The First JCA Didn’t Assess a Drug. It Documented an Absence.

…Europe’s first Joint Clinical Assessment did not lower the bar. It published the gap.

On 9 June 2026, the European Commission published the first Joint Clinical Assessment under the EU HTA Regulation — one of eighteen initiated since the rules began applying in January 2025. The framing was celebratory: a milestone, a single European evaluation, less duplication, faster access, a more predictable process for companies. After years of preparation, joint EU health technology assessment had become operational.

Read the actual report, and a different story appears.

The subject was tovorafenib — Ipsen’s Ojemda, an orphan medicine for children with low-grade glioma carrying a BRAF alteration, who have progressed after previous treatment. The assessment was led by Ireland’s NCPE with Germany’s IQWiG as co-assessor. And what it produced is not, in any meaningful sense, an assessment of how well the drug works. It is a forensic, twenty-page record of how little evidence exists to say.

The first Joint Clinical Assessment did not grade a medicine. It graded an absence.

Start with what the assessment demanded. The scope defined three patient populations and eight separate PICOs — eight distinct combinations of population, comparator and outcomes the company was expected to address. The comparators ranged from single chemotherapies to combination regimens to “individualised treatment comprising multiple options,” a deliberately moving target. The outcomes list was longer still: overall survival, progression-free survival as a time-to-event endpoint and as six- and twelve-month rates, objective response defined under two separate criteria sets, multiple quality-of-life instruments, symptom composites, fatigue, duration of response. A complete comparative picture, specified in detail.

Now consider what arrived. Of the eight PICOs, results were included for exactly one. For four of them, the company submitted no comparator data at all. For two more, the same — nothing. For the last, data were submitted and then excluded, because the information on the comparator study was insufficient to assess.

Seven of eight questions came back empty.

This is not a report about a drug. It is a report about a gap.

The single PICO that survived — tovorafenib versus the combination of dabrafenib and trametinib, in the subgroup with a BRAF V600E mutation — does not rescue the picture. The pivotal trial, FIREFLY-1, is a single-arm, open-label, non-randomised study. It contains no comparator. So there is no head-to-head evidence, and no connected network of trials from which an anchored indirect comparison could be built.

What remained was an unanchored matching-adjusted indirect comparison: the company’s single-arm data, statistically reweighted to resemble the population of a separate published study. It is the weakest standard form of comparative evidence there is, and the assessors said so. After reweighting, the effective sample size across the analyses ranged from 5.81 to 14.26 — a reduction of between 34% and 52% from the original numbers. Europe’s first comparative-effectiveness verdict, where it could be reached at all, rests on the statistical equivalent of a handful of children.

And the assessors did not wave it through. They rejected one response analysis outright for using methods they called flawed and inappropriate. They discarded the company’s own sensitivity analysis because it contained errors and reached inaccurate conclusions. They flagged inadequate justification for excluding available patient data, and a risk of bias from selective reporting of outcomes. The tone is unmistakable to anyone who has read German benefit assessments: this is IQWiG’s hand, applied at European scale.

The scope asked for a library. The dossier delivered a footnote. And the assessors annotated the footnote in red.

It would be easy to read this as a failure — of the company, of the drug, or of the JCA. It is none of those things, and the misreading is dangerous, because it hides the lesson.

The evidence gap here is structural, not negligent. Tovorafenib treats a rare cancer in children. Randomising those children against an active comparator is difficult and, in places, ethically fraught; single-arm accelerated approval exists precisely for situations like this. The absence of robust comparative data is a feature of orphan oncology, not a sign that anyone cut corners. The drug may well help the patients it was built for.

But the JCA does not care why the evidence is thin. It asks the comparative question regardless, and it publishes the answer regardless — including when the answer is that the data cannot tell us. That has three consequences every market access and evidence team should internalise now.

First, the JCA is an evidence-exposure machine, not an evidence-friendly one. Whatever gap exists between the questions it asks and the data you hold, it will find, document, and publish — in a permanent report endorsed by every member state. The comfortable assumption that a single European assessment means a lighter touch is exactly backwards.

Second, the bar is the toughest national standard, not the average. The first report reads like an IQWiG assessment because IQWiG co-wrote it. Companies accustomed to a more forgiving national body have just been upgraded to German-grade methodological scrutiny — and not in one market, but across the entire bloc at once. The methods that passed in the gentlest jurisdiction will not pass here.

Third, and most actionable: the evidence that satisfies the JCA cannot be retrofitted. By the time a product reaches assessment, the comparator was chosen years earlier, the trial was designed years earlier, the outcomes and response criteria were fixed years earlier. HTA cannot be reverse-engineered at the end of development — and the JCA scope, with its eight PICOs and its demand for the right comparator in the right population under the right criteria, is now the specification you build toward from the first protocol. Design for it late and you will produce, at best, an unanchored comparison on fourteen patients, and a report that says so.

The first Joint Clinical Assessment was billed as proof that European HTA can move faster. What it actually proves is that European HTA can now see, in one place, exactly where the evidence runs out.

The JCA is not a gentler assessment. It is the most demanding one, made continental.

Share

Beyond Approval publishes weekly strategic intelligence on market access, HTA, pricing, and regulatory shifts that reshape how medicines reach patients.