Why is France becoming a major hub for digital health deep tech startups?
By Benoît Georis with the help of Samantha Jérusalmy, Marc Rougier, Louisa Mesnard & Saish Rane
In recent years, France has seen a significant growth (60+ startups every year) in the number of health tech startups (1,700+ so far) incorporated all over its regions. When zooming further, a fraction of them are receiving more attention than ever — digital health deep tech startups. At Elaia, we have a clear definition of what a deep tech startup is: a startup that has strong links with academic or research labs.
In this article, we explain that it is not the result of chance, but the careful mix of several key ingredients of success: people, technology, the ecosystem and the market.
At Elaia, we have been convinced since 2017 through our investments in Sim&Cure, Pixyl and Therapixel that these ingredients were nascent but present. In 2022, looking back, the thesis is right, Elaia has become a key player with 20–25% of its portfolio exposure in this space, including recent investments in inHeart, Gleamer, Sonio and i-Virtual. We love this space, and by the way we have dry powder so do come and see us!
At the heart of startup creation, there are people who have the drive, the resilience and the ambition to change the world. Furthermore, for deep tech startups, these people from the founding team often come from a research background and have a PhD.
France is clearly leading the way when it comes to fundamental research and scientific excellence. Two recent and brilliant examples illustrate this statement: French mathematician Hugo Duminil-Copin was awarded a Fields medal in July 2022 and French scientist Alain Aspect received the Nobel Prize in October 2022 for his career on quantum physics and exceptional discoveries on quantum entanglement. Closer to the subject at hand, in 2008, we had the Nobel Prize in Medicine awarded to professors Françoise Barré-Sinoussi & Luc Montagnier for their work in identifying the retrovirus responsible for HIV-AIDS. The ecosystem has evolved quite a bit since. Globally, France counted more than 70,000 PhD students in 2020, out of which roughly 40% focused on subjects like IT, information & communication, computer science, medicine or mathematics (source). Makes for a growing pool of potential future entrepreneurs!
The growing number of digital health deep tech startups is the consequence of two major and long-lasting trends. First, health-related research topics are by far the most preferred by PhD students as they offer numerous open issues and thus lots of possibilities to make discoveries, breakthroughs, international publications or … file for patents! On top of it, PhD students and full-time researchers care a lot about contributing through their research to a major societal impact (e.g., cure a disease, provide healthy elderly living, etc.).
Second, not so long ago, doing business after a PhD was not the default or desirable option, but now it’s increasingly becoming the norm thanks to a cultural shift propelled by entrepreneurial chairs in research universities, startup studios and tech transfer officers to push post-docs to search for viable business applications of their research.
The underlying tech in health tech startups has greatly improved in recent years. For example, you have the advent of digital twins (e.g., inHeart), and robots increasingly being used in the operating theater (e.g. Da Vinci Surgery) driving digitalization in hospitals. We discussed this in our previous tribune on Digital Life Sciences.
Finally, and importantly, the arrival of AI and deep learning can, a posteriori, be seen as a true inflexion point in terms of market adoption of such technologies. There are three main reasons why AI has been a paradigm shift for medical software development:
First, execution speed is now proven to be truly real-time — opening access to routine markets (eg, the physician or specialist can get his/her report by the time the patient is back from the MRI scan, which means within a few minutes). Before that, these technologies were limited to clinical research where time was not the prime objective and where experiments could last for days or weeks. Most of the complexity is now pre-processed during the training phase (i.e., before the product is shipped to the customer) and not at runtime anymore.
Second, accuracy of the KPIs (sensitivity and precision) are better as specific cases or ‘frontier cases’ can be dealt with during the training phase of these AI algorithms. Before AI, software developers would either ignore these frontier cases or find a specific set of parameters to manage them in a long, tedious, back and forth development cycle.
Third, there is a significant increase in alignment and cooperation between the software development teams and the regulatory teams, within a startup. Working in an agile environment is one thing for sure, but the key element comes from AI again, as deep learning algorithms by essence require tons of data to be trained onto. This precisely matches the requirement of the regulatory process to demonstrate the product efficacy (and safety) through a clinical study on large patient cohorts. As a consequence, startups focus from day one on data collection (medical records, patient MRI or CT scan images…) whereas previously, early startups were falling short on being able to demonstrate validity on sufficiently large patient cohorts.
The same trend is true for other health tech categories such as genomics. The arrival of AI has opened a brand new avenue of care with personalized medicine. A great example is SeqOne Genomics which has created a platform to analyze patient genomes and provide accurate and fast results to accelerate the adoption of personalized medicine apart from outlier cases and to mainstream use cases.
Market acceleration often starts with the emergence of an ecosystem. We won’t pretend to be exhaustive but rather illustrative in this part. Let’s focus on examples across the value chain we know well but there are many others, of course!
Medtech is clearly at the crossroads between two disciplines: computer science and medical/health sciences. Well, guess what? The good news is that research organizations are increasingly developing strong ties amongst themselves to foster multi-disciplinary teams and innovation.
INRIA, France’s premier research center on computer science has been leading the way for decades. They spun out their first medtech startups in the early 2000s with Qantificare followed by Insimo in 2011, Pixyl in 2015 whose technology is the result of years of collaborative research between INRIA and Inserm in Grenoble as well as in 2017, with inHeart that started with a collaboration between INRIA, CHU Bordeaux as well as IHU Liryc. In 2020, INRIA went a step further and inked this strategy, among others, in its objectives and performance agreement signed with the French government (source) to serve and support universities on software technological development. In other words, it means more cooperation between physicians, clinicians and software developers.
At Université PSL, we are proud to support the innovators of tomorrow through the consolidation and facilitation of an ecosystem through the PSL Valorisation office covering the entire innovation value chain: awareness and detection, intellectual property, maturation, transfer, and support for startup creation. This support continues well after the creation of the startup with the PSL Tech Seed scheme that gives a privileged access to seed capital through the PSL Innovation Fund managed by Elaia.
A concrete example of the fruits that this strategy brings is Gleamer, fast-growing startup that develops medical-grade AI for radiologists and cuts down the time taken to analyze a case and increases accuracy of decision making. Gleamer started off at Université PSL where Christian Allouche, co-founder & CEO was a student entrepreneur. Currently, Gleamer has received its CE Marking & FDA clearance to operate in the US, has struck key partnerships with Fujifilm & Aidoc to expand the use of AI in medical imagery and is expanding to the US.” adds Cédric Denis-Remis, Vice-President at Université PSL
This observation goes even further as grad programs within universities and high schools are increasingly offering either dual track programs (major in medicine + minor options in computer science for instance) or immersive programs where a student can spend a few months in a different discipline from his/her graduating program to learn about a specialty field.
Another interesting example comes from SATTs (Société d’Accélération de Transfert Technologique) whose role is to provide operational support (eg, negotiating IP rights) and funding (eg, enabling technological maturation, raising the TRL from 1–2 to 4–6) in creating spinoffs and deep tech startups. In February 2022, the Montpellier University Hospital (CHU) and the Montpellier Cancer Institute (ICM) announced that they became the latest shareholders of SATT AxLR, with a clear objective to increase the number of new medtech startups in the Mediterranean area of France’s Occitanie region.
Over the years, many initiatives have grown up to help medtech startups from ideation to market access.
One of the oldest health tech incubators in France is Eurasanté, incorporated in 1994 in Lille in the Northern part of France. Eurasanté was followed quickly by Paris Biotech Santé (PBS) in 2000. PBS has been founded by several premier research institutions: Université Paris-Descartes, ESSEC, Ecole Centrale Paris and INSERM. This first generation of incubators were instrumental for the development of the sector, not only through the number of startups created but also by raising awareness of the industry and investors.
Over time, these regional, standalone initiatives have evolved towards more cooperation between actors to finally become structured ecosystems. The French Tech ecosystem has now its own health tech thematic network. Future4care, initiated by Capgemini, Generali, Sanofi and Orange is another great example of such an evolution.
Aside from these networks, there are specialized startup studios emerging such as iBioNext.
3. Business Angels
Medtech startups can start benefiting from two very attractive sources of early money: serial medtech entrepreneurs and physicians/specialists.
Indeed, it’s a virtuous cycle that is created and France has already had the first prominent exits in that space. The most recent example being Philips acquiring Cardiologs. These entrepreneurs are now investing in and giving back to new startup opportunities and assisting them with operational help (e.g., how to best define a predicate for a clinical study, how to accelerate a regulatory filing, etc).
More frequently, physicians and/or specialists are becoming business angels too (e.g., in the case of Sonio, they had a group of obstetricians as business angels)!
They bring at least two pieces of value to the table:
- Insights and feedback on UX and how one startup’s product would best fit in their daily workflow and
- They are the best advocates of the market appetite to buy these new technologies and convince VCs to invest during later stage financing rounds.
From a regulatory and market access point of view, several barriers have been lifted in recent years throughout the regulatory pipeline, thus promoting risk-taking by the entrepreneur and VCs, for the benefits of patients!
Further upstream, soon arose the question of medical data access. As awareness of the importance of data grew and following repeated requests from startups to access medical data, some research organizations began to implement various ways for startups to access their data. To give an example, in 2020, APHP organized the second AI for Health challenge where teams were invited to compete on oncology matters with access to specific sets of data.
In parallel, an important public initiative was brewing and in 2019 the Health Data Hub was created. The objective of this association that currently includes 56 partners is to democratize the access to health data to startups, among others.
Moving downstream, when entrepreneurs address the critical question of the business model and of defining a valid economic equation, they often had to rely on private insurers for the US market and to find a payer for the EU market (either the physician or the hospital or big pharma). This may be overly simplistic, but you get the point.
Entrepreneurs can now assess the option of a reimbursement by healthcare state systems. Indeed, there are now more supportive legislation & policies. A recent European directive on digital health technologies has been adapted at country level. We now have Article 51 in France as well as the ETAPES programme which will soon come under the aegis of the Haute Autorité de Santé, DIGA in Germany and Belgium to cite a few. Some great examples of this new trend are Tilak Healthcare in France, HelloBetter in Germany and moveUP in Belgium.
In between, each medtech startup has to obtain its Holy Grail: the market access authorization. Obviously, French deep tech health tech startups aren’t created to stay in Europe, they have to have global ambitions, and address the US, Chinese or Japanese markets for instance. When it comes to regulatory clearance, there are still lots of disparities between the three major markets. In our opinion, the USA is clearly leading the way as they recognized very early the wave of medtech startups and adressed the need by adapting the US FDA regulation that now has a category called “Software as a Medical Device”.
On the other hand, obtaining the CE marking in France is currently a major burden!
There are two reasons for this difficulty. The first one must be found on the notified body side. There is a real bottleneck of startups waiting to be audited (due to the hyper fast growth in the number of new medtech startups) and on top of it, the new MDR (Medical Device Regulation) worsens the situation as they have to upgrade their audit process and the MDR is more stringent on who can run an audit (people having a minimum of 5 years experience or more). As a result, some startups now consider entering the US market even before their domestic market. This question has become so mission critical that according to several sources, the French government is currently preparing to ask the EU Commission for an additional delay or transition period before the MDR enters in full force!
The second reason must be found on the QARA (Quality Assurance and Regulatory Affairs) profiles side. There is a glaring lack of QARA profiles on the market. The positive side of this situation is that it undeniably illustrates the maturity of the market but the negative side is that their salaries are skyrocketing.
All these initiatives described above also translated down the line into a growing appetite from the Venture Capital space to invest in that new vertical. Some of the recent funding rounds in France show a maturing of the digital health ecosystem with Dental Monitoring raising $150 million, Synapse Medicine raising a €25 million Series B and SeqOne raising a €20 million Series A. According to La Tribune, in 2019 alone, health tech startups have raised 1.8B€ globally (including medtech, biotech and e-health), with a record 60% from VCs (80% increase from 2018). In 2021, the figure climbed to €2.3 billion and since, there has been no sign of it slowing down in 2022 with a new record at 1B€ for the first quarter alone, a 233.5% increase compared with the same quarter in 2021.
Globally, venture capital funding and IPOs by European and US companies amounted to €55 billion in 2021, up 41% year-on-year. Another proof of this funding appetite, we start to see private hospitals in the US starting their own VC arm and even becoming acquisitive with Radnet acquiring Quantib for example.
This wave will soon make its way to Europe and Elaia will be on the right track to partner with such upcoming hospital funds, etc.
Besides the massive investments, this market has become dynamic and acquisitive. These transactions are not the result of chance. There is a rising attention from global players that recognize three strategic pillars in this market:
- Health market opportunities are often evaluated in $ billions, if not 10s or 100s of billions
- The health market is super resilient across crises as pandemics, viruses, cancers and the like don’t care about economics
- Technologies are now robust and proving efficient to bring value to the whole value chain, starting with the end-user: the patient!
As a single proofpoint here, Amazon acquiring One Medical this summer for $3.9B is a sign that technology is transforming healthcare at an unprecedented pace.
We have seen that in recent years, the French ecosystem and entrepreneurial base for healtech startups have made a giant leap forward. Now, in order to mature even further, this thriving ecosystem needs at least two additional things for these startups to become worldwide champions:
- More late stage/growth funds to help them scale and resist the temptation to engage in a trade sale too early. Of course, they can already get funded by Anglo-Saxon funds, which is not a bad option at all, but which is clearly not enough!
- More acquisitive French/European public companies to keep and nurture this growing pool of talents on the European ground.
At Elaia, we are more convinced than ever that this will happen and that is why we have defined a strong investment thesis on medtech and we have prepared ourselves to support this well-differentiated positioning. For instance, we work with top-notch medtech entrepreneurs as venture partners and they bring invaluable help during deal analysis as well as portfolio management and CEO mentoring.