Additively manufactured reinforced polymeric stents for peripheral artery disease: design, manufacture and performance evaluation

Abstract

Stent failure in the peripheral arterial system remains a major clinical challenge due to high mechanical demands caused by limb movement. Unlike coronary arteries, below-the-knee vessels are exposed to continuous flexion, torsion and axial movement, increasing the risk of structural failure, stent collapse, fracture or migration. Due to their tuneable mechanical properties, polymeric stents enabled by advanced additive manufacturing emerge as a potential candidate to overcome the current limitations of metallic stents. This study examines whether polymeric stents made with the ST3DT process and reinforced photoresin can endure all phases of manufacturing and deployment — including printing, rinsing, post-curing, crimping, and sterilisation — while retaining suitable mechanical properties for vascular use. Nine different stent geometries were produced by varying strut angle and printing feedrate, thus achieving different strut profiles. Curing was assessed by DSC and FTIR and mechanical properties were evaluated with radial and axial compressions tests, three-point bending and crimping. Finally, a CT-scan was performed to analyse radiopacity of reinforced polymeric stents with promising results. The findings indicated that although decreasing strut angles and reducing printing speeds led to better mechanical performance, they also made the crimping process more challenging. Two stents were selected with a balance of all criteria to continue the development of novel photopolymerised polymeric stents.

Introduction

In the early twentieth century, cardiovascular diseases (CVD) are the leading cause of death in economically developed countries. The World Health Organization (WHO) reports that in 2019, vascular diseases caused around 17.9 million deaths, accounting for 32% of all deaths worldwide. Of these, 85% were attributed to heart attacks and strokes [1]. Peripheral artery disease (PAD) is the atherosclerotic narrowing or blockage of arteries, most commonly affecting vessels supplying the legs. Between 1990 and 2019, the global prevalence increased by 72%, reaching over 110 million individuals [2]. Peripheral artery disease (PAD) may progress to critical limb ischemia (CLI) if not managed appropriately. CLI is characterised by chronic and insufficient blood flow to the affected limb, leading to pain at rest and tissue loss [3]. The healthcare costs associated with PAD are substantial [4], primarily driven by hospital admission and lower limb revascularisation procedures [5]. Current metallic drug-eluting stents (DES) have not succeeded in reducing in-stent restenosis (ISR) in the below-the-knee (BTK) area [6, 7]. Moreover, the physiological response to a metallic foreign body may be adverse, often resulting in immune system activation and chronic inflammation [8]. Recent clinical trials have shown that bioresorbable stents (BRS) may be a viable option for treating PAD [9]. However, several important factors still need to be addressed before this can become an effective treatment for such a complex region of the body [10, 11].

DES are metallic stents with drug coatings developed to improve upon bare-metal stents (BMS) for treating coronary artery disease. However, their high radial strength — particularly in stents made from cobalt-chromium (CrCo) alloys — often reduces flexibility and limits the device’s ability to conform to the arterial lumen. Besides, the high rates of femoropopliteal artery (FPA) interventions are often attributed in part to severe mechanical deformations that occur due to limb movement [12].

Many studies have focused on enhancing the polymer coatings of DES or designing BRS stents that dissolve over time, but the use of a permanent polymer core has not been widely studied. Previous work has demonstrated the feasibility of 3D printing bioresorbable and permanent polymeric stents with tight manufacturing tolerances across a range of material formulations, highlighting the potential of additive manufacturing for precise and customisable stent fabrication [13]. Early this decade, in 2021, De Oliveira et al. used a commercial DLP Cartesian printer to fabricate bioresorbable stents, achieving a production output of 180 stents per print [14]. This demonstrated both the reproducibility and potential industrial scalability of small-scale medical prostheses.

In 2022, Ding et al. manufactured a bioresorbable stent with a µCLIP technology with a biocompatible photoresin doped with iodixanol to foster device visualisation during the procedure [15]. Two years later, Ding et al. developed a citrate-based bioresorbable stent and successfully delivered it into a swine artery, showing that polymeric stents can be produced using vat photopolymerization (VPP) technology [16].

A common limitation of the studies cited above lies in the need to laminate the 3D model prior to printing [17]. This step negatively impacts the dimensional accuracy, surface roughness [18], and — most critically — the mechanical performance of the resulting stents. To address these issues, Bosch et al. recently introduced a novel tubular VPP method, termed ST3DT (Stereolithography 3D Tubular). This method produces stents in less than five minutes, using only 5 mL of resin for devices 1–8 mm wide and up to 60 mm long. The tubular architecture of the ST3DT system supports continuous printing via a scanning laser, thereby eliminating the layer-by-layer lamination inherent to Cartesian-based printers. Moreover, being able to adjust the impregnation layer makes it possible to fine-tune stent wall thickness without the need for multiple radial layers. This approach shortens print times and reduces radial lamination, which, though less significant, still influences process efficiency.

Photoresins are highly promising materials for manufacturing medical devices because of their inherent versatility and ease of processing [19–21]. Their low density and tuneable formulation enable precise adjustment of mechanical, chemical, and biological properties to meet specific clinical requirements [22, 23]. When processed via VPP techniques, photoresins can produce complex, high-resolution geometries that are essential for next-generation implantable devices [24, 25]. These materials also exhibit excellent corrosion resistance and long-term stability [26], which are critical in the demanding physiological environment. Furthermore, photopolymerization is a cost-effective manufacturing approach that supports rapid prototyping and scalable production [27, 28], which are key aspects when developing new manufacturing technologies to produce stents [29]. The intrinsic adaptability of specific resin formulations enhances manoeuvrability within complex vascular structures, offering a significant benefit for minimally invasive procedures. However, improving elongation and toughness remains a challenge, especially during stent crimping on to the balloon. The low elongation and inherent brittleness of photoresins make it difficult to control radial recoil following crimping. As a result, an external sleeve is often required to maintain the stent in a low-profile configuration prior to deployment, making the delivery system thicker. A well-known limitation of polymers used in implants and medical devices is their lack of visibility under X-ray or computed tomography. However, recent advances and the high tuneability of photosensitive resins have enabled the development of radiopaque formulations without compromising processability — and in some cases, even enhancing compressive mechanical properties.

This study seeks to determine whether ST3DT implantable stents can be printed and remain functional throughout every stage of manufacturing — such as fabrication, rinsing and cleaning, post-curing, crimping, and sterilisation — while still providing sufficient mechanical performance. A photoresin was reinforced to enhance mechanical properties and provide of radiopacity. Mechanical evaluations, including radial force (crush test), axial compression, and three-point bending assessments, were conducted to determine the performance of the manufactured stents. In addition, the degree of curing was assessed with two different techniques. To this end, various stent geometries and 3D printing parameters were investigated to identify a suitable candidate for subsequent in vivo validation in a swine model.

Materials and method

Materials

The photosensitive PMMA-based photoresin was blended with a radiopaque component, specifically barium sulphate, at a ratio of 65/35 w/w%. The mixture was used as fresh material to fabricate the stents. The resin was selected due to its biomechanical properties evaluated in previous works [30] and based on the reports available on the literature [31–33].

Distilled H₂O MiliQ (Research Technical Services, Girona, Catalonia, Spain), Glycerol-based Cleaning Fluid (BC) (Sigma-Aldrich, Sant Louis, USA) and 96% Isopropyl alcohol (IPA) (Quimimont, Barcelona, Catalonia, Spain) were used to rinse and clean the manufactured stents.

Stent model and printing file preparation

A hybrid opened- and closed-cell design was designed in SolidWorks 2023 (Dassault Systèmes, France), defined by the following parameters: 39 mm length (L), 3 mm inner diameter (), 6 radial cells and 17 longitudinal cells (Fig. 1a-b). A full factorial design (3²) was used to evaluate the influence of two independent variables at three levels each: strut angle (: 25°, 45°, and 60°) and laser scan feed rate (F: 150, 225, and 300 mm/min). This experimental setup resulted in 9 unique combinations, with each condition tested in triplicate (n = 3), leading to a total of 27 mechanically tested stents. As photopolymerization is defined by the light exposure, that is inversely related to feedrate, strut width and stent thickness is dependent on the printing feedrate [34]. Prior to manufacturing, the printing file was prepared with a custom Python script following the trajectories shown in Fig. 1c.

Fig. 1.

a) Cell detail with principal geometrical features b) 3D render of a full stent design c) Manufacturing planar trajectory and stent design. A two-stage strategy was prepared: first, the rings were printed (purple) and then the bridges connecting them (turquoise), ensuring a proper union between rings d) Render of tubular vat photopolymerization scheme, ST3DT

Two commercially available metallic stents, Angiolite RX and Architect RX (iVascular, Sant Vicenç dels Horts, Barcelona, Spain), were used as reference devices for comparison with our novel polymeric stent. Angiolite RX and Architect RX are CoCr stents produced by laser cutting, each measuring 3 mm in diameter and 39 mm in length. A polymeric peripheral stent, Esprit BTK (Abbott, Illinois, USA) was also used as reference device for comparison. This stent measures 3 mm by 38 mm and is made by laser-cutting extruded PLLA tubes.

Stent manufacturing, rinsing and post-curing

Polymeric permanent stents were manufactured using the novel method ST3DT, introduced by Bosch et al. [34] in 2023. ST3DT uses a rotary-based VPP process that allows tubular devices and models to be made without support structures, which often result in inferior surface finishes. This approach allows the printing of high complex geometries avoiding the lamination of the part, both radially and longitudinally. Unlike ST3DT technology, current Cartesian SLA and DLP technologies slice the object longitudinally, compromise mechanical properties due to layer adhesion, and require support structures, meaning the medical device must be handled once printed to remove them. By rotating, the tubular shaft is impregnated with resin (see Fig. 1d). Adjusting the vat’s temperature changes the viscosity, allowing control over impregnation thickness to achieve the desired S_T. The machine was equipped with a 50 mW, 24 V and 405 nm direct laser (OSELA Inc., QC, Canada) with a spot size at focus of 35 μm and a 60 mm focal distance. Laser beam power was measured using PM16121 (ThorLabs, Newton, NJ, USA), a standard photodiode power meter ranging from 400 to 1100 nm. The laser spot size was estimated using a DMK 37BU273 monochrome industrial camera (The Imaging Source, Bremen, Germany) attached to a NMV HF-XA fixed focal length lens (Navitar, Rochester, NY, USA). From previous studies [30, 34], printing parameters were selected to achieve the desired dimensions, resulting in a laser power of 1.08 mW for the rings and 0.73 mW for the bridges connecting the rings. Vat was heated to 40 °C and monitored with a control system.

After printing, samples were demoulded and rinsed in a bath containing Fluid BioCleaner (3dresyns, Barcelona, Catalonia, Spain) for 10 min in a 35 W and 40 MHz Vevor ultrasonic cleaner (Indepsale Technology Limited, Ireland) and finally cleaned in a bath containing distilled H₂O MiliQ for another 5 min. The samples were then carefully dried manually with compressed air. Manual air drying allows for instant inspection of cleaning quality and stent integrity. A FormLabs UV Station was used for 60 min at 70 °C to post-cure the samples. The equipment allowed a batch of 9 stents to be cured using 13 multi-directional LEDs, with a total power of 39 W at 405 nm wavelength. After post-curing, stents were kept individually in a dark watertight container to protect them from ambient light.

Stent crimping

Utilising a Blockwise RSS Buddy manual crimper, an automatic crimping machine was developed by configuring Marlin firmware version 2.0.5 on an MKS Gen v1.2 controller. Sensorless homing was used to stop the crimp, depending on the size of the shaft or balloon. The hand crank was replaced with a NEMA 17 stepper motor, which serves as the actuator in the control system (see Fig. 2a). The process was performed at constant speed (1 mm/min) from an internal diameter of 3 mm to 1.5 mm (Fig. 2b), and the samples were left inside the crimp for 30 min. In Fig. 2c-e, the changes on the geometry for every strut angle can be seen.

Fig. 2.

a) RSS Buddy converted to an automatic crimper b) Dies of the crimper at 3 mm (expanded) and 1 mm (crimped), performed at 1 mm/min and planar CAD simulations of the expanded and crimped views of c) 25° d) 40° and e) 60°

Stent sterilisation

Subsequently, crimped stents were sterilised and two different methods were studied. Stents were split in two equal parts, one was sterilised in formaldehyde (Institute for Health Science Research Germans Trias i Pujol (IGTP), Badalona, Spain) and the other with gamma radiation (Aragó Gamma, Barcelona, Spain). Formaldehyde sterilisation was performed with a 10 mg/l gas concentration at 80 °C and a relative humidity of 75% for 2 h. For the gamma radiation, a total dose of 25 kGy was applied to polymeric stents (validated dose for sterility assurance level of 10^− 6) at room temperature.

Stent manufacturing and performance characterisation

Differential scanning calorimetry

The curing degree and the glass transition temperature (Tg) of the manufactured stents were determined through DSC in a Mettler-Toledo DSC3 + 700/970 (TA Instruments, Columbus, OH, USA) calorimeter with a RC-90 refrigerated module. The device was calibrated using indium standard (heat flow calibration) and an indium-lead–zinc standard (temperature calibration). Samples of 5 mg taken from along the stent were heated from − 20 to 220 °C twice at 20 °C/min under flow of N₂ at 50 ml/min. DSC consists of the observation of the enthalpy variation in different thermals transitions that a polymeric material can present. The calorimetric curve is determined, and the most relevant transitions are analysed. DSC analysis allowed for optimisation of printing and post-curing parameters by measuring heat flow to identify incomplete curing or residual components from printing or rinsing.

Fourier-transformed infrared analysis

The curing of the photoresin was evaluated by FTIR spectroscopy using a Bruker Alpha FT-IR spectrometer with an ATR accessory. Spectra were recorded in the range of 4000–600 cm⁻¹ with a resolution of 4 cm⁻¹ and 32 scans per sample. Resin samples were drop-cast on a glass slide and analysed before and after UV curing. The reduction in absorbance of the C = C stretching band at approximately 1635 cm⁻¹ served as an indicator for assessing both the efficiency and completeness of the curing process.

Morphological and dimensional analysis

The length and diameter of the manufactured stents were characterised with a Micromar 40 EWV digital micrometer (Mahr GmbH, Germany). For gross morphological evaluation and surface inspection, a Nikon SMZ 800T stereomicroscope (Nikon Corp., Japan) was employed under magnifications ranging from 1x to 8x. Images were taken to measure the strut width and the stent thickness. Five measurements were taken at random locations, measured orthogonally to the strut centreline, to calculate the mean value for S_W and S_T.

Radial strength test

A parallel planar plates compression test was performed to measure the radial strength of the stents produced with ST3DT (Fig. 3a), following the methodology other researchers proposed [18, 35]. Stents were compressed with a Ta.XT Plus Texture Analyser (Stable Micro Systems, Godalming, UK) equipped with a 50 N load cell giving a force resolution of 0.01 N. The lower plate was fixed and the upper was moving at 1 mm/min until the 50% of the nominal diameter was reached, following ISO 25539-2 (n = 3) [36].

Fig. 3.

a) Parallel plates double compression to 50% of deformation test sequence, b) Three-point bending up to 8 mm of deflection test sequence and c) Axial compression test sequence. For the bending and the axial test, a support was designed and manufactured to carry the tests

Axial strength test

Axial compression testing is conducted to evaluate the mechanical stability and longitudinal strength of stents under compressive loads that mimic physiological forces. The test assessed the axial stiffness, peak load, and deformation behaviour. To prevent undesired buckling during axial compression testing, a custom support fixture was designed to maintain the stent in a straight and aligned position throughout the procedure (Fig. 3c). The test was conducted at 1 mm/min until the 15% of the length was reached, the mean axial compression in the femoropopliteal zone under a demanding flexion [12].

Three-point bending test

A three-point bending test was performed to assess the flexibility of polymeric vascular stents using a Ta.XT Texture Analyser (Stable Micro Systems, Godalming, UK), following the ISO 25539-2 [36]. Each stent 39 mm in length, 3 mm inner diameter post-expansion) was positioned horizontally between two rounded supports set 16 mm apart. A 3 mm in diameter polished and rounded indenter, designed following the ISO, was applied at the midpoint to apply a compressive load vertically at a constant displacement rate of 2 mm/min (Fig. 3b). All tests were performed under ambient laboratory conditions, and force-displacement data were recorded to evaluate bending stiffness and deformation behaviour. Kinking was also evaluated at 8 mm with:

1

Where is the diameter of the stent when deflection is 8 mm and is the initial diameter.

Radiopacity

Stent radiopacity, or X-ray visibility, was assessed using a CT-Falcon from Novadep NDT Systems SL, Valladolid, Spain. Data was acquired under the following parameters: X-ray tube voltage of 43 kV, intensity of 175 µA, 37.5 μm pixel size, 1 × 1 binning and 2 s exposure time, resulting in a 2D image in the greyscale. Stents were embedded in a chicken thigh muscle to imitate in-vivo conditions. Stainless steel tubes were inserted to separate the stents during image post-processing. Angiolite RX served as the positive control, while a stent made from resin without any reinforcing component was used as the negative control sample.

Statistical analysis

Data were analysed using GraphPad Prism software (Version 10.2.1; GraphPad Software, La Jolla, CA, USA). Due to the limited sample size (n = 3 per group), the analyses proceeded under the assumptions of normal distribution and homogeneity of variance to facilitate parametric statistical methods. Comparisons between multiple factors were performed using two-way or three-way ANOVA, followed by Tukey’s post-hoc test for pairwise multiple comparisons. Significance levels were defined as p < 0.050 and were indicated as follows: p < 0.050 (*), p < 0.010 (**), and p < 0.001 (***). Where no asterisks are present in the charts, the differences were found to be not statistically significant (p > 0.050). Data are presented as mean ± standard deviation (SD).

Results and discussion

Differential scanning calorimetry

Many studies have listed the problems in the final device related to cleaning with IPA. The supplier recommends using BioCleaner (BC) solvent to maintain the biomechanical properties of the resins. In Fig. 4, thermal analysis using DSC showed a distinct exothermic peak between 100 °C and 220 °C in the NC sample during the first heating, which indicates the absence of a cure reaction. The DSC confirms that NC was the only sample that was not post-cured after manufacturing. In contrast, the other three samples (MiliQ-BC, BC, and IPA) had much lower residual enthalpy in that temperature range, suggesting post-curing was largely complete but still improvable.

Fig. 4.

DSC. Heat flow of different cleaning and post-curing methods. First heating (straight lines) and second heating (dashed lines) plot indicated a high conversion, and similar results compared to IPA cleaning, studied in previous works [25]

Despite the small enthalpy differences between these post-cured samples, the variation is minimal, suggesting that the cleaning solvent used before post-curing does not substantially affect the curing kinetics. In the second heating scan, all samples exhibited smoother calorimetric profiles with negligible enthalpy changes. This behaviour reflects the elimination of thermal history during the first heating, and the occurrence of further thermal curing induced by the initial temperature ramp, despite the absence of photoinitiation. Specifically, the DSC curves showed that cleaning with BC and distilled water can be just as effective as using IPA for cleaning.

FTIR analysis

FTIR analysis was performed alongside DSC to assess polymer conversion (Fig. 5). Other researchers in the field have suggested [37]:

2

Fig. 5.

a) FTIR analysis of the fresh resin and once cured. Also, fresh resin reinforced with radiopaque component and once cured. b) Detail of the zone of interest, plotting the C = C stretch and C = O stretch, used to approximate the degree of curing

Where refers to the area under the curve in the bond C=C and refers to the area under the curve in the bond C=O, taken as reference bond as it is not affected by the UV light.

The test was performed with the resin without additives and reinforced with the radiopaque component. The results differed slightly, and better results were achieved with the pure resin. In the first case, without additives, a degree of conversion of 69.4% has been approached using (2), and in the reinforced sample, 73.7% of polymerisation conversion. The results indicated that the radiopaque product mixed with the resin did not affect the curing of the medical device with the selected manufacturing and post-curing parameters. Besides, the results are good compared to the literature and very close to the requirements for medical devices. FTIR analysis allows for dependable measurement of conversion rates, but it only presents a stationary, surface perspective on polymerisation and can be influenced by overlapping peaks or changes in the spectral baseline.

Morphological and dimensional analysis

After printing, stents were evaluated visually during the rinsing and drying to detect anomalies or defects to the scaffold structure. Following post-curing, stents were examined using a stereoscope, and various measurements of S_W and S_T were recorded. As expected by the results from previous works [34, 38], differences between printing parameters were significant. The ST3DT method controls stent impregnation thickness by adjusting vat temperature and shaft rotation speed. Process variability was generally low, and increasing printing speed reduced variability and enhanced stability. Furthermore, reduced strut width measurements were obtained, aligning more closely with established commercial and clinical standards. As can be seen in Fig. 6a-i, union points between rings and bridges were clean, flawless and visually present a good bond. In the thinnest cases, manufactured at 300 mm/min (Fig. 6g-i), laser spot can be identified on the starting point, when the laser is switched on. Careful adjustment of exposure times can address this issue and achieve improved surface finishes in future projects. Minor fibres or particles were occasionally observed on the stent surface; these are attributed to the extensive handling required during mechanical characterisation and physical measurement. Such artifacts are not inherent to the manufacturing process and are strictly prevented under the standard cleanroom and sterilisation protocols used for clinical-grade devices.

Fig. 6.

Cell detail of the different geometries manufactured with different parameters a) 25° at 150 mm/min b) 40° at 150 mm/min c) 60° at 150 mm/min d) 25° at 225 mm/min e) 40° at 225 mm/min f) 60° at 225 mm/min g) 25° at 300 mm/min h) 40° at 300 mm/min i) 60° at 300 mm/min. Scale bar is 1000 μm

ST3DT protocol has been improved, compared to previous studies [34], showing its capabilities to produce polymeric stents with high dimensional accuracy and CAD fidelity. Scanning the desired geometry with a laser with the possibility to vary the focal distance allows ST3DT to manufacture stents with different dimensional parameters without needing to prepare several CAD files. Several different stents can be printed by scripting the g-code file to modify the power, feedrate and height of the laser (Fig. 6a-i). The dimensional analysis of the manufactured stents reveals a clear trend influenced by both strut angle (25°, 40°, 60°) and feedrate (150, 225, 300 mm/min). The thinnest struts were achieved with a feedrate of 300 mm/min (Fig. 7a), lowering the resin exposure to UV light and achieving values between 185 and 210 μm. As the feedrate decreased to 225 and 300 mm/min, S_w and S_T progressively increased, reaching values between 250 and 275 μm and 330–360 μm, respectively. S_T yields more stable results under the studied parameters. The thicker strut was achieved with the lowest feedrate, over 150 μm and decreasing down to 130 μm (Fig. 7b). Centrifugal forces caused by rotating velocity when impregnating the shaft can be the source of differences on stent thickness at same temperature of the vat. To obtain the desired stent thickness, the process should be adjusted by managing both the resin’s feed rate and viscosity, which are regulated through the vat temperature during impregnation.

Fig. 7.

Bar plots depicting a) strut width and b) stent thickness for different stents

Geometry does not significantly affect the S_w and S_T. For S_w, manufacturing different geometries shows minimal variation; smaller strut angles increase laser path length and exposure time. The differences are small but quantifiable in the case of feedrate 150 mm/min. Variability of S_w and S_T may be caused by human-error while taking measurements. All results can be seen on Table 1.

Table 1.

Strut width (S_w) and stent thickness (S_T) were measured for each manufactured stent. Results given as mean [µm] of five measurements at random locations and the standard deviation [%] (SD)

	150 mm/min				225 mm/min				300 mm/min
	SW		ST		SW		ST		SW		ST
Angle	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD	Mean	SD
25°	352.60	15.76	167.40	10.05	265.73	16.19	161.00	3.29	201.20	2.11	139.60	11.82
40°	329.80	5.83	161.40	7.88	259.67	9.49	151.67	9.08	204.80	15.23	135.00	6.50
60°	338.73	5.30	176.60	11.41	243.87	5.50	149.27	5.18	191.80	6.61	130.73	0.42

Future work should focus on optimising manufacturing to produce stents with thinner struts for improved arterial adaptation and fewer post implantation complications.

Mechanical performance

Mechanical performance is a critical aspect of vascular stents, especially in the BTK zone, they must be strong enough to support the arterial walls and maintain lumen patency, while also being flexible enough to withstand the demanding bending motions of the limbs. This section presents the mechanical test results and highlights the distinctions observed among the samples.

In the first instance, a double compression between parallel plates has been performed to approximate the radial strength of the different geometries, both in strut angle and strut width. To be comparable with other commercial stents, radial force measured was normalised by unit of length. As expected and consistent with other research, wider struts lead to a higher radial stiffness (Fig. 8a). Regarding strut angle, no significant differences were found when manufacturing at higher feedrate. 60°-angled struts presented an improvement when manufacturing at 150 mm/min. Additionally, after experiencing plastic deformation, each stent returned to its original 3 mm diameter once the force was removed (Fig. 8b). Both Angiolite and Architect metallic stents exhibited plastic deformation during the first compression cycle, preventing a second test from being conducted. In the BTK region, this limited radial recovery may be problematic, as the inability of the stent to return to its original shape can lead to device migration or malapposition and can cause complications. Esprit BTK stent showed improved elastic behaviour compared to metallic stents but still cannot fully recover its original diameter after compression. Although polymeric stents aim to match the radial strength of metallic benchmarks, the lower intrinsic mechanical properties of polymers necessitate significantly thicker struts to achieve comparable radial support. However, these increased dimensions can lead to clinical complications, such as reduced flexibility, increased thrombogenicity, and higher rates of in-stent restenosis. Specifically, with the stent with struts at 60° and manufactured at 150 mm/min, an average radial force of 0.28 N/mm was obtained with a strut width of more than 320 μm. The Angiolite stent, on the other hand, withstands the same load per length with struts of less than 100 μm. Stents with strut widths closer to clinical values have withstood values above 0.15 N/mm, close to the minimum required for the most common pathologies in the BTK zone.

Fig. 8.

Crush tests. a) Normalised radial force at 50% of deformation in a parallel plate compression, b) Radial force over time of the most promising stents (40° struts), c) Bar plot with the differences once sterilised with gamma radiation and d) Bar plot depicting the differences once sterilised with formaldehyde. Note: Commercial stents are shown as a benchmark for clinical context, no formal statistical comparisons were performed between ST3DT-manufactured stents and commercial references due to fundamental material and design differences

Once the stents were sterilised, another compression test was carried out to validate the feasibility to sterilise the medical device without affecting its mechanical properties. Both sterilisation methods, widely used in the medical sector, were proven to be a suitable option for polymeric stents in terms of mechanical properties. Gamma radiation enhanced the radial strength of the stents by 10% to 40%, with a greater effect observed in stents featuring thicker and wider struts (Fig. 8c), which correspond to increased mass. In the case of the 25° stent manufactured at the lowest feedrate, there was an increase of 37.7% ± 9.5 compared to the same stent prior to sterilisation. At the other end of the spectrum, with an angle of 60° and the highest feedrate — resulting in the thinnest struts — the effect was approximately 7.21% ± 4.83.

While the findings do not provide conclusive evidence of increased value, they indicate a potential enhancement in supported force because of the extra crosslinking caused by the treatment. This uncertainty is probably caused by batch sterilisation and over radiation in some zones. The results are coincident to recent research of other groups [39]. Further investigation to study the effect of the radiation on the stent’s properties should be performed to achieve more stable and better prothesis.

Formaldehyde also affected the final properties by increasing the radial force in most of the cases, but with less effect (2% to 15%) nor repeatability (Fig. 8d). Since the sterilisation treatment was performed at 80 °C, it is likely that additional thermal curing occurred, contributing to further consolidation of the polymeric network. This hypothesis is supported by FTIR results, which also indicate that post-curing process can be fine-tunned and enhanced.

Thirdly, flexibility of the different stent geometries was assessed by measuring the force need to bend it up to 8 mm deflection. Also, kinking during bending was proven to be below the 5% of the diameter in all cases with (1), critical aspect for stents at BTK area. All stents had a lower bending strength than the metallic ones (Fig. 9a), probably being less harmful for BTK arteries during limb motion [40]. Moreover, none of them broke during the test, so all 3 geometries can be subjected to at least 8 mm deflection. Although all stents were tested, only the most relevant ones according to the previous results are included in the graph to make the graph more understandable and consultable. Esprit BTK is also a flexible stent (0.058 N) providing vessel scaffolding without a kinking superior to 5%. For both metallic stents, the test produced significant noise and resulted in inconsistent bending behaviour. On the other hand, the metallic stents could not recover the straight shape once the force was withdrawn and presented near 10% of kinking, exhibiting plastic deformation. Stents with lower profile and greater strut angle exhibit higher capacity to bend, which is a desirable feature for tortuous BTK vessels. Moreover, the stent design permits demanding bending without critical kinking (Fig. 9c). With thicker struts, the differences between geometries were more significant, with the stent with the smallest strut angle (25°) showing the highest bending strength. With the potential candidates, the stent manufactured at 225 mm/min and with struts width and thickness around 150 μm, it is observed that the case of 25° in the geometry presents a higher resistance, 0.086 N ± 0.021, compared to the 40° which is 0.080 ± 0.004. For stents manufactured at a higher speed, thus achieving thinner struts, the bending resistance was 0.060 N ± 0.006 and 0.057 ± 0.012, respectively. These values were clearly lower than those obtained with the metallic Architect stent (0.194 N) and Angiolite (0.112 N).

Fig. 9.

a) Three-point bending plot for four stents compared to metallic ones b) Axial force of all stents c) Stent exhibiting flexible capabilities, manufactured at 225 mm/min with a 25°-angled strut

Axial compression was assessed to verify that stents tolerate a 10% length deformation, as recommended for BTK arteries. None of the stents tested collapsed or experienced plastic deformation. For the axial test, S_w and S_T also played an important role in the force required to deform the stents. It was found that no stent suffered any wear or break at any strut. Given the demanding mechanical environment of below-the-knee arteries, axial compression tests highlight that stents manufactured at 150 mm/min exhibit superior strength and axial robustness (Fig. 9b). Besides, strut angle also influences as lower struts are more aligned in the axial direction, thus supporting directly the longitudinal force, than greater struts angles. Stents below the 0.25 N threshold may risk structural failure in vivo, which is the case of stents manufactured at 300 mm/min.

Results indicate that stents manufactured at 225 mm/min exhibit the best balance between radial force, axial compression strength and flexibility and are potential candidates for BTK applications.

The three mechanical tests were used to conduct an initial evaluation of the three geometries, enabling a selection for further development. A clear limitation is the fact that the tests were done separately, as in the BTK area the mechanical solutions are usually a combination of two, or even all three. It is necessary to be able to test the stents in bioreactors that simulate a leg, or in an in vivo environment.

Stent crimp

This worked attempted to crimp stents on a balloon catheter as the final, and critical, step in the stent manufacturing process and essential feature to allow navigation and deployment inside the blood vessels. Following the mechanical characterisation, crimping behaviour was studied. Metallic stents are deformed plastically while crimping with an automatic set of dies that apply radial force uniformly. In the same way, polymeric stents were crimped. Thinner struts exhibited superior crimping capabilities, reaching smaller diameters without sings of cracking. No stent manufactured at 150 mm/min with struts over 300 μm could crimp to target diameter of 1.5 mm (Fig. 10b). Therefore, despite being potential candidates due to their radial force, they were discarded due to their excessive strut size and poor crimping performance. For other printing feedrate, 60°-angled struts also broke during crimping. In this instance, the inhomogeneous crimp resulted from the stent’s geometry, excessive angulation, and a high friction coefficient between the stent and the dies. Stents with struts angles of 25° and 40° and printed at 225 mm/min and 300 mm/min were successfully crimped on a balloon reaching a crimp diameter of 1.5 mm Fig. 10b-c. The manufactured polymeric stents have a significant radial recovery after crimping, so it is necessary to have a sheath to keep them crimped to the balloon. A transparent sheath was used, and images were taken (Fig. 10a, Table 2).

Fig. 10.

a) Stent from 3 mm to 1.5 mm and covered with transparent sheath b) Lollipop plot for every geometry of stent, crimping diameter achieved (dot) and break point (x), in green the stents successfully crimped and in red stents fractured before reaching 1.5 mm c) Stent crimped in a balloon

Table 2.

The table details the manufacturing diameter and target crimping diameter versus the actual diameter at which structural failure occurred for each design configuration. Success was defined as reaching the target diameter of 1.5 mm or smaller without structural failure nor break point at any strut or ring

Stent	Manufacturing diameter [mm]	Target diameter [mm]	Fracture diameter [mm]	Success
25_F150	3,00	1,5	1,75	33%
40_F150	3,00	1,5	1,75	0%
60_F150	3,00	1,5	2,00	0%
25_F225	3,00	1,5	1,25	100%
40_F225	3,00	1,5	1,50	66,7%
60_F225	3,00	1,5	1,50	0%
25_F300	3,00	1,5	1,00	100%
40_F300	3,00	1,5	1,00	100%
60_F300	3,00	1,5	1,50	0%

The results suggested that stents with 25° and 40° strut angles are promising candidates for the intended application. These configurations provided a balanced compromise between mechanical performance and manufacturability, exhibiting sufficient radial strength while maintaining good crimping behaviour. In contrast, stents with 60° angled struts — although potentially offering increased radial stiffness due to their more vertical alignment — consistently showed poor crimping capabilities. Several of them fractured before reaching the target crimping diameter of 1.5 mm, indicating excessive brittleness or geometric instability during the crimping process. Due to these limitations, the 60° configuration was considered less viable for clinical deployment and thus discarded from further consideration.

Radiopacity

XR visibility in a simulated in-vivo conditions have been evaluated using micro-CT equipment. All ST3DT-manufactured stents showed lower visibility than commercial CoCr stent. Nevertheless, high concentrations of radiopaque agent samples can be localised within a chicken breast with optimal definition (Fig. 11). As the concentration decreases, so does visibility, since fewer light rays are captured by the radiopaque agent particles. Compared to non-reinforced stents, reinforced stents exhibit a detectable contrast to chicken thigh, confirming the contribution of the filler. It is not possible to quantitatively determine visibility under X-ray using the scout image from a CT scan. Tests in a more relevant environment should be performed to assess the radiopacity of the stent with the minimum filling concentration.

Fig. 11.

2D radiographic projection of the stents within the chicken breast, showing different concentrations of the radiopaque component

Study limitations

While this study offers useful insight into the stent performance, several limitations should be acknowledged. First, the crimping procedure was carried out at room temperature and without industrial grade equipment or precise force control. Consequently, non-uniform structural stresses may have been introduced, which would not necessarily occur under standardized manufacturing conditions. Mechanical characterization was performed using a parallel plate radial compression test. Although this approach is widely reported in the literature, it does not fully reproduce the complex and multidirectional forces that a stent experiences once deployed in an arterial environment. In addition, the effect of sterilization methods, particularly ethylene oxide, on the polymer matrix has not yet been comprehensively evaluated. Further work is required to determine whether these processes could affect the material long term stability. Finally, the X ray imaging assessment was preliminary and relied on the group’s internal equipment. Additional studies in more clinically representative settings, including the presence of surrounding biological tissues, will be necessary to confirm the device radiopacity under realistic conditions.

Conclusions

The ST3DT approach demonstrated high dimensional accuracy and proved to be a viable manufacturing workflow, encompassing fabrication, cleaning, post-curing, and crimping stages, confirming the method’s capability to produce functional devices with reproducible results. Photosensitive polymers offer significant potential for developing innovative stents with improved properties and functions. DSC and FTIR indicated a good degree of curing but with possibility to improve it by extending post-curing time. The stents with the best mechanical properties have been found to be those designed with a smaller angle due to a better distribution of forces. In addition, the steeper struts proved to be very difficult to crimp without breaking due to their low elongation and low capacity for plastic deformation.

The stents showed similar mechanical properties to current commercial stents, especially their polymeric competitor. Radial strength with thinner struts is an aspect to be improved with new formulations and slight design changes, which will be done in future work. The flexibility of the manufactured stents has proven to be compatible with leg movements with a kinking in diameter of less than 5%. Given that some geometries failed during the crimping process, the most viable designs were shortlisted for improvement and subsequent evaluation in application-relevant environments.

Furthermore, exposing the stents to standard gamma and formaldehyde treatments showed that the devices maintain structural integrity post-processing; mechanical characterisation revealed a slight increase in radial strength, likely due to radiation-induced post-curing. Studies on the effects on the microstructure should be carried out.

The manufactured stents are visible under XR, although samples with 35% of radiopaque component can be positioned and localised easily in a chicken thigh, radiopacity should be improved to track stents with efficacy in a relevant environment while performing the intervention. Reinforcing photoresins to enhance the final medical device is a key aspect in the biomedical sector and should be studied further to exploit its full potential.

The two-stage printing strategy — sequentially manufacturing the rings and then the bridges — enables the potential for multi-material fabrication. Although not implemented in the present study, this approach allows for an intermediate resin exchange between stages. To ensure structural integrity and consistent curing kinetics, a cleaning step would be required during the material transition to prevent cross-contamination from non-polymerised resin.

Abbreviations

AM

Additive Manufacturing

ANOVA

Analysis of variance

3DP

3D Printing

BC

BioCleaner

BMS

Bare metal stent

BRS

Bioresorbable stent

BTK

Below-the-knee

CAD

Computer-aided design

CLI

Chronic limb ischemia

CT

Computerised Tomography

CVD

Cardiovascular diseases

DES

Drug-eluting stent

DLP

Digital light processing

DSC

Differential scanning calorimetry

FPA

Femoropopliteal artery

FTIR

Fourier-transformed infrared

IPA

Isopropyl Alcohol

ISR

In-stent restenosis

Inner diameter

MiliQ

Distilled Water MiliQ

L

Length

PAD

Peripheral artery disease

ROC

Radiopaque component

S_T

Stent thickness

ST3DT

Stereolithography 3D Tubular

S_W

Strut width

UV

Ultra-violet

VPP

Vat photopolymerization

WHO

World Health Organization

XR

X-Ray

µCLIP

Micro continuous liquid interface production

Author contributions

A.B. led the conceptualization of the study, developed the methodology, conducted the experimental work, and drafted the original manuscript and created the visualizations. P.C. contributed to data acquisition, image analysis, and assisted in reviewing and editing the manuscript. J.C. provided supervision and project coordination and revised the manuscript critically. A.J.G. was responsible for funding acquisition, supported the interpretation of results, overall supervision of the research activities, and contributed to manuscript revision and validation of findings.

Funding

This work was financially supported by the Catalan Government through the funding grant ACCIÓ-Eurecat (Project TRAÇA2025-DESTOFU). A. Bosch is a fellow of Eurecat’s “Vicente López” PhD grant program.

Data availability

No datasets were generated or analysed during the current study.

Declarations

Competing interests

Antonio J. Guerra has patent #PCT/EP2021/080095 pending to Eurecat. Other authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.