For a decade, the brain was where pharmaceutical money generally went to die. Too complex, too poorly understood, too many failed trials. Nicolas Roberti, co-founding partner of the European BrainTech Fund, calls it a graveyard.
His argument is that the graveyard is about to reopen for business, and that the reason is not a single breakthrough but a pile-up of them arriving at once: AI that can read multi-omic patient data, carrier molecules that can finally get bigger drugs across the blood-brain barrier, electrodes with new designs and new purposes. The idea he hears repeated across the industry is that neurology is becoming the new oncology, following the same arc from investment desert to funded frontier that cancer took ten years ago.
Roberti and colleagues are raising now to back companies across the whole field: therapeutic biotech, neurotech and neurostimulation, brain-computer interfaces, and the digital biomarkers and clinical-trial tools that sit underneath all of it. His thesis is convergence. Nobody, he argues, becomes a leader in this field by staying in one silo. His own track record runs through life-science venture, including a board seat and lead-investor role at a EEG-software company he uses as his worked example of the model paying off.
Roberti spoke to Unfiltered at HLTH Europe in Amsterdam. The fund is at what he calls the initiation stage: team assembled, research-centre partnerships forming, first conversations with investors underway.
The conversation spans the sober to the speculative, and Roberti is candid about both ends. He’ll tell you the brain is still too poorly understood to model on a chemical level, and that our overall understanding remains, in his words, too bad. Then he’ll tell you what he’d personally sign up for if the technology existed. Not a chip that shapes how he thinks. He wants to keep an independent brain until the end of his days. But a faster one, with perfect memory? He’s open minded.
How important do you think the brain is, as a subject within health, and how important will it be over the coming years?
Very good question. I think everyone agrees it’s the last frontier. It’s where science has not [had] all these answers. And so that’s where the most incurable diseases are located. So it’s very important to be able to understand it better, to define better treatments and better outcomes for the patients at the end of the day.
Regarding the trends, when engaging with people from the medtech world or the biotech world, you have that sense that there is something changing. There’s something changing because there’s a lot of innovation in the toolbox, be it technological revolutions, AI, be it on the science side, with brain carriers that are able to get bigger molecules through the blood-brain barrier. So you’ve got new things coming, and that suddenly makes pharma companies come back, makes medtech companies investigate more into the field. It was a graveyard, but it’s about to change.
How do you see the space currently?
So the big trend that medtech people and biotech people are sharing is that neurology is likely to become the new oncology from a pharma perspective. Like, 10 years ago, suddenly the transcription possibilities in genomics could define better targets. And suddenly it brought gigantic amounts of investment, and now, 10 years later, big results for patients, good outcomes.
On the medtech side, you’ve got a bit of that same assumption, that neurology is becoming what cardiology was 20 years ago, where suddenly better understanding of all the flows brought a lot of innovation, a lot of funding, and a lot of better patient outcomes, with the stent modelling and all these types of things. Well, for the brain, the assumption is the same: that it’s going to change. So it’s becoming, again, an attractive investment proposal.
And that’s the reason [for] launching a fund on that investment theme. I’m at a really initiation stage. So we have a team, we are building partnerships, [with] research centres, and we start only now engaging with investors.
So we’re really at the beginning of that road. But our investment thesis is that we would invest in therapeutic [assets], so biotech, neuroscience, in medtech, neurotech companies, neurostimulation, brain-computer interface, non-invasive stimulation, new designs of electrodes that can be used for different purposes. Digital health, development of new biomarkers, development of new tools for clinical trials. So it’s across the field.
The reason is, we believe that there is more and more convergence, that tomorrow you cannot become a leader in that field without breaking the silos. You need a better understanding of how, technologically, you can read the brain to better treat it, either with chemicals or through electrical stimulation. AI is definitely part of the equation. The convergence of all these types of technologies is going to make the champions of tomorrow.
Is there anything so far that you’re aware of that gives you confidence we’re entering a new phase of possibility?
It’s interesting when you listen to pharma companies in different panels. They say, oh, we’re all coming back because the toolbox is becoming better and better, and they state, like, biomarkers, it’s where it’s gonna happen. But then you ask them: which biomarkers? And there you suddenly have a lot of long silence. Everyone has got somehow an idea, but we don’t know what the business model is.
Pharma companies before were investing in biomarkers, they were internalising these types of technologies, but that’s not their core business. Is it the CROs that are going to develop that, or [is it] smaller companies that are going to consolidate? We know that the technology is there, but who is going to finance it? That’s still a bit the question mark.
But I got a great example of where it somehow works, because I was a lead investor and I’m on the board of a company [Clouds of Care] that developed some great knowledge and insights about EEG signals. They are a software company. They are hardware agnostic. And they build algorithms where they’re basically denoising all these signals, doing quality checks, and that provide a very good base for clinical trials.
That was their stage one. But stage two of the company is that they developed digital biomarkers [from] all the insights and the data they were able to provide. So they’re now able to give very worthful insights to the pharma companies that allow them to adjust their clinical trial design, by either stratifying more of the patients, or doing retrospective studies; say, okay, there is an EEG signature for this type of patient, let’s get them out or let them in. You can now use these types of biomarkers for increasing the probability of success of these clinical trials. They were initially in epilepsy. Today they are developing a biomarker on Alzheimer’s, super promising.
How well do you think we currently understand the brain?
Very hard question. It’s a bit like the Black Swan, you know? Do we know everything? Every day we’re gaining more and more understanding, because we’re able to do now longitudinal [studies].
There are a lot of foundations or pharma conglomerates financing cohorts of tests on large cohorts of patients, that are measured on a multi-omic side. So you have blood, you have a CSF test, you have PET scans, you have CT scans, you have MRI. You have the genomics of these patients, and they are measured across different years. When you run AI on it — you still have [problems] to solve, because it’s complex, but we gain a much better understanding. And so the speed at which we understand it better is tremendous. Now, how good is our understanding of the brain? Well, still too bad.
Is that a bigger risk to what you’re doing, or a bigger potential reward, because there’s so much still to discover?
Let’s say it’s a calculated risk. Because the opportunity is so, so, so big, and the technological revolutions are so big. Probably some other people would have said the same 20 years ago on certain other types of technologies, but I think the quantum leap of this technological evolution is much bigger. So it’s still calculated. If we don’t know the [unknown], it’s not calculated — it’s a risk, an unquantifiable risk, but one that we want to take.
We believe that there are some very interesting business models that are going to rise around all of that. And at the end, we’re gonna measure the success of the fund with the return. But I think it will be quite important, because when you bring it to patients successfully, it’s going to be a nice return.
Christian Larsen at Netholabs is working on a whole-brain emulation project, a digital replica of the brain. Will you be looking at some of those frontier ideas?
You’ve got various established companies that have been working on digital twins of the brain that are then used to make the designs of medtech devices. I think it’s more on the chemical side that it’s very hard, because we still don’t know a lot about it — and [at] which level do you stop your model? Do you go on cell level? Do you go on molecular level, do you go on an ion level? These are very different dimensions, and today we cannot grasp the complexity of that into a model. We will likely be able to, but then it’s like physics, you know, you can always go one level deeper.
We would look at it from a business perspective. What can it bring? How can it play its role, with an economic model, into bringing these new therapies to market? Is it on a diagnostic side, we can model a replica of each patient’s brain to understand better how we can do a surgical intervention? You can do a foundational model of a disease. These are also new business models. But what’s the moat? If AI can support you doing that, what would impede a competitor doing exactly the same in a month’s time? So these are the things we will be looking at.
We’ve got interventions that might solve a medical issue. And we’ve got performance and enhancement on top of that. Will you be looking at both of those areas?
Our entry point is, we want to see therapeutic impact or symptom management. But sure, you have the other, especially when you go to mental health, sometimes it’s a thin frontier between the two, a fine line. Especially about depression. You feel better, but what does that mean? Is there a chemical signature that says yes or no?
Sometimes it’s not that easy to define. So it’s not the focus of the fund, but it’s on the periphery. So we’ll definitely have a look at it, but more like a positive side effect of an initial entry point that we would have in a company. On brain-computer interfaces, we’ve got all these ethical questions. That’s less easy, because it’s still blurred over there, how a company is going to develop into that.
If technology came along that doesn’t yet exist, would you upload a new language to your brain? Give yourself perfect memory retention? What would you personally be up for?
Yeah, it’s a good question. Do you want a code that would somehow push you to think in a certain way or another? I don’t think so. From a human point of view, I like to have an independent brain until the end of my days. But then it would be more on the neuroplasticity side, you know, being able to think super quickly, super fast, having a memory that works perfectly. That would be more than ideal. I would love to have that.
What do you think humans will embrace, and what do you think they will never accept, when it comes to brain interventions?
What they’ll never accept, I think, will depend on the type of society we decide to live in. Do you like to have the big American companies, that we like very much, but with their leaders that sometimes are not that democracy-prone, that would like to write into your brain? Would you like that? I think for Europe, at least, that would not be acceptable. But what are the things that would be acceptable? It’s also a fine line with the psychedelics, and, you know, it’s somehow related to drugs without too many side effects. So, also from an ethical perspective, if you’ve got drugs but without the addiction, is that fine? We can love each other so much more and have a super fast brain, without side effects. Is that okay? Do we want that? Good question.
