Just more than a decade after
CMR Surgical
‘s founding, the surgical robotics developer won
FDA de novo authorization for its multi-port, soft tissue Versius platform in October 2024
.
When announcing the regulatory milestone, CMR Surgical said Versius (which had a CE mark since 2019) was already the world’s second-most-utilized surgical robotic system, with more than 26,000 procedures completed. CMR Surgical sought marketing authorization in the U.S. after conducting clinical trials in Poland and India.
“It’s a great relief as you can imagine, and a lot of work — my goodness,” CMR co-founder Dr. Mark Slack said in a recent interview with Medical Design & Outsourcing.
We asked Slack for advice that could help other device developers navigate the FDA de novo pathway, meet new cybersecurity standards and successfully launch their own products. The following has been lightly edited for clarity and length.
MDO: What did you learn taking your Versius surgical robotics system through the de novo platform that you and your team wish you knew when you started on that journey?
Slack:
“Pick very carefully who you work with, and you have to pick very carefully somebody who’s got the right background to do it. And then you need to absolutely match it with in-house capability as well. You cannot just go out to consultants and leave out a real skill in your internal, because otherwise you just get caught. To me, the biggest problem with regulation at the moment is that they are trying to prevent things that occurred due to criminality rather than things that occurred just by chance. Theranos is a good example. They were lying, making it up. Regulation’s not going to catch that. We have the same with the doctors. If you have a bad doctor who maims and kills people, then they change the appraisal process and they change the paperwork, but that won’t catch them. We are at the very critical point where regulation expenses are becoming an issue, and it is an issue for us in medtech, because Medtronic — they
make $30 billion a year
, they’ve got 800 products. An average drug [maker might bring in] $20-$30 billion a year for the blockbusters. They can afford the regulatory process, and I think we can’t. So one, you’ve got to pick your person really, really well. Number two, you mustn’t self-complicate. I have a a line with my team: ‘We’re not here to set the standards. We’re here to meet them.’ … We have absolute chaos in Europe with the MDR. It is not fit for purpose. It is harming the industry more than you can imagine. At the moment, it is easier to work with FDA than with Europe and and a lot of companies in Europe will be making a decision to go to the FDA.”
MDO: What are the advantages you would note of working with the FDA?
Slack: “I can do a Q-Sub (Q-Submission), I can get to know the team. I can get meetings, I can ask questions, I get answers. The Q-Sub process is a very good one … but you’ve got to be very careful to make sure that you create Q-Subs correctly, because otherwise you waste a hell of a lot of time. Sometimes you also have to take some risk as well, and sometimes just do it and and depending on the size of it or what that particular issue is. The de novo process is so big, it was scary.
MDO: What about cybersecurity standards and regulations?
Slack: “The other area that I think probably we weren’t anticipating was cybersecurity, but we did do that very, very well, and have learned a lot, and are now very well prepared going into the future for that. And we did that largely in-house, using external test houses as well. You’ve got to get the people to come and try and hack you. But our software people largely did our cyber themselves, which was really good. We did very well. We got a very high score and it’s just such a delight to see.”
MDO: Do you have tips for finding a cybersecurity partner? How would you vet somebody to hack into your systems?
Slack: “We went through legitimate companies, and there are plenty springing up at the moment. I would be very scared of using a hacker. I wouldn’t know where they would be coming from, or where they’d be going and so on. … The software team went up pretty quickly. We got one or two involved. They came in with their teams. It was a very set process for cyber, but anything cyber is money. It’s how many hours can you afford.”
MDO: Any other advice for navigating the regulatory process?
Slack: “The other bit of advice I’ll give with the FDA is working out how much they will tolerate outside of the U.S., and that remains a problem. FDA want U.S. data, and it cost us quite a lot of effort trying to justify that our surgical population was similar to the U.S. one and that the patient population was similar to the U.S. And that’s a very difficult one, because what is the normal American in 2024? I think the white population is going to be in the minority relatively soon. You’ve got people of Indian origin, people of Far Eastern origin, you’ve got your traditional Black Americans, you’ve got your traditional white Caucasian. It’s not dissimilar to Europe. But we definitely got pushed quite hard on matching. They are not keen on OUS (outside of the U.S.) data still. So I think going into the U.S. you need — depending on the depth of your pocket — to think very carefully about doing U.S. studies.”
MDO: If you had to do it over again, would you have paid to do trials in the U.S.?
Slack: “Yeah, 100%. An even bigger question — it depends on your depth of pocket, you know — would it not be better to, from the outside, do an IDE? I think the U.S. finds that very comforting. I suspect it’s congressional pressure to make people do de novos and so on. But you know, if you can afford it and you come and you do an IDE, well you’ve got all the same questions. You’re getting the data. You’re getting it in the right place. You’re getting it with the right surgeons. I think that would solve, for many people, quite a lot of problems.”
MDO: What are your thoughts on real-world data?
Slack: “It’s frustrating, and I’m not sure what the way forward is. At the moment, people mustn’t be fooled. If you go into the FDA and you get to use the term real-world data, make absolutely certain that your real-world data matches the criteria stipulated in the FDA document from around 2018. It’s out of date, it’s not accurate, but if you don’t use that they will argue with you about your real-world data as well. The FDA is very concerned about clinical acceptance criteria. And when you go to real-world data rather than a curated trial, the chance that you will get to those conversion rates and so on are quite small. The fact that instead of dealing with 50 people, you have a data set of 12,000 doesn’t seem to impact them. I’m a great believer that the future is real world data, but you have to curate your real world data particularly well. You have to have auditable standards.”
MDO: Any other advice you’d like to share with device developers?
Slack: “When you start a company like this, be very careful not to over commercialize too quickly. Get yourself solid, get your basis done. And if I did it again, I’d probably sell 50 systems and then stop, and then work those 50 systems to massively improve, get it to as strong as you can, and then go back out commercially. I think that’s a sensible move. Unfortunately, you get pressure from investors and so on. And we’ve basically done that now. We’ve matured the system and it’s giving performance to doctors that are coming back saying, ‘This is as good as, if not better than.’ That’s what I’ve always wanted. Co-founder Luke Hares used to say to me, ‘Mark, you’re always whinging, you’re always moaning. You always want things better. So when are you going to be happy?’ I said, ‘It’s got nothing to do with me. It’s when the surgeons tell me they’re happy.'”
We’ll have more from our interview with Slack in the coming weeks. Subscribe to our free email newsletter to make sure you don’t miss it.