In vivo comparison of implantation behavior and laboratory analysis of two preloaded intraocular lens injectors

Abstract

Purpose

We evaluated implantation behavior and injector damage of two different IOL injector systems, the Multisert and the AutonoMe.

Design

Prospective, randomized, comparative study with laboratory investigation.

Methods

We examined used injectors from 48 bilateral cataract cases and assessed video recordings of each Implantation. All eyes were intraindividually randomized for treatment with one of the two IOL injectors. Implantation videos were reviewed for inadvertent events and the time for different implantation steps was determined. The injector nozzle tips were examined using light and scanning electron microscopy (SEM). Damage was graded using the Heidelberg Score for IOL injector damage (HeiScore). Three months postoperatively, IOLs were assessed for material changes.

Results

Implantation was without critical events in 96 of 96 eyes. Mean implantation time was 41.90 ± 7.11 s with the Multisert and 52.22 ± 12.06 s with the AutonoMe. In the AutonoMe group, we observed 4 eyes (8.3%) with a failed docking attempt, 28 eyes (58.3%) with a haptic adherence, one case (2.1%) of straight leading haptic and 2 cases (4.2%) of intrawound IOL manipulation. There were no events observed in the Multisert group. The mean HeiScore values were 0.87 ± 0.61 and 3.68 ± 0.47 for the AutonoMe and Multisert. 3 months postoperatively, IOL material changes were absent.

Conclusions

Both injectors allowed safe and controlled implantation. Using Multisert, implantation behavior was more consistent. The injectors showed different damage profiles with a higher damage score for the Multisert.

The study is registered at the German Clinical Trials Register (Deutsches Register Klinischer Studien; reference number: DRKS00007837).

Introduction

An important step in cataract surgery is the implantation of the IOL, and numerous developments in instrumentation and IOL design have improved and standardized this step. Nowadays, an injector rather than a forceps is used for this purpose. In manual injectors, the first step is manually loading the IOL into the injector system to prepare it for implantation. A significant achievement was the development of preloaded IOL injectors which eliminates the need for the surgeon or assisting staff to touch the IOL with a forceps and manually load it into the injector. This not only reduces preparation time and consequently surgery time, but also can minimize or eliminate some surgical risks.

Manual loading errors or loading variability and IOL damage resulting from incorrect IOL configuration during injection are avoided. Apart from that, the risk of contamination of the IOL with foreign bodies or microbial contamination is reduced.^1,2 As preloaded injectors proved to be safe and effective, they gained popularity and are nowadays preferred by many surgeons.^3,4 A further development of preloaded injectors are automated injectors, in which advancement of the IOL via an automated mechanism allows a standardized implantation procedure. This type of injector has the characteristics of a preloaded injector, but is additionally characterized by an automated implantation mechanism. ⁵

One trend in the recent development of preloaded IOL injectors is to make injectors more suited to small-incisions surgical techniques, making nozzles which are compatible with these incision sizes and designs that aim at improved wound protection.^6,7 Such a new preloaded injector for which published data are currently sparse is the Multisert (Hoya). In this study, we investigated Multisert's implantation behavior, checked for injector damage after IOL implantation, and compared it to another IOL delivery system from a different manufacturer, the AutonoMe (Alcon Laboratories Inc.), which is an automated implantation system.

Materials and methods

Preloaded IOL injectors

The Multisert injector (Hoya, Singapore) is preloaded with a single-piece, glistening-free, hydrophobic acrylic monofocal IOL, the Vivinex model XY1. ⁸ It allows for either a single-handed push injection or a two-handed screw-injection. In this study, all implantations were performed using the push-mode. The injector features the so called ‘insert shield’ (IS), which can be kept in the default position for IOL delivery into the capsular bag or moved to the advanced position for IOL delivery through the incision wound allowing better wound protection.

The AutonoMe injector (Alcon Laboratories Inc., Fort Worth, Texas, USA) is preloaded with a single-piece, glistening-free, hydrophobic acrylic monofocal IOL, the Clareon model CNA0T0, and designed for single-handed IOL implantation. ⁹ It features an automated CO₂-powered delivery mechanism and a depth guard design to limit injector advancement.

Patient enrolment and randomization

The study was approved by the local Ethics Committee of the University of Heidelberg and is registered at the German Clinical Trials Register (Deutsches Register Klinischer Studien; reference number: DRKS00007837). The study was performed in accordance with the tenets of the Declaration of Helsinki. Written informed consent for the treatment and participation in the study was obtained.

In this prospective, randomized, comparative study, we enrolled 48 patients with bilateral clinically significant cataracts who were scheduled for phacoemulsification extraction and posterior chamber IOL implantation. The minimum age was 18 years, and all provided signed informed consent. Patients were excluded if they had ocular comorbidities that were considered to potentially influence the outcome (e.g., pupil abnormalities, previous ocular surgery, corneal or intraocular pathologies) or intraoperative complications that could compromise IOL stability. Uncontrolled or acute systemic disease as well as pregnancy and lactation were also exclusion criteria. Only patients with a calculated IOL power within the commercially available spherical power range and with a difference of no more than 1.5 diopters between both eyes were enrolled. All eyes were intraindividually randomized for treatment with one of the two IOL injector systems using a randomization list generated before the start of the study. For eyes in the Multisert group, the IS was used in 23 randomly selected patients in the advanced position. Patients were blinded to the intervention they received in each eye until the study was completed; blinding of the investigators was not possible because of the examinations required.

Surgical procedure and video analysis

All surgeries were performed by the same experienced surgeon (R.K.) and with the patient under local or general anesthesia, according to the patient's preference.

A Centurion phacoemulsification machine and a 0.9-mm 45° ABS Intrepid Balanced phaco tip with a NanoSleeve (Alcon Laboratories, Inc) were used to allow for micro-coaxial phacoemulsification. For all patients, the main corneal incision was 2.2 mm for both IOL injectors, and it was placed at the 12 o’clock position, these being standard procedures at our clinic.

The injector was prepared following the manufacturer's instructions using a qualified ophthalmic viscosurgical device. After IOL folding and advancement immediately prior to injection of the lens through the incision, the leading and trailing haptic configurations were inspected and misconfigurations such as looped or straight leading haptic were recorded. In cases with critical misconfigurations such as straight trailing haptic, the injector was exchanged. In cases with acceptable haptic configurations, implantation was performed using a through-the-wound technique.

The surgery of all patients was recorded, and the videos were reviewed to check for unwanted events during the implantation by an independent observer (I.D.B) following criteria suggested by Oshika et al. ¹⁰ and previously published work of our group. ¹¹ The evaluation included noncritical events (haptic adherence, failed docking attempt, intra-wound IOL manipulation) as well as potentially critical events (sudden IOL ejection, overriding plunger, gross IOL damage, flipped IOL orientation, trapped trailing haptic). The videos were also used to note the duration of each surgical step: determining the time of docking of the injector, IOL injection and IOL unfolding and the overall implantation time.

Quantification of injector damage

All the used injectors were inspected and photographed with a light microscope and camera. (Olympus BX50 and Olympus C-7070, Olympus Optical Co. Ltd, Shinjuku, Tokio, Japan). Scanning electron microscopy (SEM) imaging was performed at the Max Planck Institute for Polymer Research under low-voltage (< 1 kV) conditions with an SU8000 microscope (Hitachi, Chiyoda, Tokio, Japan).

The Heidelberg Score for IOL injector damage (HeiScore) ¹² was used for damage evaluation, but modified to allow inclusion of the SEM images. Six grades of damage were distinguished:

• Grade 0: No damage.

• Grade 1: Slight scratches or deformation.

• Grade 2: Deep scratch or moderate deformation of the nozzle tip.

• Grade 3: Extension of the injector tube but not a full thickness crack.

• Grade 4: Full thickness Crack.

• Grade 5: Burst of the injector tube.

Deviating from the scale developed by Fang et al. ¹² slight or moderate deformation were observed for the AutonoMe injector using SEM and were graded the same as slight scratches (grade 1) and deep scratches (grade 2). Figure 1 shows representative images for the different grades. Each injector was classified based on light and scanning electron microscopy images.

Figure 1.

Representative images of injector damage for different damage profiles. The top row shows examples for the Multisert and the bottom row for the AutonoMe. (a) (b) Examples for injector damage classified as ‘extension’. (c) Example for injector damage classified as ‘crack’. (d) No damage. (e) Example for injectors classified as ‘slight scratches/slight deformation’. (f) Example for injector damage classified as deep scratches/moderate deformation’.

Postoperative follow-up

Three months postoperatively, slit-lamp examination and dilated funduscopy were performed with the intention of identifying changes in the IOL material and the presence of intraocular foreign bodies.

Statistical analysis

Sample size calculation was based on the results we obtained in a previous study and was performed using G*Power v. 3.1.9.6. ¹¹ Power calculation indicated that 44 implantations were necessary to detect a difference between groups at a 5% significance level with a power of 95.5%.

All data were collected in an Excel file (version 16.5; Microsoft Corporation). For numerical outcome measures we calculated mean values and standard deviations. The Shapiro-Wilk test was used to test for normal distribution. The Wilcoxon signed-rank test for dependent parameters or a t-test for dependent parameters were used when appropriate to test for differences between groups. A p-value of less than .05 was considered statistically significant. Jamovi (version 1.6.23.0) was used for all statistical tests.

Results

In total, 58 patients were assessed for eligibility. One patient did not meet the inculsion criteria, three patients declined to participate. Of 54 patients randomized, four declined second eye surgery and were therefore excluded from the analysis. One patient was lost to follow-up and one patient discontinued participation due to other health issues. The data of 48 patients could be analyzed.

The patients’ mean age was 72 ± 13 years. 22 of 48 patients (45.8%) were male.

Implantation behavior

Mean overall implantation time with the AutonoMe was 52.55 ± 12.06 s and 41.90 ± 7.11 s with the Multisert (p < .05). Implantation time did not differ between patients treated with or without the IS (p > 0.05). Figure 2 shows the time needed for different steps of the procedure in both groups.

Figure 2.

Time of IOL implantation. The implantation time was longer for the AutonoMe injector, which was mostly due to a slower IOL unfolding. * Statistically significant difference.

In the AutonoMe group, we observed a failed docking attempt in 4 eyes. One case of uncritical misconfiguration, a looped leading haptic, was observed in the AutonoMe group as well. An adherence of the leading haptic to either the IOL optic or the trailing haptic was seen in 28 eyes. The adherence either was resolved spontaneously or after gentle manipulation. We also observed 2 cases where part of the IOL haptic was trapped in the corneal wound. In both cases, intrawound IOL manipulation with a second instrument was necessary to complete the IOL implantation. In all cases, IOL implantation could be completed without complications, and none of the injectors used had to be discarded. Figure 3 illustrates the occurrence of inadvertent events during IOL injection.

Figure 3.

Percentage of systems showing inadvertent events during IOL injection for both injectors.

For the Multisert group, we did not observe any unwanted events during IOL implantation.

The supplementary video illustrates the implantation behavior of both injector systems.

Nozzle tip damage

IOL power ranged from 16.00 to 25.00 diopters (D) in the AutonoMe group and from 16.00 to 26.00 D in the Multisert group. Mean values did not differ significantly between both groups (p > .05).

The distribution of different grades of injector damage is depicted in Figure 4.

Figure 4.

The distribution of injector damage profiles for both injectors.

The mean value for the HeiScore was 0.87 ± 0.61 for the AutonoMe and 3.18 ± 0.38 for the Multisert.The median was 1 for the AutonoMe and 3 for the Multisert. For the AutonoMe no damage could be identified using light microscopy or SEM in 25.00% of cases. Most AutonoMe injectors showed slight scratches or deformation. For the Multisert, most injectors showed an extension and 18.75% showed cracks. There was no statistically significant difference in injector damage for injectors used with the insert shield in the advanced position vs. the insert shield in the default position. The highest level of injector damage following the HeiScore (grade 5: burst) was not observed in any of the injectors that we evaluated.

Discussion

As they are intuitive to use and time-effective, many surgeons prefer preloaded IOL delivery systems over manually loaded systems.^3,4 It is reported that the risk of microbial or foreign body contamination is reduced when using preloaded injectors.^1,2 The implantation is more predictable and less prone to complications arising from incorrect IOL configurations in the injector.^2,13

Although preloaded injectors generally increase the safety of cataract surgeries, e.g., by reducing the risk of microbial contamination, one cannot rule out complications completely. Hayashi et al., for example, reported one case of endophthalmitis after uneventful cataract surgery, which occurred despite the use of a preloaded injector. ¹⁴

Similarly, the damage to IOLs reported when using manually loaded injectors, such as stress fractures, haptic fractures, and intraocular foreign bodies appearing during injection^14,15 is also observed when using preloaded injectors. Oshika et al. studied different preloaded injector models in an ex vivo setting and reported a case of gross IOL damage related to a delayed trailing haptic when using the preloaded Ultrasert long tip injector for a sclerocorneal incision. ¹⁰ Ong et al. conducted a clinical study with 85 patients examining the Acrysert injector (Alcon Laboratories Inc.). They observed one case (1% of eyes) of trauma to the optic edge, that did not necessitate further treatment. ¹⁵ Although IOL damage appears to be a rare event when using a preloaded injector, it can have far-reaching consequences. When gross damage of the IOL occurs during injection, immediate explantation of the IOL may be required, which may go along with complications. In our study, we did not notice IOL damage and intraocular foreign bodies in any of the cases.

Another complication when using hydrophobic acrylic IOLs as performed in this study is the formation of so called ‘glistenings’, fluid-filled microvacuoles within the IOL material, that become visible upon slit lamp examination over time.^16–21 In our study, we did not find any alterations of the IOLs used including formation of glistenings during the follow-up period of 3 months.

As well as the IOL, the injector itself can also incur material damage during the lens injection. Nanavaty et al. compared the damage to the nozzle tip in different injector models during IOL implantation in porcine cadaver eyes. They found that models with a longer and more parallel nozzle and a more obtuse tip angle were associated with a lower level of nozzle tip damage. ²²

Our evaluation of injector damage was based on a scale developed by Fang et al. ¹² for light microscopy evaluation of IOL injectors, to which we made a minor adaptation that allows inclusion of damage types seen on SEM. Fang et al. also investigated the AutonoMe and found similar results for the damage profile of this injector with a median of 1 and a range from 0 to 2. Our results are also in good agreement with the results of a previously published study comparing the AutonoMe to the iSert injector. ¹¹ Very similar to our study, the AutonoMe was found to show either no damage (23%) or only slight deformation (73%). In our study, 25% of the injectors were free of damage and 62.5% showed slight deformation or scratches. For the iSert injector, the predecessor of the Multisert, 80% showed cracks of different lengths, for the Multisert we observed a lower rate of 18.8% which we attribute to changes in the injector manufacturing method. Oshika et al. examined different preloaded models using porcine cadaver eyes including the iSert and Ultrasert, which are predecessors of the injectors we studied. They found that the iSert was the only model showing splitting of the nozzle tip, which occurred in all cases. ¹⁰ The previously mentioned in vitro study by Nanavaty et al. also included both the iSert and Ultrasert. The iSert showed more nozzle tip damage than the Ultrasert. ²² The results in both studies are consistent with our findings for the successor models. Compared to the iSert the Multisert showed a lower damage profile than its predecessor. The extension of the nozzle tip in a predetermined area of this injector model is intentional to reduce the force of the IOL during injection. Due to this mode of operation, the material of the nozzle tip is thinner compared to the AutonoMe. Although both injectors showed damage of the nozzle tip, we did not observe any intraocular foreign bodies in either group.

As study protocols differ, a direct comparison of IOL implantation time is difficult. Results published for different IOL delivery systems range from 30 to 60 s, which is comparable to what we found for both IOL injector models.^15,22 In a study previously published by our group, IOL implantation time was longer for the AutonoMe compared to the iSert, which could be attributed to a slower IOL unfolding. ¹¹ We could confirm this result comparing the AutonoMe to the Multisert. The use of the Multisert's insert shield in the advanced position did not lead to a significantly longer implantation time compared to treatment with the insert shield in the default position. In contrast to the study using the iSert injector that only allows screw-injection, we performed all implantations using push mode in this study. The different implantation technique did not affect total implantation time. Although IOL implantation took longer with the AutonoMe, this result may be of limited clinical relevance. The surgeon can proceed to the following surgical steps, such as removing the ophthalmic viscosurgical device while the IOL unfolding completes.

Ong et al. studied the implantation behavior of the Acrysert injector, one of the first preloaded delivery systems. They observed a variety of inadvertent delivery behavior with only 45% of the IOLs implanted in the desired orientation. In 55% of the cases, a rotation of the injector was necessary to achieve correct IOL orientation. In 6 out of 85 cases, a trapped trailing haptic occurred and one case of an overriding plunger was observed. Therefore, they judged the performance of the injector as too unpredictable and inconsistent. ¹⁵ In an ex vivo study by Oshika et al. comparing different preloaded IOL injectors that are currently available, including the Ultrasert and iSert, no sudden IOL ejection or overriding plunger were observed in a sample of 110 porcine cadaver eyes. They did observe cases of adherence of the IOL to the plunger, delayed trailing haptic and intrawound IOL manipulation, however. These events were not limited to a specific implantation system. ¹⁰ Advancements in the design seem to have reduced unwanted interactions of IOL and injector in the newer delivery systems.

In our study, we did not observe any critical events like trapped trailing haptic or overriding plunger. While no misfolding of the IOL was observed in a previous study, ¹¹ in this study a straight leading haptic occurred in one case of the AutonoMe group. A consistency between both studies was seen in haptic adhesions (60%), which occurred only in the AutonoMe group. ¹¹ We observed adhesions of the leading haptic to either the IOL optic or the trailing haptic in 58.3%. Haptic adhesions were not observed for the Vivinex IOL in both studies, as the surface of the haptic is designed to prevent such events. Intrawound IOL manipulation with a second instrument was necessary in 2 eyes (4.2%) of the AutonoMe group but in none of the eyes of the Multisert group. All events observed were resolved without sequalae. We therefore consider both delivery systems studied to be safe and efficient.

A sometimes mentioned disadvantage of preloaded IOL injectors is, that the injector is discarded after use, which could generate more waste. A relatively recent trend is toward reusable or recyclable injectors to increase sustainability. With reusable injectors, only the cartridge needs to be disposed of after use, while the injector itself can be sterilized and reused.^23,24 However, resources are also needed in non-preloaded IOLs (e.g., packaging of the IOL, sterilization process of the injector etc.)

Therefore, it is not very easy to quantify differences regarding the impact on the environment and studies evaluating this topic are lacking.

The steps required for preparation and implantation when using preloaded injectors from different manufacturers may differ considerably. Therefore, the surgeon and assisting staff need to be trained how to use each injector system. In addition, not all IOL models are available in preloaded injectors, therefore manual systems must still be used in certain cases.

Conclusion

Our study showed that safe and reproducible IOL implantation was possible with both investigated IOL injectors. There were no cases in which the implantation had to be aborted or the injector discarded for safety reasons. The implantation was quicker and more consistent with the Multisert injector. The injectors showed different damage profiles with the Multisert injector showing a higher damage score than the AutonoMe. IOL material changes including glistenings and intraocular foreign bodies were absent during the follow-up period of 3 months in all groups.

Supplemental Material

Video 1: The video illustrates the differences in implantation behavior of the two IOL injector models.