Flapless and Conjunctiva-Sparing Technique for Transscleral Fixation of Intraocular Lens to Correct Refractive Errors in Eyes without Adequate Capsular Support

Purpose To evaluate refractive outcomes, intraocular lens (IOL) power calculation, and IOL position following a novel conjunctiva-sparing transscleral fixation technique. Methods Forty-one eyes of 40 patients managed with a flapless transscleral-sutured technique were included. Preoperative and postoperative refractive errors (spherical equivalents, SE) were compared. IOL position was assessed on the Scheimpflug images. IOL power was calculated by SRK/T, Holladay 1, and Hoffer Q formulas. Results The mean age was 57.39 ± 14.83 years (range: 26 to 79 years), and the mean follow-up was 7.46 ± 6.42 months (range: 1 to 24 months). Surgical indications were aphakia (n = 14), subluxated lenses (n = 3), and IOL dislocation (n = 24). The SE was 4.50 ± 6.38 diopter (D) (range: −3.75 to 13.75 D) preoperatively and −1.68 ± 1.57 D (range: −5.50 to 1.13 D) postoperatively (P < 0.001). The mean tilt angle and decentration were 2.90° ± 1.93° (range: 0.39° to 9.10°) and 0.23 ± 0.19 mm (range: 0.02 to 0.94 mm) vertically, and 1.75° ± 1.41° (range: 0.24° to 7.65°) and 0.18 ± 0.19 mm (range: 0.02 to 1.06 mm) horizontally, which were clinically insignificant. All three IOL formulas produced myopic errors (range: −0.29 to −0.50 D). The SRK/T had the lowest median absolute error (0.55 D), followed by the Holladay 1 (0.70 D) and the Hoffer Q (0.74 D). The three formulas had the same percentage of prediction errors (PEs) within ±0.5 D (43.48%), while the Hoffer Q had the highest percentage of PEs within ±1.0 D (82.61%). Conclusion The present technique can serve as an alternative approach for transscleral IOL fixation and refractive correction in eyes with compromised capsular support, ensuring the stability of IOLs and reasonable IOL power calculation accuracy.


Introduction
Cataract surgery has evolved from the restoration of visual function to refractive correction. Refractive correction in eyes with inadequate capsular support can be accomplished using anterior chamber IOL, iris-fxated IOL, or scleralfxated IOL [1][2][3][4][5]. Implanting a posterior chamber IOL via a scleral-fxated method has several inherent advantages over other techniques, allowing the implanted IOL to be positioned closer to the original crystalline lens with a reasonable distance from anterior segment structures. Transscleral-sutured fxation, a commonly performed scleral-fxated technique, typically requires the large opening of conjunctival tissues and the creation of scleral faps to bury the suture ends and knots [6][7][8]. Te most commonly used 10-0 polypropylene in sutured-fxation methods is prone to loosening and breakage, leading to IOL tilt, decentration, or even up to a dislocation rate of 18% to 28%, and thus the occurrence of signifcant postoperative refractive errors, of which IOL tilt angle larger than 15°cannot be corrected with spectacles [9,10]. We have previously described a fapless and conjunctiva-sparing technique for transscleral IOL fxation using the 8-0 polypropylene [5]. We herein present the results of refractive correction and IOL position after this minimally invasive surgical technique. followed throughout the study. Te surgical indications were 1) aphakia, 2) subluxated lenses with more than 8-clockzonulodialysis that had less chance of preserving the capsular bag, and 3) dislocated IOLs. Before surgery, surgical procedures and potential risks were explicitly explained; meanwhile, informed consent was acquired from all patients.

Surgical Techniques.
Te surgeries were performed under retrobulbar anesthesia by one of the authors (J. H). Te supplemental video (see Videos 1 and 2, Supplement Digital Content 1 and 2) demonstrates the procedures. A twin-armed single 8-0 polypropylene suture (Prolene, Polypropylene Suture; Ethicon, Johnson & Johnson, New Brunswick, New Jersey) was cut at its middle. Te suture was introduced into the eye using a 30-gauge needle from the fxation site, 2.0 mm posterior to the limbus, with the aid of a suture-in-needle technique [11]. Te suture loop inside the globe was grasped and externalized from the eye by an endopen forceps, through the clear corneal incision. Te doublestrand suture was twined around the haptic on the centripetal side. Ten, the loop was lassoed around the end of the haptic. Te suture was further pulled to fasten the modifed cow-hitch knot ( Figure 1). Te same set of manipulations was performed to fxate the opposite haptic. For cases with subluxated lenses, the present technique was performed after phacoemulsifcation by the aid of temporary capsule retractors. A limbal vitrectomy was conducted if necessary. For cases with a dislocated IOL, the haptics were externalized through corneal paracentesis before the knotting technique was adopted. After introducing both haptics into the eye and adjusting the suture tensions to center the IOL, the conjunctiva-sparing and fapless fxation technique described in our previous publication was adopted to fxate the suture to the scleral wall [5]. In short, performing the back-and-forth intrascleral suture pass for defnitive fxation. Te two ends of the suture were knotted into the sclerotomy. Another overhand knot 2.0 to 3.0 mm from the fxation knot was anchored intrasclerally by the aid of a 30gauge needle.

IOL Tilt and Decentration
Measurement. Te Pentacam examination was performed by one experienced technician on all patients under scotopic conditions. Two Pentacam Scheimpfug images on 90°and 180°meridians were analyzed using ImageJ software (version 1.8.0) to measure postoperative IOL position (IOL tilt and decentration). Te anterior and posterior IOL surfaces were frst plotted to determine the IOL axis, the line passing the IOL midpoint and perpendicular to the line of intersection of the IOL surfaces. Te pupillary axis was defned as the line passing the pupil center and connecting the anterior corneal center of curvature, the IOL tilt angle as the angle between the IOL axis and the pupillary axis, and the IOL decentration as the distance between the two axes ( Figure 2).

Refractive Prediction Error.
As IOL formulas using ACD as variables to calculate IOL power were impractical for aphakia, three formulas (SRK/T, Holladay 1, and Hofer Q formulas), independent of preoperative ACD, were used. Te User Group for Laser Interference Biometry website (https://ocusoft.de/ulib/index.htm) was browsed to determine the optimized lens constant values for each formula. Te three formulas were calculated online (https://www. eyecalcs.com). Prediction error (PE) was calculated as the postoperatively measured refractive spherical equivalent (SE) minus the predictive SE of the implanted IOL; the negative PE indicated that the formula produced myopic errors and the positive PE hyperopic errors. Te arithmetic PE, the mean absolute error (MAE), the median absolute error (MedAE), and the percentage of eyes with a PE within ±0.50 diopters (D), ±1.00 D, and ±2.00 D were evaluated.

Statistical Analysis.
Te statistical analyses were executed using SPSS software (version 26; IBM Corp, New York, NY), Prism (version 8.0.1, GraphPad Software, San Diego, CA), and Excel spreadsheet (Microsoft Corp). Te normality of data distribution was determined using the Kolmogorov-Smirnov test. Te preoperative and postoperative parameters were compared by the paired t-test or Wilcoxon signed-rank test based on the data distribution. Te Friedman test with Dunn's post-test was applied to compare the arithmetic PEs and AEs between the formulas. Te Cochran Q test was utilized to assess the percentage of eyes with PEs within ±0.50 D and ±1.00 D. For multiple comparisons, the Bonferroni adjustment was used. Statistical signifcance was defned as a P value less than 0.05. Clinically signifcant tilt and decentration were defned as tilt angles larger than 7°and decentration larger than 0.4 mm, respectively [12].
Te accuracy of IOL power calculation was assessed in 23 eyes that underwent IOL implantation. All three IOL formulas produced myopic PEs, and the SRK/T formula had a more signifcant myopic error than the Holladay 1 formula    Te SRK/T formula had the lowest MedAE (0.55 D), which was statistically lower than the Hofer Q formula (0.74 D, P � 0.02) ( Table 4). Te percentage of PEs within ±0.5 D was the same for the three formulas (43.48%). Te Hofer Q formula had the highest percentage of PEs within ±1.0 D (82.61%) than the other two formulas (SRK/T: 69.57%; Holladay 1 : 73.91%) (Figure 3). However, there was no statistical signifcance among the three formulas concerning the percentage of PEs within ±0.5 D, ±1.0 D, and ±2.0 D (all P > 0.05).
At the time of ciliary sulcus penetration, a mild and temporary hemorrhage was noted in two eyes. Tere were no other observed intraoperative complications. Postoperatively, hypotony occurred in two eyes. With close observation and routine medications, the IOP normalized within the frst week. Transient IOP spike was detected in six eyes, of which fve eyes were managed with antihypertensive medications, and the IOP returned to normal after one week. Te other one with elevated postoperative IOP, diagnosed with traumatic angle recession glaucoma before surgery, received the implantation of a glaucoma drainage device two months later. During the follow-up period, two cases of IOL pupillary capture and one case of suture exposure were observed, which were managed by paired suture. Te IOLs remained well centered (Figure 4). No scleral atrophy, suture lack, chronic corneal edema, hyphema, vitreous hemorrhage, retinal tear, retinal detachment, or IOL redislocation was detected.

Discussion
Refractive correction in eyes with insufcient capsular support remains challenging. First, current formulas assume the accomplishment of in-the-bag IOL implantation; however, due to diferent IOL types and surgical techniques, the position of IOLs placed in such eyes can be unpredictable. Second, the accuracy of IOL power calculation would be compromised by factors contributing to ocular biometry measurement errors, including poor visual acuity for eye fxations, comorbidities, and altered refractive index following vitrectomy [13][14][15]. Moreover, surgeons typically use anterior chamber IOL, iris-fxated IOL, and scleral-fxated IOL to correct refractive errors in the absence of adequate capsular support, all of which are more technically difcult than routine cataract surgery [1][2][3][4][5]. Currently, the fanged fxation technique for three-piece IOLs is a reliable and widely used technique; however, it has limitations on the types of IOLs that can be used. Te fanged technique is only practicable in certain types of three-piece IOLs [16,17]. As a fange created by thermos-cauterization is the key point of the technique, certain types of IOLs, whose haptics are made from materials that cannot be reshaped by thermosplasticity, are impractical for this technique [16,17]. Transscleral-sutured fxation remains a vital method to fxate IOLs; compared with anterior chamber IOL and iris-fxated IOL, it places an IOL closer to the original crystalline lens and has the advantages of lower demand for corneal endothelium, iris structure, and angle status [4,6]. Furthermore, compared with the fanged method, transscleralsutured fxation is suitable for a broader range of IOL types. However, most transscleral-sutured methods are time-consuming and traumatic because of the need to create scleral fap(s)/pocket(s)/groove(s). In addition, concerns about the long-term stability of the conventionally used 10-0 polypropylene are growing due to the risk of suture breakage or degradation over time [6][7][8][9][10].
To fxate the suture to the scleral wall with minimal invasiveness and to reduce suture-related long-term instability, we proposed a conjunctiva-sparing transscleral suture fxation technique [5]. Te present technique has several advantages over traditional transscleral-sutured fxation. First, it simplifes surgical procedures with minimal invasiveness, enabling the introduction of a suture loop, securing the suture to the haptics, and burying the suture free ends and knots without the need for creating faps [10,18]. Second, the 8-0 polypropylene used in this technique has greater durability and fatigue resistance than the conventional 10-0 polypropylene, which is promising for establishing long-term stability between the suture and the haptics and preventing suture breakage. Tird, the modifed cow-hitch knot used to anchor the suture to the haptics is a non-free-end fxation technique. Te incarceration ability of the fxation technique without a free end is secured by the      Journal of Ophthalmology 5 friction between the suture and the haptic. Te friction of diferent types of lasso techniques used to fxate IOLs is mainly determined by two factors: the contact areas between the suture and the haptic; and the overlapping areas of the suture. Compared with the conventional 10-0 polypropylene suture, the 8-0 polypropylene used in this technique has a wider diameter, providing more contact areas to enhance the knot's friction and reduce its loosening and slippage. Te postoperative position of implanted IOLs varies with intraocular conditions and surgical techniques [19,20]. Previous studies have confrmed that a certain degree of IOL tilt occurs even in patients undergoing routine cataract surgery, which is generally well tolerated [19,21]. Others reported that the mean tilt angle after an in-the-bag IOL implantation ranged from 1.5°to 4.8° [19,22,23]. A transscleral-sutured IOL is expected to have a greater tilt angle due to the lack of capsular support. In a study comparing IOL position between conventional scleral suture fxation and primary in-the-bag implantation, Hayashi et al. found that the IOL fxated to the scleral wall exhibited a more pronounced mean tilt degree (scleral-sutured: 6.35°v s. in-the-bag: 3.18°) [24]. However, recent advancements in IOL fxation techniques in the absence of capsular support have led to a decrease in the degree of IOL tilt. Yamane et al. reported a mean IOL tilt degree of 3.4°after fanged IOL fxation [3]. Kumar et al. found that the tilt angle after glued fxation was 3.2°and 2.9°in horizontal and vertical axes, respectively [8]. More recently, the study using a novel scleral anchored IOL revealed that the tilt angle was approximately 1° [25]. Te mean tilt angle was 2.33°± 1.36°in our series, which is lower than 7°, indicating a nonsignifcant tilt and is comparable to that of routine cataract surgery. Decentration is another critical parameter infuencing longterm visual outcomes. It has been confrmed that patients with clinically signifcant decentration (decentration > 0.4 mm) would experience a worse visual quality than those without it [12]. Previous fndings have demonstrated that IOL within the capsular bag exhibited less decentration amount than conventional transscleral-sutured one (in-thebag: 0.29 mm vs. transscleral-sutured: 0.62 mm) [24]. Others showed that the range of postoperative IOL decentration following transscleral-sutured fxation was 0.31 to 0.45 mm [7,26]. Te introduction of intrascleral sutureless fxation has simplifed the surgical procedures and reduced the potential risks of suture-related issues; however, reportedly, the decentration amount of IOL fxated via intrascleral sutureless approaches was approximately 0.35 to 0.42 mm, which was still larger than that of routine cataract surgery [19,[26][27][28]. In our study, the mean decentration was 0.21 ± 0.15 mm, consistent with routine cataract surgery [19,28]. Possible explanations for the diferent extents of decentration between the intrascleral and the present technique were as follows: frst, intrascleral sutureless methods require inserting the haptics into scleral tunnels, which are not securely fxated, and the haptics might slide before reaching its fnal position. In contrast, the present technique was a defnitive knotting approach to secure the haptics to the scleral wall. Second, after sclerotomy, an angled intrascleral pass of the needle to create an intrascleral tunnel is needed to embrace the leading and trailing haptics in fanged fxation; therefore, there is variability in the location and/or length of the intrascleral tunnel compared to the in situ knotting fxation in the present technique. Regarding IOL power calculation accuracy, all three formulas produced myopic errors, ranging from −0.29 to −0.50 D, in agreement with previous transscleral-sutured results, as the IOLs were implanted outside the capsular bag [24,29]. Te principle of the three formulas might account for the results because these formulas were introduced to predict IOL power within the capsular bag; however, the fxation sites in the present technique were 2.0 mm posterior to the limbus, anterior to the intracapsular implantation position, leading to myopic errors. Te reported MAE following intrascleral or transscleral fxation using the thirdgeneration formulas was 0.61 to 0.86 D, which agreed with our results [30,31]; whereas the values of MAE in our series were larger than those of routine cataract surgery, suggesting the compromised performance of formulas when the IOL was placed outside the capsular bag [32]. Previous studies reported that 30% to 45% of eyes with a PE within ±0.5 D in patients receiving fanged or transscleral-sutured fxation, in accordance with our study (43.48%) [30,33]. Furthermore, for cases following fanged or transscleral-sutured fxation, the percentage of a PE within ±1.0 D was approximately 60% to 75% [30,33]. Te U.K. National Health Service guidelines recommended that at least 85% should have a PE within ±1.0 D [34]. In the current study, nearly 85% of eyes had a PE within ±1.0 D using the Hofer Q formula, indicating that the present technique can ofer satisfactory calculation accuracy even in such complicated surgeries. However, given that nearly 95% of cases can achieve a PE within ±1.0 D in routine cataract surgery, large sample studies are needed to investigate contributors to calculation errors in patients receiving the present technique [32].
Te present technique can safely manage eyes with inadequate capsular support. Tere were no intraoperative complications besides a mild and temporary hemorrhage. Te postoperative complications were transient IOP abnormalities, IOL pupillary capture, and suture exposure. Transient IOP abnormalities were noted in 8 cases. Tey were successfully managed with topical medications except for one case who was diagnosed with traumatic angle recession glaucoma and received antihypertensive surgery according to the preoperatively determined surgical plan-fxating an IOL frst and implanting a glaucoma drainage device afterward. Previous studies reported that IOL pupillary capture and suture exposure rates after scleral fxation were 2% to 8% and 2.5% to 10%, respectively [3,[35][36][37]. In our study, two cases of IOL pupillary capture and one case of suture exposure were observed, and the incidence rates were within the abovementioned ranges. Previous studies with a mean follow-up of 6 to 7 months reported that the rate of IOL subluxation/ dislocation after transscleral-sutured fxation was 3.0% to 7.8% [38,39]. Also, due to suture degradation, IOL redislocation with a long-term incidence rate of 18% to 28% remains a major concern after transscleral-sutured fxation [9]. Tough it was not observed in the present study, a longer follow-up is needed. 6 Journal of Ophthalmology We also acknowledged the limitations. Tis is a singlearm clinical trial whose IOL position and IOL power calculation accuracy were compared with those of previously published studies. Moreover, the sample size of this study is small, and the follow-up period for some cases is relatively short. Nevertheless, the present technique provides a reliable alternative to manage refractive errors in eyes without adequate capsular support.

Data Availability
Te data analyzed in this study are available from the corresponding author upon reasonable request.

Ethical Approval
Te registration number of the study was ChiCTR2000035707.

Conflicts of Interest
Te authors declare that they have no conficts of interest.