Introduction
Small-incision lenticule extraction (SMILE) is a refractive surgery procedure currently used to correct refractive errors in patients with myopia with or without astigmatism while the correction of presbyopia and hyperopia is still in development. It involves the creation of a corneal tissue disk called a lenticule and its extraction through a minimally invasive incision, utilizing only a femtosecond (FS) laser. This enables rapid visual recovery with very little discomfort to the patient. Therefore one can consider SMILE as laser in situ keratomileusis (LASIK) without flap, and photorefractive keratectomy (PRK) without pain.
The history of SMILE as a refractive procedure began in 1996, when a picosecond laser–based system was first used to generate an intrastromal lenticule that was manually removed after lifting a superficial flap similar to LASIK. At this early stage, significant manual dissection of stromal tissue bridges generated an irregular surface. In 1998, the FS laser replaced the early picosecond laser–based system to perform lenticule extraction in rabbit corneas. FS laser–based lenticule extraction was taken further to perform extractions in partially sighted eyes in 2003 ; however, these preliminary studies were not followed up by clinical trials. Four years later, in early 2007, femtosecond lenticule extraction (FLEx) was presented in sighted patients that can be thought of as the prototype of the current SMILE procedure as an alternative to LASIK to correct high myopia. The refractive outcomes of FLEx were similar to that of LASIK. Further refinements led to the development of a procedure to extract a corneal lenticule involving only a small incision of 2 to 4 mm, without creating a flap. September 2011 saw the commercial launch of this SMILE procedure.
Principles Behind SMILE
Unlike an excimer laser used in LASIK to remove corneal tissue by a process called photoablation , the principle by which an FS laser works to create a corneal lenticule and incision in SMILE is a process called photodisruption . Photodisruption involves vaporizing a small volume of tissue by creating plasma (free electrons and ions), carbon dioxide, an acoustic shockwave, thermal energy, and a cavitation bubble. The millions of plasma bubbles created through photodisruption are aligned in a laminar manner at the target location, where they expand and subsequently stretch and separate the surrounding tissues. The mechanical separation occurring along the laminar structures of the tissue results in the effect of tissue cleaving ( Fig. 16.1 ). The amount of tissue cleaving depends on the amount of laser pulse energy: the higher the pulse energy, the greater the effect. Often, cleaving occurs owing to the concerted effects of about 10,000 to 100,000 pulses/mm 2 . The laser that is used for SMILE is a neodymium:yttrium aluminum garnet (Nd:YAG) solid-state laser that emits energy into a focal point with a wavelength of 1043 nm. In the United States, the US Food and Drug Administration (FDA) has approved the VisuMax Femtosecond Laser (Carl Zeiss Meditec Inc.) for the reduction or elimination of nearsightedness using SMILE. The limit values as approved by the FDA for the correction of myopia with SMILE are presented in Table 16.1 .
| Parameter | Unit | Range |
|---|---|---|
| Laser Parameter | ||
| Laser energy | nJ | 125–170 |
| Track distance | µm | 2.0–3.0 |
| Spot distance | µm | 2.0–3.0 |
| Surgical | ||
| Cap diameter | mm | 7.0 or 7.5 |
| Cap thickness | µm | 120 |
| Lenticule diameter | mm | 6.0 or 6.5 |
| Lenticule edge minimum thickness | µm | 15 |
| Residual bed minimum thickness | µm | 250 |
| Incision position: opening position | deg | 90 |
| Incision angle: cap opening size | deg | 90 |
| Side cut angle: opening cut | deg | 90 |
| Side cut angle: lenticule cut | deg | 90 |
| Refractive | ||
| Intended spherical corrections | D | -100 to -8.00 * |
| Intended cylindrical corrections | D | — |
| Intended cylindrical axis | deg | — |
* Treatment of −8.01 to −10.00 D will present a flagged warning to the users so that the user understands that correction of these powers has not been substantiated by an adequate data set.
Surgical Techniques for SMILE ( )
Preoperative Considerations
Patient Selection
In selecting patients for SMILE, the surgeon must screen surgical candidates for myopia and myopic astigmatism. Currently, in the United States, the FDA has approved SMILE for the treatment of myopia of −1.00 diopter (D) to −8.00 D, and astigmatism of less than or equal to −0.50 D. Outside the United States, SMILE can be used to treat myopia of up to −10.00 D combined with astigmatism of up to −5.00 D. SMILE can be harder to perform for very low myopia chiefly owing to difficulty in managing such a thin lenticule. The difficulty can be overcome by making the lenticule thicker using a wider optical zone. The FDA has experimented with a minimum peripheral lenticule thickness of 15 µm in all lenticules, and surgeons have observed good results. However, with increasing surgeon experience, we recommend a minimum thickness of only 10 µm to reduce the refractively neutral removal of precious stromal tissue (only outside the United States). SMILE for hyperopia is not in routine use yet, but some studies have been published of its initial success in monkey corneas and small-diopter correction in hyperopic patients. A prospective multicenter trial is underway.
Centration
Unlike in other refractive laser procedures, currently, no eye tracking system is used in SMILE; this is perceived as a major drawback by some colleagues. However, the patient is instructed to fixate coaxially on a fixation light before the application of suction ( Fig. 16.2A ). This coaxial fixation causes the refractive lenticule to be auto-centered on the corneal vertex of the eye. The surgeon confirms the correct centration by comparing the relative positions of the pupil center and the corneal reflex (which may not coincide) to the Placido eye image obtained from the Atlas topography scan.
The coaxial fixation on the fixation light follows a number of steps in SMILE. First, the patient is raised to appropriate height to make contact with the contact glass of the FS laser. The contact glass is curved, allowing for comfortable docking. When contact is made between the cornea and the contact glass, a tear film meniscus appears; simultaneously, the patient is able to see the fixation target clearly. This happens because the vergence of the fixation beam is adjusted according to the individual refraction of the patient’s eye. The surgeon instructs the patient to focus directly on the target green light. Once the patient is focusing on the green light, the surgeon activates the corneal suction ports to fixate the eye in this position in order to align the visual axis. The surgeon can release suction and repeat the docking procedure if the centration of docking is not satisfactory.
The physiologic location of the visual axis is only approximated, however, through the corneal light reflex method because a coaxially aligned light reflex corresponds to the center of the optical system as opposed to the true visual axis. The lack of an eye tracking system to compensate for alignment and cyclotorsional errors in SMILE is a weakness of this procedure.
Preoperative Considerations
Patient Selection
In selecting patients for SMILE, the surgeon must screen surgical candidates for myopia and myopic astigmatism. Currently, in the United States, the FDA has approved SMILE for the treatment of myopia of −1.00 diopter (D) to −8.00 D, and astigmatism of less than or equal to −0.50 D. Outside the United States, SMILE can be used to treat myopia of up to −10.00 D combined with astigmatism of up to −5.00 D. SMILE can be harder to perform for very low myopia chiefly owing to difficulty in managing such a thin lenticule. The difficulty can be overcome by making the lenticule thicker using a wider optical zone. The FDA has experimented with a minimum peripheral lenticule thickness of 15 µm in all lenticules, and surgeons have observed good results. However, with increasing surgeon experience, we recommend a minimum thickness of only 10 µm to reduce the refractively neutral removal of precious stromal tissue (only outside the United States). SMILE for hyperopia is not in routine use yet, but some studies have been published of its initial success in monkey corneas and small-diopter correction in hyperopic patients. A prospective multicenter trial is underway.
Centration
Unlike in other refractive laser procedures, currently, no eye tracking system is used in SMILE; this is perceived as a major drawback by some colleagues. However, the patient is instructed to fixate coaxially on a fixation light before the application of suction ( Fig. 16.2A ). This coaxial fixation causes the refractive lenticule to be auto-centered on the corneal vertex of the eye. The surgeon confirms the correct centration by comparing the relative positions of the pupil center and the corneal reflex (which may not coincide) to the Placido eye image obtained from the Atlas topography scan.
The coaxial fixation on the fixation light follows a number of steps in SMILE. First, the patient is raised to appropriate height to make contact with the contact glass of the FS laser. The contact glass is curved, allowing for comfortable docking. When contact is made between the cornea and the contact glass, a tear film meniscus appears; simultaneously, the patient is able to see the fixation target clearly. This happens because the vergence of the fixation beam is adjusted according to the individual refraction of the patient’s eye. The surgeon instructs the patient to focus directly on the target green light. Once the patient is focusing on the green light, the surgeon activates the corneal suction ports to fixate the eye in this position in order to align the visual axis. The surgeon can release suction and repeat the docking procedure if the centration of docking is not satisfactory.
The physiologic location of the visual axis is only approximated, however, through the corneal light reflex method because a coaxially aligned light reflex corresponds to the center of the optical system as opposed to the true visual axis. The lack of an eye tracking system to compensate for alignment and cyclotorsional errors in SMILE is a weakness of this procedure.
Patient Selection
In selecting patients for SMILE, the surgeon must screen surgical candidates for myopia and myopic astigmatism. Currently, in the United States, the FDA has approved SMILE for the treatment of myopia of −1.00 diopter (D) to −8.00 D, and astigmatism of less than or equal to −0.50 D. Outside the United States, SMILE can be used to treat myopia of up to −10.00 D combined with astigmatism of up to −5.00 D. SMILE can be harder to perform for very low myopia chiefly owing to difficulty in managing such a thin lenticule. The difficulty can be overcome by making the lenticule thicker using a wider optical zone. The FDA has experimented with a minimum peripheral lenticule thickness of 15 µm in all lenticules, and surgeons have observed good results. However, with increasing surgeon experience, we recommend a minimum thickness of only 10 µm to reduce the refractively neutral removal of precious stromal tissue (only outside the United States). SMILE for hyperopia is not in routine use yet, but some studies have been published of its initial success in monkey corneas and small-diopter correction in hyperopic patients. A prospective multicenter trial is underway.
Centration
Unlike in other refractive laser procedures, currently, no eye tracking system is used in SMILE; this is perceived as a major drawback by some colleagues. However, the patient is instructed to fixate coaxially on a fixation light before the application of suction ( Fig. 16.2A ). This coaxial fixation causes the refractive lenticule to be auto-centered on the corneal vertex of the eye. The surgeon confirms the correct centration by comparing the relative positions of the pupil center and the corneal reflex (which may not coincide) to the Placido eye image obtained from the Atlas topography scan.
The coaxial fixation on the fixation light follows a number of steps in SMILE. First, the patient is raised to appropriate height to make contact with the contact glass of the FS laser. The contact glass is curved, allowing for comfortable docking. When contact is made between the cornea and the contact glass, a tear film meniscus appears; simultaneously, the patient is able to see the fixation target clearly. This happens because the vergence of the fixation beam is adjusted according to the individual refraction of the patient’s eye. The surgeon instructs the patient to focus directly on the target green light. Once the patient is focusing on the green light, the surgeon activates the corneal suction ports to fixate the eye in this position in order to align the visual axis. The surgeon can release suction and repeat the docking procedure if the centration of docking is not satisfactory.
The physiologic location of the visual axis is only approximated, however, through the corneal light reflex method because a coaxially aligned light reflex corresponds to the center of the optical system as opposed to the true visual axis. The lack of an eye tracking system to compensate for alignment and cyclotorsional errors in SMILE is a weakness of this procedure.
Incision Technique
Incision technique comprises the use of the appropriate laser pulse parameters, the shape of the incision, and the centering accuracy of the incision.
Appropriate Laser Pulse
The FDA-approved laser pulse energy for SMILE is between 125 to 170 nJ (see Table 16.1 ). However, even lower energy settings may be preferable because they have been shown to produce smoother interfaces. The energy settings of 140 and 170 nJ have been reported to be effective for the correction of myopia and myopic astigmatism, respectively. No significant differences have been observed in the optical quality, including the corneal light scattering in eyes undergoing SMILE procedure with the laser settings of 140 nJ, spot distance 3.0 µm vs 170 nJ, and spot distance of 4.5 µm. The tissue cutting precision results from the average of the axial precisions of a large number of laser pulses. The optimization of laser settings for SMILE depends on 2 factors: (1) the level of energy per area and (2) optimized distances between adjacent spots and adjacent spiral tracks to reduce the roughness of the stromal surfaces after the lenticule is removed. These 2 factors can be responsible for delayed visual recovery. Spiraling-in application of the femto-laser spots is used to create the refractive posterior surface of the lenticule ( Fig. 16.2B ), while the spiraling-out technique is used to create the anterior surface of the lenticule that is parallel to the anterior surface of the cornea ( Fig. 16.2C ). A complete laser application with superotemporal incision is shown in Fig. 16.2D .
Shape of the Incision
The current conventional design of the incision technique for lenticule preparation involves cap cutting that is parallel to the corneal surface and a lenticule cut connected to the cap cut by a side cut on the edge of the lenticule (side cut is approximately 10–15 µm in height typically; Fig. 16.3 ). The side cut ( Fig. 16.2E ) is refractively neutral and is required to give the surgeon some tissue to grasp and manipulate. After opening the side cut with a semi-sharp tip, the upper lenticular surface is entered first ( Fig. 16.2F ) and then separated using a blunt spoon-shaped SMILE spatula ( Fig. 16.2F ). The lower interface is separated following the upper interface separation ( Fig. 16.2G ), and lenticule is extracted with micro-forceps ( Fig. 16.2H ). A finished SMILE procedure is shown in Fig. 16.2I .
The incision shape can be spherical or ellipsoidal depending on whether there is need for astigmatism correction along with spherical correction. The thickness of the lenticule to be extracted during SMILE depends on the degree of the required refractive error based on the theoretical limit value according to Munnerlyn. The usual rule of thumb for 1 D correction using a 6-mm correction zone is 13 µm lenticule thickness corresponding to the equivalent ablation depth in excimer laser surgery. This maximum lenticule thickness is central lenticule thickness in cases of myopia correction and the thickness at the optical zone edge in cases of hyperopia correction, respectively. Note that the refractively neutral lenticule side cut adds to lenticule thickness.
Centering Accuracy of the Incision
The centration of the lenticule cut is important to achieve accurate refractive correction in SMILE. However, the issue is not as serious for lenticule preparation as it is for ablative surgery because, comparatively, lenticule decentration induces a smaller amount of higher-order aberrations owing to the use of an FS laser ( Fig. 16.4 ). The excimer laser used in LASIK is more sensitive, as it typically causes ablation errors around the edge of the working area owing to projection errors and thus a variation of ablation efficiency. For SMILE, the visual outcomes can still be good if the lenticule diameter is sufficiently large. Better refractive outcome is reported for lenticules centered near the corneal vertex normal. Incidentally, eyes that undergo the SMILE procedure have been reported to show less mean centration offset compared to eyes that undergo LASIK.
Other Important Considerations
Thickness of Lenticule Cap
The cap thickness of the lenticule appears to have no effect on visual acuity or refractive outcomes after the SMILE procedure. Guell et al. conducted a retrospective, comparative, nonrandomized clinical study in which they performed myopic SMILE with four different cap thicknesses: 130, 140, 150, and 160 µm. They found no statistically significant differences between uncorrected distance visual acuity (UDVA), corrected distance visual acuity (CDVA), and objective scattering index (OSI) between the four different scenarios. El-Massry et al. conducted a prospective comparative interventional clinical trial in patients who underwent SMILE with lenticule creation at two different depths, 100 µm in the right eye and 160 µm in the left eye. They evaluated the manifest refraction, UCVA, total higher-order aberrations (HOAs), and corneal biomechanical properties in both eyes at 1 month postoperatively. Although no statistically significant differences were found between the former three parameters in the two eyes, the left eye with 160 µm–deep lenticule creation displayed less damage to the corneal biomechanics. The predictability of the cap thickness in SMILE is reported to be consistent with the flap thickness in femtosecond-LASIK (FS-LASIK) with the use of same FS laser platform.
Appropriate Laser Pulse
The FDA-approved laser pulse energy for SMILE is between 125 to 170 nJ (see Table 16.1 ). However, even lower energy settings may be preferable because they have been shown to produce smoother interfaces. The energy settings of 140 and 170 nJ have been reported to be effective for the correction of myopia and myopic astigmatism, respectively. No significant differences have been observed in the optical quality, including the corneal light scattering in eyes undergoing SMILE procedure with the laser settings of 140 nJ, spot distance 3.0 µm vs 170 nJ, and spot distance of 4.5 µm. The tissue cutting precision results from the average of the axial precisions of a large number of laser pulses. The optimization of laser settings for SMILE depends on 2 factors: (1) the level of energy per area and (2) optimized distances between adjacent spots and adjacent spiral tracks to reduce the roughness of the stromal surfaces after the lenticule is removed. These 2 factors can be responsible for delayed visual recovery. Spiraling-in application of the femto-laser spots is used to create the refractive posterior surface of the lenticule ( Fig. 16.2B ), while the spiraling-out technique is used to create the anterior surface of the lenticule that is parallel to the anterior surface of the cornea ( Fig. 16.2C ). A complete laser application with superotemporal incision is shown in Fig. 16.2D .
Shape of the Incision
The current conventional design of the incision technique for lenticule preparation involves cap cutting that is parallel to the corneal surface and a lenticule cut connected to the cap cut by a side cut on the edge of the lenticule (side cut is approximately 10–15 µm in height typically; Fig. 16.3 ). The side cut ( Fig. 16.2E ) is refractively neutral and is required to give the surgeon some tissue to grasp and manipulate. After opening the side cut with a semi-sharp tip, the upper lenticular surface is entered first ( Fig. 16.2F ) and then separated using a blunt spoon-shaped SMILE spatula ( Fig. 16.2F ). The lower interface is separated following the upper interface separation ( Fig. 16.2G ), and lenticule is extracted with micro-forceps ( Fig. 16.2H ). A finished SMILE procedure is shown in Fig. 16.2I .
The incision shape can be spherical or ellipsoidal depending on whether there is need for astigmatism correction along with spherical correction. The thickness of the lenticule to be extracted during SMILE depends on the degree of the required refractive error based on the theoretical limit value according to Munnerlyn. The usual rule of thumb for 1 D correction using a 6-mm correction zone is 13 µm lenticule thickness corresponding to the equivalent ablation depth in excimer laser surgery. This maximum lenticule thickness is central lenticule thickness in cases of myopia correction and the thickness at the optical zone edge in cases of hyperopia correction, respectively. Note that the refractively neutral lenticule side cut adds to lenticule thickness.
Centering Accuracy of the Incision
The centration of the lenticule cut is important to achieve accurate refractive correction in SMILE. However, the issue is not as serious for lenticule preparation as it is for ablative surgery because, comparatively, lenticule decentration induces a smaller amount of higher-order aberrations owing to the use of an FS laser ( Fig. 16.4 ). The excimer laser used in LASIK is more sensitive, as it typically causes ablation errors around the edge of the working area owing to projection errors and thus a variation of ablation efficiency. For SMILE, the visual outcomes can still be good if the lenticule diameter is sufficiently large. Better refractive outcome is reported for lenticules centered near the corneal vertex normal. Incidentally, eyes that undergo the SMILE procedure have been reported to show less mean centration offset compared to eyes that undergo LASIK.
Other Important Considerations
Thickness of Lenticule Cap
The cap thickness of the lenticule appears to have no effect on visual acuity or refractive outcomes after the SMILE procedure. Guell et al. conducted a retrospective, comparative, nonrandomized clinical study in which they performed myopic SMILE with four different cap thicknesses: 130, 140, 150, and 160 µm. They found no statistically significant differences between uncorrected distance visual acuity (UDVA), corrected distance visual acuity (CDVA), and objective scattering index (OSI) between the four different scenarios. El-Massry et al. conducted a prospective comparative interventional clinical trial in patients who underwent SMILE with lenticule creation at two different depths, 100 µm in the right eye and 160 µm in the left eye. They evaluated the manifest refraction, UCVA, total higher-order aberrations (HOAs), and corneal biomechanical properties in both eyes at 1 month postoperatively. Although no statistically significant differences were found between the former three parameters in the two eyes, the left eye with 160 µm–deep lenticule creation displayed less damage to the corneal biomechanics. The predictability of the cap thickness in SMILE is reported to be consistent with the flap thickness in femtosecond-LASIK (FS-LASIK) with the use of same FS laser platform.
Thickness of Lenticule Cap
The cap thickness of the lenticule appears to have no effect on visual acuity or refractive outcomes after the SMILE procedure. Guell et al. conducted a retrospective, comparative, nonrandomized clinical study in which they performed myopic SMILE with four different cap thicknesses: 130, 140, 150, and 160 µm. They found no statistically significant differences between uncorrected distance visual acuity (UDVA), corrected distance visual acuity (CDVA), and objective scattering index (OSI) between the four different scenarios. El-Massry et al. conducted a prospective comparative interventional clinical trial in patients who underwent SMILE with lenticule creation at two different depths, 100 µm in the right eye and 160 µm in the left eye. They evaluated the manifest refraction, UCVA, total higher-order aberrations (HOAs), and corneal biomechanical properties in both eyes at 1 month postoperatively. Although no statistically significant differences were found between the former three parameters in the two eyes, the left eye with 160 µm–deep lenticule creation displayed less damage to the corneal biomechanics. The predictability of the cap thickness in SMILE is reported to be consistent with the flap thickness in femtosecond-LASIK (FS-LASIK) with the use of same FS laser platform.
Corneal Wound Healing After SMILE
The way a cornea heals has a major effect on visual outcomes after refractive surgery. Corneal haze, myopia regression, and epithelial ingrowth after FS-LASIK have been associated with less than ideal corneal wound healing. Complex cascades of molecular and cellular pathways involving cytokines, growth factors, and tumor necrosis factors are involved in corneal wound healing. After an insult to the cornea, cytokines and growth factors are released from the injured epithelium and mediate apoptosis of the stromal keratocytes. Subsequently, proliferation and migration of the remaining keratocytes occur within a few hours in order to restore the cellularity of the stroma. Simultaneously, inflammatory cells migrate to phagocytize the apoptotic cells and to enhance the transformation of keratocytes to fibroblasts within 24 hours of the injury. A good balance between the development of myofibroblasts and the apoptosis of myofibroblast precursors is decisive in determining whether the cornea heals ideally or develops haze after wound healing.
SMILE appears to induce minimal corneal wound healing responses after the surgery. A study of early corneal wound healing responses after SMILE in rabbit eye models showed that significantly fewer terminal deoxynucleotidyl transferase dUTP nick end labeling (TUNEL)–positive stromal cells, Ki67-positive cells, and CD11b-positive cells were found after SMILE than after FS-LASIK. The degree of early corneal wound healing response depends proportionally on the degree of correction of the refractive error. In vivo confocal microscopy immediately after SMILE showed greater corneal reflectivity in corneas after −8.00 D of correction compared to lower myopia corrections. The greater corneal reflectivity was observed at all planes of the procedure—the anterior, posterior, and extracted lenticular planes. In lower degrees of refractive corrections, there were no differences in the expression of CD11b, fibronectin, and HSP47 observed. If any increased amount of inflammation was observed in lower refraction error correction, it was attributed to less surgical experience of the surgeon.
Mixed sensory and autonomic nerves enter the cornea by approaching the corneoscleral limbus radially from all directions. These arise from two long ciliary nerves that enter the posterior globe adjacent to the optic nerve and course forward in the suprachoroidal space at the nasal and temporal meridians ( Fig. 16.5 ). The central corneal nerve fiber density is reduced after SMILE owing to the transection of the nerve fibers that intersect the lenticule-cap plane by the FS laser cut. However, the damage is significantly less compared to the flap-based and surface ablation techniques in which almost total fiber resection takes place. Surgical denervation is also significantly less in SMILE compared to other refractive procedures, along with faster nerve regeneration, based on in vivo studies. Therefore in theory, there should be less severe and less frequent postoperative dry eye after SMILE compared to other corneal refractive procedures. Only mild to moderate keratocyte apoptosis, secondary keratocyte activation, and stromal inflammation occur in any intrastromal technique.
At times, microdistortions can be observed in the Bowman layer after SMILE, especially for high myopia owing to the longer arc length of the cap than the stromal bed after removing the lenticule. Quantitative evaluation of the microdistortions on optical coherence tomography (OCT) has found that there is a positive correlation between the magnitude of the refractive error corrected and the Bowman Roughness Index (BRI). The BRI is defined as the enclosed area between the actual and ideal smooth layer to quantify microdistortions. The microdistortions, however, are reported to remain stable after a week, with no impact on long-term visual performance. The postoperative BRI in SMILE eyes is also reported to be similar to the preoperative BRI in comparison to LASIK eyes, indicating earlier wound healing after SMILE.
A study has reported milder ocular surface changes after SMILE compared to LASIK, especially pertaining to the tear inflammatory mediators, such as interleukin 6 (IL-6) and nerve growth factor (NGF), which are thought to play important roles in the healing of the ocular surface. The authors measured IL-6, NGF, tumor necrosis factor-α (TNF-α), and intercellular adhesion molecule-1 (CAM-1) in the tears of myopic patients. Although TNF-α and CAM-1 concentrations were similar between both groups at any follow-up time (1 day, 1 week, 1 month, and 3 months postoperatively), IL-6 and NGF levels were observed to be lower in the SMILE group. The levels of IL-6 and NGF were seen to correlate with the changes of the ocular surface after SMILE or LASIK, and the SMILE group showed faster recovery of IL-6 and NGF levels.
Corneal Biomechanics
Corneal shape determines the refraction of the eye, which, in turn, depends on its inherent biomechanical properties. Corneal biomechanics is dependent on several factors, such as extracellular matrix components, collagen lamellae organization, osmotic pressure, corneal layers, systemic diseases (e.g., diabetes mellitus), hormonal fluctuations, and to some extent environmental factors (although only little is known). Stable corneal biomechanics after refractive surgery is important in order to maintain satisfactory visual recovery and to avoid any postoperative complications.
Corneal biomechanics is typically thought not to be compromised too badly after SMILE because of the minimally invasive nature of the procedure and the absence of the flap cut. Although some studies have shown the differences in biomechanical properties between flap-based and flapless procedures to be insignificant and inconsistent, other studies have shown some differences in biomechanical strength of the cornea between flap-based and flapless procedures. Based on the tomographic data of 10 eyes, Seven et al. reported a mean reduction of 49% in the effective stiffness of the collagen fiber of the corneal stroma in the eyes that underwent the flap-based procedure (FLEx) compared to eyes that underwent the flapless procedure (SMILE). The mean reduction ranged from 2% to 87%, and the eyes that underwent the flap-based FLEx procedure also showed higher stresses and deformations within the residual stromal bed than the eyes that underwent SMILE.
Knox Cartwright et al. reported that corneal strain after delamination only is lower than the strain after LASIK flap and side cut; they also reported that corneal strain tends to increase with a thicker flap cut. Randleman et al. reported that stromal cohesive tensile strength is strongest in the anterior 40% of the central cornea and at least 50% weaker in the posterior 60% of the corneal stroma. The tangential tensile strength also has been reported to be greater for the anterior than the posterior stroma using different methodologies. Therefore theoretically, the absence of the anterior flap cut in SMILE would not compromise the integrity of the cornea as much as flap cut procedures would. Additionally, in contrast to LASIK, in which deeper ablations contribute to lower tensile strength postoperatively (because of the minimal contribution of the flap to the corneal biomechanics after wound healing), performing SMILE at greater stromal depth (of the weaker posterior stroma) would also leave the cornea with greater tensile strength.
A paper by Shih et al. compared the hoop stresses of the cornea under tension and bending for patients who undergo radial keratotomy (RK), PRK, LASIK, and SMILE, and mapped the stress concentration, potential creak zones, and potential errors in intraocular pressure (IOP) measurements for all four surgical procedures. The stress and potential creak zones were mapped based on finite element analysis (FEA), in which the roles of the stroma, Bowman membrane, and Desçemet membrane under physiologic tension and nonphysiologic bending of the cornea were determined. The authors did the first principle stress (FPS) analysis of the cornea under four surgical models (FPS is defined as the component of stress tensors when the shear stress component is reduced to zero by rotating the basis ). Based on the FPS analysis, both SMILE and LASIK appeared to have potential creak zones near the edge of ablation with cracks at 45 degrees spreading radially, but SMILE appeared to have less maximum stress than LASIK ( Fig. 16.6 ). Study of the FPS upon eye rubbing applied to the 4 surgical models showed that SMILE, LASIK, and PRK caused stress concentrations around the Bowman membrane near the ablation zone, whereas the RK incisions caused stress concentrations around the Desçemet membrane ( Fig. 16.7 ). A mathematical model developed by Reinstein et al. also predicted considerably greater stromal tensile strength after SMILE compared to LASIK.
Literature Review
Several clinical studies have recognized the success and effectiveness of SMILE. However, there have been reports of several complications during and after SMILE. As it is a relatively new procedure, we want to know how it compares to the standard LASIK procedure in terms of the visual and refractive outcomes. We also want to know how SMILE compares to LASIK in terms of the onset of the commonly reported postoperative complications after LASIK, such as dry eye, ectasia, and others.
In this chapter, we have systematically compiled a list of papers published on SMILE in PubMed using the search words small incision lenticule extraction from the beginning of time until October 23, 2017. The search yielded 305 papers, which were divided into English and foreign language. Only the papers published in English or with an English abstract were used. The selection included 280 papers.
Inclusion Criteria
The relevant papers on SMILE were selected with the following inclusion criteria in mind: (1) retrospective case series, prospective randomized controlled trials, and nonrandomized comparative trials; (2) studies that compared the preoperative and postoperative visual outcomes and reported either spherical equivalent (SE) within ± 0.5 D, SE within ± 1.0 D, or safety and efficacy indices; (3) the follow-up period could be as short as 1 month for visual and refractive outcomes, and as short as 1 week for the dry-eye outcome measurements; (4) patients aged 18 to 60 years with any degree of myopia and myopic astigmatism; (4) patients treated with SMILE alone or in combination with another method and, for comparison studies, patients treated with FS-LASIK; and (5) studies reporting intraoperative or postoperative complications.
Results
Predictability, Safety, and Efficacy
The abstract of each of the 280 papers was read and categorized into multiple categories, such as “visual and refractive outcomes after SMILE,” “intraoperative SMILE complications,” and “postoperative complications after SMILE.” The “postoperative complications” category was further divided into “dry eye,” “ectasia,” and “other complications.” In the “visual and refractive outcomes” category, only the papers that reported either the SE within ± 0.50 D, SE within ± 1.00 D, safety index, or the efficacy index were chosen. This inclusion criteria yielded 31 papers. Four of these papers were separated from the primary “visual and refractive outcomes” group because these papers reported predictability and/or safety and efficacy indexes based on the comparison between the traditional SMILE technique and variations of the traditional SMILE technique. Of the 27 papers included in the primary “visual and refractive outcome” group, one paper was entered twice because it reported predictability, safety, and efficacy indices separately for low myopia (< −6.00 D), and high myopia (≥ −6.00 D). The weighted averages of SE with ± 0.50 D, within ± 1.00 D, and weighted averages of safety and efficacy indices were calculated for both “visual and refractive outcome” tables ( Table 16.2 and Table 16.3 ).
| Study | Year | Country | No. of Eyes | SE Within ± 0.50 D (%) | SE Within ± 1.00 D (%) | Follow-up (mo) | Safety Index | Efficacy Index |
|---|---|---|---|---|---|---|---|---|
| Kanellopoulous et al. | 2017 | Greece | 44 | 77.3 | 3 | |||
| Khalifa et al. | 2017 | Egypt | 110 | 81.54 | 0.98 ± 0.08 | 0.92 ± 0.11 | ||
| Zhang et al. | 2017 | China | 9 | 88.9 | 1 | 1.10 ± 0.24 | 1.08 ± 0.16 | |
| Pedersen et al. | 2017 | Denmark | 101 | 74 | 93 | 12 | ||
| Kobashi et al. | 2017 | Japan | 30 | 100 | 24 | |||
| Burazovitch et al. | 2017 | France | 496 | 87 | 48 | |||
| Ganesh et al. | 2017 | India | 120 | 97 | 3 | |||
| Hyun et al. | 2017 | Korea | 69 | 84 | 6 | |||
| Chan TC et al. | 2017 | China | 66 | 100 | 6 | |||
| Wong et al. | 2017 | Singapore | 164 | 83.8 | 97.2 | 3 | ||
| Hansen et al. | 2016 | Denmark | 722 | 88 | 98 | 3 | ||
| Han et al. | 2016 | China | 47 | 89 | 48 | 1.16 ± 0.14 | 1.07 ± 0.16 | |
| Ang et al. | 2016 | Singapore | 50 | 94 | 12 | 1.17 ± 0.17 | 0.98 ± 0.20 | |
| Ang et al. | 2015 | Singapore | 172 | 82.5 | 3 | 1.13 ± 0.19 | 0.91 ± 0.21 | |
| Kamiya et al. | 2015 | Japan | 52 | 100 | 12 | |||
| Pedersen et al. | 2015 | Denmark | 87 | 78 | 90 | 36 | ||
| Kim et al. * | 2015 | Korea | 58 | 87.9 | 96.6 | 12 | 1.27 ± 0.17 | 1.04 ± 0.19 |
| Kim et al. ** | 2015 | Korea | 125 | 88 | 97.6 | 12 | 1.24 ± 0.17 | 0.99 ± 0.19 |
| Xu Y et al. | 2015 | 52 | 90.4 | 12 | ||||
| Kamiya et al. | 2014 | Japan | 26 | 100 | 6 | |||
| Vestergaard et al. | 2014 | Denmark | 35 | 88 | 97 | 6 | ||
| Agca et al. | 2014 | Turkey | 20 | 95 | 6 | |||
| Ang et al. | 2014 | Singapore | 88 | 78.4 | 95.5 | 3 | ||
| Kim et al. | 2014 | Korea | 447 | 86.1 | 97.9 | 6 | ||
| Miao et al. | 2014 | 66 | 3 | 1.12 ± 0.17 | ||||
| Vestergaard et al. | 2012 | Denmark | 144 | 77 | 95 | 3 | ||
| Hjortdal et al. | 2012 | Denmark | 670 | 80 | 94 | 3 | ||
| Sekundoet al. | 2011 | Germany | 91 | 80.2 | 95.6 | 6 | ||
| Weighted average of SE within ± 0.50 D = 85.1% | ||||||||
| Weighted average of SE within ± 1.00 D = 96.1% | ||||||||
| Weighted average of safety index = 1.17 | ||||||||
| Weighted average of efficacy index = 0.93 | ||||||||
| Study | Year | Country | No. of Eyes | SE Within ± 0.50 D (%) | SE Within ± 1.00 D (%) | Follow-up (mo) | Safety Index | Efficacy Index |
| Kanellopoulous et al. | 2017 | Greece | 44 | 77.3 | 3 | |||
| Khalifa et al. | 2017 | Egypt | 110 | 81.54 | 0.98 ± 0.08 | 0.92 ± 0.11 | ||
| Zhang et al. | 2017 | China | 9 | 88.9 | 1 | 1.10 ± 0.24 | 1.08 ± 0.16 | |
| Pedersen et al. | 2017 | Denmark | 101 | 74 | 93 | 12 | ||
| Kobashi et al. | 2017 | Japan | 30 | 100 | 24 | |||
| Burazovitch et al. | 2017 | France | 496 | 87 | 48 | |||
| Ganesh et al. | 2017 | India | 120 | 97 | 3 | |||
| Hyun et al. | 2017 | Korea | 69 | 84 | 6 | |||
| Chan et al. | 2017 | China | 66 | 100 | 6 | |||
| Wong et al. | 2017 | Singapore | 164 | 83.8 | 97.2 | 3 | ||
| Hansen et al. | 2016 | Denmark | 722 | 88 | 98 | 3 | ||
| Han et al. | 2016 | China | 47 | 89 | 48 | 1.16 ± 0.14 | 1.07 ± 0.16 | |
| Ang et al. | 2016 | Singapore | 50 | 94 | 12 | 1.17 ± 0.17 | 0.98 ± 0.20 | |
| Ang et al. | 2015 | Singapore | 172 | 82.5 | 3 | 1.13 ± 0.19 | 0.91 ± 0.21 | |
| Kamiya et al. | 2015 | Japan | 52 | 100 | 12 | |||
| Pedersen et al. | 2015 | Denmark | 87 | 78 | 90 | 36 | ||
| Kim et al. * | 2015 | Korea | 58 | 87.9 | 96.6 | 12 | 1.27 ± 0.17 | 1.04 ± 0.19 |
| Kim et al. ** | 2015 | Korea | 125 | 88 | 97.6 | 12 | 1.24 ± 0.17 | 0.99 ± 0.19 |
| Xu et al. | 2015 | 52 | 90.4 | 12 | ||||
| Kamiya et al. | 2014 | Japan | 26 | 100 | 6 | |||
| Vestergaard et al. | 2014 | Denmark | 35 | 88 | 97 | 6 | ||
| Agca et al. | 2014 | Turkey | 20 | 95 | 6 | |||
| Ang et al. | 2014 | Singapore | 88 | 78.4 | 95.5 | 3 | ||
| Kim et al. | 2014 | Korea | 447 | 86.1 | 97.9 | 6 | ||
| Miao et al. | 2014 | 66 | 3 | 1.12 ± 0.17 | ||||
| Vestergaard et al. | 2012 | Denmark | 144 | 77 | 95 | 3 | ||
| Hjortdal et al. | 2012 | Denmark | 670 | 80 | 94 | 3 | ||
| Sekundo et al. | 2011 | Germany | 91 | 80.2 | 95.6 | 6 | ||
| Weighted average of SE within ± 0.50 D = 85.1% | ||||||||
| Weighted average of SE within ±1.00 D = 96.1% | ||||||||
| Weighted average of safety index = 1.17 | ||||||||
| Weighted average of efficacy index = 0.93 | ||||||||
* Denotes mild to moderate myopia;
| Study | Year | Country | No. of Eyes | SE Within ± 0.50 D (%) | SE Within ± 1.00 D (%) | Follow-up (mo) | Safety Index | Efficacy Index |
|---|---|---|---|---|---|---|---|---|
| Taneri et al. | 2017 | Germany | 100 | 0.87 ± 0.23 | ||||
| Taneri et al. | 2017 | Germany | 100 | 0.87 ± 0.23 | ||||
| Kim et al. | 2016 | Korea | 52 | 1 | 1.12 ± 0.14 | 1.09 ± 0.17 | ||
| Kim et al. | 2016 | Korea | 60 | 1 | 1.09 ± 0.15 | 1.02 ± 0.11 | ||
| Ng et al. | 2016 | China | 32 | 94 | 6 | 1.00 ± 0.00 | 0.97 ± 0.06 | |
| Ng et al. | 2016 | China | 21 | 89 | 6 | 0.96 ± 0.06 | 0.88 ± 0.13 | |
| Zhao et al. | 2015 | 16 | 1.12 | 1.06 | ||||
| Zhao et al. | 2015 | 15 | 1.09 | 1.09 |
Intraoperative Complications
Among the 305 SMILE papers, there were 16 that reported some form of intraoperative complications. The reported intraoperative complications were opaque bubble layer, suction loss, incision bleeding, incision tear, incision abrasion, epithelial defect, difficult lenticule extraction, subconjunctival hemorrhage, lenticule tear, unintended posterior plane dissection, inaccurate laser placement, and cap perforation. Of these reported intraoperative complications, cap perforation was excluded from Table 16.4 because this complication was rare and no paper reported its occurrence in percentage. Of the 16 papers, 10 were included to calculate the weighted average of each of these 11 complications and six were excluded because they reported complications either as case reports or were in a different language. Of the 10 included papers, Ivarsen et al. reported some intraoperative complications, such as cap perforations (four eyes) and lenticule tear (one eye) instead of percentages because these complications were so rare (out of 1800 eyes). Thus only the intraoperative complications that were reported in percentages were included in Table 16.4 and in calculating weighted average percentages. We also could not find any other paper reporting the complication of cap perforation.
| Study | Eyes | Suction Loss (%) | Opaque Bubble Layer (%) | Black Spots (%) | Difficult Lenticule Extraction | Incisional Bleeding | Subconjunctival Hemorrhage | Unintended Posterior Plane Dissection | Lenticule Tear | Incision Abrasion | Inaccurate Laser Pulse Placement | Epithelial Defect |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Tityal et al. | 100 | 2 | 19 | 11 | 9 | 2 | ||||||
| Wang et al. | 3004 | 0.93 | 0.73 | 0.33 | 0.93 | 0.67 | 0.33 | 0.27 | 0.17 | 0.1 | ||
| Park and Koo | 11,762 | 0.2 | ||||||||||
| Gab-alla | 12 | 2.7 | ||||||||||
| Son et al. > * | 208 | 51.82 | 3.84 | |||||||||
| Liu et al. | 8490 | 0.41 | ||||||||||
| Osman et al. | 3376 | 2.1 | ||||||||||
| Ramirez-Miranda et al. | 160 | 3.13 | 4.38 | 3.75 | 3.13 | 11.3 | ||||||
| Wong et al. | 183 | 4.4 | ||||||||||
| Ivarsen et al. | 1800 | 0.8 | 6 | 1.9 | 1.8 | 6 | ||||||
| Weighted average of each complication (%) | 0.65 | 0.83 | 4.49 | 6.23 | 2.27 | 1.26 | 0.67 | 0.33 | 0.50 | 0.32 | 0.10 | |
* Son et al. included eyes from a retrospective case series, but the incidence of opague bubble layer is 51.82%, which is quite high.
| Study | Year | Eyes | Preop | 1 wk | 1 mo | 3 mo | 6 mo |
|---|---|---|---|---|---|---|---|
| Xia et al. | 2016 | 69 | 10.7 | 20.34 | 14.91 | 12.11 | |
| Denoyer et al. | 2015 | 30 | 19.7±12.7 | 7.5±4.5 | |||
| Li et al. | 2013 | 38 | 12.26±12.45 | 23.95±13.54 | 16.72±10.96 | 12.05±9.38 | 8.47±7.89 |
| Weighted average OSDI | 11.25 | 23.95 | 19.2 | 13.89 | 10.09 |
| Study | Year | Eyes | Preop | 1 week | 1 mon | 3 mon | 6 mon |
|---|---|---|---|---|---|---|---|
| Xia et al. | 2016 | 59 | 10.40 ± 2.87 | 26.03 ± 4.01 | 20.63 ± 3.66 | 16.00 ± 2.82 | |
| Denoyer et al. | 2015 | 30 | 23.9 ± 14.8 | 20.6 ± 20.8 | |||
| Li et al. | 2013 | 33 | 11.59 ± 16.92 | 18.78 ± 19.01 | 17.77 ± 16.64 | 16.22 ± 15.29 | 15.50 ± 14.00 |
| Weighted average OSDI | 10.83 | 18.78 | 23.27 | 19.05 | 17 |
We also identified five case reports of ectasia, one report of diffuse lamellar keratitis and one report of interface fluid collection. All reported intraoperative complications in the literature are shown in Fig. 16.8 .
Postoperative Complications
Among the 305 SMILE papers, SMILE complications papers were categorized separately into several groups, such as dry eye, ectasia, diffuse lamellar keratitis, interface lamellar fluid, and others. This categorization included 21 papers. Among 21 papers, 14 were sorted into the dry-eye category. The papers dealing with meta-analysis of dry eye were excluded, and only the papers that reported original studies on dry eye were included. These inclusion criteria yielded nine papers. From these papers, dry-eye indicators—such as Ocular Surface Disease Index (OSDI) score, tear breakup time (TBUT), Schirmer test values, corneal sensitivity values, and tear osmolarity values—were collected preoperatively and postoperatively at 1 week, 1 month, 3 months, 6 months, and 12 months when reported. These values are organized in Tables 16.5 through 16.8 and weighted average of each of these dry-eye test values are calculated. The weighted average of each of these test values is shown in linear graphs to give a visual representation of the dry-eye parameters after SMILE vs after FS-LASIK.
| Study | Year | Eyes | Preop | 1 wk | 1 mo | 3 mo | 6 mo | 12 mo |
|---|---|---|---|---|---|---|---|---|
| Xia et al. | 2016 | 69 | 6.8 ± 3.0 | 6.4 ± 3.1 | 9.7 ± 6.5 | 6.0 ± 2.2 | 6.3 ± 2.1 | |
| Denoyer et al. | 2015 | 30 | 5.9±1.7 | 7±1.8 | ||||
| Wang et al. | 2015 | 47 | 9.87 ± 1.57 | 6.28 ± 1.35 | 8.21 ± 0.95 | 9.57 ± 0.93 | 9.83 ± 0.99 | |
| Xu and Yang | 2014 | 81 | 10.35 ± 3.28 | 6.79 ± 2.25 | 5.79 ± 2.38 | 7.39 ±2.36 | ||
| Li et al. | 2013 | 38 | 8.58 ± 4.42 | 4.32 ± 3.57 | 5.68 ± 4.84 | 5.03 ± 3.83 | 7.06 ± 3.85 | |
| Demirok et al. | 2013 | 28 | 9.0±1.2 | 9.9 ± 4-14 | 10.9 ± 2.8 | 11.6 ± 2.8 | 11.1 ± 2.7 | |
| Weighted average TBUT | 8.93 | 6.54 | 7.74 | 6.7 | 7.75 | 9.83 |
| Study | Year | Eyes | Preop | 1 wk | 1 mo | 3 mo | 6 mo | 12 mo |
|---|---|---|---|---|---|---|---|---|
| Xia et al. | 2016 | 59 | 7.8 ± 3.3 | 4.5 ± 3.2 | 4.2 ± 3.4 | 5.1 ± 2.2 | 6.6 ± 1.6 | |
| Denoyer et al. | 2015 | 30 | 5.1 ± 1.9 | 5.2 ± 1.8 | ||||
| Wang et al. | 2015 | 43 | 9.56 ± 1.35 | 6.53 ± 1.24 | 7.42 ± 0.96 | 8.19 ± 1.45 | 9.30 ± 0.89 | |
| Xu and Yang | 2014 | 97 | 11.09 ± 3.15 | 6.41 ± 2.96 | 5.67 ± 2.14 | 7.13 ± 2.56 | ||
| Li et al. | 2013 | 33 | 7.88 ± 5.57 | 4.70 ± 3.65 | 3.77 ± 2.91 | 4.43 ± 4.22 | 4.97 ± 3.57 | |
| Demirok et al. | 2013 | 38 | 9.1 ± 1.0 | 10.1 ± 2.3 | 10.7 ± 2.4 | 10.9 ± 2.8 | 10.4 ± 2.5 | |
| Weighted average TBUT | 9.46 | 6.19 | 6.11 | 6.41 | 7.16 | 9.3 |
| Study | Year | Eyes | Preop | 1 wk | 1 mo | 3 mo | 6 mo |
|---|---|---|---|---|---|---|---|
| Xia et al. | 2016 | 69 | 11.8 ± 5.5 | 9.1 ± 3.9 | 9.7 ± 6.1 | 12.6 ± 5.5 | 9.5 ± 4.1 |
| Denoyer et al. | 2015 | 30 | 13.2 ± 6.1 | 17.3 ± 8.2 | |||
| Xu and Yang | 2014 | 81 | 17.49 ± 7.48 | 16.98 ± 6.43 | 17.46 ± 9.25 | 17.13 ± 6.73 | |
| Li et al. | 2013 | 38 | 14.63 ± 7.51 | 13.51 ± 10.96 | 12.11 ± 7.58 | 14.14 ± 9.38 | 13.28 ± 8.72 |
| Demirok et al. | 2013 | 28 | 17.5 ± 6.5 | 17.3 ± 4.8 | 15.8 ± 5.9 | 16.6 ± 4.0 | 17.3 ± 4.4 |
| Weighted average Schirmer Test | 15.17 | 12.04 | 13.6 | 15.21 | 14.44 |
| Study | Year | Eyes | Preop | 1 wk | 1 mo | 3 mo | 6 mo |
|---|---|---|---|---|---|---|---|
| Xia et al. | 2016 | 59 | 11.8 ± 5.5 | 5.6 ± 3.5 | 7.6 ± 3.8 | 10.4 ± 5.5 | 9.3 ± 2.6 |
| Denoyer et al. | 2015 | 30 | 19.9 ± 10.5 | 16.9 ± 7.8 | |||
| Xu and Yang | 2014 | 97 | 18.55 ± 7.75 | 17.35 ± 7.72 | 18.22 ± 9.82 | 17.00 ± 7.20 | |
| Li et al. | 2013 | 33 | 15.36 ± 9.47 | 10.00 ± 8.28 | 10.90 ± 7.99 | 13.73 ± 9.54 | 13.17 ± 9.32 |
| Demirok et al. | 2013 | 28 | 18.5 ± 5.5 | 17.93 ± 6 | 17.5 ± 5.1 | 16.5 ± 4.4 | 16.9 ± 3.9 |
| Weighted average Schirmer Test | 14.25 | 14.01 | 14.53 | 15.19 | 14.63 |
| Study | Year | Eyes | Preop | 1 wk | 1 mo | 3 mo | 6 mo |
|---|---|---|---|---|---|---|---|
| Xia et al. | 2016 | 69 | 59.5 ± 1.7 | 57.2 ± 6.4 | 58.7 ± 4.1 | 59.6 ± 1.6 | 59.8 ± 0.8 |
| Li et al. | 2013 | 38 | 58.16 ± 3.37 | 29.59 ± 17.73 | 30.00 ± 16.37 | 37.92 ± 15.42 | 46.94 ± 11.73 |
| Demirok et al. | 2013 | 28 | 56.8 ± 4.7 | 45.6 ± 11.5 | 45.3 ± 10.5 | 49.3 ± 9.9 | 55.9 ± 4.9 |
| Li et al. | 2013 | 32 | 58.2 ± 4.5 | 23.2 ± 14.6 | 28.4 ± 13.9 | 34.2 ± 15.7 | 43.7 ± 11.7 |
| Wei and Wang | 2013 | 61 | 56.6 ± 4.5 | 47.5 ± 12.1 | 51.1 ± 10.5 | 57.3 ± 5.1 | |
| Weighted average corneal sensitivity | 57.98 | 44.5 | 45.99 | 50.54 | 53.13 |
| Study | Year | Eyes | Preop | 1 wk | 1 mo | 3 mo | 6 mo |
|---|---|---|---|---|---|---|---|
| Xia et al. | 2016 | 59 | 58.7 ± 2.6 | 21.0 ± 11.1 | 27.8 ± 13.6 | 36.1 ± 10.7 | 46.9 ± 8.9 |
| Li et al. | 2013 | 33 | 57.27 ± 6.26 | 20.61 ± 15.50 | 21.45 ± 15.34 | 27.50 ± 17.46 | 39.17 ± 16.09 |
| Demirok et al. | 2013 | 28 | 56.2 ± 5.0 | 30.3 ± 15.3 | 31.2 ± 14 | 37.5 ± 14.8 | 53.7 ± 5 |
| Li et al. | 2013 | 42 | 58.0 ± 3.8 | 15.4 ± 7.9 | 15.8 ± 9.5 | 25.6 ± 15.1 | 36.4 ± 16.1 |
| Wei and Wang | 2013 | 54 | 58.1 ± 4.3 | 22.1 ± 12.8 | 26.2 ± 17.2 | 37.9 ± 14.4 | |
| Weighted average corneal sensitivity | 57.87 | 21.33 | 24.54 | 28.55 | 43.78 |
| Study | Year | Eyes | Preop | 1 wk | 1 mo | 3 mo | 6 mo |
|---|---|---|---|---|---|---|---|
| Denoyer et al. | 2015 | 60 | 305.1 ± 12.5 | 300.3 ± 11.4 | |||
| Demirok et al. | 2013 | 28 | 303 ± 10 | 304 ± 11 | 303 ± 10 | 302 ± 6 | 306 ± 9 |
| Weighted average tear osmolarity | 303 | 304 | 304.36 | 302 | 301.91 |
| Study | Year | Eyes | Preop | 1 wk | 1 mo | 3 mo | 6 mo |
|---|---|---|---|---|---|---|---|
| Denoyer et al. | 2015 | 30 | 316.3 ± 11.6 | 315.0 ± 11.9 | |||
| Demirok et al. | 2013 | 38 | 298 ± 11 | 300 ± 8 | 302 ± 10 | 303 ± 6 | 304 ± 8 |
| Weighted average tear osmolarity | 298 | 300 | 308.31 | 303 | 308.85 |
Explanation of the Graphs (Dry Eye)
OSDI is a series of questionnaires designed to evaluate the reported severity of dry eye by the patient. The OSDI score ranges from 0 to 100, with score of 13 to 22 categorized as mild, 22 to 33 categorized as moderate, and greater than 33 categorized as severe. The weighted OSDI scores after both SMILE and FS-LASIK were comparatively similar, with SMILE showing a higher OSDI score at 1 week than FS-LASIK (23.95 vs 18.78). Over the follow-up period of 1 month, 3 months, and 6 months, however, SMILE showed slightly better dry-eye outcome in the OSDI score index than FS-LASIK (10.09 for SMILE at 6 months vs 17 for FS-LASIK; Fig. 16.9 ). Considering the weighted OSDI scores, SMILE does not exacerbate any dry-eye problem, but FS-LASIK does at least in the relatively short follow-up period of 6 months. Further studies with a longer follow-up time are warranted.
Both SMILE and FS-LASIK showed comparable outcomes in TBUT assessment. Remarkably, initial TBUT was in both groups slightly shorter than 10 seconds, indicating mild dry-eye syndrome before surgery. Both procedures shortened TBUT in the postoperative period but values returned to baseline after 12 months with both procedures ( Fig. 16.10 ).
The Schirmer test uses a paper strip to measure the ability of the eyes to produce tears typically in 5 minutes, whereby moisturing of the strip of greater than or equal to 15 mm is considered normal, 14 to 9 mm is considered mild dry eye, 8 to 4 mm is considered moderate dry eye, and less than 4 mm is considered severe dry eye. According to the weighted Schirmer test values, both SMILE and FS-LASIK had little effect on tear secretion during the postoperative period (14.44 for SMILE vs 14.63 for FS-LASIK; Fig. 16.11 ).
The corneal sensitivity test uses an aesthesiometer (typically Cochet-Bonnet) to measure the degree of tactile corneal sensation by retracting the metal filament in an increment of 0.5 cm (full extension length is 6 cm) until the patient is able to feel it contact the cornea. As the length of the filament extension decreases from 6 cm to 0.5 cm, the pressure changes from 11 mm/g to 200 mm/g. Patients seem to have lower corneal sensitivity values after FS-LASIK than after SMILE at all postoperative follow-up periods, which may be explained with the lesser trauma to the innervation associated with SMILE (as described earlier; Fig. 16.12 ).
Tear film osmolarity testing is meant to be used in conjunction with other dry-eye signs and symptoms to provide quantitative information mainly on the inflammatory component of dry-eye suspects. Abnormality in tear film is indicated either by an elevated reading of greater than 300 mOsm/L (indicating loss of homeostasis) or an intereye difference of greater than 8 mOsm/L (indicating tear film instability). Only two studies were found to report tear film osmolarity measurements after SMILE and FS-LASIK; thus the results may not be as valid. However, tear film osmolarity appears to show greater loss of homeostasis after FS-LASIK procedure than after SMILE (301.91 at 6 months for SMILE vs 308.85 at 6 months after FS-LASIK; Fig. 16.13 ).
