Bridging the innovation-regulation gap

We're in the year 2038. The European Union’s General Pharmaceutical Legislation entered into force almost 10 years ago. For the 3rd consecutive year, 7 of the 10 most innovative products in BioWired’s BioHealth Impact ranking* emerged from companies tapping into Europe’s scientific and data science excellence.

* Illustrative only

Europe is still not the largest single biopharma market but it leads the way in the development, testing and embedding of healthcare innovation into clinical practice. The innovation ecosystem in Europe has become adept at fostering the development of quantum leap products, balancing an openness to significant innovations on the one hand with the regulatory obligation to ensure product quality, safety, and efficacy on the other.

It was different back in 2023. At the time, Europe was losing ground to other large mature and developing markets. For established companies, the fragmented nature of its regulatory system was complex to navigate, and daunting for SMEs and start-ups.  It also inhibited vibrant innovation. This resulted in a deterioration of Europe's attractiveness for pioneering science and clinical trials. Companies chose other locations for the translation of research advances into clinical studies and approved, applied solutions for patients. This, despite the priority of public health to European values and the strategic importance of a robust ecosystem and health union to society at large.

How did the European regulatory system turn around its laggardly reputation for the regulation of medicines and related healthcare products? As in science, so also in regulation. Through experimentation. By having the foresight to adopt a regulatory sandbox approach to figure out how best to regulate the truly novel.

There were multiple drivers to the needed change: Fast advancing, new and disruptive technologies as well as the blurring of boundaries between interdependent dimensions of care.

Multiple triggers leading to the adoption of the sandbox approach:

  • Regulatory challenges at the interplay of drug, medical devices and in vitro diagnostics due to healthcare solutions drawing on multiple disciplines

  • Fewer clinical trials conducted in Europe with risks to the knowhow of the investigator community and consequently to the access of patients to innovations addressing unmet needs

  • Artificial Intelligence (AI), accelerating at an unprecedented pace, affecting not only R&D processes but also the nature of medical solutions offered to healthcare professionals and patients

  • The EMA’s own horizon-scanning report identified upcoming trends in innovations that would challenge the regulatory system1

Decision makers in Europe realised that the modernisation of the regulatory framework called  for it to be inherently able to anticipate future shifts in technology, in order to enable innovation uptake, while ensuring the safety and efficacy of the future’s incoming novel, medical technologies. The regulatory sandbox was the missing mechanism in the toolbox of regulators

A regulatory sandbox provides the means to clarify and calibrate

As of 2038, the sandbox approach is used more systematically, in conjunction with formalised horizon scanning and consultation mechanisms, to define which innovation requires the setting up of such a sandbox and which ones can be assessed via other means. The sandbox mechanism helps to clarify the pathway where none exists and to expedite regulatory decision making in the face of multidisciplinary complexity. 

By initiating a sandbox approach for truly novel innovations Europe worked around the  "pacing problem", the lag between the advent of those innovations and clarity on how to regulate them in a timely manner and fit-for-purpose way.  Through sandbox experimentation, it reinforced both its regulatory science and fostered the uptake of those innovations as they arose. 

The framework provided a means of reaching certainty, and with that certainty, developers were enabled to advance investment and progress their plans with confidence.  The accommodation within a regulatory sandbox of the weave of multiple disciplines and technologies has been decisive and a boost to the dynamism of healthcare startups in Europe. It has added credence to the sense of an openness in Europe to healthcare innovation leadership. 

What if the sandbox mechanism had not been introduced or been of too narrow a scope? Europe may have closed out a great deal of innovation potential, and healthcare systems and patients might have missed from their armamentarium the eventual products/ solutions from the following blue sky case opportunities: 

Illustrative examples of the regulatory challenges ahead

3D-printing technology has the potential to create highly customized products to fit each patient's body weight, lifestyle, needs, and display better surface, mechanical properties, and biocompatibility, compared to traditional manufacturing methods. Applications include dose, pharmaceutical form, personalized complex combination products such as bone cements with antibiotics or drug delivery strategies, blurring more and more the line between what is a drug and what is a device.  And it does not stop there.

As an alternative approach to current bone grafting and permanent implants, biofabrication combines the body’s own regenerative capacity with bioactive factors and biodegradable biomaterials that are formed into the complex shapes required to restore tissue form and function. This leads to the creation of living 3D constructs for regenerative medicine falling in scope of Advanced Therapy Medicinal Products (ATMP) and or potentially Substance of Human Origin (SoHO) regulations, adding further complexity over the challenges brought by the 3D-printing technology. 3D-bioprinting of products starts to be a reality as well with the first clinical trial running in the US in 20222.

Products using 3D-printing severely challenge the current regulatory system paradigm. The current system has been put in place for mass manufactured therapies and not for bespoke solutions. With 3D-printing, digital processing becomes central to the manufacturing of custom made solutions. Will it be regulated as a medical device? Is a regulatory system focused on the end product still the best approach or will a more process-focused framework be more adapted? How best to evaluate the performance of “a batch of one” from a quality perspective? The future of medicine is moving further towards personalization and custom-made solutions will no longer be exceptional. Existing regulatory exemption mechanisms are incompatible with this dynamic and are also open to risk in terms of required standards.

At Roche we are developing an AI/ML-based imaging solution to identify lung cancer lesions and measure the tumor burden in clinical trials, and increasing evidence generation in R&D. Companies are initiating AI applications across all stages of a medicine lifecycle, in particular to optimize R&D.

AI /ML use to diagnose or monitor patients along their care journey (a clear medical purpose) falls under Medical Device Regulation (MDR). It is a different story when used in clinical trials to measure outcomes (medicine development). In this case it would fall under the responsibility of the EMA, and even while it is not in fact a medical device, it will require technical expertise from that field. In Europe, it may be overseen to a degree by the AI Regulation. While the AI Regulation is cross-sectorial it does not anticipate the specificities of the healthcare sector. AI-based medical devices fall in the high risk category but there is ambiguity around AI/ML based imaging solutions. What we do know is that the impact of the use of such tools on drug development - and the associated evidence to be assessed - will be significant. 

There must be a risk that such undefined tools fall between the cracks, generating uncertainty which inhibits their uptake. Last but not least, EMA procedures are not adaptive enough to assess and evaluate an AI-based tool whose performance evolves over time with machine learning. Expansion of the tool’s use to another category of solid cancer beyond lung would bring further regulatory complications. New regulatory approaches need to be experimented and a regulatory sandbox provides the mechanism to do so.

A regulatory framework during which it is possible to develop, validate and test in a controlled environment innovative or adapted regulatory solutions that facilitate the development and authorisation of innovative medicinal solutions, pursuant to a specific plan and for a limited time under regulatory supervision. The approach has been implemented in other geographies3 in fintech, energy and education sectors among others. The concept is getting traction globally (OECD4, World Economic Forum, United Nations Organization) as an enabler to foster innovation uptake.

Nothing defines a bottleneck as clearly as the past years’ experience with implementation of IVDR (In-Vitro Diagnostics Regulation). That this has been unduly challenging is widely understood and even acknowledged by companies and regulators alike. There have been many pain points emerging with its implementation, ranging from a lack of coordination across the various supervising regulators to difficulties for Notified Bodies5 in absorbing the workload that came with further requirements6. It has caused erosion in the attractivity of Europe as a clinical trial destination and has delayed the start of several studies. 

What if a regulatory sandbox had already been in place in the EU? A sandbox approach could have been useful to anticipate the concrete implications of the emergent changes in the field of clinical trials with drug and devices/IVDs, and take measures to inform and adjust the regulatory provisions, as needed, based on evidence. 

Looking ahead, technology will drive the rise of more and more complex solutions, drawing on several frameworks (including new ones such as AI, SOHO) and will face further complexity at interfaces. It will take multi-disciplinary regulatory support and oversight, and mechanisms that can anticipate such developments, to resolve the open questions on how best to regulate such future products and solutions. 

As illustrated above, innovation is increasingly arising from inter-disciplinary approaches. It can take a very long time to formalize into regulation the changes brought about by advances in science and technology. And such formalization is an essential condition of a dynamic environment that stimulates and enables the uptake of related innovations. On one hand developers need regulatory certainty on the path to follow to secure authorisation. On the other hand Regulators need to calibrate what’s right in terms of regulatory oversight of emerging, novel clinical products. And it’s important to get it right, so that even for the most experimental medicinal solution, the principles of quality, safety, and efficacy, as well as of benefit risk assessments, continue to guide regulatory decisions. 

If we peer only a little into the future, we can see several trends that may require regulatory adaptation, including gene editing platform medicine, multi-intended computational tools, the rise of synthetic data, AI/ML, and the Internet of Medical Things (IoMT). Chat GPT has surprised society and AI applications are exploding. What will surprise us 10 years from now? Synthetic biology? Implantable drug factories? Digital twins?

Future proofing the regulatory system means equipping it with mechanisms to better anticipate, adapt and accompany the timely uptake of innovation. Today the provisions allowing the use of regulatory sandboxes are being debated in the context of the EU’s revision of the General Pharmaceutical Legislation. The essential multidisciplinary nature of incoming waves of innovation makes it imperative that the regulatory sandbox mechanism extends beyond pharmaceuticals.


  1. Health horizons:Frontiers in Medicine journal, December 2022.

  2. Pullen LC. Non-transplantable organs and tissues: a golden opportunity. Am. J. Transplant. 22(9), 2127–2128 (2022)

  3. Notified bodies are government accredited, mostly private companies, that take over sovereign tasks on behalf of national authorities. Related article:


Six bold actions for regulatory reform

More stories

See all stories

This website contains information on products which is targeted to a wide range of audiences and could contain product details or information otherwise not accessible or valid in your country. Please be aware that we do not take any responsibility for accessing such information which may not comply with any legal process, regulation, registration or usage in the country of your origin.

ContactLocationslinkedinfacebooktwitterinstagramyoutubeCovid-19Pharma solutionsRoche careersMedia libraryAnnual Report 2023Privacy policyLegal statement