). Figure 1; Table 1 show the median postoperative spherical error (A), cylindrical error (B) and visual acuity (C) of the 3 groups at all reading distances. At distance, the SILK group had zero spherical error, and this was significantly different with respect to emmetropes (P < 0.001) and LASIK eyes (p < 0.01). At intermediate, all groups showed similar spherical error (P > 0.05). At near, the spherical error in SILK eyes was similar to LASIK eyes (P > 0.05), However, both groups differed significantly from emmetropes (P < 0.01). Furthermore, cylindrical error (Fig. 1-B) at distance was significantly higher in SILK eyes than LASIK eyes (P < 0.01) and emmetropes (P < 0.05). A similar trend was observed at intermediate and near reading distances. Visual acuity (Fig. 1-C) at distance and near reading distances in the SILK eyes was equal in magnitude as compared to the LASIK eyes but statistically different (P < 0.01). At intermediate accommodation, SILK eyes showed similar visual acuity as compared to emmetropes (P > 0.05) but were better than LASIK eyes (P < 0.001) (Table 2).

Table 1 Median and 95% confidence interval of pre-operative (distance measurement only) and post-operative refractive errors and visual acuity at distance (20 ft), intermediate (60 cm) and near (40 cm) accommodation of the 3 groups: Emmetropes (n = 50); SILK (n = 69) and LASIK (n = 95).
Full size table
Fig. 1
figure 1

Post-operative sphere (A), cylinder (B) and uncorrected visual acuity (C) at distance, intermediate and near reading distances.

Full size image
Table 2 Median and 95% confidence interval of post-operative ocular aberrations at distance (20 ft), intermediate (60 cm) and near (40 cm) accommodation of the 3 groups: Emmetropes (n = 50); SILK (n = 69) and LASIK (n = 95).
Full size table

Figure 2 shows the distribution of postoperative spherical error at all reading distances (A–C) between the 3 groups. At distance (Fig. 2-A), the proportion of eyes having spherical error greater than − 0.5D (light green columns) was higher in the SILK group but was statistically similar between SILK eyes and emmetropes (P = 0.8). Both emmetrope and SILK eyes fared better than the LASIK eyes (P < 0.001). This trend was similar in the proportion of eyes having spherical error between − 0.5D and − 1D (purple columns) where SILK eyes performed better than LASIK eyes (P < 0.001). The proportion of eyes having spherical error greater than − 1D (yellow columns) was smaller in the SILK group as compared to LASIK eyes but was statistically insignificant (P > 0.05). At intermediate (Fig. 2-B), the proportion of eyes having spherical error below − 1D was statistically similar to the LASIK eyes (P > 0.05). However, the proportion of eyes having residual spherical error higher than − 1D (yellow columns) was higher in SILK eyes as compared to emmetropes and was marginally similar (P = 0.05). At near accommodation (Fig. 2-C), the proportion of eyes in all the spherical error ranges was similar between the groups (P > 0.05). Overall, the proportion of eyes among the ranges of postoperative spherical error in the SILK group were closer to emmetrope eyes at distance and intermediate accommodation as compared to the LASIK group.

Fig. 2
figure 2

Proportion of eyes with postoperative sphere > − 0.5D, − 0.5D to − 1D and < − 1D at distance (A), intermediate (B) and near (C) distances in the three groups of eyes.

Full size image

Figure 3 shows the distribution of postoperative cylindrical error at all distances (A-C) across the 3 groups. At distance (Fig. 3-A), the proportion of eyes having cylindrical error greater than − 0.5D (light green columns) was significantly lower in the SILK eyes as compared to LASIK eyes (P < 0.01) and emmetropes (P = 0.03). The proportion of eyes having cylindrical error between − 0.5D to -1D (purple columns) was statistically similar between the groups (P > 0.05). The ELITA eyes showed significantly higher proportion of eyes having cylindrical error between − 1 to -1.5D (yellow columns) as compared to emmetropes (P = 0.04) but was statistically similar as compared to LASIK eyes (P > 0.05). However, SILK eyes showed a significantly higher proportion of eyes having cylindrical error higher than − 1.5D (red columns) as compared to LASIK eyes (P = 0.04). Similar trend was observed at intermediate (Fig. 3-B) and near accommodations (Fig. 3-C) where the SILK eyes showed significantly higher proportion of eyes having cylindrical error greater than − 0.5D (light green columns) as compared to LASIK eyes (P < 0.01) and emmetropes (P = 0.04). Further, at intermediate accommodation (Fig. 3-B), the SILK eyes showed significantly higher proportion of eyes having cylindrical error greater than − 1.5D (red columns) as compared to LASIK eyes (P < 0.01) and emmetropes (P = 0.04). At near accommodation (Fig. 3-C), the SILK eyes showed a significantly greater proportion of eyes having residual cylindrical error between − 1 to -1.5D (yellow columns) as compared to LASIK eyes (P < 0.01).

Fig. 3
figure 3

Proportion of eyes with postoperative cylinder > − 0.5D, − 0.5D to − 1D and < − 1D at distance (A), intermediate (B) and near (C) distances in the three groups of eyes.

Full size image

Figure 4 shows postoperative ocular aberrations at all reading distances for the 3 groups. At distance (Fig. 4-A), the SILK eyes showed slightly higher defocus as compared to LASIK eyes (p < 0.05) and emmetropes (p > 0.05). However, LORMS was comparable between the groups (P > 0.05). Similarly, RMS higher orders aberrations and Coma RMS were also similar between the groups (P > 0.05). However, primary SA was significantly lower in LASIK eyes as compared to ELITA eyes and emmetropes (P < 0.01). At intermediate accommodation (Fig. 4-B), the ELITA eyes showed significantly higher defocus and LORMS than LASIK eyes (P < 0.01). Amongst the higher order aberrations, Coma RMS and primary SA in the SILK eyes were significantly higher than LASIK eyes (P < 0.01). However, HORMS was similar between both the groups (P > 0.05). At near accommodation (Fig. 4-C), similar trend was observed where defocus, LORMS and primary SA were significantly higher in the SILK eyes as compared to the LASIK eyes (P < 0.01). However, HORMS and Coma RMS were similar between the groups (P > 0.05).

Fig. 4
figure 4

Postoperative defocus, spherical aberration, root mean square of coma, root mean square of lower order and higher order aberrations. Unit of aberration term is µm.

Full size image

Discussion

The settings of the femtosecond laser can play an important role in ensuring the surface quality of lenticules. One contributing factor is the energy per pulse. The higher the pulse energy, the bigger the cavitation bubble and rougher the tissue surface. Another set of contributing factors are the force and the amount of mechanical dissection involved in the lenticule extraction. More tissue bridges will require more manipulation during dissection, leading to rougher lenticular surface13. In this study, SILK procedure used a much lower energy (40–90 nJ) per pulse compared to SMILE (100–180 nJ) and Schwind ATOS (100–120 nJ)14,15. In addition, SILK used a high precision scanning system with contiguous spot pattern to generate an accurate lenticule shape with minimum tissue bridges. Thus, the lenticule removal was easier than other LALEX procedures with minimal corneal deformation. Furthermore, SILK uses a biconvex shaped lenticule compared to a plano-convex shaped SMILE lenticule. This may allow better apposition of the residual stromal surface lying immediately above and below the extracted lenticule due to similar arc length of the anterior and posterior lenticular surfaces. We hypothesized that these design features enabled a lenticule interface with less light scattering and a corneal shape with less higher order aberrations, leading to outcomes that are either better than or equivalent to, or similar in comparison with emmetrope eyes and eyes treated with wavefront optimized LASIK11.

The greater depth of field at intermediate distance in this study could be due to the higher postoperative cylinder in SILK eyes because a cylinder always provides a wider range of focus for optical rays in the planes of incidence that are at different meridians. The following were the key outcomes of the study:

  1. 1.

    We measured postoperative UDVA, sphere, cylinder, defocus, LORMS, HORMS, Coma, and SA at distance (20 ft), intermediate (60 cm), and near (40 cm) distances. All these measurements of the SILK eyes were either better than, or equivalent to, or very close to the emmetrope eyes and the eyes treated with wavefront Optimized LASIK. These excellent results provided a clinical validation of the SILK procedure.

  2. 2.

    At distance, the postoperative sphere of SILK eyes (91.3% eyes within 0.5 D) was similar to emmetrope eyes (90.0% eyes within 0.5 D), and much better than the LASIK eyes (65.3% eyes within 0.5 D). At intermediate and near distances, the distribution of sphere of SILK eyes was closer to emmetrope eyes than the LASIK eyes (Fig. 2). This implied that the SILK nomogram for sphere adjustment was accurate. Since there was no cylinder adjustment in this study, the SILK eyes showed a higher postoperative cylinder at all reading distances compared to LASIK eyes (Fig. 3). At distance and near, the SILK eyes had clinically similar visual acuity compared to emmetrope and LASIK eyes. However, the intermediate visual acuity of SILK eyes was similar to the emmetrope eyes, but statistically significantly better than the LASIK eyes.

  3. 3.

    Except SA, all other ocular aberrations at distance were similar between SILK and LASIK eyes. At intermediate and near, ELITA SILK showed significantly higher defocus and LORMS than LASIK eyes. This again implied a better depth of field of the SILK eyes compared to the LASIK eyes postoperatively.

In our previous work, we reported a consistent difference in postoperative cylinder of approximately 0.25 D between SMILE and LASIK eyes at all reading distances11. This suggested that both lenticule extraction procedures (SILK and SMILE) yielded a similar postoperative cylinder as compared to LASIK despite different lenticule shapes. This postoperative cylinder is the net cylinder refractive error present in the eye after the surgery as there was no nomogram adjustment for cylinder treatment. Furthermore, the intermediate visual acuity of SILK eyes was marginally better than LASIK eyes (P < 0.05) and comparable to emmetrope eyes.

The slight amount of postoperative lower order myopic defocus wasn’t the intended objective of refractive error correction with SILK and could be compensated by cylinder correction nomogram. Higher order aberrations were invariant at different reading distances which demonstrated that the depth of field response was primarily determined by the lower order ocular aberrations. As patients age, the eye tends to develop presbyopia. This slight amount of postoperative myopic defocus may help overcome the early-stage loss in near visual acuity as the patients get older. Further, lenticule shapes differ widely between manufacturers. Yet we observed a similar level of myopic defocus between SILK and SMILE eyes11. An earlier study showed minimal impact of SILK surgery without any nomogram adjustment upon longitudinal follow-up of higher order aberrations post-operatively11. Thus, potential presbyopic treatments with lenticule extraction and customized lenticule shapes may be plausible and require further study. All the higher order aberrations including HORMS, coma, and SA at all reading distances were all very similar and small for the three groups even though there may be statistically significant differences between the groups (Fig. 4)11. Further, except at distance, defocus and LORMS were also significantly different between SILK and LASIK eyes (Fig. 4).

Compared to the SMILE eyes in our earlier study, the SILK eyes had similar postoperative sphere, cylinder and ocular aberrations postoperatively11. However, the lenticule shapes may have a bearing on the postoperative shape of the Epithelium-Bowman’s interface and the resultant epithelium remodeling17,18. Thus, these proposed future studies may yield valuable insights into further customization of lenticule shapes to achieve selective modulation of aberrations such as wavefront customized LASIK and presbyopic laser surgery. Limitation of this study were the use of only Asian-Indian eyes which generally tend to have smaller pupil size than North-American and European eyes and a short follow-up time. Smaller pupil size will impact the magnitude of induced aberrations to achieve optimal visual acuity at near and intermediate distances. Further evaluation of depth of field in other populations is required to understand the differences between the different LALEX procedures.

Conclusions

The post-SILK eyes had unique depth of field at different reading distance relative to post-LASIK and healthy unoperated eyes. Nomogram adjustment to sphere was adequate for refractive accuracy of distance vision while zero cylinder adjustment resulted in better reading performance compared to post-LASIK eyes.