Secondary Logo

Journal Logo


Performance of a temperature-controlled shape-memory pupil expander for cataract surgery

Tan, Royston K.Y. PhD; Perera, Shamira A. FRCOphth; Tun, Tin A. MD; Boote, Craig PhD; Girard, Michaël J.A. PhD

Author Information
Journal of Cataract and Refractive Surgery: January 2020 - Volume 46 - Issue 1 - p 116-124
doi: 10.1016/j.jcrs.2019.08.042
  • Free

Cataract surgery is the most performed surgery worldwide, with this disease affecting more than 20 million people.1 This number is estimated to increase to over 30 million by 20202,3 driven by an increase in the global elderly population. The surgery is performed by replacing the cloudy crystalline lens with an artificial intraocular lens.4 To do so requires a sufficiently large pupil for unobstructed surgical maneuvers. Therefore, pharmacological drugs such as phenylephrine, tropicamide, and cyclopentolate are used to relax the sphincter muscle and constrict the dilator muscle before surgery.5,6 Despite this, small pupils may persist due to reduced muscle accommodation from aging or as a result of ingestion of drugs (eg, tamsulosin), long-term miotic drug usage (eg, pilocarpine), and pseudoexfoliation.7–10

To remedy persisting small pupils, surgeons may deploy techniques such as mechanical stretching and sphincter cuts to stretch the iris.10–12 Pupil expander devices may also be deployed to provide external mechanical support. These devices include iris hooks (MicroSurgical Technology), Malyugin ring (Malyugin Ring 2.0' MicroSurgical Technology), B-HEX pupil expander (Med Invent Devices), OASIS iris expander (6.25 mm and 7.00 mm, Oasis Medical Inc), Perfect Pupil (Milvella Limited), APX dilator (Assia Pupil Expander, APX Ophthalmology Ltd.), and i-Ring (Beaver-Visitec International, Inc.).13–21 They function by engaging the iris margin and providing support to keep the pupil enlarged during cataract surgery.

An issue with many pupil expanders lies in the method of iris margin engagement, where focal points of iris contact induce high stress and potentially increase the risk of iris damage.22 Iris hooks and the APX dilator engage the iris at 4 distinct locations to form a quadrilateral pupil, forming a nonphysiological opening with high localized stresses. The OASIS iris expander, B-HEX pupil expander, and Malyugin ring also form nonphysiological openings with 6 or 8 contact points that reduce these point forces. The ideal expansion requires full circumferential iris margin engagement, which is only currently adopted by the i-Ring.20,23 However, the i-Ring, like the OASIS iris expander, B-HEX pupil expander, and Malyugin ring, requires additional surgical maneuvers for positioning.12,16 By stretching the spring-like devices across the anterior chamber at multiple engagement points, large tissue stress beyond the physiological range are generated that could potentially distort and tear the iris tissue.20,24 In addition, the need for current mechanical devices to be dragged across the pupil for iris engagement in cases of a small pupil may induce trauma25 and iridodialysis.26,27 We previously conducted a theoretical finite element modeling study showing reduced stress on the iris tissue predicted by a uniform circular expansion design.28 In the current study, we applied this design experimentally and developed a novel pupil expander to improve on the cumulative shortcomings of existing devices.

We propose the use of shape-memory technology to enhance the cataract procedure.29 Shape-memory material is able to configure and “memorize” a specific shape at a specific transition temperature. At a lower temperature, this material is flexible and can be compacted. A heat stimulus, such as that from the eye provides the energy for the shape-memory polymer to deform back to its configured shape upon reaching the transition temperature in a controlled manner. Implementing this material in a pupil expander allows for insertion into smaller incisions while retaining its ability to mechanically induce a large pupil. Moreover, expansion of the pupil occurs gradually, slowly stretching the pupil to avoid sudden tissue enlargement.

In this article, we aim to (1) describe the design and construction of an optimized shape-memory material to expand the pupil, (2) validate its performance in ex vivo porcine and in vivo monkey experiments, and (3) compare our pupil expander with commercially available devices using ex vivo porcine eyes.


Experiments were conducted at the SingHealth Experimental Medicine Centre at the Singapore Eye Research Institute (SEMC). All experiments were performed in accordance with the Association for Research in Vision and Ophthalmology Statement for the Use of Animals in Ophthalmic and Vision Research and were approved by the Institutional Animal Care and Use Committee of the SEMC located in the SingHealth General Hospital. The SEMC has accreditation by the Association for Assessment and Accreditation of Laboratory Animal Care International.

Molding and Manufacturing the Shape-Memory Pupil Expander

Shape-memory material was purchased from SMP Technologies Inc. (MP Resin and Hardener). To determine the maximum transition temperature allowed, we measured the in vivo anterior chamber temperature in a non human primate (NHP) under surgical conditions using a small custom-made temperature sensor. The measured temperature reading was 34.0°C after filling the anterior chamber with viscosurgical solution.

Our custom shape-memory material was manufactured with a glass transition temperature (Tg) of 30.0°C. Below Tg, the polymer can be folded and physically manipulated into compact shapes. Heating the polymer above Tg will supply the required energy for it to return to its programmed shape. The target shape was set by polymerizing the shape-memory material in a custom mold with the desired shape and dimensions of our pupil expander.

Three-dimensional (3D) printing was used to manufacture molds using MakerBot 2.0 (Stratasys) with acrylonitrile butadiene styrene as the printing material. The mold was 3D printed with a resolution of 50 μm for the center insert and 100 μm for the top and bottom molds. Dimensions of the mold and pupil expander were optimized to minimize the thickness of the device (300 μm) (Figure 1, A, B, and D), while ensuring full engagement at the iris margin and sufficient force for mechanical pupil dilation.

Figure 1
Figure 1:
A: Computer-aided design cross-section drawing of the mold design. B: The device is designed to be 300 μm, with the thickness of the device approximately 80 μm. C: Processing of the shape-memory pupil expander after allowing it to set overnight. The polymer is separated from the mold and initially contains excess material. It is manually cut and trimmed down using Vannas scissors until a satisfactory shape is obtained. D: The final thickness of the device is measured to be approximately 300 μm.

The shape-memory polymer was prepared by potting. The resin and hardener were first placed under vacuum (<200 mTorr) for 1 hour to evaporate the water within the polymers. The resins were then mixed and stirred for approximately 1 minute and placed under vacuum again for 1 minute to remove the effervescence. The final mixture was poured into the mold and left to set overnight. After removal from the mold, the device was trimmed using Vannas scissors (Ref:1-111, Duckworth & Kent Ltd.) to remove the excess material before testing (Figure 1, C).

Ex Vivo Validation in Enucleated Porcine Eyes

Eleven enucleated porcine eyes were purchased from Primary Industries Ltd. (Singapore Food Industries Pte Ltd.). Fresh porcine eyes were transported to the laboratories where experiments were conducted immediately and completed within 6 hours postmortem.

To ensure that the tissues maintained their properties similar to those in vivo, the enucleated eyes were kept in a modified Krebs–Henseleit buffer solution (Product Number K3753; Merck KGaA) similar to the protocol performed by Whitcomb et al.30 The buffer solution was composed of the following: 10.0 mM D-glucose, 1.2 mM MgSO4, 1.2 mM KH2PO4, 4.7 mM KCl, 118 mM NaCl, and added with 25 mM NaHCO3 and 1.25 mM CaCl2. The solution was oxygenated with 95% O2 and 5% CO2 to maintain a pH of 7.5. This kept the sphincter and dilator muscle tissues active to provide pupil constriction. Thus, we were able to induce pharmacological constriction for small pupil expander insertion to provide validation of our device's function.

Fresh eyes were placed in a warm and moist medium above a rubber heating pad (12 V/10 W Silicone Rubber Flexible Heating Pad, O.E.M Heaters). A temperature sensor was used to maintain a steady temperature of 34.0°C (±1°C) (TE333 Temperature Controller, XCSource) as was measured in vivo. A power transducer (72-10505 DC Bench Power Supply, TENMA Corporation) was used to power the heating pad (10 W) and temperature sensor (9 V, 0.1 A). At the Singapore National Eye Centre, an ophthalmic microscope (OPMI 1 FR Pro, Zeiss) was used to enhance surgical vision, and a DSLR camera (Canon EOS 800D) was used to record the experiments. Pilocarpine was administered to obtain a small pupil (2 drops of 2% Isopto Carpine, Alcon Laboratories, Inc.). The shape-memory pupil expander prototypes were manufactured to provide optimal specifications for the porcine eyes: compact width of under 2.0 mm, expanded circular diameter of 7.0 mm, and 300 μm overall thickness.

Insertion of the shape-memory pupil expander was performed in a similar manner to existing techniques and consisting of several important steps9,31 (Figure 2). First, a triangle blade (Ref: 72-2661, Surgical Specialties México) was used to make a 2.65 mm incision at a 30 to 40° angle near the cornea periphery. Viscosurgical solution is usually injected to maintain the shape of the anterior chamber, but was not used in these ex vivo porcine eyes to prevent dilation from its use.

Figure 2
Figure 2:
Ex vivo porcine eye validation of the shape-memory pupil expander. A: Insertion of the pupil expander into the anterior chamber using a Malyugin ring injector. B: The ambient temperature slowly opens the device to a more circular shape. C: A Sinskey hook is used to position the device to engage the iris margin. D: The device deforms instead of overstretching the iris for engagement. E: Complete iris margin engagement to provide a 7.0 mm pupil. F–H: Removal of the pupil expander. The Sinskey hook is used to flip up and disengage one section of the device, and the Malyugin ring injector is used to grab and swiftly retract the device, revealing an atraumatic pupil.

Second, an injector was used to deliver the compacted circular shape-memory pupil expander into the anterior chamber (Figure 2, A). Because we did not have a custom-made injector, we used the Malyugin ring injector (Ref: MAL-1002-1, MicroSurgical Technology), although similar injectors from devices such as the OASIS iris expander would perform the same function. Retraction of the circular device flattened it to a hyperellipse shape to fit the injector lumen. The 34.0°C ambient physiological temperature within the anterior chamber provided energy for the flattened device to deform back to a circular shape (Figure 2, B).

Third, a Sinskey hook (Ref: 0105109, John Weiss & Son Ltd.) was used to maneuver and adjust the device into position (Figure 2, C). To engage the iris margin, the Sinskey hook was used to pull and deform the pupil expander and to push the polymer to enlarge the pupil after iris engagement. This action was performed 3 to 5 times, between 20 to 30 seconds depending on the individual condition of each pupil. Because the device deforms only from user manipulation, the iris tissue would not be overstretched from engagement (Figure 2, D). The result was a 7.0 mm circular expanded pupil that was protected at the iris margin from any external manipulations (Figure 2, E).

Last, the Sinskey hook was used to disengage the pupil expander from the iris margin at the incision site. This was performed by run-in to previous paragraph the device at the edge and flipping it upward. The Malyugin ring injector was then inserted to hook the disengaged corner and retract the device. This process was performed swiftly and required no additional surgical maneuvers, leaving an atraumatic pupil (Figure 2, F–H).

In Vivo Validation in NHPs

After optimization and successful validation in ex vivo porcine eyes (Figure 3), the pupil expander was tested in specific pathogen-free NHPs (Macaca fascicularis) of approximately 6 to 7 years of age. Because the NHP's eye is significantly smaller, we scaled our device to the following specifications: compacted diameter of 1.5 mm, expanded diameter of 5.0 mm, and 300 μm overall thickness.

Figure 3
Figure 3:
Images from the ex vivo porcine study for each of the devices tested. Fully engaged pupils from the (A) Malyugin ring, (B) OASIS iris expander, (C) iris hooks, and (D) the shape-memory pupil expander. All devices are expanded to a maximal diameter distance of 7.0 mm.

Intraoperative optical coherence tomography (OCT) imaging (RESCAN 700; Carl Zeiss Meditec) was used in conjunction with the surgical microscope to validate the position of the device.

The monkeys were anaesthetized with ketamine. The periocular area was cleaned with povidone–iodine 10%. A wire lid speculum was placed to separate the eyelids, and topical povidone–iodine 5% was instilled onto the ocular surface for a few minutes before the surgery. An operating microscope was positioned over the eye undergoing surgery. The same surgeon operated on all the monkeys, using a standardized aseptic surgical technique: two self-sealing wounds were made with a blade into the anterior chamber temporally and for a right-handed surgeon 90 degrees away. The temporal wound allowed insertion of the devices, whereas the other was for manipulation. Viscosurgical solution was injected into the anterior chamber, and the expander device was used to open the pupil to 5.0 mm. The procedure for insertion and deployment in the NHP's eye was identical to that of the ex vivo porcine eye. The compact device was inserted using straight conjunctival forceps (Ref: 2-500-4N, Duckworth & Kent Ltd.) after application of the viscosurgical solution. The choice of forceps was to insert the device into a smaller incision. A Sinskey hook was used, when necessary, to position and adjust the device. After the device was fully deployed, iris cross-sectional images were taken using the intraoperative OCT.

Performance Comparison Shape-Memory Expander with Commercially Available Devices

To evaluate the efficacies of the shape-memory pupil expander, we selected 3 commercially available pupil expanders for comparison. We selected iris hooks because they are used by surgeons internationally. We also selected the OASIS iris expander and the Malyugin ring expander, the latter being recognized as one of the best devices currently available. In addition to the 11 porcine eyes used for the pupil expander validation, another 34 eyes were purchased for performance comparison. Eyes were placed in the aforementioned Krebs–Henseleit buffer, and pilocarpine was used to obtain a small pupil before pupil expansion. All experiments were conducted within 6 hours postmortem.

Malyugin Ring

The technique used for deploying the Malyugin ring was similar to that described in the recent literature. The device was retracted into the injector and delivered into the anterior chamber. Using the Malyugin ring manipulator (Ref: MAL-1003; MicroSurgical Technology), opposite ends of the loops were engaged by flexing and dragging these loops to engage the iris margin to obtain a final pupil diameter of 7.0 mm (Figure 3, A). The reverse procedure was performed for removal of the device from the iris margin, and the injector was used to remove the Malyugin ring from the anterior chamber.

OASIS Iris Expander

The technique used for deployment of the OASIS iris expander was similar to the Malyugin ring and described in the provided manual. The injector hooks onto the straight connectors for retraction within the injector lumen. The Sinskey hook was used to engage the opposite ends of the 4 engagement points with the iris margin to obtain a final pupil diameter of 7.0 mm (Figure 3, B). The reverse procedure was performed for removal of the device from the iris margin, and the injector was used to remove the iris expander from the anterior chamber.

Iris Hooks

The technique used for deploying the iris hooks was similar to that described in the current literature. Four stab incisions were made to the cornea, and the hooks were inserted to engage the iris margin. Tightening was performed individually until the maximal diameter of the pupil reached 7.0 mm (Figure 3, C). The reverse was performed to remove the iris hooks.

A total of 45 eyes were used for this comparison: 10 eyes for the Malyugin ring, 13 for the iris expander, 11 for the iris hooks, and 11 for the shape-memory pupil expander. After experimentation, the eyes were immersion-fixed in 10% neutral buffered formalin solution for 24 hours. The irides were then isolated under a fume hood and stored again in the 10% neutral buffered formalin solution. All eyes tested were included in the results, without any exclusions.

Images of the fixed irides were taken under a microscope with a DSLR camera and lens (Canon EF-S 10-18mm f/4.5-5.6 IS STM). A primary image of the iris was taken, followed by zoomed in sections of each quadrant of the tissue for a more accurate analysis (Figure 4). To remove bias during results analysis, the iris samples were blind-graded by randomization and evaluated separately by an independent clinician. Analysis was performed by manually marking the areas of tissue affected during the procedure, and image analysis was performed to measure the marked images. Analysis was classified into two complications: iris pigment epithelial (IPE) loss (defined as a section of missing dark pigment of the IPE at the iris margin) and sphincter tear (defined as a discontinuity of the circular shape of the sphincter tissue at the iris margin). Using ImageJ32 (v1.50i, National Institutes of Health), the circumferential lengths of tissue damage at the iris margin were measured (Figure 4, D). The pupil diameters before device insertion were also measured using the smaller radii because the porcine pupil is elliptical rather than circular.

Figure 4
Figure 4:
Images of the isolated porcine irides taken from the microscope for processing. (A) First, a x2 zoom image is taken, followed by (BF) multiple x4.5 zoom images for clear image analysis. (D) Loss of iris pigment epithelium is noted and measured at the locations marked by the red boxes, performed by blind grading. The iris has been isolated from the eye and is therefore not regularly shaped.


Manufacturing of Shape-Memory Device Prototype

The prototypes were made using potting, requiring custom-designed molds to encase the polymer. Three-dimensional printed molds were successfully manufactured with the 50 μm resolution of the 3D printer. The resultant molds were able to provide the 300 μm overall thickness desired but lacked smoothness in the U-shaped curvature. The cross-sectional thickness of the device was approximately 80 μm, with an opening measuring approximately 140 μm for engagement of the iris margin.

Ex Vivo and In Vivo Validation

The ex vivo porcine iris experiment was performed in accordance with standard cataract surgery protocol. When inserted into the anterior chamber using forceps, the device was able to deform upon reaching the transition temperature within the anterior chamber. This deformation was slow and controlled because of the inherent shape-memory polyurethane properties. This prevented any sudden external forces, which may cause trauma to surrounding tissues. Only a Sinksey hook was needed to manipulate the pupil expander. The device was disengaged from the iris margin with the Sinskey hook and was easily retracted into the injector for all 11 samples tested, leaving the minimally traumatized pupil (Figure 2).

The in vivo experiments were performed by an experienced senior consultant. The device was optimized following the initial experiments with the NHPs. The device was successfully delivered into the anterior chamber and guided to the iris margin with a Sinskey hook. After deployment, we were able to verify that the pupil expander engaged the iris margin using intraoperative OCT (Figure 5). A 6-month follow-up examination showed no complications such as inflammation or ocular hypertension on the primate, indicating biocompatibility using the polyurethane material.

Figure 5
Figure 5:
(A) Intraoperative optical coherence tomography image from a monkey undergoing insertion of the pupil expander device. (B) The cross-sectional image of interest is the subimage across the blue arrow. The device successfully engaged the iris margin. The outline of the pupil expander is represented by the white dotted lines.

Performance Comparison

Comparison of the 4 pupil expander devices showed mostly IPE loss and minor sphincter tears (Table 1). Sphincter tears are always accompanied with IPE loss at the same location. Iris hooks fared the worst: of the 11 tested samples, 3 exhibited small sphincter tears and 4 exhibited IPE loss. Of the 10 Malyugin ring samples tested, 1 exhibited a small sphincter tear, and 2 exhibited IPE loss. Of the 13 iris expander samples tested, zero exhibit sphincter tears, but 4 exhibited IPE loss. The shape-memory pupil expander performed the best, with no observable sphincter tears or IPE loss in the 11 samples tested. The mean pupil diameters before device insertion were 5.50 ± 0.876 mm for the Malyugin ring, 5.35 ± 0.576 mm for the OASIS iris expander, 5.27 ± 0.768 mm for the iris hooks, and 5.10 ± 0.743 mm for the shape-memory expander.

Table 1
Table 1:
Complications arising from the use of pupil expanders.


Previously, using numerical biomechanical methods,28 we identified the unmet need of a well-optimized pupil expander for cataract surgery, capable of uniformly engaging the iris margin and smoothly increasing the pupil diameter to avoid potentially deleterious stress and strain on the iris tissue. The present study, to our knowledge, provides the first proof of concept for such a device that uses a shape-memory polymer-based smart material. We demonstrated here its successful application via ex vivo and in vivo experimental testing in porcine enucleated eyes and NHPs. Our novel pupil expander could potentially accommodate even smaller pupil sizes than other commercially available devices.

With devices such as iris hooks and the APX dilator, multiple parts must be deployed individually. For iris hooks, 4 to 5 hooks are inserted, creating high-stress points that greatly increase the risk for tissue tearing.14,22,33 Similarly, the scissor-like claws of the APX dilator contact the iris at 4 locations. Although the pupil is enlarged directly, both devices create additional corneal incisions, creating further tissue damage.

In the case of the Malyugin ring, i-Ring, and OASIS iris expander, the opposite issues were observed. Although only requiring the standard primary and secondary corneal incision for cataract surgery, and only deployed into the anterior chamber, there are additional manipulations. Both devices need to engage the iris margin, stretching the iris tissue excessively to engage the opposite ends. Especially when engaging the final corner, the pupil has already been enlarged significantly. Pushing the device to the opposite corner creates significant stress that is clinically suboptimal.

We designed our shape-memory pupil expander to address these two main issues. By adopting a more flexible design, the device is able to deform instead of overstretching the iris tissue. With the U-shaped cross-section, the pupil expander can engage the entire iris margin, exerting uniform stresse on the iris tissue while protecting it from external forces such as accidental tears from surgical tool manipulation.34 The circular shape provides minimum distributed stress on the tissues, with full expansion not exceeding the designed maximum diameter, avoiding unnecessary stress.28

Although the type of pupil expansion is important, the speed at which the tissue is stretched also plays a role in determining whether damage is induced.22,33 Like most tissues in the human body, the iris tissue behaves in a viscoelastic manner.35,36 Fast expansions can create significant stress, which may result in tears. Existing devices mostly use the flexible, spring-like properties of a plastic-like polypropylene. The use of shape-memory material close to the Tg allows for a slower deformation speed that can avoid sudden pupil stretching.

By optimizing the Tg of the polymer, it is possible to control and adjust how fast the device uncoils. Our clinician feedback revealed that the ideal duration to deploy the device is between 20 seconds and 30 seconds after device insertion into the anterior chamber. We designed our shape-memory material to slowly deform over 10 to 20 seconds after insertion; thereafter, simple manipulation is conducted to position the device. The material will remain sufficiently rigid to hold the iris in place for cataract surgery, with stiffness akin to that of a harder silicone rubber.

Femtosecond laser–assisted cataract surgery has been gaining popularity in recent years.37 The use of pupil expanders could enhance the safety of the procedure by maintaining a dilated pupil for extended durations.38,39 Before the laser is used, there is a waiting period of about 15 minutes after the pharmacological drug is administered. The drug can wear off in a shorter duration for some patients, resulting in a smaller pupil. Further 1% atropine drops can be administered to limit pupil constriction, but this is not a fail-safe solution.38 With the use of the custom circular pupil expander, an optimal 7.0 mm pupil could be maintained throughout the procedure to ensure patient safety and surgical success. This is not optimal with noncircular devices such as the Malyugin and B-HEX rings, and completely impossible for devices with external protrusions such as iris hooks and the APX dilator because they would interfere with the suction cup placed on the cornea.40,41

In addition, it is believed that the anterior capsulotomy is the main trigger for an increase in prostaglandins in the aqueous with femtosecond laser–assisted cataract surgery. The resulting miosis has been somewhat but not completely mitigated by the use of nonsteroidal anti-inflammatory drugs. The longer the wait between the laser portion and the phase emulsification portion of the surgery, the worse the miosis.42 The use of an optimized mechanical device may be helpful in alleviating this problem.

Usually, in a hospital, the variety of pupil expanders available is limited to focus on perfecting the technique in one or two devices. Comparison between multiple devices is therefore uncommon and impractical. The versatility of the current device circumvents some disadvantages of existing alternatives. The method of incision and the size of the small pupil are two areas of concern with currently limited viable solutions.10 For this study, porcine irides were used to obtain a larger pool sample, and because the pupil expander experiments were all performed by the same person, it is possible to provide an unbiased comparison of the various devices.

Iris hooks take the longest to deploy and remove,13 and the small contact points have the greatest potential to damage soft tissue. Although it allows for flexibility in positioning and varying pupil size, it is less practical in providing a sufficiently large pupil unless the tissue is retracted significantly.43 Multiple stab incisions are not ideal because healing after corneal incisions can be slow and incomplete.44–46 More recent devices, including the shape-memory expander, have been more efficient in this regard by eliminating additional incisions.

The OASIS iris expander works similarly to the many variations of ring devices on the market. However, the rigidity in material could be a concern. The connectors between the 4 loops can be weak and break easily, as happened during the first attempt to retract the expander into the retractor. Subsequently, care was taken to assist the device retraction by flattening the sides using forceps. In addition, once in the anterior chamber, the device did not retain its square shape, but remained slightly deformed in a rectangular shape from the bent curved connectors. The material construction is a hard polypropylene that may require excessive force to flex to engage the loops. The hard plastic against the soft iris tissue may have caused iris chafing and IPE loss in several samples. Thus, we chose a soft polymer that can be more easily deformed to reduce the risk for iris-tissue damage.

Although the Malyugin ring may be very popular because of the ease of deployment, removal is significantly more challenging. The OASIS iris expander has specific shielded holes where the Sinskey hook is positioned, and the Malyugin ring relies on a custom manipulator tool to hook onto the expander. The manipulator tool contacts the iris tissue during removal, and it is easy for the iris margin to get caught between the devices. At these 4 loops, the iris tissue may accidentally be dragged and torn. In addition, the Malyugin ring is designed to be a continuous loop glued at the ends. During removal, one of the loops often gets stuck during retraction into the injector. Because it is a one-use device, forceful retraction is possible, but that might bend the device upward or downward, potentially contacting the corneal endothelium and inducing further trauma. Our shape-memory pupil expander encompasses a continuous U-shaped cross-section, eliminating the risk of getting caught by the iris. It is also easily removed, taking less time and effort in comparison to existing methods.

There are several limitations of this study. The intended design consisted of a U-shaped cross-section that can engage the iris margin. However, because of the low resolution of our 3D printer, the curved edges were instead right angled. This resulted in a rectangular cavity for the cross-section. In addition, the surface finish for the completed device was imperfect, with rough edges and surfaces. However, as a first proof-of-principle, this laboratory-made device was successful for both ex vivo and in vivo testing.

In addition, the polymer used for testing would ideally be manufactured differently from the prototypes tested. We used potting to mold the device individually, whereas injection molding pellets would be better used for large-scale production. With injection molding, the resolution and surface finish for the prototypes would be within acceptable tolerances (±5 μm).

Comparison of the various pupil expanders would benefit from a larger sample size. This should allow for a greater pool of data for analysis and an accurate representation of the complication percentages. However, that would require a large number of devices, which is not practical with porcine data.

Finally, because testing of the device was in ex vivo porcine and healthy in vivo cynomolgus eyes, we have yet to follow-up the procedure with phacoemulsification. Therefore, further studies are needed to ensure that there are no potential complications of our shape-memory pupil expander.

We developed an optimized pupil expansion device designed to minimize and limit the amount of stress exerted onto the iris tissue. The in vivo and ex vivo experimental validations presented herein provide proof of concept of the device's efficacy and further highlight the translational potential of smart materials in the development of other ophthalmological implants to improve patient healthcare.


  • Current pupil expander devices are made of hard plastic materials, and ring expanders use the tension–spring effect of the plastic during iris engagement, overstretching the iris in the process.
  • Removal of pupil expanders can sometimes be more difficult than their deployment.


  • A novel pupil expander that is made of a shape-memory polyurethane could deform to prevent overstretching of the iris tissue during device deployment.
  • Ex vivo and in vivo animal experimental validation of the novel shape-memory pupil expander, as well ex vivo comparison with commercially available pupil expander devices to quantitatively and qualitatively compare each device.


1. Foster A. Cataract—a global perspective: output, outcome and outlay. Eye (Lond) 1999;13:449–453
2. Taylor HR. Cataract: how much surgery do we have to do? Br J Ophthalmol 2000;84:1–2
3. Foster A. Cataract and “Vision 2020—the right to sight” initiative. Br J Ophthalmol 2001;85:635–637
4. Desai P, Minassian D, Reidy A. National cataract surgery survey 1997–8: a report of the results of the clinical outcomes. Br J Ophthalmol 1999;83:1336–1340
5. Cionni RJ, Barros MlG, Kaufman AH, Osher RH. Cataract surgery without preoperative eyedrops. J Cataract Refract Surg 2003;29:2281–2283
6. Amini R, Whitcomb JE, Al-Qaisi MK, Akkin T, Jouzdani S, Dorairaj S, Prata T, Illitchev E, Liebmann JM, Ritch R. The posterior location of the dilator muscle induces anterior iris bowing during dilation, even in the absence of pupillary block. Invest Ophthalmol Vis Sci 2012;53:1188–1194
7. Winn B, Whitaker D, Elliott DB, Phillips NJ. Factors affecting light-adapted pupil size in normal human subjects. Invest Ophthalmol Vis Sci 1994;35:1132–1137
8. Chang DF, Osher RH, Wang L, Koch DD. Prospective multicenter evaluation of cataract surgery in patients taking tamsulosin (Flomax). Ophthalmology 2007;114:957–964
9. Linebarger EJ, Hardten DR, Shah GK, Lindstrom RL. Phacoemulsification and modern cataract surgery. Surv Ophthalmol 1999;44:123–147
10. Akman A, Yilmaz G, Oto S, Akova YA. Comparison of various pupil dilatation methods for phacoemulsification in eyes with a small pupil secondary to pseudoexfoliation. Ophthalmology 2004;111:1693–1698
11. Chang DF, Campbell JR. Intraoperative floppy iris syndrome associated with tamsulosin. J Cataract Refract Surg 2005;31:664–673
12. Malyugin B. Small pupil phaco surgery: a new technique. Ann Ophthalmol (Skokie) 2007;39:185–193
13. Novák J. Flexible iris hooks for phacoemulsification. J Cataract Refract Surg 1997;23:828–831
14. Merriam JC, Zheng L. Iris hooks for phacoemulsification of the subluxated lens. J Cataract Refract Surg 1997;23:1295–1297
15. Santoro S, Sannace C, Cascella MC, Lavermicocca N. Subluxated lens: phacoemulsification with iris hooks. J Cataract Refract Surg 2003;29:2269–2273
16. Bhattacharjee S. Pupil-expansion ring implantation through a 0.9 mm incision. J Cataract Refract Surg 2014;40:1061–1067
17. Auffarth GU, Reuland AJ, Heger T, Völcker HE. Cataract surgery in eyes with iridoschisis using the Perfect Pupil iris extension system. J Cataract Refract Surg 2005;31:1877–1880
18. Kershner RM. Management of the small pupil for clear corneal cataract surgery. J Cataract Refract Surg 2002;28:1826–1831
19. Lee BS, Chang DF. Management of small pupils. Expert Rev Ophthalmol 2016;11:49–58
    20. Tian JJ, Garcia GA, Karanjia R, Lu KL. Comparison of 2 pupil expansion devices for small-pupil cataract surgery. J Cataract Refract Surg 2016;42:1235–1237
    21. Agarwal A. Visitec® I-Ring® Pupil Expander. EyeWorld Weekly 2016;21:77–78
      22. Masket S. Avoiding complications associated with iris retractor use in small pupil cataract extraction. J Cataract Refract Surg 1996;22:168–171
      23. Ha-Eun L, Joon-Young K, Da-Eun L, Jin-Gu K. Two cases of phacoemulsification in the presence of a small pupil using an iris expander. Turkish J Vet Anim Sci 2019;43:159–166
      24. Chang DF. Use of Malyugin pupil expansion device for intraoperative floppy-iris syndrome: results in 30 consecutive cases. J Cataract Refract Surg 2008;34:835–841
      25. Mattox C. Management of the small pupil. In: Cataract Surgery in the Glaucoma Patient. New York, NY, Springer; 2009:23–34
      26. Narang P, Agarwal A. Small Pupil Phacoemulsification. Step by Step Phacoemulsification. New Delhi, India, Jaypee Brothers Medical Publishers (P) Ltd; 2015:57
      27. Vollman DE, Gonzalez-Gonzalez LA, Chomsky A, Daly MK, Baze E, Lawrence M. Intraoperative floppy iris and prevalence of intraoperative complications: results from ophthalmic surgery outcomes database. Am J Ophthalmol 2014;157:1130–1135.e1
      28. Tan RKY, Wang X, Perera SA, Girard MJA. Numerical stress analysis of the iris tissue induced by pupil expansion: comparison of commercial devices. PLoS One 2018;13:e0194141
      29. Huang WM, Yang B, Fu YQ. Polyurethane Shape Memory Polymers. Boca Raton, FL: CRC Press; 2012
      30. Whitcomb JE, Amini R, Simha NK, Barocas VH. Anterior–posterior asymmetry in iris mechanics measured by indentation. Exp Eye Res 2011;93:475–481
      31. Minassian DC, Rosen P, Dart JK, Reidy A, Desai P, Sidhu M, Kaushal S, Wingate N. Extracapsular cataract extraction compared with small incision surgery by phacoemulsification: a randomised trial. Br J Ophthalmol 2001;85:822–829
      32. Abràmoff MD, Magalhães PJ, Ram SJ. Image processing with ImageJ. Biophotonics Int 2004;11:36–42
      33. Morris B, Cheema RA. Phacoemulsification using iris-hooks for capsular support in high myopic patient with subluxated lens and secondary angle closure glaucoma. Indian J Ophthalmol 2006;54:267–269
      34. Graether JM. Graether pupil expander for managing the small pupil during surgery. J Cataract Refract Surg 1996;22:530–535
      35. Zhang K, Qian X, Mei X, Liu Z. An inverse method to determine the mechanical properties of the iris in vivo. Biomed Eng Online 2014;13:1
      36. Jouzdani S. Biomechanical Characterization and Computational Modeling of the Anterior Eye [dissertation]. University Of Minnesota, Minneapolis, MN; 2013
      37. Roberts TV, Lawless M, Chan CC, Jacobs M, Ng D, Bali SJ, Hodge C, Sutton G. Femtosecond laser cataract surgery: technology and clinical practice. Clin Exp Ophthalmol 2013;41:180–186
      38. Donaldson KE, Braga-Mele R, Cabot F, Davidson R, Dhaliwal DK, Hamilton R, Jackson M, Patterson L, Stonecipher K, Yoo SH. Femtosecond laser–assisted cataract surgery. J Cataract Refract Surg 2013;39:1753–1763
      39. Roberts TV, Lawless M, Hodge C. Laser-assisted cataract surgery following insertion of a pupil expander for management of complex cataract and small irregular pupil. J Cataract Refract Surg 2013;39:1921–1924
      40. Nagy ZZ, Takacs AI, Filkorn T, Kránitz K, Gyenes A, Juhász É, Sándor GL, Kovacs I, Juhász T, Slade S. Complications of femtosecond laser–assisted cataract surgery. J Cataract Refract Surg 2014;40:20–28
      41. Dick HB, Schultz T. Laser-assisted cataract surgery in small pupils using mechanical dilation devices. J Refract Surg 2013;29:858–862
      42. Schultz T, Joachim SC, Stellbogen M, Dick HB. Prostaglandin release during femtosecond laser-assisted cataract surgery: main inducer. J Refract Surg 2015;31:78–81
      43. Birchall W, Spencer AF. Misalignment of flexible iris hook retractors for small pupil cataract surgery: effects on pupil circumference. J Cataract Refract Surg 2001;27:20–24
      44. Jester JV, Petroll WM, Cavanagh HD. Corneal stromal wound healing in refractive surgery: the role of myofibroblasts. Prog Retin Eye Res 1999;18:311–356
      45. Ernest P, Tipperman R, Eagle R, Kardasis C, Lavery K, Sensoli A, Rhem M. Is there a difference in incision healing based on location? J Cataract Refract Surg 1998;24:482–486
      46. Gasset AR, Dohlman CH. The tensile strength of corneal wounds. Arch Ophthalmol 1968;79:595–602
      © 2020 Published by Wolters Kluwer on behalf of ASCRS and ESCRS