2023 Volume 46 Issue 10 Pages 1461-1467
Since three-dimensional (3D)-printed tablets were approved by the United States Food and Drug Administration (FDA), 3D printing technology has garnered increasing interest for the fabrication of medical and pharmaceutical devices. With various dosing devices being designed for manufacture by 3D printing, 3D-printed ophthalmic formulations to release drugs have been one such target of investigation. In the current study, 3D-printed contact lenses designed for the controlled release of the antibiotic azithromycin were produced by vat photopolymerization, and the effect of the printer ink composition and a second curing process was investigated. The azithromycin-loaded contact lenses were composed of the cross-linking reagent polyethylene glycol diacrylate (PEGDA), PEG 400 as a solvent, a photoinitiator, and azithromycin. The 3D-printed contact lenses were fabricated successfully, and formulations with lower PEGDA concentrations produced thicker lenses. The mechanical strength of the PEGDA-based contact lenses was dependent on the amount of PEGDA and was improved by a second curing process. Drug release from 3D-printed contact lenses was reduced in the samples with a second curing process. The azithromycin-loaded contact lenses exhibited antimicrobial effects in vitro for both Gram-positive and -negative bacteria. These results suggest that 3D-printed contact lenses containing antibiotics are an effective model for treating eye infections by controlling drug release.
Three-dimensional (3D) printing, which is a type of additive manufacturing process, is a method of constructing a 3D object by the sequential addition of layers. Since this concept was invented in the 1970s and 1980s, 3D printing technology has been used extensively and practically in the industrial field. Its application has also been extended to other fields including medicine, dentistry, and healthcare.1–4) The production of pharmaceuticals using 3D printers has attracted global interest,5,6) particularly with the 3D-printed swallowable tablet, SPRITAM®, being approved by United States Food and Drug Administration. One of the advantages of 3D printing is that it allows the creation of drug products in a variety of sizes and shapes, with different internal structures and spatial arrangements.5,7) 3D-printed medicine can allow flexible dosing and a controlled drug release rate and timing. 3D printing can shorten the manufacturing process and saves manufacturing space, and on-demand manufacturing and extemporaneous compounding by 3D printing could allow the on-site preparation and dispensation of medicine in a clinical setting by medical staff.8) The fabrication of various dosage forms using 3D printing has been reported (e.g., tablets, suppositories, films, capsules, stents, microneedles).9) It is expected that many future pharmaceutical products will be created by 3D printing technology, and there is growing discussion around the regulation of the preparation of personalized medicine using 3D printing.10)
ASTM International has defined seven types of 3D printers: material extrusion, binder jetting, material jetting, powder bed fusion, vat photopolymerization (VPP), sheet lamination, and directed energy deposition.11) The first five of these have been considered the most suitable for the preparation of drug products because of the properties of materials. For example, material extrusion 3D printers to use drug-loaded polymers by fused deposition modeling (FDM), and pharmaceutical pastes produced by semi-solid extrusion have been well studied for these purposes.12,13) In the current study, we focused on VPP, to explore whether this type of 3D printing is feasible for biomedical applications and drug formulation. VPP 3D printers produce 3D objects by polymerizing a monomer layer by layer via irradiation with light. VPP fabrication has a relatively high accuracy in comparison with typical material extrusion. Biocompatible materials have been studied for the preparation of medicine using VPP. For example, polyethylene glycol diacrylate (PEGDA) is a common cross-linking reagent composed of a polyethylene glycol (PEG) derivative and can be used for 3D printing.14) 3D-printed tablets have been prepared by VPP with PEGDA-based printer ink for drug formulation.15,16) The tablet in these studies was composed of PEGDA, a solvent, an active pharmaceutical ingredient (API), and a photoinitiator. PEGDA-based 3D hydrogels have been tested in tissue engineering,17) and in vivo applications of PEGDA for corneal inlay implantation18) and a mixture with other polymers19) have been studied. These studies demonstrate the high biocompatibility of PEGDA-based objects.
Various dosage forms and devices for ophthalmic application have been developed for efficient drug delivery.20) Eye drops are the most common and preferred choice because of their ease of use and safety. However, the volume of depot to pool drug formulation in the eyes is limited. Conjunctival sacs can only retain a low volume of fluid, and most of the drug flows out into nasal cavity because of the anatomical structure at the delivery site. Multiple doses of eye drops lead to the ejection of fluid from the eye, and the administration of eye drops many times at short intervals may lead to poor drug compliance. Thus, various ophthalmic formulations have been developed to increase therapeutic efficacy and convenience. For example, conjunctival resorbable formulations, intraocular insertions, lacrimal plugs, and contact lens formulations have been reported as viable alternatives to eye drops.20) Regarding the 3D printing of ophthalmological devices, previous studies have reported the development of medical implants,21) punctal plugs,22) drug-releasing patches,23) and drug-loaded contact lens24); however, drug formulation in this area is still in its infancy. Advancement of these formulations are expected to improve not only drug retention but also bioavailability.
The purpose of this study is to investigate the characteristics of a 3D-printed PEGDA-based medical device shaped like a contact lens that can administer antibiotics with a controlled release rate. Contact lenses without drugs have been manufactured by several conventional methods, including cast molding, spin casting, and lathe cutting.25) 3D printing of contact lenses is a newly emerging method. Because customized production is an advantage of 3D printing, the drug-releasing device could be manufactured to have variable controlled drug release rates, flexible dosing, and a shape specific to each patient, making it a promising technology for personalized therapy. VPP 3D printers also provide higher resolution than typical FDM printer. Although the common materials used to manufacture soft and hard contact lenses are polymethyl methacrylate, hydroxyethyl methacrylate, and silicon hydrogel,25,26) we focused on a PEGDA-based contact lens manufactured by VPP. As PEGDA-based contact lens formulations containing dexamethasone have been fabricated in a previous study,27) there is still little information about the 3D printed PEGDA-based ophthalmic formulation. In the present study, a PEGDA-based contact lens formulation containing azithromycin (AZI) was fabricated using VPP, and the 3D-printed PEGDA-based contact lens formulation was assessed to determine its mechanical properties and drug release profile.
PEGDA (Mn = 700, Sigma-Aldrich, St. Louis, MO, U.S.A.) was used as the cross-linking agent in the printer ink, and diphenyl (2,4,6-trimethyl benzoyl) phosphine oxide (TPO) (Tokyo Chemical Industry, Tokyo, Japan) was used as the photoinitiator. These were prepared together with the solvent PEG 400 (Nacalai Tesque, Kyoto, Japan). AZI (Tokyo Chemical Industry) was used as the API loaded into the printer ink. Phosphate buffered saline (PBS) tablets were purchased from Takara Bio (pH 7.35–7.65, dissolved in water; Shiga, Japan). HPLC-grade acetonitrile was purchased from Sigma-Aldrich. Potassium dihydrogen phosphate was purchased from Wako Pure Chemical Corporation (Osaka, Japan). Staphylococcus aureus (NBRC 12732) and Pseudomonas aeruginosa (NBRC 3445) were supplied from NITE Biological Resource Center (Chiba, Japan). Commercially available 1% AZI eye drop (AZIMYCIN ophthalmic solution 1%) was obtained from a Japanese pharmaceutical company (Senju Pharmaceutical Co., Ltd., Osaka, Japan).
Preparation of Printer InkThe printer ink (drug formulation) was prepared as described previously.28) The compositions of the printer ink used in the present study are given in Table 1. PEGDA as the cross-linking reagent, PEG 400 as the solvent, AZI as the API, and TPO as the photoinitiator were added to a beaker at room temperature and stirred until the mixture become transparent. The mixture was further stirred under light shielding for at least 1 h.
| Formulation | PEGDA (%) | PEG400 (%) | AZI (%) | TPO (%) |
|---|---|---|---|---|
| A | 20 | 78 | 1 | 1 |
| B | 40 | 58 | 1 | 1 |
| C | 60 | 38 | 1 | 1 |
| D | 80 | 18 | 1 | 1 |
| E | 98 | 0 | 1 | 1 |
As a typical experiment, the shape of the contact lens was designed using 3D computer-aided design (CAD) software (Fusion 360, Autodesk Inc., San Rafael, CA, U.S.A.) (Fig. 1A). The estimated dimension of contact lens was 3.525 mm (height), 13.783 mm (diameter), and 0.10 mm (thickness). The designed object was imported into slicer software (Phrozen Tech Co., Ltd., Hsinchu, Taiwan). In this case, the support was attached in the design (Fig. 1B). After the printer ink shown in Table 1 was placed into a vat, the lens was fabricated using a VPP 3D printer (405 nm, Phrozen Shuffle 2019, Phrozen Tech Co., Ltd.). Each printed layer had a thickness of 50 µm, and the curing time was set to 7.5 s. Several 3D printed samples were placed in a secondary curing machine (ELEGOO MERCURY, Shenzhen ELEGOO Technology Co., Ltd., Shenzhen, China) and were exposed for 5 min. The weight of the 3D-printed samples was measured using an electronic balance. The height, diameter, and thickness of the 3D-printed samples were measured with a digital caliper.
The surface of the samples was coated with Pt–Pd for 1.5 min using an ion sputtering system (E-102, Hitachi, Tokyo, Japan). The contact lens formulation was placed in the ion spatter to coat the entire surface of concave side by fixing the convex side of contact lens with carbon tape. The concave side of 3D printed contact lens was observed by scanning electron microscopy (SEM; S-4300, Hitachi).
Hardness TestA hardness test was performed using a digital force gauge (ZTA-50N, IMADA, Aichi, Japan) equipped with a motorized test stand (EMX-500N-FA, IMADA). Tablets were printed with a height of 5 mm and a diameter of 10 mm and were vertically placed on a stage. The load against the probe (V-type attachment) was measured until the tablet was broken (Method 1).
Because the digital force gauge can measure up to 50 N, a Kiya digital hardness tester (Kiya Seisakusho, Tokyo, Japan) was used (Method 2). The tablets were inserted into the hardness tester with the side facing vertically. The hardness of the tablets was measured when they were broken.
Next, the contact lens formulation prepared by the VPP 3D printer was placed with the convex side of the lens facing up, and the load against the flat probe was measured until the device was broken (Method 3). The comparison of three methods were shown (Fig. 2).
The detail of each method was described in Materials and Methods.
The contact lens formulation was placed in a 20 mL glass vial and 5 mL of PBS was added into the tube. The sample was incubated with continuous shaking at speed 4 using a Thermo Bath Shaker SN-60SD (NISSIN, Tokyo, Japan) at 37 °C. After 15, 30, 60, 90, 120, 240, and 480 min of incubation, 0.5 mL of the eluate was collected and replaced with 0.5 mL of PBS solution. The collected solution was filtered through a 0.20 µm membrane filter and used as the sample solution.
The drug concentration was determined by HPLC (LC-20 series, Shimadzu, Kyoto, Japan) with an Inert Sustain C18 column (5 µm, 4.6 × 150 mm, GL Science Inc., Tokyo, Japan). The column temperature was 40 °C, and the mobile phase was 40 mM potassium dihydrogen phosphate (pH 11)/acetonitrile (2 : 8 [v/v]). The flow rate was 1.0 mL/min, and the wavelength for the measurement was 215 nm.
Weight MeasurementThe samples were weighed before and after the release test, following incubation for 8 h. The weight change was calculated as
 | (1) |
where W8 is the sample weight after the drug release test following 8 h of incubation and W0 is the initial sample weight before the drug release test (time 0).
Antimicrobial ActivityAntimicrobial activity was evaluated using the disk diffusion method as described previously.23) In brief, 100 µL of 1 × 109 colony forming unit (CFU)/mL bacterial solution was smeared on soybean-casein digest agar medium DAIGO (Nihon Pharmaceutical, Tokyo, Japan). A circular patch (9 mm diameter, 1 mm height) with the same composition as listed in Table 1 was prepared by 3D printing and placed in the center of the agar medium. To compare the antimicrobial effect of the lens to that of an eye drop formulation, the corresponding eye drop volume was deposited onto a circular paper with a diameter of 9 mm. After incubation at 35 °C for 24 h, the diameter of the inhibition zone was measured with a digital caliper.
In the present study, we focused on AZI as a model antibiotic, and AZI-loaded PEGDA-based contact lenses were produced by VPP. AZI eye drops are broadly applicable for eye infections such as conjunctivitis, blepharitis, hordeolum, and dacryocystitis. However, AZI is a hydrophobic molecule and thus requires specialized techniques to be solubilized and retained on ocular surfaces after administering eye drops or ophthalmic formulations.29) In the present study, AZI was dissolved into printer inks composed of PEG 400, and the shape of the contact lenses was produced by cross-linking PEGDA, which is known to be biocompatible. The effect of the second curing process on the properties was investigated. The second curing process after 3D printing by VPP is expected to strengthen the cross-linking and thus also improve the mechanical properties of the printed object,30) resulting in changes to pharmaceutical qualities such as drug release.
The appearance of the printed contact lenses with different formulations is shown in Fig. 3, and the corresponding weight, height, diameter, and thickness are given in Table 2. Formulations A–E have progressively increasing concentrations of PEGDA (Table 1), which yields a corresponding progressive decrease in weight, height, diameter, and thickness (Table 2). The degree of cross-linking of PEGDA may not be sufficient at low concentrations of PEGDA; in this case, an incomplete cross-linked product of formulation in the deposition of one layer may affect the production of the subsequent layer, resulting in a thicker product. The designed height, diameter, and thickness were 3.525, 13.783, and 0.10 mm, respectively. Our results demonstrate that the VPP 3D printer can produce contact lenses with the desired size, though compensation is necessary to produce lenses with accurate height and thickness. The deviation of the height, diameter, and thickness of the printed products from the corresponding design values was small (generally <10%), which suggests that AZI-loaded PEGDA-based formulations are suitable for 3D printing by VPP.
| Formulation | Weight (mg) | Height (mm) | Diameter (mm) | Thickness (mm) |
|---|---|---|---|---|
| A | 251.1 ± 31.6 | 4.46 ± 0.26 | 14.23 ± 0.13 | 0.48 ± 0.05 |
| B | 217.9 ± 38.9 | 4.25 ± 0.22 | 13.87 ± 0.07 | 0.46 ± 0.02 |
| C | 170.5 ± 25.1 | 4.10 ± 0.35 | 13.59 ± 0.11 | 0.43 ± 0.04 |
| D | 145.3 ± 24.1 | 4.47 ± 0.14 | 13.96 ± 0.10 | 0.43 ± 0.03 |
| E | 107.1 ± 14.1 | 3.96 ± 0.17 | 13.99 ± 0.09 | 0.36 ± 0.03 |
The data represent the mean ± standard deviation (S.D.) (n = 6).
SEM images of the surface of the 3D-printed contact lenses are shown in Fig. 4. We found that contact lenses with a higher PEGDA content show a marked pattern of contour lines (Formulations C–E). This is a byproduct of the additive manufacturing process: stronger cross-linking produces a structure that more clearly shows the layer-by-layer construction. Many of the contact lenses display a partially rough structure in magnified images (Formulations B–E). This indicates that cross-linking of PEGDA was facilitated, depending on the amount of PEGDA. The staircase effect that occurs in additive manufacturing as a result of the layer-by-layer deposition process, becoming a potential issue.25) Because this rough structure can adversely impact how the contact rests on the eye, an additional polishing process is necessary. A method to completely remove the supports from the lens is necessary for practical use in the future.
In the present study, three methods were used to assess the hardness of AZI-loaded 3D-printed tablets (see “Hardness Test” for a description of the methods). First, a tablet was 3D printed for each formulation, and the hardness was measured using a texture analyzer for compressive testing (Table 3, Method 1). Formulations A and B exhibited lower hardness (<10 N). Kadry et al.31) reported a lower hardness for a formulation with 20% PEGDA (approximately 3 N), which is consistent with our results. Formulations C–E exhibited stronger hardness (>50 N). The difference in hardness between these tablets is a result of the influence of the incorporated PEGDA used for photopolymerization. Next, the stronger 3D-printed tablets composed of Formulations C–E were measured by setting them vertically in a hardness tester, which can apply a greater load and is commonly used for pharmaceutical hardness testing (Table 3, Method 2). The hardness was proportional to the amount of PEGDA. The samples that underwent a second curing process could undergo higher loading than those without a second curing process in Method 2. In contrast, the samples prepared with Method 1 did not show a remarkable influence of second curing process (Formulations A, B). The interlayer mechanical strength of 3D-printed materials is generally low, which results in mechanical anisotropy. A second curing process can improve the mechanical anisotropy of some products printed by VPP.32)
| Formulation | Method 1 (N) | Method 2 (N) | Method 3 (N) |
|---|---|---|---|
| A | 1.48 ± 0.28 | N.D. | 5.65 ± 1.42 |
| A (SC) | 1.27 ± 0.50 | N.D. | 8.46 ± 3.08 |
| B | 7.54 ± 1.39 | N.D. | 14.00 ± 7.62 |
| B (SC) | 6.68 ± 0.70 | N.D. | 32.00 ± 16.85 |
| C | 50 < | 53.25 ± 9.48 | 3.86 ± 2.33 |
| C (SC) | 50 < | 85.83 ± 19.15 | 5.17 ± 2.16 |
| D | 50 < | 91.30 ± 14.36 | 21.52 ± 6.27 |
| D(SC) | 50 < | 122.75 ± 29.55 | 32.55 ± 19.35 |
| E | 50 < | 147.00 ± 34.87 | 39.25 ± 10.87 |
| E(SC) | 50 < | 168.97 ± 18.83 | 50 < |
The data represent the mean ±S.D. (n = 6). Descriptions of Methods 1–3 can be found in “Hardness Test.”
In the next, the hardness of the contact lenses was assessed using a texture analyzer (Table 3, Method 3). The strength was also dependent on the printer ink composition, and the second curing process tended to increase the hardness. However, Formulation C unexpectedly exhibited a much lower strength. After the loading test, we found that the contact lenses bent without destruction under the applied loading conditions. We conclude that the hardness of 3D-printed contact lenses is dependent on the shape, while they are highly dependent on their composition and amount of PEGDA.
Drug Release Test for AZI-Loaded 3D-Printed PEGDA-Based Contact LensesThe drug release profiles from the 3D-printed contact lenses are shown in Fig. 5. We consider that the PEGDA concentration may affect the drug release rate, and that a higher degree of photopolymerization can inhibit the release of the drug from the matrix. We found that after a second curing process, the 3D-printed contact lenses showed a lower drug release rate than the corresponding sample without a second curing process (Figs. 5A, B). The current drug release test was conducted in non-sink condition to mimic the drug dissolution in eye tissue partially. We guess that the use of PEG which is water-miscible can assist dissolution of AZI which is poorly water-soluble drug. However, our results suggest no clear relationship between the release rate and the different PEGDA concentrations in Formulations A–E. The contact lens with the highest amount of cross-linked PEGDA and without PEG (Formulation E) unexpectedly exhibited the highest drug release rate. One possible reason is that the degree of cross-linking of PEGDA by light irradiation may not be sufficient to inhibit drug release under the present experimental conditions. A second reason for the rapid release from Formulation E may be that contact lenses based on this formulation contained the lowest volume of drug because they were the smallest lenses among the different formulations, which may cause them to release the drug more rapidly.
The data represent the mean ± standard deviation (S.D.) (n = 3).
The PEGDA-based 3D-printed lenses did not disintegrate in the drug dissolution process, because photopolymerization occurred. In our previous study, drug release from several formulations of PEGDA-based tablets followed the Higuchi model, in which the drug release rate decreases over time.28) Drug release from deeper sites in a 3D printed object is expected to be slower. In other studies, the geometry of PEG-DA-based tablets and the surface area-to-volume ratio were found to be factors that influence drug release,33) and increasing the surface area of tablets with perforations has been shown to enhance drug release.31) In the current study, although the thickness of the contact lenses was approximately 0.4 mm, 30–40% of the drug molecules were released within the first 60–120 min (Fig. 5), while the remainder of the drug was retained in the lens. In our previous study, drug release from PEGDA-based tablets was investigated for three different APIs; because drug release was dependent on the aqueous solubility, poorly water-soluble drugs were not completely released.28) These results are in line with our current results. In the U.S. Pharmacopoeia, AZI is categorized as being practically insoluble in water. This means that AZI might become crystallized in the contact lens during a drug release test.
Xu et al.22) showed that 3D-printed punctal plugs composed of 80% PEGDA and 20% PEG 400 released dexamethasone over a period of 7 d, whereas for those composed of 100% PEGDA, full drug release required more than 21 d.22) Whereas their device released approximately 5–60% of the drug within 1 d, ours released approximately 30–40% within 8 h in the current study. This difference may be due to the fact that their study adopted an in-house in vitro rig using a flow cell model. Thus, drug release from PEGDA-based objects may be dependent on the object composition and shape, and the physical properties of the drug. Clarifying this dependence may lead to controlled drug release, would be beneficial for the development of personalized treatments.
Weight Change for 3D-Printed PEGDA-Based Contact Lenses after Soaking in WaterThe weight of the printed contact lenses was measured before and after the drug release test to characterize their release properties (Fig. 6). We consider that the lenses release PEG 400, AZI, and unreacted PEGDA molecules, which are then replaced by water molecules. After drug release, the weight of the lens was found to increase by approximately 20–30% for each of the formulations (Fig. 6A). Although the wettability and permeability are strongly related to the thickness of the contact lens,34) the water diffusion rate reached a steady state within 1 h. The thickest contact lenses exhibited the greatest weight change, with Formulation E, which has the highest PEGDA concentration, exhibiting the lowest increase in weight. We suggest that a high degree of cross-linking may reduce water retention because of the limited space in the matrix. In contrast, the second curing process increases the overall weight of the 3D-printed contact lenses (Fig. 6B). This suggests that further cross-linking of the 3D printer ink may make additional space to be filled with water. Alternatively, the second curing process may make the 3D object strengthen and swell, resulting in increased water retention.
The data represent the mean ± S.D. (n = 3).
The antimicrobial effect of the 3D-printed contact lenses was assessed on Staphylococcus aureus and Pseudomonas aeruginosa (Fig. 7). Formulations A–E loaded with AZI exhibited remarkable antimicrobial effects for the representative Gram-positive and -negative bacteria, which cause eye infections. The antimicrobial effect differed slightly among the formulations because the drug content was dependent on the weight of the contact lenses (Table 2). Overall, the diameter of the inhibition zone was approximately 30 mm, which was comparable to that for commercially available eye drops containing AZI. Although the present results suggest that the 3D-printed contact lens formulations presented here exhibit antibacterial effects comparable to those for eye drops, we expect that the contact lens formulation may be more effective for in vivo applications. Eye drop formulations are likely to leak out within a short time, whereas the proposed contact lens formulation can achieve sustained drug release. A control group of contact lenses not loaded with AZI exhibited a weak antimicrobial effect for Staphylococcus aureus (Formulation D-). We consider that remaining PEGDA monomers may affect bacterial growth, although further investigation is necessary.
The data represent the mean ± S.D. (n = 3).
In this study, AZI-loaded contact lenses were successfully prepared by VPP 3D printing, and their mechanical properties and pharmaceutical quality were characterized. The PEGDA and PEG 400 content and second curing process affected the dimensions, strength, and drug release for the resulting 3D-printed object. Although further investigation is necessary, antibiotic-loaded contact lenses prepared by 3D printing provide useful information for ophthalmic device manufacture and personalized dosing. This contact lens formulation is applicable as a controlled drug release device and in a setting before patients sleep. We estimated that the 3D printed contact lenses can be stored without water to prevent the drug release, and that they are used soak into water before use to prevent the absorption of tear film in eye.
This work was supported by Grant-in-Aid for Research in Nagoya City University (Grant Number: 2212007), and Grant-in-Aid for Scientific Research (C) KAKENHI (Grant Number: 23K06213).
Eiichi Goto, Conceptualization, Methodology, Investigation, Data curation, Writing-original draft, Visualization, Project administration; Tatsuaki Tagami, Conceptualization, Resources, Writing-review & editing, Supervision, Project administration, Funding acquisition; Koki Ogawa, Writing-review & editing; Tetsuya Ozeki, Resources, Supervision.
The authors declare no conflict of interest.
