A Novel, Low-Cost Alternative to Traditional Glaucoma Surgeries

Abstract

Purpose: To investigate the real-world efficacy of a novel, low-cost glaucoma drainage device in canine and human patients.

Methods: A retrospective case series of 17 eyes in 14 canines and one eye of a human patient who each underwent novel drainage device implantation is described. This device was constructed by insertion and advancement of a 24-gauge cannula (canine) or 23-gauge cannula (human) perpendicularly through five adjacent tubes of a 25-mm Yeates surgical drain.

Results: Of the canine patients, the average follow-up period was 362 days (range, 27–863). The mean preoperative intraocular pressure (IOP) was 50.9 ± 17.9 mm Hg. Following tube surgery, IOP was maintained at <20 mm Hg in 81.3%, 100%, 100%, 85.7%, 100%, and 75.0% of eyes at 1, 2, 3, 6, 9, and 12 months, respectively. Bleb needling and/or revisions were required in five eyes. Enucleations and/or device explantations were performed in five eyes at mean day 140. In the human case, the device was implanted in the right eye of a 64-year-old male with refractory raised IOP (55 mm Hg) despite maximum medical therapy. IOP was well controlled until day 818, when eventual tissue breakdown necessitated device removal.

Conclusions: This design represents a novel, low-cost, effective alternative to traditional glaucoma tube devices.

Translational Relevance: This device has great potential for use in regions where the needs for glaucoma drainage devices and surgical alternatives to trabeculectomy have not been met. Further development may include tube crimping or fenestration and preoperative loading of slow-release antibiotics and/or anti-metabolite medications within the non-draining lumens.

Introduction

Glaucoma is the leading cause of irreversible blindness globally and affects approximately 76 million people worldwide. It is further estimated to rise to more than 110 million by 2040, for which developing nations in Asia and Africa are expected to be disproportionately affected.¹ Treatment modalities for glaucoma include medications, laser procedures, and surgery. Traditional incisional surgeries include both trabeculectomy and glaucoma drainage device (GDD) implantation. In developed nations, medical or laser treatments are usually the preferred first-line therapies for patients, whereas incisional surgery is often reserved for refractory and/or advanced glaucomatous disease. However, in developing nations, surgical treatments are considered particularly viable first-line alternatives, as the cumulative costs of lifelong medical therapies may be particularly burdensome for low-income individuals. In these settings, trabeculectomies are more often performed due to the increased cost, lack of distribution, and limited availability of GDDs.² This is despite the growing preference for GDD implantation over trabeculectomy among members of the American Glaucoma Society, who elected GDD implantation as the preferred surgery in seven of eight clinical settings in 2016.³ 

Here, we report our experience with a novel, low-cost alternative to other GDDs. Our goal was to create a reliable, easily reproducible, and effective GDD from widely accessible and affordable materials already used in the healthcare setting. The approximate cost of this device is US$1.50 per unit. In comparison, the Baerveldt glaucoma implant (BGI; Johnson & Johnson Vision, Irvine, CA) costs approximately US$350. 

Methods

This was a retrospective analysis of canines and a single human patient with refractory glaucoma treated with surgical implantation of our novel tube. The study adhered to the tenets of the Declaration of Helsinki, as well as the ARVO Statement for the Use of Animals in Ophthalmic and Vision Research. Ethics approval for the use of this device in animals was obtained via Western Sydney Local Health District Animal Ethics Committee (WSLHD AEC 2019/1508), and approval for implantation in the human case was granted by WSLHD on compassionate grounds. The Australian Therapeutics Goods Administration was consulted to confirm that the materials were approved for use in an “off-label” setting. Informed consent for surgery was obtained from the human patient and the dog owners. 

Data Collection

Analysis of the medical records of the first consecutive 17 canine eyes treated with surgical implantation of the drainage tube at the Small Animal Specialist Hospital in Sydney was undertaken in October 2023. The general criterion for use of the tube was uncontrolled IOP > 25 mm Hg, and, in each case, surgical intervention was deemed necessary by a veterinary ophthalmologist for the preservation of vision and/or improved comfort following a comprehensive assessment. Data extracted included age at time of surgery, breed, history of ocular disease, IOP, ocular hypotensive medications used, presence of vision, and complications reported in the postoperative period. Successful postoperative control of IOP was defined as <20 mm Hg in keeping with studies of Baerveldt tubes in canines.^4,5 In the human case, data collection parameters included age, sex, ocular history, visual acuity, pre- and postoperative IOP, and complications reported in the postoperative period. 

Device Construction

Assembly of the device occurred on a sterile field after a surgical timeout with aseptic technique to ensure sterility. The device baseplate was made from a 19-mm cut section of a 5-lumen Yeates surgical drain manufactured from silicone elastomer (Smiths Medical, Minneapolis, MN) (Fig. 1). A 24-gauge cannula manufactured from polyurethane (Becton Dickinson, Franklin Lakes, NJ) for canine cases or a 23-gauge cannula for the human case was inserted perpendicularly in the middle of the sagittal plane of the drainage tubing, maintaining midplane alignment until piercing through the far edge of the fifth lumen. The needle was then retracted, and the cannula Luer-lock segment was severed from the cannula. The distal cannula was advanced approximately 17 mm, leaving the proximal cut end lying within the middle of the third lumen. The cannula stylet was then used to make several perforations through the lumens of the baseplate to connect them. This allowed for a pressure-dependent sequential filling of tubes, thereby creating a potential functional valve effect. 

Figure 1.

 

Novel tube glaucoma drainage device construction and dimensions. (A) Schematic drawings of device construction. (A1) A 23-gauge cannula is aligned to the middle of a sagittal plane of a five-tiered multitubular 19 × 14-mm piece of Yeates surgical drainage tubing. (A2) The cannula is inserted, maintaining midplane alignment until it pierces through the far edge of the fifth lumen. (A3) The needle is retracted and the Luer-lock is severed with a scalpel. (A4) The cannula tip is advanced approximately 17 mm, leaving the proximal portion within the middle of the third lumen of the baseplate. (B) Constructed device prior to implantation. (C) Schematic drawing of the constructed device including dimensions. (D) High magnification of the distal tip of the 23-gauge cannula with (above) and without (below) needle.

Surgical Technique

In the canine cases, novel tube implantation was performed under general anesthetic by one of two surgeons (MB, KG). A fornix-based conjunctival flap was created dorsolaterally, and blunt dissection was used to fashion a pocket for the placement of the implant overlying the sclera. Mitomycin C 0.04% was applied via a hemostatic sponge held in place for 2 minutes, with care taken to avoid contact with the edges of the conjunctival incisions. The site was then flushed copiously using 0.9% saline solution. The non-draining tubing of the implant baseplate was injected with Actilyse tissue plasminogen activator (tPA; Boehringer Ingelheim, Ingelheim am Rhein, Germany), and the implant was then placed between the adjacent extraocular muscles. The implant was secured to the sclera using two interrupted 9/0 nylon sutures passing through the anterior-most tube of the baseplate. The cannula tube was then cut on an angle with the bevel facing upward such that it would extend 3.5 mm into the anterior chamber. A 23-gauge needle was used to create a scleral tunnel approximately 2 mm posterior to the limbus. Following injection of a small amount of viscoelastic (Amvisc 1.6%; Bausch & Lomb, Bridgewater, NJ) into the tunnel, the tube was then inserted into the anterior chamber. The conjunctiva was closed using 9/0 polyglycolic acid sutures in a simple continuous pattern. Postoperative management consisted of oral amoxicillin/clavulanic acid (15–25 mg/kg every 12 hours), oral prednisolone (0.25–1 mg/kg every 12 hours; Apex Laboratories, Tamil Nadu, India), and colchicine (0.02–0.03 mg/kg every 24 hours; Aspen Pharmacare, Durban, South Africa), as well as topical prednisolone 1% (Prednefrin Forte; Allergan Australia, North Sydney, Australia), ketorolac trometamol 5% (Acular; Allergan Australia), and chloramphenicol 5% (Chlorsig; Aspen Pharmacare) eye drops each 4 times daily. Ocular hypotensive agents such as dorzolamide hydrochloride 2%/timolol maleate 0.5% (Cosopt; Merck Sharp & Dohme, Macquarie Park, Australia), brinzolamide 1% (Azopt; Alcon Laboratories, Geneva, Switzerland), and/or latanoprost 0.005% (Xalatan; Pfizer Australia, Sydney, Australia) were added and/or ceased as required to control IOP in the postoperative period. 

In the human case, under local anesthesia, device implantation was carried out in a similar fashion. A conjunctival peritomy was performed in the superotemporal quadrant, and blunt dissection was used to expose the subtenon space. The device baseplate was implanted 9 mm posterior to the limbus and secured to the sclera using 8/0 nylon sutures as previously described. A 23-gauge needle was used to create a scleral tunnel 2 mm posterior to the limbus, and the tube was then advanced into the anterior chamber. A double-thickness Tutoplast patch graft (Tutogen Medical, Alachua, FL) was secured over the tube with fibrin glue to minimize the risk of tube erosion. Conjunctival closure was achieved using fibrin glue and Vicryl 7-0 sutures. Subconjunctival dexamethasone was injected (1 mg/mL) at the conclusion of the case. 

Follow-Up

IOP was measured with rebound tonometry (iCare TONOVET; iCare, Oy, Finland) in the canine cases and with Goldman applanation tonometry for the human case. Follow-up reviews generally occurred in canine cases initially daily until discharge, weekly for the first month, and then in decreasing frequency dependent on the progress and at the discretion of the treating ophthalmologist. 

Statistical Analysis

Prism 10.0 for Windows (GraphPad, Boston, MA) was used for statistical analysis. Overall significance was confirmed using a mixed-effects model (restricted maximum likelihood method). Post hoc Dunnett's multiple comparisons testing was used to confirm significant differences in IOP and topical glaucoma eye drops usage between the preoperative time point and all subsequent time points. Normality of the data was confirmed by Q-Q plotting of the residuals. 

Results

Canine Cases

A summary of the clinical and demographic characteristics of the 17 canine eyes is provided in the Table. The follow-up period varied from 27 to 863 days (mean, 362 days). In all except one case (dog 6), this surgery was performed as the primary procedure for IOP reduction. In dog 6, the new device was implanted in an eye that had previously undergone Baerveldt GDD insertion 2 years prior that had subsequently failed. 

Table.

 

Clinical and Demographic Characteristics of Canine Patients Undergoing Novel Tube Insertion

 

Preoperatively, the overall mean ± SD preoperative IOP was 50.9 ± 17.9 mm Hg. Following novel GDD implantation, the mean IOPs were 9.0 ± 3.8 mm Hg (P < 0.001), 11.5 ± 5.8 mm Hg (P < 0.001), and 12.7 ± 5.3 mm Hg (P = 0.001) at 3, 6, and 12 months, respectively, for the eyes that had follow-up to each period. The percentages of eyes with IOP < 20 mm Hg were 81.3%, 100%, 100%, 85.7%, 100%, and 75.0% at 1, 2, 3, 6, 9, and 12 months, respectively (Fig. 2). The canine patients were on an average of 6.2 ± 1.9 topical glaucoma drops daily prior to surgery. The mean numbers of daily anti-glaucoma eye drops used at 3, 6, and 12 months were 1.0 ± 1.2 (P = 0.001), 1.2 ± 1.4 (P < 0.001), and 1.4 ± 1.3 (P < 0.001), respectively, for the eyes that had follow-up to each period. 

Figure 2.

 

Canine IOP following novel glaucoma drainage device implantation. The number of eyes available for data collection at months 1, 2, 3, 6, 9, and 12 postoperatively were 16, 15, 14, 14, 12, and 8, respectively.

Transient hypotension (IOP < 5 mm Hg) occurred in three eyes (17.6%) only once at 2 months after surgery (dog 3 bilaterally and dog 7). For dog 7, this was followed by a revision surgery including debridement of the Tenon capsule and replacement of the novel tube device. Intraocular hypertension (IOP > 25 mm Hg) occurred in five eyes (29.4%), usually within the first 2 months (dogs 1, 2, 3, and 10). Five of these eyes underwent either bleb needling or revisions, with or without mitomycin C, and one dog required three revisions (range, day 7–day 64) (Table). Fibrin formation was noted in 10 of 17 eyes (58.8%) and generally resolved within the first 4 weeks. One eye developed a fungal ulceration (dog 8), initially noted at day 41, which was successfully treated topically with antifungal agents. 

Five of the 17 eyes (29.4%) underwent subsequent enucleation and/or device explantation, at an average period of 139.8 ± 89.1 days. Reasons for enucleation included one case of poor vision with elevated IOP (day 27), two cases of significant anterior fibrinous uveitis with refractory raised IOP (day 45 and day 165), and one case of a late dislodged implant (day 250). One eye required device explantation at day 212 following late conjunctival breakdown, device exposure, and secondary Pseudomonas infection. This was subsequently replaced with an Ahmed tube. 

Human Case Report

A 64-year-old man with a complex 30-year history of primary open-angle glaucoma was referred for refractory high IOP in his left eye of up to 55 mm Hg despite maximal medical therapy including oral azetazolamide. His previous ocular history included bilateral failed XEN Gel Stents followed by bilateral Baerveldt tube insertion 3 years prior. He had also undergone bilateral cataract surgery and a later left eye Baerveldt tube repair. He subsequently developed Streptococcus pneumoniae endophthalmitis necessitating vitrectomy and removal of the Baerveldt tube 2 years prior to insertion of the drainage tube. His vision before implantation of the novel device was 6/24. This was reduced to 6/48 at 1 month but recovered to 6/24 by 12 months and was 6/36 at 2 years. 

In the following 12 months, the left IOP was maintained between 9 and 15 mm Hg without any additional topical ocular hypotensive medications (Fig. 3). At day 548, dorzolamide 2% and timolol maleate 0.5% were introduced, and latanoprost 0.005% was introduced at day 634 to manage rising IOP. IOP was maintained within target until tissue breakdown eventuated, necessitating removal of the device on day 818 and subsequent cyclodiode therapy. Histopathological assessment of the excised conjunctiva (Fig. 4) demonstrated non-specific features of acute and chronic inflammation without evidence of granulomas, dysplasia, neoplasia, or infection. 

Figure 3.

 

Postoperative IOP of the human patient following implantation with the novel tube. Day 0 represents preoperative IOP. Dorzolamide and timolol were initiated at day 548 and latanoprost daily at day 634. The device was explanted at day 818 due to conjunctival erosion.

Figure 4.

 

Histopathological analysis of excised conjunctiva following removal of the novel device that demonstrates non-specific features of acute and chronic inflammation. (A) Low magnification. (B) High magnification.

Discussion

Glaucoma drainage devices are divided into valved and non-valved devices. The most commonly used non-valved implant is the BGI. More recently, the newer Aurolab Aqueous Drainage Implant (AADI; Aurolab, Tamil Nadu, India) has gained increased standing among low- to middle-income countries as a suitable alternative, as its design is a prototype of the BGI and it is produced at a significantly lower cost locally in India.⁶ Our novel GDD is simple to construct and is made from biocompatible materials (silicone and polyurethane) that are widely available even in rural healthcare settings in developing nations. The approximately cost of this device is US$1.50 per unit. In comparison, the BGI costs approximately US$350, and even the AADI costs US$50. Furthermore, the AADI may be susceptible to local price mark-ups and intermediary fees. Thus, our novel device was conceived as an ultra-low-cost GDD with significant potential for uses in areas where more traditional devices are not available or are prohibitively expensive. To further improve the cost effectiveness, coverage of the GDD stent with a long scleral tunnel and conjunctival closure with sutures is recommended in place of Tutoplast and fibrin glue. 

Although our device is quite different in design features than the BGI and the AADI, the therapeutic principles of increasing aqueous outflow via shunting, bleb formation, and episcleral venous drainage remain the same. In canines, successful IOP control (defined as IOP < 20 mm Hg) with our device at approximately 1 year (75%) was comparable to the reported outcomes of Baerveldt tubes in other studies (71%–74%).^4,5 However, the reduction in the number of daily glaucoma eye drops administered postoperatively was slightly lower with our GDD (ranging from 6.2 doses/day preoperatively to 1.4 doses/day 12 months postoperatively) than Baerveldt tubes in these same studies (ranging from 6.3–6.4 doses/day preoperatively to 0.5–0.6 doses/day 12 months postoperatively).^4,5 In our human case, novel GDD implantation was undertaken as a final surgical attempt to control IOP in an eye with a complex history, including previous XEN tube insertion, BGI implantation, endophthalmitis, and vitrectomy. IOP was well maintained and comparable to the mean IOP reported in Baerveldt tubes at 1 year.⁷ Subsequent cyclodiode laser treatment was delayed for a period of more than 2 years prior to tissue breakdown. 

Sequential filling of the draining lumens of the baseplate of our novel device may also provide a valve-like effect. Our results demonstrated low rates of hypotony in canines (17.6%) and an absence of hypotony in the human case, suggesting that this may well be the case. In comparison, postoperative hypotony is reportedly common in canines following Baerveldt tube implantation (20%–81%).^4,5,8 In terms of dimensions, our novel tube is comparable with the BGI and the AADI. For example, the baseplate of our tube measures 270 mm² in cross-sectional area, which lies between the 250 mm² and 350 mm² available sizes of the BGI. Given that the diameters of the tubes of the baseplate are greater than the thickness of the BGI baseplate, the baseplate sizing was designed such that it would fit between two adjacent recti muscles, as opposed to fitting underneath adjacent muscles as is required by the BGI, to perhaps reduce the risk of significant postoperative diplopia. 

Although the rate of complications and/or enucleations in the dogs in this series is high by human standards, it is important to recognize that glaucoma management in dogs is a vastly different enterprise. In comparison to humans, glaucomatous disease in dogs typically presents at a much later stage, often only when the second eye is affected and when the vision is severely compromised. Tissue response to glaucoma surgery is also markedly different in dogs, as surgery tends to induce more intense inflammatory reactions than those seen in humans. For example, in dogs, fibrin formation in the anterior chamber is common following Baerveldt tube surgery, occurring in 60% to 63% of cases, and often necessitates treatment with intracameral tPA injections.^4,5,8 In this study, fibrin formation was noted at a similar rate (58.8%) and did not result in any clinical sequelae such as tube occlusion and intraocular hypertension. Furthermore, enucleation rates following Baerveldt tube insertion in dogs have been reported as to be as high as 25% to 34% in similar studies and are undertaken most commonly due to inadequate IOP control or, less frequently, endophthalmitis.^4,8 Thus, our GDD demonstrates similar safety and efficacy in comparison to the existing literature on Baerveldt tube insertion in canines, although both devices do less well in dogs compared with Baerveldt tubes in human studies. 

This real-world study allowed for longer term follow-up data than may have otherwise been possible in an experimental setting, and the promising performance in an uncontrolled environment suggests that further development is warranted. Limitations to this study include its retrospective nature and the resulting minor variations in surgical technique, duration, and frequency of follow-up, as well as the protocol for bleb needling and/or revision. Furthermore, the degree and type of glaucoma were not standardized across the small case numbers of canine patients. Recruitment of canine patients was highly dependent on owner consent and cooperation, and the costs of treatment and ongoing care may be considerable. The impact of this selection process is difficult to determine. Finally, ongoing assessment and monitoring of the GDD bleb following surgery was also limited in the canine cases, as the posterior location of the endplate, as well as the large corneal diameter of dogs, makes visualization frequently impossible without anesthesia. 

Regardless, this study demonstrates the efficacy and use of a novel, ultra-low-cost GDD made of simple and easily accessible materials. The authors envisage its use particularly in low- and/or middle-income countries where the need for effective GDD and surgical alternatives to trabeculectomy have not been met. Further directions for development of this device may include tube crimping or fenestration to optimize aqueous flow rates with using three-dimensional printed standardized templates. Other possibilities include drug loading and sustained delivery within the non-draining lumens of the device baseplate, as well as the development of bioactive glass hybrid materials to improve the implant–tissue interface and minimize the associated inflammatory response. 

Acknowledgments

Disclosure: C.F. Kong, None; J. Yek, None; P. Clegg, None; K. Graham, None; R. Gupta, None; P. Boughton, (O, P); M. Billson, None; A. White, (P)