LLNL Researchers Bioprint Living Aneurysm and Watch it Heal Post-Op

Cerebral aneurysms, caused by the artery walls in the brain weakening, affect roughly one in every 50 people in the US, and are distinguished by a bulging blood vessel, which can cause brain damage, stroke, or even death if ruptured. A team of researchers from Lawrence Livermore National Laboratory (LLNL), Duke University, and Texas A&M has been working hard to improve current surgical procedures and make them more patient-specific. These scientists used 3D bioprinting to create the first living aneurysm outside of the human body, and then performed a medical procedure, watching it respond to treatment and heal just like a real brain.

The team published a paper on their work in the journal Biofabrication.

“Herein, we developed three-dimensionally (3D) printed aneurysm-bearing vascularized tissue structures using gelatin-fibrin hydrogel of which the inner vessel walls were seeded with human cerebral microvascular endothelial cells (hCMECs). The hCMECs readily exhibited cellular attachment, spreading, and confluency all around the vessel walls, including the aneurysm walls. Additionally, the in vitro platform was directly amenable to flow measurements via particle image velocimetry, enabling the direct assessment of the vascular flow dynamics for comparison to a 3D computational fluid dynamics model,” they wrote in the abstract.

The in vitro living cerebral aneurysm. (a) Illustration of the 3D printed aneurysm bioreactor. (b) The in vitro aneurysm vessel structure perfused with red fluorescent beads, demonstrating the formation of patent vessels post-evacuation of sacrificial ink.

Obviously, brain aneurysms are not easy to fix. One very invasive treatment consists of attaching a metal clip to the base in order to redirect blood flow away from it, and requires surgeons to open the skull and expose the brain. Another common, but less invasive treatment, is called endovascular metal coiling, and involves a surgeon inserting a thin metal catheter into an artery in the patient’s groin, feeding it all the way up through the body and into the aneurysm (sounds invasive to me!). Then, they pack it with stents or coils, which causes a clot, and the endothelium that line the blood vessel in question will grow over the clotted plug, basically building a wall around the aneurysm so it’s separated from the rest of the vasculature. Unfortunately, the outcome of both of these treatment methods often varies from patient to patient.

“While there are a lot of promising treatment options, some still have a long way to go. Animal models aren’t necessarily the best way to try out these options, as they lack direct observation of treatment effects and have uncontrollable aneurysm geometries,” said LLNL engineer and principal investigator Monica Moya. “Having this robust, human in vitro testing platform could help facilitate new treatments. If we can replicate aneurysms as much as we need to with these devices, we might help accelerate some of these products into the clinic and essentially provide patients with better treatment options.”

3D printing has been used before to help surgeons train for these complex procedures using models, and even monitor cerebral aneurysms in real time. But the LLNL-led research team was able to replicate a brain aneurysm in vitro by using human cerebral cells to create bioprinted blood vessels. Another LLNL engineer, William “Rick” Hynes, who was the original principal investigator, had the idea that bioprinting with human cells could help medical researchers create and validate more predictive, patient-specific, biologically relevant 3D models.

Using 3D printing, an LLNL team replicated an aneurysm in vitro and performed an endovascular repair procedure on it, inserting a catheter into the blood vessel and tightly packing platinum coils inside the aneurysm sac. They introduced blood plasma into the aneurysm and observed a blood clot form where the coils were located. The green areas depict the endothelial cells and the red indicates the formed clot. Photo by Elisa Wasson, LLNL

Hynes said, “We looked at the problem and thought that if we could pair computational modeling and experimental approaches, maybe we could come up with a more deterministic method of treating aneurysms or selecting treatments that could best serve the patient. Now we can start to build the framework of a personalized model that a surgical practitioner could use to determine the best method for treating an aneurysm.”

Hynes and Moya teamed up with former LLNL scientist Duncan Maitland, who leads a biomedical engineering group at Texas A&M and also heads Shape Memory Medical, which is developing an experimental shape memory coil for aneurysm treatment; and Amanda Randles, another former LLNL scientist and current Duke assistant professor who developed a code for simulating blood flow, for this work. Additional authors of the paper are Lindy K. Jang (Texas A&M), Javier A. Alvarado (LLNL), Marianna Pepona (Duke), Elisa M. Wasson (LLNL), Landon D. Nash (Shape Memory Medical), and Jason M. Ortega (LLNL).

Particle image velocimetry (PIV) analysis and 3D computational flow model simulations. (a) (Above) PIV measurement at the back of the aneurysm dome showing no detectable flow at a 300 µl min−1 flow rate. (Middle) A 3D flow simulation of the same geometry and flow rate at z = −0.66 mm. (Below) PIV measurements within the parent and daughter vessels captured with a 2× objective at the same flow rate. (b) (Above) PIV measurement gathered at the back of the aneurysm dome showing circular flow patterns and captured with a 4× objective at a 20 ml min−1 flow rate. (Below) Simulation of the same geometry and flow rate demonstrating that fluid motion only occurs within the dome at high flow rates. (c) High-fidelity geometric reconstruction of the printed living aneurysm constructed from image stacks gathered via confocal microscopy.

The team designed bioreactor sidewalls for the aneurysm platform in SOLIDWORKS, and used open source Slic3r software to convert the design to G-code. Using a custom extrusion bioprinter, the walls were then printed out of SE-1700 silicone onto glass slides, which were then cured and sterilized in an autoclave. The blood vessel geometry was printed using a sacrificial ink, surrounded by a protein-based hydrogel; this was cooled to dissolve the ink, which left behind the shape of the vasculature. Human brain endothelial cells coated the channels to form the aneurysm and blood vessels.

Hynes performed the repair on the bioprinted aneurysm by inserting a microcatheter and tightly packed platinum coils inside the sac. Then, the researchers introduced blood plasma, and watched a blood clot form at the coils’ location on the aneurysm, cutting it off from the fluid flow. Eight days after what LLNL believes is “the first surgical intervention ever performed on an artificial living tissue,” the team saw the post-op healing process of the endothelium within the vessels with their own eyes.

Endothelialization of aneurysm-bearing printed vessels. (a) Confocal image of actin stained endothelial cells after 7 d of perfusion culture. (b) Close-up image of actin stained endothelium (green) within the aneurysm dome, demonstrating fully confluent monolayer growth.

The researchers also used the device to show the validity of Randles’ flow dynamics model, noticing little blood movement into the aneurysm at low flow rates and a faster, circular flow when the rate was increased, like what would happen if a human patient was agitated.

When paired with computer modeling, LLNL says that this platform is a big step in creating patient-specific brain aneurysm care based on factors like blood pressure and blood vessel geometry, which could help speed up the amount of time it takes for complex surgical techniques to make their way to training clinics. Surgeons can use it as a tool to choose the best aneurysm packing coils ahead of surgery.

“Essentially a clinician could literally look at somebody’s brain scan, run it through the modeling software, and the software could show the fluid dynamics prior to treatment. It also should be able to simulate that treatment and allow the practitioner to narrow down to a certain type of coil or packing volume to ensure the best possible outcome,” Hynes explained.

Deployment of endovascular bare platinum coil (BPC) intervention treatment within the in vitro aneurysm. (a) Image of dual coil deployment with the aneurysm dome. (b) Micrographs of brightfield monitoring during BPC insertions from the endovascular microcatheter (1st BPC: 3 mm × 6 cm, 2nd BPC: 2 mm × 3 cm). (c) Maximum projection confocal image stacks of an artificial aneurysm filled with 1 µm red fluorescent beads before and (d) after BPC (2 mm × 3 cm) deployment and retraction, demonstrating no damage to the sac during insertion.

The platform can also be used to gain a better understanding of basic biology and post-op healing, as well as perform test runs ahead of time, without having to induce animals with aneurysms and then perform surgery. It can directly measure fluid dynamics inside the aneurysm and blood vessels, which can’t be done with animals.

“This is an ideal platform for an in silico model because we can make these flow measurements that would be incredibly difficult to make if you were doing this in an animal. What’s exciting is that this platform mimics blood vessel compliance and the mechanical stiffness of brain tissue. It’s also robust enough to handle a coiling procedure. You’re seeing the vessel distend and move, but it’s able to withstand the procedure — very much like you would in vivo,” Moya said. “This makes it ideal to be used as a training platform for surgeons or as an in vitro testing system for embolization devices.”

Plasma clot formation in response to BPC deployment within in vitro living aneurysm dome. Maximum projection confocal image stack of the complete in vitro aneurysm after BPC deployment and injection with bovine plasma. Clot formation is visualized via accumulation of trace fluorescently labeled red human fibrinogen included with the plasma mixture. Endothelial cells are fluorescently stained green for actin. Imaging reveals clot formation and occlusion of the aneurysm sac, with no major clot formation present elsewhere in the vessel structures.

The LLNL team says that this platform is showing promise at this early stage. Their next step is to combine a 2D blood-clotting model that LLNL computational engineer Ortega created with Randles’ 3D fluid dynamics model, in order to simulate how aneurysm-causing blood clots form in response to coils in 3D.

(Source: LLNL)