The JCA has reduced some duplication in clinical assessments across the EU, but in reality, it has also made early evidence planning more complex and timepressured for companies. Different member states ask slightly different questions about the same product, which leads to multiple versions of what evidence is needed. In some cases, a single product can face more than 10 different PICOS, and companies may only have a fraction of the evidence needed available when receiving them. This gap creates pressure to generate new data very quickly. Companies may need to rush real-world evidence and run additional analyses under tight timelines. This can stretch budgets and sometimes affect the quality of the work. Because of these challenges, some companies are even considering avoiding the JCA process for now, choosing to focus on individual country routes or delay EU engagement until the situation becomes clearer. The Joint Scientific Consultation can help by aligning expectations early, but access is limited and not all companies can get a slot when they need it. It is also important to remember that JCA results are not binding, which means countries can still make different decisions on pricing and reimbursement later. To manage this, companies should plan early by mapping likely evidence needs across key countries, seek joint advice where possible, and set aside extra time and budget for additional analyses. Preparing strong materials early also increases the chance of getting valuable scientific advice.

- Mia Malmenäs MSc, EMBA, Senior Principal, Insights, Evidence & Value – Health Economics & Epidemiology, ICON

The early promise of Joint Clinical Assessment is real, but it would be overstating things to say it has solved duplication across Europe. What it has done is create a more coherent starting point: one clinical evidence package, assessed once at EU level, rather than repeated in parallel across multiple markets. That matters, and it is a meaningful structural shift. But the reality on the ground is likely to be more nuanced. National HTA bodies will still be interpreting evidence through their own lenses, applying local comparators, subpopulation logic and decision frameworks that reflect domestic priorities rather than a single European view. In practice, that means the burden has not disappeared so much as moved. The visible duplication in clinical assessment may be reducing, but the real complexity now sits in PICO strategy and national adaptation. Manufacturers are still having to plan for multiple evidence pathways, because a JCA that works at EU level does not automatically land cleanly in Germany, France or elsewhere. So yes, JCA is moving Europe towards greater convergence in evidence generation. But when it comes to interpretation, value and access decisions, we are still very much in a decentralized world.

- Esther Nzenza, Senior Partner and Global Head, Global Value and Access, Acumetis

It is honestly too soon to tell. The first JCA report on Ipsen’s tovorafenib has been endorsed by the HTA Coordination Group, but (at the time of writing) we are awaiting completion of the European Commission’s procedural review before the dossier and report will be published. There have been developments in how local processes are being adapted to meet the requirements of the EU HTA Regulation, including the recent approval of the Spanish Royal Decree on HTA which prohibits the duplication of clinical evaluations for technologies that have undergone JCA. We also know that Germany has started its own early benefit assessment of tovorafenib ahead of publication of the JCA report. Ultimately, only after local HTA reports are published will we be able to gain a clear understanding of how member states are using JCA dossiers and reports within their own pricing and reimbursement decision-making processes.

- Suzette Matthijsse, Vice President, Strategic Advisory Services, HEOR at Lumanity

The CMS Model GENEROUS aims to align U.S. Medicaid drug prices with international MFN levels.¹ While the program is voluntary, early analyses indicate that the model forces manufacturers to fundamentally rethink their global launch sequences and domestic rebate strategies.

The threat to global launch strategies²
The GENEROUS model pegs U.S. Medicaid prices to the second-lowest GDP-adjusted net price among a reference basket of G7 countries, plus Denmark and Switzerland. For drugs with a high Medicaid patient mix (e.g. 35% of total U.S. volume), matching these lower international prices domestically would severely erode overall revenue. In general, the tipping point for withdrawing from European markets often occurs when a drug's Medicaid volume exceeds 30%, making the U.S. domestic channel too valuable to discount. To preserve lucrative U.S. profit margins, manufacturers may increasingly choose to delay or cancel product launches in the benchmark European countries entirely. By withdrawing from reference markets, manufacturers can maintain higher list prices in the U.S. and "Rest of World" markets.

The "all-or-nothing" portfolio dilemma³
Manufacturers cannot cherry-pick which drugs to subject to the model. Participation requires offering MFN pricing across their entire portfolio of covered outpatient drugs. Manufacturers must conduct complex, portfolio-wide revenue modeling before volunteering. While a company might benefit from the model's standardized, streamlined coverage criteria for some of its drugs, it could lose out significantly if forced to offer MFN pricing on its entire portfolio. The true value of this standardization depends heavily on each asset's competitive landscape.

References:
1. CMS GENEROUS Model, GENErating cost Reductions for U.S. Medicaid Model, Request for Applications from Applicable Manufacturers Version: 2, December 23, 2025.
2. Charles River Associates. Potential impact of the CMS GENEROUS model. April 2026, accessed at: Potential impact of the CMS GENEROUS model | Charles River Associates, on May 22, 2026.
3. Hogan Lovells. CMS announces voluntary 'GENEROUS' Medicaid most-favored-nation drug pricing model. November 2025, accessed at: CMS announces voluntary “GENEROUS” Medicaid most-favored-nation drug pricing model, planned to launch in 2026, on May 22, 2026.

– Lisa Cashman, Vice President, Market Access Strategy at Lumanity

Agentic AI can create clear value in two areas: getting work done faster and helping teams think ahead. Right now, the biggest and most reliable benefit is in HTA dossier and evidence synthesis work. AI can speed up writing, pull together relevant research, check facts against sources, and make outputs more consistent. This saves time and reduces manual review, so teams can focus on more strategic tasks. Using AI to predict payer reactions and shape access strategy is promising, but the value is less consistent. It depends heavily on having high-quality data and well-tuned models, and it can struggle with the real-world complexity of payer decision-making. That’s why having people involved is still essential. Humans are needed to interpret unclear or incomplete evidence, question AI outputs, apply judgment on ethical, HTA, and regulatory issues, and build relationships with stakeholders. Most importantly, people ensure the final story is credible and trusted, something AI alone cannot guarantee. The best results come from combining both. Using AI to handle speed and scale, while relying on human expertise for judgment and decision-making. To make this work, companies should set clear rules, check AI outputs against real outcomes, and track impact such as time saved and error reduction. Start with lower-risk uses like dossier and report automation, then expand carefully. In short, AI adds value, but people ensure trust and quality.

- Mia Malmenäs MSc, EMBA, Senior Principal, Insights, Evidence & Value – Health Economics & Epidemiology, ICON

In HTA submissions and HEOR modelling, we see the highest near-term ROI from agentic AI in the speed and efficiency of dossier generation. The impact is particularly visible in evidence synthesis: AI-supported workflows can materially accelerate SLR development, data extraction, targeted deep dives, review of clinical and economic materials, and preparation of first-draft summaries. Similar gains are emerging in modelling, where automated model builders (like our Acumetis Model Builder) and AI-enabled tools can support model shells development, scenario generation, documentation, and parts of quality control. In practical terms, AI is helping teams move faster from evidence identification to a submission-ready narrative. The predictive depth of access strategy is also improving, but we see this more as an enhancement of expert judgement than a replacement for it. AI can rapidly review previous HTA decisions, summarize payer pushback, identify recurring objections, and extract assumptions that have been accepted in prior appraisals. This gives teams a stronger and faster starting point for anticipating challenges around comparators, endpoints, extrapolation, uncertainty, indirect comparisons, and model assumptions. However, human-in-the-loop remains non-negotiable where judgement, accountability and strategy are required. This includes deciding when to follow precedent and when to depart from it, validating whether a model structure is clinically and economically credible, interpreting uncertainty, and shaping the value story for a specific HTA body or payer. This is especially important in rare diseases, advanced therapy medicinal products, indications with crowded treatment landscapes, and other areas where precedent may be limited, immature, highly contested, or difficult to apply directly to a new product. Many of the most important access decisions are also shaped by context that is not fully visible in public documents: prior negotiations, stakeholder expectations, local clinical practice, political sensitivity, and the specific risk appetite of a payer or committee. AI can synthesize what is written down, but experienced HEOR and market access teams are still needed to interpret what it means and decide how to act. So, the clearest ROI today is in accelerating dossier development and evidence synthesis. The strategic and predictive value will continue to grow, but only when AI outputs are combined with human expertise, methodological discipline and payer-specific judgement.

- Ewa Dlotko, Technical Director, Acumetis

We do not view speed of HTA dossier generation and predictive depth of access strategy as an either/or. The underlying data and knowledge required for predictive depth also enhance the generation of “first draft” HTA dossiers. The highest ROI comes when they are designed as complements on top of the same governed evidence foundation. This enables the same evidence foundation to both accelerate drafting/updates and improve foresight regarding where payers and HTA bodies will push back. In the near term, the most readily measurable ROI tends to show up in dossier speed, while the biggest strategic opportunity often comes from stronger prediction once you have sufficient data maturity. “Human-in-the-loop” remains non-negotiable throughout. Human expertise is critical in identifying the optimal approach, exercising scientific and strategic judgment, ensuring cross-functional alignment, implementing rigorous data validation/governance, and refining the narrative and final argumentation to ensure credibility with external reviewers.

Dan Gladwell, PhD, Senior Vice President & Chief Scientific Officer, HEOR at Lumanity