Design and preliminary evaluation of a novel,flexible endourological device for transurethral intravesical implantation of drug-coated titanium seeds

Abstract

Objectives:

To develop a functional prototype of a device capable of reliably applying drug-coated titanium seeds transurethrally into the bladder wall through a cystoscope, as part of a seed-based system for intravesical drug instillation to treat non-muscle invasive bladder cancer (NMIBC). This innovation addresses the limitations of current approaches, such as short drug dwell time in the bladder lumen and poor drug penetration into the bladder wall, while preserving the benefits.

Methods:

An interdisciplinary collaboration between biomedical engineers and clinical urologists followed a rigorous design process with theoretical evaluation of various designs, resulting in the production of a prototype. This was tested for its ability to implant seeds into different materials (gelatin, bovine muscle tissue). Model-based application by a panel of urologists resulted in the refinement of the initial prototype. The optimized prototype was finally tested in an ex vivo pig bladder model.

Results:

A functional prototype combining a reusable user interface component and a single-use component was successfully produced. In this cystoscopy-based approach, the seeds stored at the distal end of the device are applied to the bladder wall using a proximal trigger/plunger mechanism. We demonstrated efficient and accurate seed implantation in bladder-mimicking models and in an ex vivo porcine bladder experiment.

Conclusions:

We developed a device that provides proof-of-concept for the feasibility of transurethral intravesical seed application using the working channel of a flexible cystoscope. With further refinement of the design and adequate validation of corresponding drug-coated seeds, this device would be suitable for testing in patient bladders.

Level of evidence:

Not applicable

Introduction and background

Bladder cancer is the ninth most common cancer globally with a 4:1 predominance in men versus women. ¹ It can be categorized into muscle invasive bladder cancer (MIBC) and the more common non-muscle invasive bladder cancer (NMIBC).^2,3 Localized bladder cancer is characterized by a considerable recurrence and progression rate leading to a mortality rate of 3.7–4.4/100,000 in North American and European men.^2,4 MIBC requires aggressive multimodal therapy with systemic chemotherapy, bladder removal and sometimes bladder radiation. Standard treatment of NMIBC involves transurethral resection of bladder tumors (TURBT) followed by adjuvant intravesical chemotherapy (e.g. mitomycin C, epirubicin or gemcitabine) or immunotherapy (Bacillus Calmette-Guérin (BCG)) to reduce recurrence and progression risk. ²

Intravesical therapy is attractive because the therapeutic agents remaining largely to the bladder lumen, minimizing systemic toxicity. However, instilled drugs often fail to adequately penetrate the bladder wall, and the dwell time in the bladder limits their efficacy. Furthermore, each agent requires an appropriate chemical formulation for intravesical instillation, slowing drug development in this field. Novel devices to facilitate intravesical drug delivery are urgently needed. For instance, it is well known that up to 80% of NMIBC harbor activating FGFR3 (fibroblast growth factor receptor 3) mutations promoting tumor growth.^5,6 Oral FGFR inhibitors (FGFRi) are available and approved for the treatment of advanced bladder cancer, ⁷ but systemic delivery is associated with adverse effects limiting its routine use in patients with NMIBC. Only recently, the first phase I study introduced a local FGFRi delivery system (TAR-210), administering erdafitinib intravesically to NMIBC patients with FGFR mutations. ⁸

We have developed coated titanium seeds loadable with various drugs such as FGFRi with the aim to implant these seeds into the bladder wall to enable sustained drug elution over time. Clinical implementation of such a strategy necessitates designing an endourological device for seed implantation. Consequently, this project aimed to develop a functional prototype of a novel seed injection device that reliably places seeds intravesically in an outpatient setting with minimal disruption to existing workflows.

Material and methods

Design process

The design process was structured into three phases: orientation, design and construction, and optimization.

In the orientation phase, meetings with clinical stakeholders (nurses, urologists, medical-device reprocessing staff) and pharmacologists involved in drug coating, and clinical observerships (attendance at cystoscopies and demonstration of established endourological devices) led to the defining of critical device functions from an engineering perspective: accessibility to the inner bladder wall (MA: method of access), maneuverability to tumor sites (MM: method of maneuverability), provision of visualization (MV: method of visualization), storing and loading of seeds (SSL: seed storage and loading), user-activated seed ejection (MEA: method of ejection activation), isolated ejection of seeds, and implantation of a seed into the bladder wall (MEI: method of ejection and implantation). These functions guided the evaluation of concepts.

The design and construction phase comprised conceptualization, concept design, and concept construction stages. The initial conceptualization phase focused on identifying the best approach for bladder access. A cystoscopy-based approach was used as reference concept, chosen for its clinical establishment and wide range of technical solutions, and based on standard flexible cystoscopes with ⩾ 2.0 mm (6-Fr) working channels routinely used in outpatient practice. A Pugh matrix was used to evaluate the performance of each conceptualized approach for the most relevant critical functions at this stage (MA, MM, MV). Each competing concept was reviewed with respect to these functions to determine its superiority (1 point), equality (0 points) or inferiority (−1 point) compared to the reference method. Net scores were calculated, recorded in a rating table and analyzed.

In the concept design stage, various device designs were drafted and evaluated based on the initial conceptualization results. A revised Pugh matrix, now incorporating criteria for seed storage and loading (SSL), ejection activation (MEA), and implantation (MEI), alongside access (MA), maneuverability (MM), and visualization (MV), identified the top four designs. These were further assessed using a weighted decision matrix (WDM) that considered implantation depth, successive implantation, device costs, seed capacity, procedure time, technical feasibility, and implementation accuracy, with respective weights of 0.2 for the first two and 0.1 for the remaining. Five team members rated each design from 0 (poor performance) to 10 (great performance), with a team meeting held to assign a consensus rating. The best rated design was identified as the most promising solution and used as design basis. The highest-scoring design was chosen as the basis for subsequent development steps.

In the concept construction phase, a three-step approach was used. Initially, the concept’s fundamental feasibility was assessed using cost-effective materials. This was succeeded by creating of computer-aided design (CAD) models and large-scale 3D printing. The final step involved refining the design to realistic dimensions with high-quality materials and conducting iterative testing as described below. The testing results led to iterative refinements of the device’s internal mechanics and user interface. The resulting prototype, Prototype I, underwent rigorous testing by six clinical urologists using gelatin and bovine muscle models, concluding the design and construction phase.

Based on accumulated knowledge and feedback, Prototype I received several modifications during the optimization phase, resulting in the creation of Prototype II, which was subsequently evaluated on an ex vivo porcine bladder model to assess seed implantation reliability. Figure 1 provides a flow chart of the design process.

Figure 1.

Flow chart of the design process.

Applied test scenarios

The prototypes were assessed across multiple test scenarios using gelatin and bovine muscle model. Literature suggests the Young’s modulus, indicative of tissue elasticity, varies from 64 to 160 kPa across different bladder wall sections. ⁹ We chose a 20% gelatin concentration, closely mirroring the average bladder wall’s 80 kPa elasticity, for initial tests. ¹⁰ Models measuring 400 cm² (20 × 20 cm) and approximately 0.5 cm thick (similar to the average bladder wall thickness) were prepared. ¹¹ Initially, seed implantation was performed without limiting the cystoscope’s mobility, followed by restricted mobility tests where gelatin plates were placed in a polymethyl methacrylate box with a 0.8 cm tubular access channel. Implantation accuracy was visually verified, with deviations measured and penetration depth analyzed post-dissection.

Certain bladder areas exhibit a higher Young’s module, particularly those with stronger muscular components. ⁹ Therefore, the aforementioned tests were repeated on bovine muscle tissue with a thickness of 0.5 mm and an area of 100 cm², representing a Young’s modulus of 110 kPa. ¹²

These test scenarios provided the basis for an evaluation by a board of six internationally trained urologists from Switzerland, Japan, Belgium, Germany, and Canada not previously involved in the project. Following a set protocol, the evaluation was conducted individually. Initially, the urologists received explanatory videos on device operation, then they familiarized themselves with the device, including assembling it. Subsequently, they were given a preloaded device to perform targeted implantations into gelatin and bovine muscle tissue models. Standardized questions rated the device’s intuitiveness on a Likert-type Scale from 1 (not intuitive) to 10 (highly intuitive). Further questions post-testing assessed the device’s practical application, focusing on clinical workflow integration, user and patient safety, setup, handleability, and meeting clinical requirements.

Based on the results, Prototype II was produced and tested on two porcine bladders in an ex vivo model. The pigs were euthanized during surgical simulation training for trauma surgeries. Pig bladder collection adhered to guidelines from the University of British Columbia Animal Care Committee and the Canadian Council of Animal Care guidelines (protocol no. A22-011). During harvesting, the urethra was dissected to include at least 4 cm of length. The bladder was placed in a transparent polymethyl methacrylate box with a side length of 11 cm, and the internal cavity was adjusted to 500 cm³ using polyurethane to mimic pelvis placement. The urethra was connected to 0.8 diameter plastic tube for access. Seed implantation target sites were marked on the bladder’s exterior, deliberately covering different bladder regions including the anterior wall, in order to assess device performance under conditions of maximal cystoscope deflection. Bladders were filled with approximately 500 cm³ sterile saline.

Results

During the conceptualization stage, a comparative analysis of the criteria MA, MM, and MV showed that no alternative concept surpassed the cystoscopy-based approach. The 11 developed concepts varied mainly in terms of MEA, SSL, and MEI criteria. After narrowing down to the top four concepts, the design below emerged following a WDM evaluation. Pilot testing with large-scale 3D prints of CAD models confirmed its operability, leading to its production in realistic dimensions.

Components, materials and sizes of prototype II

Prototype II consists of two main components: a reusable user interface featuring a handle and trigger, and a single-use component comprising a cassette, tubing with an internal plunger, and a distal needle. The handle, trigger, and cassette are made from 3D-printed polylactic acid (PLA). The tubing is polytetrafluoroethylene (PTFE), and the internal plunger, crafted from stainless steel wire, attaches to a commercial 14G stainless steel needle via a biocompatible, bicomponent adhesive. An internal distal silicone layer enhances the needle’s friction, preventing premature seed displacement.

The curved handle measures 95.47 mm in height, 113.54 mm in length, and 30.00 mm in width. The trigger, fitted into the handle’s rail system, has widths of 67.90 mm and 21.00 mm, and a length of 23.00 mm. The closing rail cap measures 39.37 mm high, 30 mm wide, and 12.50 mm long. The inserted cassette reaches a total length of 70 mm. Tubing dimensions include an outer diameter of 1.9 mm, an inner diameter of 1.5 mm, and a total length of 855.0 mm. The internal plunger extends 893.2 mm with a diameter of 0.9 mm. The needle is 23 mm long, with an additional 5 mm for the needle tip. Figure 2 illustrates the design of the device.

Figure 2.

Design of prototype II. (a) Schematic model of the device: (1) user interface with handle, trigger and rail cap; (2) cassette and tubing connector; (3) tubing with plunger inside; (4) needle with loaded seeds. (b) Drawings of the re-usable user interface (left) and single-use component (right). (c) 3D printed parts of the device including user interface (top), cassette with tubing and plunger (bottom) and assembled device (right). Note: needle not attached.

Handling and functionality of prototype II

Assembly begins with the user interface, consisting of the handle, trigger, and rail cap. The trigger is inserted into the handle’s rail system and secured with the rail cap. Next, the cassette gets positioned within this reusable segment. Using a plug-in mechanism, the tubing connector is inserted into the designated notch on the trigger.

After insertion through the cystoscope’s working channel, intravesical seed implantation is performed. The needle is intended to penetrate into the bladder wall to allow seed deposition in the submucosal to superficial detrusor layer, analogous to the target depth used for intravesical botulinum toxin injections. User-initiated bladder wall penetration is achieved by manually advancing the tubing, similar to botulinum toxin injections for neurogenic and overactive bladders. During this process, the tube’s outer segment retracts relative to the inner plunger with each trigger pull, facilitating the ejection of one of the five seeds stored in the device’s distal section. The slots provide haptic feedback upon seed ejection. Figure 3 illustrates the device’s application.

Figure 3.

Anticipated application of the device. (a) Insertion of the single-use cassette in the re-usable user interface. (b) Schematic drawing of seed implantation with illustration of the processes in the device (top) and general steps of the implantation (bottom).

Modifications from prototype I to prototype II

Several modifications were implemented from Prototype I to Prototype II based on feedback from clinical urologists. The cassette’s length was reduced from 65 to 60 mm, requiring adjustments to the handle’s rail and trigger, enhancing usability for individuals with smaller hands. To improve the fit and alignment between the plunger and trigger, facilitating easier ejection, the plunger connector’s hole diameter was decreased and its length increased. A hook was added to restrict the cassette’s lateral movement within the handle. Ergonomic enhancements also included shortening the handle slightly and adding an edge fillet to increase user comfort.

Performance in test scenarios

Initially, the basic design’s efficacy was evaluated using Prototype I in gelatin and bovine muscle tissue models, with clinical urologists conducting the tests. Following these outcomes, the refined Prototype II was developed and tested on porcine bladders.

In the evaluation of Prototype I, urologists rated its intuitiveness at 9/10 on the Likert-type scale. On average, the trigger was activated twice before feeling confident with the device’s handlebar. Criticisms included the lack of a protective cover for the implantation needle, posing a risk to the cystoscope’s working channel coating. The handlebar design received an 8/10 Likert-type score. Positive feedback was given for the tactile feedback upon advancing to the next cassette notch, though improvement was suggested. Urologists with smaller hands found the handlebar too large, affecting comfort, while its lightweight nature was viewed favorably. The overall device assembly scored 6/10. A noted area for enhancement was the cassette’s insertion into the handlebar, with a recommendation for an audible click to indicate correct placement.

In functional tests, the gelatin model showed an average successful implantation rate of 3/5 seeds, while the bovine muscle tissue model achieved about 2.33/5 seeds. Both models consistently reached the required implantation depth, with seeds fully implanted in 4/5 cases. Precision-wise, both models managed successful implantation within a 1–2 mm margin. The implantation time for each seed ranged between 8.5 and 10.92 seconds. Each device effectively stored and implanted 5 seeds sequentially without premature loss, provided the trigger was correctly used. Initial tests on the gelatin model indicated occasional excessive retraction of the trigger beyond a single slot, causing multiple seeds to eject simultaneously, with usually only the first seed properly placed. This observation was attributed to inadequate resistance between cassette slots. Maneuverability was affected by the seed device’s tubing in the cystoscope’s working channel, particularly during implantations at the bladder’s anterior wall which require extensive cystoscope flexion. Despite this, cystoscopic visualization of the bladder was mostly unaffected, although the lack of visible markings on the needle made assessing the exact depth of implantation challenging.

Prototype II was tested in an ex vivo pig bladder study to evaluate its implantation efficacy in authentic bladder tissue. This testing confirmed the limitations seen in earlier gelatin and bovine muscle tissue models. Maneuvering the cystoscope to the bladder’s anterior regions was notably challenging with the device inserted, making seed implantation in these areas impractical. Initially, the cystoscope could deflect 210° upward and 140° downward, but this range decreased to 79.3° upward and 54° downward with the device inserted. In addition, the lack of depth markings on the needle led to perforations in the porcine bladder wall after insertion. Despite these challenges, repeated seed implantations into the bladder wall were achievable, albeit with a reduced success rate (2/10 seeds implanted). Figure 4 shows images of the pig bladder experiment.

Figure 4.

Ex vivo pig bladder experiment. (a) The porcine bladder is placed inside a transparent polymethyl methacrylate box. The internal cavity was reduced to a size of 500 cm³ to simulate placement within the pelvis. The urethra was fixed to an accessing plastic tube with a diameter of 0.8 cm. (b) Placement of 2 seeds inside the bladder wall (black circle), caused by trigger retraction over two slots. Only the first seed is completely inserted inside the bladder wall. (c) Cystoscopic images show successful complete seed implantation (left), incomplete implantation (middle) and an urothelial defect after seed injection with subsequent dislodgement (right).

Discussion

Implantation of drug-coated titanium seeds, or those made from an alternative biodegradable substance, represents a novel concept for local drug delivery in patients with early-stage bladder cancer. Standard intravesical chemotherapy at this disease stage is limited by rapid drug dilution, washout and restricted urothelial penetration, and although device-based approaches such as thermochemotherapy or drug-retaining systems (e.g. the Pretzel device) aim to prolong luminal drug exposure and penetration, they remain dependent on repeated catheter-based instillation and passive tissue diffusion. In contrast, intramural implantation of drug-eluting seeds enables direct tissue-level delivery with sustained local release independent of intravesical dwell time. Our group is currently investigating such an approach using FGFRi-coated titanium seeds endoscopically placed close to the tumor, but the concept could be applied to a broad range of compounds, including established chemotherapeutic agents. Our group is currently investigating such an approach using FGFRi-coated titanium seeds endoscopically placed close to the tumor, but the concept could be considered for many different compounds. In this project we have developed an endoscopic instrument designed for minimally invasive seed implantation that can be seamlessly and cost-effectively integrated into existing clinical workflow in an outpatient setting. In an iterative process involving numerous stakeholders, we have produced a prototype device suitable for further clinical development if the implantable drug-eluting seed concept proves successful.

We have developed a novel endourological device capable of implanting five consecutive titanium seeds into the bladder wall via the working channel of a standard flexible cystoscope. This method offers a practical and cost-effective solution, requiring only minor modifications to existing clinical workflows. Our design features two main components, which, due to early development stages and budget constraints, are currently made from economically viable materials. The vision is to upgrade these to higher-quality materials in future iterations. A key aim in our design process was to minimize waste, leading to the creation of a reusable user interface. The current prototype uses 3D-printed polylactic acid (PLA) for rapid prototyping and is therefore not intended for autoclave sterilization. In a clinical-grade version, the reusable user interface would be manufactured from medical-grade autoclavable metal alloys, allowing standard hospital reprocessing and repeated use, and its ergonomic and mechanical properties were already iteratively refined from Prototype I to Prototype II. The second component (the so-called cassette) connects the distal device parts (needle tip, tubing, plunger) to the handle and trigger mechanism. Engaging the trigger retracts it to the next slot, pulling the tubing back relative to the stationary plunger and ejecting the seed at the distal end. The prior penetration of the bladder wall by the needle ensures the seed’s placement within the bladder wall. The intended target zone for seed deposition is the intramural compartment of the bladder wall, corresponding to the submucosal to superficial detrusor layer, which provides proximity to urothelial tumors while minimizing the risk of transmural perforation. Indeed, superficial placement resulted in urothelial disruption and subsequent seed dislodgement in the porcine model, highlighting the importance of precise implantation depth. This directly informs future design requirements, including optimized needle length and bevel geometry, visible depth markings, and potentially modified seed geometry or anchoring mechanisms to ensure stable intramural retention. In the current stage, the cassette is made from 3D-printed PLA, the plunger from stainless steel wire, and the tubing from PTFE, connected to a standard 14G stainless steel piercing needle using a bicomponent adhesive. These materials will require upgrades in future development phases. The cassette should be designed from a more robust material to withstand the forces exerted by a metallic user interface, ensuring stability. The tubing and plunger materials are of particular importance as they must offer enough rigidity to transmit the force needed for seed ejection while maintaining flexibility for cystoscope mobility. Possible materials include polyurethane or silicone. Future designs could emulate the flexible mechanics of established endoscopic tools like biopsy forceps, which use a central core and spiral steel wire. It’s crucial to note that the current prototype’s limited flexibility restricts seed application to the anterior bladder regions.

In addition, the device’s distal tip needs optimization to ensure a seamless connection between the tubing material and the tip itself. The internal silicone coating requires enhancements due to observed premature seed losses. An external valve mechanism could be implemented. Importantly, a visible marking on the needle’s exterior during cystoscopy is crucial to gauge optimal penetration depth. There is also a need to develop a protective mechanism for the needle tip to prevent potential damage to the cystoscope’s working channel coating.

The design process yielded an advanced, operational proof-of-concept prototype, laying the groundwork for future enhancements. This prototype has successfully implanted seeds in various models, including gelatin, bovine muscle tissue, and an ex vivo pig bladder, affirming our confidence in its efficacy. The outlined limitations are predominantly attributed to time and budget constraints, which affected the acquisition of more suitable and flexible materials.

Conclusion

In our collaborative project, we successfully designed, manufactured and tested a novel endourological device for minimally invasive transurethral implantation of drug-eluting titanium seeds into the bladder wall through the working channel of a conventional flexible cystoscope. These promising results serve as proof-of-concept and basis for the continued evolution of our device. The current prototype necessitates further refinements to enhance its functionality and improve reliability prior to clinical implementation.