3D bioprinting takes major step forward with realistic blood vessel structures

There are more than 100,000 people on organ transplant lists in the U.S., some of whom will wait years to receive one – and some may not survive the wait. Even with a good match, there is a chance that a person's body will reject the organ. To shorten waiting periods and reduce the possibility of rejection, researchers in regenerative medicine are developing methods to use a patient's own cells to fabricate personalized hearts, kidneys, livers, and other organs on demand.

Ensuring that oxygen and nutrients can reach every part of a newly grown organ is an ongoing challenge. Researchers at Stanford have created new tools to design and 3D print the incredibly complex vascular trees needed to carry blood throughout an organ. Their platform, published June 12 in Science, generates designs that resemble what we actually see in the human body significantly faster than previous attempts and is able to translate those designs into instructions for a 3D printer.

The ability to scale up bioprinted tissues is currently limited by the ability to generate vasculature for them – you can't scale up these tissues without providing a blood supply. We were able to make the algorithm for generating the vasculature run about 200 times faster than prior methods, and we can generate it for complex shapes, like organs."

Alison Marsden, the Douglas M. and Nola Leishman Professor of Cardiovascular Diseases, professor of pediatrics and of bioengineering at Stanford in the Schools of Engineering and Medicine and co-senior author on the paper

Organ-scale vasculature

When blood is pumped to an organ in the body, it moves from a large artery into smaller and smaller branching blood vessels, where it can exchange gases and nutrients with the surrounding tissues. In most tissues, cells need to be within a hair's width of a blood vessel to survive, but in metabolically demanding tissues such as the heart, the distance is even smaller – there may be more than 2,500 capillaries in a millimeter-sized cube. All of these tiny blood vessels eventually join back together before leaving the organ.

These vascular networks aren't standardized; organs come in many shapes, and there is a lot of variety even between two similarly sized hearts. Up to this point, generating a model of a realistic vascular network that fits a unique and complex organ has been difficult and incredibly time-consuming. Many researchers have instead relied on standardized lattices, which work well in small engineered tissue models but don't scale up well.

Marsden and her colleagues built an algorithm to create vascular trees that closely mimic native organ blood vessel architectures, and have made the software available for anyone to use via their SimVascular open-source project. They incorporated fluid dynamics simulations to ensure that the vasculature would evenly distribute blood and successfully shorten the time needed to generate the network while still avoiding collisions between blood vessels and creating a closed loop with a single entrance and exit.

"It took about five hours to generate a computer model of a tree to vascularize a human heart. We were able to get to a density where any cell in the model would have been about 100 to 150 microns away from the nearest blood vessel, which is pretty good," said Zachary Sexton, a postdoctoral scholar in Marsden's lab and co-first author on the paper. The design contained one million blood vessels. "That task hadn't been done before, and probably would have taken months with previous algorithms."

While 3D printers aren't yet up to the task of printing such a fine-scale and dense network, the researchers were able to design and print a vascular model with 500 branches. They also tested a simpler version to ensure that it could keep cells alive. Using a 3D bioprinter – which prints with living cells instead of resin or metal – the researchers created a thick ring loaded with human embryonic kidney cells and built a network of 25 vessels running through it. They pumped a liquid loaded with oxygen and nutrients through the network and successfully kept a high number of cells in close proximity to the vascular network alive.

"We show these vessels can be designed, printed, and can keep cells alive," said Mark Skylar-Scott, an assistant professor of bioengineering and co-senior author on the paper. "We know that there's work to do to speed up the printing, but we now have this pipeline to generate different vascular trees very efficiently and create a set of instructions to print them."

A bioprinted heart

The researchers are quick to note that these vascular networks are not yet functional blood vessels – they're channels printed through a 3D matrix, but they don't have muscle cells, endothelial cells, fibroblasts, or anything else that they would need to work on their own.

"This is the first step toward generating really complex vascular networks," said Dominic Rütsche, a postdoctoral scholar in Skylar-Scott's lab and co-first author on the paper. "We can print them at never-before-seen complexities, but they are not yet fully physiological vessels. We're working on that."

Turning these designs into functioning blood vessels is just one of the many aspects of bioprinting a functioning human heart that Skylar-Scott and his colleagues are working on. They're also exploring how to encourage the tiniest blood vessels – those that are too small or too closely spaced to print – to grow on their own, improving the capabilities of 3D bioprinters to make them faster and more precise, and growing the massive amounts of cells that they will need to print a whole heart.

"This is a critical step in the process," Skylar-Scott said. "We have successfully generated enough heart cells from human stem cells to print the whole human heart, and now we can design a good, complex vascular tree to keep them fed and living. We are now actively putting the two together: cells and vasculature, at organ scale."

Marsden is a professor of pediatrics, of bioengineering, and, by courtesy, of mechanical engineering; and a member of Stanford Bio-X, the Stanford Cardiovascular Institute, the Wu Tsai Human Performance Alliance, the Maternal & Child Health Research Institute, and the Institute for Computational and Mathematical Engineering.

Skylar-Scott is a member of Bio-X, the Stanford Cardiovascular Institute, and the Wu Tsai Neurosciences Institute.

Additional Stanford co-authors of this research include medicine Professor Sean M. Wu; postdoctoral researchers Jianyi Du, Jonathan Pham, and Jason M. Szafron; and graduate students Jessica E. Herrmann, Soham Sinha, Anastasiia Masaltseva, and Fredrik Samdal Solberg.

Other co-authors on this work are from Carnegie Mellon University.

This work was funded by the National Science Foundation, the National Institutes of Health, the Swiss National Science Foundation, Berg Scholars Program, Alpha Omega Alpha Carolyn L. Kuckein Student Research Fellowship, Dorothy Dee and Marjorie Helene Boring Trust Award, Medical Scholars Research Program, National Heart, Lung, and Blood Institute, Stanford Graduate Fellowship, Parker B. Francis Fellowship, American Heart Association, Hoffmann/Schroepfer Foundation, Joan and Sanford I. Weill Scholar Fund, Breakthrough T1D, National Heart, Lung, and Blood Institute, Advanced Research Projects Agency for Health, Additional Ventures Cures Collaborative, and Stanford Cardiovascular Institute.

Research reported in this publication was supported by the Advanced Research Projects Agency for Health (ARPA-H) under Award Number AY1AX000002. The content is solely the responsibility of the authors and does not necessarily represent the official views of the Advanced Research Projects Agency for Health.