The Joint Clinical Assessment (JCA) changes the question behind evidence planning. The starting point used to be whether the available evidence could support a sufficiently strong HTA submission. It is now whether companies are generating the evidence that assessors are likely to need.
As of September 2026, the European Commission has published its first four JCA reports: tovorafenib for paediatric low-grade glioma, lurbinectedin and tarlatamab, both for extensive-stage small cell lung cancer, and onasemnogene abeparvovec, a gene therapy for spinal muscular atrophy. The sample is small, yet it covers a single-arm case, a direct RCT case and a broad multi-PICO case, which is enough to test assumptions that have shaped evidence planning for a decade.
Each report speaks to one of them. Tovorafenib, a rare paediatric tumour drug with a single-arm design, met only one of eight PICOs with supported comparative evidence, and its MAIC was challenged on small effective sample size, residual confounding and statistical uncertainty. Rarity, in other words, intensifies the scrutiny of indirect methods. Lurbinectedin met its single PICO with a direct RCT, the strongest design available, and assessors still pressed on certainty of evidence, maturity of outcomes and trial representativeness. Tarlatamab reached full coverage of seven PICOs through a mix of RCT, network meta-analysis, anchored and unanchored MAIC, and all were accepted because each was appropriately justified. For onasemnogene abeparvovec, Ireland’s NCPE completed its Rapid Review before the JCA report was public and stated that the report had informed its assessment, so the two levels of assessment do not have to run in sequence.
The practical consequence is that evidence has to be built into trial design years in advance. The pivotal programme must answer the comparative question as well as the efficacy question, anticipate a set of comparators rather than a single one, and run alongside the regulatory submission, with the assessment dossier drawing on evidence already in hand. Documentation and timing belong to readiness as well: four JCA were halted as of September 2026 because of incomplete dossiers and missing evidence. The JCA also brings fixed assessment timelines, a single evidence lock at dossier submission and the need to coordinate functions and EU affiliates. Meeting that model calls for centralised governance, portfolio-level planning and standardised dossier development.
For this reason readiness cannot remain a submission-stage activity. It rests on four capabilities: anticipating evolving JCA expectations, aligning evidence priorities across Clinical Development, Regulatory, HEOR/RWE, Medical and Market Access, orchestrating gaps and submission readiness, and learning from assessor feedback. Sustaining this across Member States, assets, indications and changing PICO scenarios is a scale problem that manual processes are unlikely to handle, so advanced analytics and AI become enabling tools. They also become a subject of assessment: on 15 July 2026 the HTACG adopted principles for the use of AI in JCA dossiers, built around accountability, transparency and traceability. A governed AI capability is therefore part of readiness itself.
The paper describes five connected capabilities that digital platforms and AI can support: continuous JCA intelligence, PICO anticipation and scenario modelling, dynamic evidence-gap management, AI-assisted synthesis and submission readiness, and enterprise learning. Together they shift preparation from a reaction to the submission date to a continuous, data-driven process across the product lifecycle.
Four strategic imperatives follow for pharmaceutical companies. Planning moves upstream into trial design, since pivotal design, comparators and endpoints are often locked three to five years before submission. Evidence planning is multifunctional from the outset, with clear governance and ownership in place of sequential handoffs. Portfolio-level planning is embedded well before submission. Technology is built into the evidence system with oversight and auditability. Access is now decided by choices made long before anyone files a dossier, and the advantage goes to the companies that build evidence early and own it together.
