Refractive error induced by intraocular lens tilt after intrascleral intraocular lens fixation

To investigate the spherical shift of intraocular lens (IOL) tilt after intrascleral fixation. We retrospectively reviewed the medical records of patients who underwent flanged intrascleral IOL fixation with transconjunctival 25- or 27-gauge pars plana vitrectomy at the Department of Ophthalmology of the Jikei University Hospital. The minimum follow-up duration was 3 months. Second-generation anterior segment optical coherence tomography (CASIA2; TOMEY) was used to obtain the values of tilt and decentration of the intrasclerally fixated IOL and postoperative anterior chamber depth. We investigated the relationship between refractive error and various parameters, such as IOL tilt and decentration, axial length, and keratometry. In addition to our clinical investigation, we conducted optical simulations using Zemax to evaluate the spherical shift of the IOL tilt by means of the through-focus response and change in spherical equivalent power. The study involved 72 eyes of 67 patients. The degree of IOL tilt was correlated with the amount of refractive error (Spearman’s rank correlation coefficient [CC] = − 0.32; P = 0.006). In particular, a tilt angle greater than 10° strongly affected the refractive error. The postoperative anterior chamber depth also correlated with the refractive error (CC = 0.50; P < 0.001), as opposed to decentration (CC = − 0.17; P = 0.15), axial length (CC = − 0.08; P = 0.49), and keratometry (CC = − 0.06; P = 0.64). Optical simulations also revealed a myopic shift that exponentially increased as the tilt became greater. IOL tilts that are greater than 10° induce refractive error.


Introduction
Intraocular lens (IOL) implantation in eyes that lack or have insufficient capsular support is challenging. In such cases, the treatment options include anglesupported anterior chamber IOL, iris-fixed IOL, and transscleral IOL suture [1,2]. However, each of these techniques presents several potential problems. For instance, angle-supported anterior chamber IOL is associated with a high rate of endothelial cell loss [3] and uveitis-glaucoma-hyphema syndrome [4]. Although the implantation of an iris-fixed IOL can be an easy technique with low intrusiveness, issues relating to endothelial cell loss remain, and this technique is also limited by the condition of the iris and the anterior chamber depth (ACD) [5]. Transscleral IOL suture is also known to present problems, including suture-related complications, such as endophthalmitis due to suture exposure [6] and IOL dislocation induced by suture breakage [7].
Intrascleral fixation of both haptics in the ciliary sulcus was first introduced by Gabor et al. [8].
Although this method has lower risks of suture-related complications, it still requires suturing to close the scleral incision. Since then, various methods of intrascleral fixation have been reported [9][10][11][12]. The flanged intrascleral IOL fixation with double-needle technique reported by Yamane et al. [11,12] has become popular owing to its minimal invasiveness and ease of use. The tilt and decentration of IOL following intrascleral fixation have been reported to be rather small [11], but there are certain cases reported in which severe tilting and refractive error simultaneously occur [13]. Thus, we hypothesized that IOL tilting can cause refractive errors.
In 1988, Uozato et al. [14] quantitated the IOL decentration and tilt using Purkinje reflections and reported that they cause spherical and astigmatic refractive changes. Since then, however, several clinical studies have concluded that IOL tilt does not induce any refractive errors [11,15].
In 2015, second-generation anterior segment optical coherence tomography (CASIA2; TOMEY) was developed, and it was found to measure postoperative tilt and decentration with good repeatability [16]. This study aimed to evaluate the tilt, decentration, and ACD of scleral-fixated IOL using CASIA2 and investigate the potential relationship between these parameters and refractive error.

Methods
This study was conducted with approval from the institutional review board and in accordance with the ethical standards of the Declaration of Helsinki 1975, as revised in 2000 and 2008. The study protocol was approved by the Institutional Review Board of the Jikei University School of Medicine (approval number: 31-421[10003]), and all clinical procedures were performed according to the principles of the Declaration of Helsinki.
The medical records of consecutive patients who underwent intrascleral IOL fixation between September 2017 and October 2020 at the Department of Ophthalmology of the Jikei University Hospital were retrospectively assessed. The study protocol was disclosed, and the patients were given the opportunity to refuse participation in the study. The inclusion criteria were as follows: 20 years of age or older, treatment with intrascleral IOL fixation for aphakia; a subluxated crystalline lens, or a dislocated IOL; and a postoperative follow-up duration of longer than 3 months.
The exclusion criteria were as follows: patients with corneal disease (such as keratoconus or bullous keratopathy) and patients with incomplete operative or postoperative medical records. Although most postoperative complications were early and transient, iris capture and macular edema were observed several months following surgery in some cases. Hence, we collected post-treatment data in these cases (Supplementary Information).
The data collected included patients' age, gender, preoperative lens status, axial length, pre-and postoperative refractive power measured by autorefractometry, the best-corrected visual acuity, postoperative intraocular pressure (IOP), postoperative IOL status, ACD, IOL power, and intra-and postoperative complications. Postoperative hypotony and IOP elevation were defined as an IOP value of less than 6 mmHg and greater than 25 mmHg, respectively. Refractive error was determined by the difference between the predicted spherical equivalent calculated using the Barrett Universal II formula and the actual objective refraction value at the final visit.
The characteristics of the 72 eyes of 67 patients that were included in this study are presented in the Supplementary Information. The study group included 59 male and eight female patients with a mean age 61.3 ± 11.9 years (range, 37-86 years). The mean axial length was 25.28 ± 1.56 mm (range, 21.89-29.02 mm). We registered 6 cases of aphakia, 60 dislocated IOLs, and 6 subluxated crystalline lenses. The mean follow-up period was 14.3 months.

Surgical procedure
Intrascleral IOL fixation was performed as previously described [12], with some modifications. A 25-or 27-gaugepars plana vitrectomy was performed using a Constellation Vision System (Alcon Laboratories, Inc. Fort Worth, TX USA) under retrobulbar anesthesia. Phacoemulsification cataract extraction was performed for the subluxated crystalline lens. If the dislocated IOL was made of soft material, it was cut into two or three pieces and extruded from the 3.5-mm sclerocorneal incision. If the dislocated IOL was made of polymethyl methacrylate, it was removed through a 3 9 3-mm L-shaped scleral pocket incision [17].
A scleral tunnel incision was created using a trocar through the conjunctiva, sclera, and ciliary surfaces at the 2 o'clock, 10 o'clock, and temporal lower positions. A three-piece IOL (X-70, Santen Pharmaceutical, Co. Ltd. Japan) was inserted into the anterior chamber using an injector. An angled sclerotomy was made through the conjunctiva using a 30-gauge thinwall needle (TSK Ultra-Thin-Wall Needle; Tochigi Seiko, Japan). The scleral penetrating point was 2.0 mm posterior from the limbus at the 2 o'clock position, aimed 20°inward to the posterior chamber. The leading haptic was pushed into the lumen of the needle using forceps. A second sclerotomy was then performed using a 30-gauge thin-wall needle at the 8 o'clock position. The same procedure was performed for the second haptic. Both haptics were extracted, and their ends were cauterized using an ophthalmic cautery device (Accu-Temp Cautery; Beaver-Visitec International, Inc., USA) to make a flange. The flange of the haptics was pushed back and fixed into the scleral tunnel, and the IOL position was centered. A peripheral iridotomy was performed using the vitrectomy cutter after miosis to avoid iris capture of the IOL. Corneal suture was not required in all cases.

IOL tilt, decentration, and ACD measurements
Myopic shifts associated with the forward IOL movement persisted in eyes with in-the-bag fixated three-piece IOLs until 3 months post-surgery [18,19]. Therefore, the tilt and decentration of the IOL and ACD were measured using CASIA2 at least 3 months following surgery. The CASIA2 uses a 1310-nm swept-source laser wavelength, producing 16 anterior segment optical coherence tomography (AS-OCT) images from 16 different angles, and also provides a three-dimensional analysis of the results. Furthermore, it automatically measures the tilt and decentration of the IOL relative to the corneal topographic axis, which connects the fixation point on the corneal topographer to the corneal vertex.

Analysis
The SPSS software (version 20; IBM, USA) was used for all statistical analyses. The normality of all data samples was first assessed using the Shapiro-Wilk test. Spearman's rank correlation coefficient was calculated to evaluate the relationships of IOL tilt, IOL decentration, axial length, and average keratometry with the refractive error since the tilt and decentration data did not fulfill the criteria for normal distribution. The data were expressed as mean ± standard deviation.

Optical simulations with ZEMAX
We conducted optical simulations in addition to clinical research. A schematic eye [20] was modeled in the Zemax optical design program (Zemax, LLC, Washington, USA) to study the spherical shift of the IOL tilt by means of through-focus response (TFR) [21] and visual acuity (VA) estimated from the intersection point of modulation-transfer and modulation-threshold function. The modulation-threshold function was the reciprocal of the retinal-contrast sensitivity adapted from the results of Campbell and Green [22]. The values of every 5°from 0°to 40°were evaluated and subsequently interpolated to obtain smooth curves. 10 D, 20 D, and 27 D of spherical IOL X-70 (Santen Pharmaceutical, Co. Ltd., Japan) were used for this simulation. Anterior and posterior radii, central thickness, and a refractive index of X-70 were provided by Santen Pharmaceutical.
Four cases exhibited excessive IOL tilts of 35.8°, 34.2°, 28.8°, and 24.2°, with refractive errors of -10.42 D, -4.7 D, -5.82 D, and -2.03D, respectively. In 61 cases with a tilt less than 10°, the mean tilt was 5.6 ± 1.6°, and the mean refractive error was -0.46 ± 0.49 D. In 25 cases with a tilt less than 5°, the mean tilt was 3.0°± 1.2°, and the mean refractive error was -0.29 ± 0.52 D. Therefore, we determined that in cases in which the IOL tilt was less than 10°, the refractive error was relatively small. However, a tilt angle greater than 10°could significantly induce a refractive error.

Optical simulations with ZEMAX
The peak of TFR and VA revealed an exponential myopic shift as the tilt became greater at all powers of IOL (Fig. 2). These simulation results were consistent with the results of our clinical research.
To date, several studies have evaluated the tilt and decentration of the IOL. The Purkinje and Scheimpflug methods were reported in the 1980s [14,23,24]. AS-OCT and ultrasound biomicroscopy (UBM) have been used to demonstrate the position of implanted IOLs since the 2000s [25,26]. In fact, tilt and decentration measurements using AS-OCT have demonstrated higher repeatability than either the Purkinje method or the Scheimpflug method [27]. These studies have used the pupil center as a reference point, but their processes may not be optimal because the pupillary axis is affected by the shape of the pupil and because surgical invasion may cause deformation of the pupil. In previous studies, the corneal vertex was reported to be a safer and more effective reference point for fixating the center of the excimer laser beam in refractive surgery for myopia compared with the center of the pupil [28]. Because the corneal vertex is not affected by the shape of the pupil, the corneal topographic axis is a better reference line to evaluate IOL tilt and decentration compared with the pupillary axis.
CASIA2 is a second-generation AS-OCT developed in 2015. It has a 1310-nm swept-source laser wavelength. CASIA2 measures the tilt of the IOL as a reference to the corneal topographic axis, produces 16 meridional scan images from 16 different angles, and defines the greatest angle between the corneal topographic axis and the IOL in these 16 images as the IOL tilt.
Kimura et al. [16] reported that tilt and decentration measurements of the in-the-bag IOL obtained using CASIA2 had high repeatability under both nonmydriatic and mydriatic conditions, and they are not affected by pupil shape. The authors reported that the IOL tilt obtained by CASIA2 was higher than the value presented in previous studies.
Certain studies have referred to IOL tilt after intrascleral fixation. Kumar et al. reported that the mean angle between the IOL and the iris was 3.2°± 2.7°on the horizontal axis and 2.9°± 2.6°o n the vertical axis using a straight line passing through the iris pigment epithelium on either side as the reference line [29]. Other studies used a straight line passing through the iridocorneal angles on either side as a reference line and found that the mean IOL tilt angle was 2.5°-3.8°on the horizontal axis and 2.2°-2.8°on the vertical axis [11,12,16]. Recently, Jujo et al. reported that the IOL tilt measured by CASIA2 after intrascleral fixation was 3.5°± 3.0° [ 30].
In our series, the IOL tilt after intrascleral fixation was greater than the values reported in previous studies. There are two possible reasons for this. First, intrascleral fixation is a relatively new technique, and the introductory period for this process in our hospital was included in the investigation period. Furthermore, the learning curve of the surgical procedure should be considered. Of the 72 cases who underwent surgery, four had a very high tilt (7th, 16th, 29th, and 45th), which unavoidably influenced the average tilt (Supplementary Information; cases 7, 16, 29, and 45). When IOL is positioned within the lens capsule, the IOL and capsular are supported by lens zonules, thus maximizing the chances of optimal surgical and refractive outcomes. Even if the posterior capsule is violated, the IOL may be placed in the ciliary sulcus and maintain excellent visual outcomes [31]. However, transscleral-sutured IOL and scleral-fixated IOL may have extensive tilt and decentration values, which will subsequently change the value of ACD due to the lack of a zonular support.
Despite the fact that it is well known that longitudinal IOL displacement induces refractive errors [32,33], there have been a limited number of reports on the potential correlation between tilt, decentration, and refractive error. In 1988, Uozato et al. [14] reported that IOL tilt induces spherical and astigmatic changes, and IOL power exponentially increases as the tilt becomes larger. The authors concluded that tilt should be less than 5°. In 1990, Erickson et al. [33] simulated spherical and cylindrical refractive errors induced by IOL tilt and demonstrated that the 3°, 7°, 11°, and 15°tilting of 19 D IOL can induce exponentially increasing myopic shift of 0.04, 0.20, 0.52, and 0.98 D, respectively. Since then, studies that used either AS-OCT or UBM have reported that IOL tilt does not have any effect on astigmatism and refractive error, unlike previous studies [11,16]. This discrepancy was possibly caused by the relatively small IOL tilts measured using AS-OCT and UBM. For instance, Kurimori et al. [13] reported on two patients after intrascleral fixation who exhibited significantly high values of tilt, i.e., 38.1°and 25.3°, and a strong myopic shift of -4.50 and -2.75 D, respectively. A second surgery to reduce the length of both haptics by 2-3 mm simultaneously decreased the excessive tilt and the myopia-shifted refraction in both cases. In our study, no patients, including the four cases with extreme postoperative IOL tilt, desired second surgery.
In this study, the mean refractive error was -0.29 ± 0.52 D in cases of less than 5°tilt, and the mean refractive error was -0.46 ± 0.49 D in cases of tilt values that were between 5°and 10°. The refractive error exponentially increased as the tilt became greater. An extremely large tilt and strong refractive error were simultaneously identified in four cases as shown in Fig. 1a. Our findings are consistent with previous reports, but also with the optical simulations conducted by Zemax (Fig. 2). Therefore, we determined that 10°of IOL tilt was an acceptable cutoff value when investigating the effect of IOL tilting on refractive error.
Snell's law describes the relationship between the angles of incidence and refraction when light passes through a boundary between two different isotropic media. As the IOL tilt affects the angle of incidence from the aqueous humor to the IOL, the angle of refraction from the IOL to the aqueous humor changes, thus leading to refractive errors. Our study did not find any correlation between decentration and refractive error. However, Uozato et al. [14] found that this relationship was even smaller than that between tilt and refractive error. Thus, a study with more cases would be desirable in the future. Although axial length and keratometry have also been reported to cause postoperative refractive error, our study found no correlations. This result may have been caused by the use of the Barrett Universal II formula, which has a low prediction error [34].
The limitations of our study were its retrospective nature and a learning curve for scleral-fixated IOL implantation.
In conclusion, intrasclerally fixated IOL tilts that are greater than 10°strongly induce refractive errors.
Author contributions The authors thank Tjundewo Lawu for conducting optical simulation using ZEMAX.
Funding No funding was received for this study.
Data availability Clinical data supporting the findings of this study are available within the article and its Supplementary Information.

Declarations
Conflict of interest Financial interests: Author T. Nakano has received Grant from CREWT Medical Systems, Kyowa Medical, Kuribara Medical Instruments, Kowa, Tomey, Otsuka Pharmaceutical, Senju Pharmaceutical, MSD, Pfizer, Alcon Japan, Santen Pharmaceutical, NIDEK, AMO Japan Bayer, and IOL MEDICAL. Author T. Nakano has also received Personal Fees from Kowa, Otsuka Pharmaceutical, Senju Pharmaceutical, Santen Pharmaceutical, Nitto Medical, and Nikon. Consent for participation The study protocol was disclosed, and patients were given the opportunity to refuse participation in the study.
Consent to publishing The study protocol was disclosed, and patients were given the opportunity to refuse participation in the study.