In an interview with Interventional News, John Kaufman—Frederick S Keller chair of interventional radiology at Oregon Health and Science University, in Oregon, USA, president-elect of the American Board of Radiology (ABR) board of governors, and chief medical officer for Cook Medical—looked back to see ahead, addressing what is next for device development in the interventional arena.
IN: Earlier this year you gave a talk at the Society of Interventional Oncology (SIO) 25 Years Past and Future session during this year’s meeting (30 January–3 February, Las Vegas, USA)— what were your main points of discussion?
The topic that was assigned to me was the kind of overview subject matter you receive when you reach the later period of your career. My talk focused on how we in interventional oncology (IO) evolved from very focused, limited interventions—either the early transvascular, which was all embolization and initially just ischaemic to kill the tumour, then to the delivery of therapies and the evolution of said therapies. That has been one trend that continues to grow, particularly concerning transvascular intervention in the liver and kidney.
I also spoke on ablation becoming much more mainstream than it was 25 years ago. Although there was ablation back then, a lot of it was ill-directed alcohol ablation. Today, ablation technologies have taken off and expanded to new approaches such as the delivery of energy to tumours. What I think will be interesting, is the development of ablation to target malignant disease states that we haven’t had a significant role in, such as pancreatic and lung cancer.
One of the nice things about IO is that the multidisciplinary approach will continue to be important as our interventions are never in isolation. We can use a combined approach with other modalities, immunologic modalities, and oncologic radiotherapy, to expand in a more synergistic direction, so that’s going to be really important.
On the device front, I’m very interested in devices that do more than just deliver treatment. Devices that can help to monitor or share information, can tell you the flow, or give you viability information. It sounds crazy, but a device that can tell you about the local genetic environment it’s sitting in, such as whether you’ve got tumours that are changing — devices that have functionality in addition to delivering therapy, I believe that’s the direction we’re going in.
IN: Artificial intelligence (AI) is playing a larger role in medical device development and function—do you think AI will be integrated with IO?
In oncology, AI may support screening—figuring out which patients appear to be on a trajectory that is approaching a threshold for treatment. With AI as a background system that can collect data on lung nodules, for example, you can compare data from the entire programme. Radiomics is very important for cancer and the digital information management that AI can provide is crucial at pre- or post-treatment to understand what we’re doing and what results we’re getting. In the visualisation world AI is going to be huge. I think we’re only just getting to the most interesting part of IO, particularly in the image-guided oncology space.
In terms of clinical information processing, in the end we need to realise that programmes like ChatGPT are an echo chamber—it’s just what’s available in the world and all you’re going to get back is information based on what is already accessible. We will always have a level of scepticism, and you have to be a little bit critical.
IN: Do you see 3D-printed medical devices playing a role outside of prototype creation in the future?
Outside of prototypes, 3D printing medical devices is a sort of magical concept that says—I’m going to place a stent, here’s the computed tomography (CT) scan, it goes to your printer and out pops the stent tailored to that patient. That’s the holy grail, but the regulatory pathway for that would be unbelievable—impossible, although it has happened in the aortic space through custom-made devices. The process is becoming more mechanised with inputs from the imaging actually felt throughout the device manufacturing process.
But, will you end up being able to print patient-specific devices on the fly? Where a patient is on the table and you say ‘I need this stent—print if for me!’, That would be awesome, but I’m not sure we’ll get there. In terms of managing inventory, 3D-printed devices would be incredible, but maintaining and certifying them is unlikely.
“I believe smart devices that give us feedback, pain scores and procedural follow-up are going to change people’s lives immensely”
IN: How often have you seen new devices be applied to treat disease states that have long been managed medically rather than cured?
Portal vein reconstruction is a great example of this. Tumour invasion or adhesion to the portal vein is a disease state that has been sitting in front of us forever. We have been able to image and diagnose the problem, but it is what it is, you just have to live with it, and we manage the complications. Then taking tools from peripheral intervention, taking knowledge from liver access that we use in the biliary area, and portal vein embolization, taking knowledge of access and recanalisation, we’ve been able to open up veins no one ever thought could be opened. There’s a tremendous amount of interest because there are many patients with this issue.
Now it’s one of these situations in which you have nothing to offer— you’re not looking for a solution so you’re not paying attention to how often you’re seeing it. When you suddenly have the ability to treat something, you see patients with this all the time.
Another area where I have seen this development is pulmonary embolism (PE). Fifteen years ago, how often were interventional radiology (IR), cardiology and vascular surgery involved in PE? Almost never. The patient was managed medically. Yet now, with the thrombectomy tools at our disposal, we treat patients with PE up to three or four times a week. This is something IR does a lot, where you are always taking tools and skill sets and looking for new applications, which I believe will also play out in the IO world.
IN: US President Trump revealed plans to impose tariffs on imported goods which will affect a significant proportion of available medical devices that are manufactured outside of the USA. What are your thoughts on this?
There is a lot of concern, certainly for any of the international corporations, not over what’s going to come in but what’s going to happen to devices going out. I am concerned that the proposed tariff will drive up the already excessive cost of healthcare. Many companies have moved device manufacturing overseas or across borders, and if this suddenly becomes a much more expensive proposition, it ends up trickling down to the price of available medical devices. It remains to be seen how real this will be.
IN: What has been the most significant area of device development in the last five years, and in the next five, what development will be the most influential?
For me, the biggest change has been within PE intervention. We now have very good devices for dealing with large embolisms, and these devices have been pivoted to in endovascular IR, vascular IR and elsewhere. In the next five years, I believe smart devices that give us feedback, pain scores and procedural follow-up are going to change people’s lives immensely.
