The Children’s Hospital of Philadelphia (CHOP) bought its first 3D printer primarily to model hearts. It did not expect to find such a compelling use case in melting them – metaphorically, that is. The story starts in 2020, with the birth of Lilianna and Addison Altobelli, twin girls joined at the chest and abdomen. Since 1957, CHOP has separated more conjoined twins than any other hospital in the US. Practice makes perfect, but, with such a rare condition, it’s very difficult to be able to. Before the Altobellis, the hospital had still only performed 28 procedures in more than 60 years. Or that would be the number – unless you count all the rehearsals made possible by the hospital’s onsite 3D printing facility.

In preparation for the operation, the Children’s Hospital Additive Manufacturing for Pediatrics (CHAMP) lab created full-size 3D models of the Altobelli twins’ shared anatomy – which spanned their chest wall, diaphragm and liver. Designed to be assembled and disassembled like Lego, these models helped the surgical team solve the unique challenges of safely separating Addison and Lilianna, and then enabled them to repeatedly walk through the procedure for doing so. The results speak for themselves. Ten months after they were born, in a ten-hour operation, Addison and Lilianna were successfully separated.

“We’re starting a new book – it’s not even a new chapter, it’s a new book,” said the twins’ father, Dom, after bringing them home to Chicago in late 2021. “We started a brand-new book for the girls, and there’s an Addy book, and there’s a Lily book.”

But the CHAMP lab’s involvement didn’t end there. Dom and his wife Maggie wanted a keepsake, something that their daughters could return to as they grew up. That meant there was one more model left to make. For that, CHOP medical illustrator Brittany Bennett took a full surface scan of the twins prior to their operation, refined it, and removed all the medical tubes, lines and trachs. Then the engineers printed it.

“We were able to give the parents a little model of the two girls in their conjoined state,” explains Elizabeth Silvestro, the lab’s lead engineer. “And the mother was over the moon.” CHAMP had given her more than she thought was possible: her children, twice over. “She had never seen them without the tubes; the only time was in this model.”


Before the Altobelli twins were separated, CHOP had only performed 28 heart printing procedures in more than 60 years.


Double vision

It’s often the exceptional cases that introduce hospitals to the potential of 3D printing at the point of care. “One of the running jokes is most people either start with kidneys, hearts or conjoined twins,” says Silvestro. “Every printing lab has a conjoined twin story because it’s such a unique case for what printing can do.” That’s not to say it’s the only one. Implemented properly, in-house AM can help hospitals take better care of almost any patient.

As well as surgical models like those used for the Altobelli twins, the CHAMP lab also produces educational toys for young patients and their families; simulation and training tools for trainee physicians; and replacement and patient-specific devices that can’t be sourced from external manufacturers. The other main application for onsite AM is in the creation of surgical guides – plastic templates that enable surgeons to carry out joint replacements and corrective bone surgeries according to plans determined on computer screens. Most impressively, these guides enable off-the-shelf components like replacement joints (which come in a range of sizes) to be personalised for specific patients. “Each guide is produced to cut the bone in a specific way, so the implant fits exactly as planned,” explains Andy Christensen, current chair of the Radiological Society of North America (RSNA) 3D Printing Special Interest Group. As he puts it, “it’s an elegant way of doing personalisation without having to personalise the implant,” bringing all the advantages of custom implants for patient outcomes at a fraction of the time and cost.

At first, guides and models were almost always provided by external 3D printing specialists like Medical Modeling, the company Christensen founded in 2000. That process was expensive and could take time, so as certain types of 3D printers became more affordable, what Christensen calls “surgeon-tinkerers” began to explore what they could achieve by bringing the capabilities into their workplaces. “They get complicated anatomies and they love to see things in their hands,” Christensen says, referring in particular to maxillofacial and orthopaedic surgeons, as well as cardiologists dealing with heart deformities.

And that brings us back to CHOP in 2011, when a group of radiologists and cardiologists decided to invest in a 3D printer that they could use in planning for heart surgeries. It could have gone better. “A very nice printer was bought and put in a room six floors away from everybody with the idea that the physicians would have time to run it,” says Silvestro. “A few years later, they realised that wasn’t the most productive plan.” CHAMP only began to transition to what it is today when Silvestro, and her radiologist colleague Dr Raymond Sze, were given the full-time jobs of running it in the mid-2010s.

A rarity

Since then, every hospital in the US has been exposed to a different type of “extreme, once-in-a-lifetime case”: Covid-19. Although Silvestro and Sze have always used the CHAMP lab’s position in the radiology department to avoid being siloed, even they weren’t prepared for the sudden spike in interest. “We were able to put devices all over the hospital,” Silvestro says. “Even making simple ear-savers for masks got the word that there was a printing lab out there further than we ever thought we could reach.”

That’s particularly important in a paediatric hospital like CHOP because so few medical devices are specifically tailored to the needs of children – masks and PPE included. “A lot of paediatric devices are just scaled adult devices, which isn’t a good representation of the paediatric anatomy,” explains Silvestro, “but by having the lab inside the hospital, we’re able to help fill some of those gaps.”

Indeed, with the whole hospital newly aware of what the CHAMP lab could offer, Silvestro and her team received a flood of requests about replacing child-friendly devices that were either unavailable because of supply chain issues or had simply been discontinued. Their work on creating mask adapters, airway devices and guidewire tools for young patients even began to attract attention from outside the hospital’s walls. The Philadelphia Paediatric Device Consortium, which is funded by the FDA to improve the supply of devices suitable for children, has met with the CHAMP team to talk about facilitating partnerships that might enable their designs to be mass-produced by external manufacturers.

That could be a real boon for less well-resourced hospitals. They can’t all afford their own labs, and, at present, even those that purchase 3D printed guides and models from specialists do so without much hope of reimbursement. “The hospitals are doing this because they know that it helps provide better patient care,” explains Christensen. “But there isn’t a direct financial connection between the provision of this service and some funding that would come back to pay for it. And it’s been that way forever.”

He should know. Having repeatedly tried and failed to get US reimbursement codes for 3D-printed surgical tools established while at Medical Modeling. Christensen’s now leading the RSNA’s attempts to put things right. So far, the 3D Printing Special Interest Group has secured a set of temporary codes, which means certain 3D-printed tools no longer need to be filed as miscellaneous (and completely ignored by insurers), but there is still work to be done before they become permanent, paid-for parts of the system. “In order for these labs to stay afloat, they have to have some reimbursement,” says Christensen. “If there’s no money, it’ll always be one person in a corner with a maker bot trying to make it work. To do it right, you need the right kit, and you need the right people, and you need time. And all of that means you need money.”

The right call

As manufacturers will happily explain, 3D printing can get very expensive, very quickly. Many of them still outsource it themselves. Oddly, hospitals in the US have a bit of an advantage there. “If you’re going to sell an anatomic model, you would have to clear it by the 510K pathway as a class two medical device,” explains Christensen. “But if you’re a hospital – even if you want to make your own implants – you can do it without any oversight by the FDA.”

At least, they can for now. As the prospect of hospitals creating their own implants has become less fanciful (the Mayo Clinic has implemented metal AM, for instance), regulators, too, have begun to pay more attention. According to a recent discussion paper it published on the topic, the FDA is considering three models for regulating point of care AM. The first would see manufacturers selling certified medical device production systems that hospitals could install for on-label use. Otherwise, manufacturers could take on the regulatory burden by co-locating small facilities in hospitals, or hospitals could themselves choose to be regulated as medical device manufacturers.

Whatever happens, Silvestro welcomes the prospect of greater regulatory involvement. Hospitals today – with their legal departments, review boards and device committees – do not use the ‘practice of medicine’ concept as a cover for undercutting medical device manufacturers while evading regulatory scrutiny.

“The biggest challenge right now is not having a defined pathway,” Silvestro explains. “If you look at the literature of different hospitals on how they get devices through development, it’s a completely different route for almost every single device.”

As might be expected, what Christensen calls the “democratisation of manufacturing” isn’t happening all that smoothly. Nonetheless, it’s opened a new book on how hospitals care for their patients – and we’re not going back to the old one. “Think of all the regenerative medicine technologies for replacing human tissue,” says Christensen. “That isn’t here yet, but when it is, technology-wise, I think one of the ways it will happen most often is at the point of care.” CHOP started by modelling hearts; one day, it might make them.