Poster M20 - Accurate In Silico 3-Point Bend Test to Predict 3D Printed Polymeric Biliary Stent Behavior When Deployed

Introduction: The improved resolution of 3D printers makes it feasible to fabricate medical devices tailored to the features and dimensions of each patient’s unique anatomy. Our lab is focused on improving in silico and in vitro methods to vet customizable, 3D-printed biliary stent designs and material options. We have developed an FEA 3-point stent bend test simulation to predict biliary stent behavior under conditions comparable to deployment in a patient. Simulation accuracy is evaluated by comparison to experimental 3-point stent bend testing data.

In recent studies, after determining that Formlabs Durable (Dur) was not a suitable polymer for biliary stents1, we evaluated copolymers of Dur and Polyethylene Glycol 2-Dimethacrylate (PEG) as more elastic options for stents. Following analysis of the mechanical properties of the copolymer Dur + 15% v/v PEG (DurPEG15) after exposure to physiological conditions, we narrowed our FEA material model focus to three, non-linear hyperelastic material models: Neo-Hookean, Mooney-Rivlin 3 Parameter, and Yeoh 3 Parameter.

In addition, to further characterize DurPEG15, we have tested the ability of a DLP to produce DurPEG15 samples with equivalent mechanical properties to those printed on a Formlabs Form 3B. Our goal is to efficiently generate drug elution samples; the resin tank and resin volume for the Form 3B are not conducive to generation of < 20 samples per concentration.

Herein, we compare the accuracy of the FEA material models in predicting stent behavior, we share our findings to date on DLP-printed DurPEG15 samples versus SLA-printed DurPEG15 samples, and we discuss our next steps.

Materials and

Methods: Experimental and in silico Stent 3-Point Bend Testing

Patmas Weave Stents (US patent D987826) were printed and post-processed as previously described. Stents (N ≥ 8 per condition) were submerged in water, phosphate buffered saline, or bile acid solution at 37 ℃ for 1 week. We then performed 3-point bend tests to failure on all stents according to ASTM D790. The bend test fixture was modified with a custom, 3D-printed stent holder that fixed the ends of the stents in place during the test. Force was applied at 10 mm/min with a 6 mm rod in the center of each stent.

The stent geometry was imported into Ansys as a static structure. Material properties for the Neo-Hookean (NH), Mooney-Rivlin 3 Parameter (MR3), and Yeoh (YP3) models were determined using absolute error, uniaxial tensile data, and volumetric test data. Stent geometry was meshed, and a 3-point bend test was simulated by applying a unidirectional remote force on the stent center with the end faces fixed. Ansys solved for stress, strain, and total deformation at each respective force (0.5 N - 2.5 N in increments of 0.5 N).

Printing with DurPEG15 on a DLP

Samples were printed on an AnyCubic Photon D2 DLP with DurPEG15 six at a time or individually (N ≥ 5). Print settings were standard aside from a 15 sec layer exposure time; resin volume started at 160 mL for each print. Samples were post-processed and tensile tested as previously described. Statistical significance was determined via ANOVA.

Results, Conclusions, and Discussions: Accuracy of In Silico Stent 3-point Bend Testing

The extrahepatic bile ducts have thin, highly elastic walls. In turn, biliary stents must be flexible and bent no more than 30o upon deployment. For a 60 mm stent that angle results in 7.76 mm deformation. Experimental deformation of DurPEG15 stents was just over 8 mm at 1.5 N force (Figure 1). Therefore, we focused on the accuracy of our FEA from 0.5 N to 1.5 N to reliably predict the peak stress the stent will experience when deployed.

The three non-linear, hyperelastic material models that we evaluated were all accurate in the 1 N to 1.5 N range with percent errors of < 14% (at 1.395 N; NH = 10.1%, MR3 = 10.1%, YP3 = 6.8%). In turn, we can evaluate the suitability of DurPEG15 for use in biliary stents via the simulation-generated peak stress values. Target peak stress values should be well below the peak strength of the stent fabrication material. For DurPEG15, the predicted peak stress at 1.5 N was < 6 MPa; as this is < 60% of the peak strength of DurPEG15 (10.6 MPa) indicating that the stent will not fracture when deployed.

Mechanical properties of DurPEG15 printed with SLA and DLP 3D printers

It is costly and inefficient to test drug loading of 3D printed polymers in 1 L batches. Thus, we tested the ability of an inexpensive DLP printer, AnyCubic Photon D2 ( < $500) with a small resin vat volume (140 mL to 200 mL) to generate DurPEG15 samples with comparable mechanical characteristics to those printed with a Formlabs Form3B SLA printer. We tested group-printed and individually-printed samples to eliminate light scattering as a variable. Samples printed on the DLP were significantly stiffer than those printed on the SLA but had very similar peak strength values (Table 1).

Future work

Next, we plan to expand our in silico experiments and develop an FEA simulation for radial compression of self-expanding polymeric stents of various architectures. In addition, we will evaluate post-processing modifications for DLP DurPEG15 samples to increase their elasticity while maintaining peak strength.

Acknowledgements (Optional): Mercer Provost office and School of Engineering for funding. Dr. Chris Basciano for his advice.