The Chessboard Problem: Why Europe's ATMP Rulebook Still Isn't One Rulebook

The plenary panel, "Europe's ATMP rulebook reset: the 12-month pathway from fragmented rules to trusted EU execution," brought together a genuinely rare mix: a sovereign regulator, a nonprofit developer who's just been through EU marketing authorisation, and a former EMA Committee for Advanced Therapies member now working the policy side. Moderated by Robert Allen (Dark Horse Consulting), the panel featured Jacob George (Chief Medical and Scientific Officer, MHRA), Sean Russell (Head of Regulatory Affairs, Fondazione Telethon), and Esteve Trias (ATMP expert, former CAT member).

What came out of it wasn't reassurance. It was a working diagnosis of exactly where Europe is blocking itself, and what the next twelve months would need to look like to fix it.

The UK is moving. The question is whether Europe can keep up

George opened with a genuinely useful update on where the MHRA's flagship principles, technology neutrality, risk proportionality, and patient centricity, actually stand in practice, not just in policy documents. The AI and machine learning oversight framework is live. The mRNA consultation has closed and guidance is expected within weeks. The rare disease platform framework has also just closed consultation.

None of that is abstract. George pointed to real numbers behind it: the MHRA's aligned pathway with NICE, launched in March 2026, puts the health technology assessment body in the room from the very start of development, so the questions developers get asked are the ones that actually matter for adoption later. The ILAP pathway goes further still for genuinely paradigm-shifting therapies, bringing the whole healthcare system in as early as Phase 1.

The clearest proof point: Casgevy (exagamglogene autoleucel) was approved by both the MHRA and NICE, and the UK commissioned its first treatment centre within two days. That's not a regulatory win in isolation, it's a regulatory-plus-adoption system working in lockstep.

So what's actually blocking Europe?

Asked directly for the single biggest blocker to European execution, all three panellists gave different answers, and together they build a fuller picture than any one of them alone.

Russell's answer was structural: Europe is supposed to have harmonised clinical trial regulation, but individual member states apply it differently in practice. Some use the 50-day extension mechanism, some don't. The result is unpredictability that developers have to plan around country by country, inside a bloc that's meant to be a single market.

Trias went further, describing a specific and genuinely striking example of internal fragmentation: hospital exemption. In principle, hospital exemption data could act as a bridge toward centralised authorisation, a low-cost proof of concept before a full submission. In practice, the EMA won't accept it, because the way hospitals collect and manage that data isn't standardised across member states. The tool that could reduce fragmentation is itself a casualty of fragmentation.

George's answer looked outward rather than inward: the UK's own biggest constraint isn't the regulator, it's healthcare adoption. Approving a therapy quickly means little if the wider system, hospital capacity, workforce, finance, isn't ready to actually deliver it. That's precisely why Scotland runs a Horizon Scanning Advisory Board, chaired by George, bringing NHS planners, finance directors, clinicians and patients to the same table before a drug even reaches the market.

The uncomfortable comparison nobody in Brussels wants to hear

Russell was blunt about where the real competitive pressure is coming from, and it isn't just the UK's post-Brexit flexibility. It's China's speed in generating clinical data, and an increasingly aggressive US approach to opening INDs. Measured against both, Europe isn't just slow on its own terms, it's falling behind on relative terms too.

Trias added a pointed anecdote from his time observing the FDA-EMA exchange programme: an FDA colleague was reportedly shocked that the EMA doesn't pick up the phone and ask a company directly why they've chosen a particular development approach, rather than expecting the company to divine the "correct" interpretation of a regulation on its own. His conclusion, stated plainly: Europe needs to listen more, and listen better.

George's response to the China question was carefully calibrated but unambiguous: geography isn't the arbiter of data quality. The UK is willing to work with strong preclinical data regardless of where it originates, provided the quality and veracity hold up. But he was equally clear that the UK isn't trying to out-scale China on volume. Its competitive bet is speed, flexibility, and the calibre of its regulatory science, not sheer capacity.

Conditional approval: the right tool, with real strings attached

The panel spent real time on conditional marketing authorisation, and the nuance here matters more than the headline concept.

George described the MHRA's investigational marketing authorisation for rare diseases, which combines clinical trial approval and marketing authorisation into a single mechanism. It works, he was careful to stress, only because it's paired with genuinely rigorous post-marketing pharmacovigilance: without careful ongoing data collection, conditional approval isn't actually lower risk, it's just risk deferred.

Russell backed the concept but flagged the operational reality: resourcing, sustained data collection, and a workable pricing framework all need solving, or the burden lands disproportionately on whichever company or regulator is least equipped to carry it. Trias added a sharper, more commercially uncomfortable point: conditional authorisation can look, to investors, like an admission that the therapy isn't really approved yet. For companies whose valuation depends on a clean regulatory outcome, "conditional" is sometimes the opposite of what they're hoping to hear.

The UK's answer to the resourcing problem is concrete: a £600 million commitment to Health Data Research Service UK, intended to bring healthcare and social care data together under one infrastructure, so the burden of proof doesn't fall unevenly on individual developers.

Approval isn't the same as access

Trias raised a point that's easy to overlook amid all the regulatory-speed conversation: Europe's real access problem isn't only about therapies that never get authorised. It's about therapies that are authorised and never actually reach patients in a given country, because a company decides that market isn't commercially worth the pricing negotiation or the pharmacovigilance infrastructure required to operate there. A regulatory win on paper and a patient access win in practice are two entirely different milestones, and Europe's fragmented national systems mean plenty of therapies clear the first hurdle and stall at the second.

The chessboard, and why "one region at a time" thinking fails

Russell's central metaphor for the whole session was a chess analogy, and it's one worth remembering: a developer isn't playing one game of regulatory chess. They're playing two, three, or four simultaneously, against different regulators, and the smart move in one game is knowing how to leverage it into the next, rather than playing each board to conclusion in isolation before starting the next one.

His practical advice for founders followed directly from that: don't walk into a regulator meeting with an open book asking to be told what to do. Work backwards from where you ultimately want your therapy to reach, which regions and which patients, and use that to decide the right questions to ask, and the right regulator to ask them of, at each stage.

George's advice sat comfortably alongside this rather than against it: engage early, but engage at the genuine pinch points and decision-critical moments in development, not as a substitute for having a plan. Trias, only half-joking, pushed back on the idea of going to a regulator too early at all: plan your risk mitigation and key control points first, then bring the regulator in to negotiate the detail, rather than treating the regulator as a free consultancy for a plan that doesn't exist yet.

The synthesis, from all three: begin with the end in mind. If a therapy will be genuinely difficult to administer or monitor once approved, that's a problem to solve at the design stage, not a surprise to discover after launch.

Real-world data, open platforms, and the optimism nobody quite trusted

Trias made a case for two structural fixes with real momentum behind them, even if none of the panel expected either to be finished within the next year. The first is greater use of real-world data and digital twin methodologies in place of repeating full head-to-head trials against gold-standard treatments, something the UK already permits more readily than the EU. The second, more ambitious, is the idea of open technological platforms: shared, non-competitive infrastructure (distinct from any single company's proprietary platform) that could let developers prove a therapy's core technology once, then compete only on what's genuinely novel, lowering cost, timeline, and investor risk all at once.

Where the panel expects Europe to actually be in 12 months

Asked to look a year ahead, the answers ranged from measured to genuinely optimistic, but none pretended the fundamentals will have shifted by then.

Russell was the most cautious: Europe will likely have a clearer direction on its pharmaceutical legislation reform, but nothing will be fully implemented or active. He expects the UK to generate more concrete case studies in that window, China's landscape to sharpen further, and the US IND pathway to get more aggressive still, leaving Europe, in his words, still in "the slow burn."

Trias was more hopeful, pointing to regulatory sandboxes as a genuinely underused tool that could let regulators and developers test proposed changes in practice rather than debating them on paper indefinitely. George was confident specifically about UK momentum: decentralised, modular manufacturing frameworks have already generated "tremendous" industry interest, and the constraint now is resourcing the rollout, not appetite for the idea.

Why this session matters beyond the room

What made this panel land wasn't consensus, it was three people with genuinely different vantage points (sovereign regulator, EU-facing developer, former EMA insider) converging on the same underlying diagnosis: Europe's problem isn't a lack of good intentions, it's execution fragmented across borders that were never meant to fragment it. The therapies aren't the bottleneck. The rulebook is.

That's exactly the kind of unfiltered, cross-perspective conversation that's hard to get anywhere outside this room, and it's a preview of the depth on offer at Advanced Therapies Europe 2027.

Want the rest of the story? Register your interest in attending Advanced Therapies Week 2027 and be first to hear when sessions and speakers go live.