Three JCA dossiers were discontinued. What was missing?

The European Commission’s tracker lists four discontinued Joint Clinical Assessments. One, for sasanlimab, stopped because the developer withdrew its marketing authorisation application. The other three were discontinued because their dossiers still did not meet the required standard after a second request for information.[1]

The medicines were catequentinib, Tacquell and sintilimab.[1]

This does not mean the medicines were found to be ineffective. It is not a refusal of marketing authorisation or reimbursement. The tracker states that these discontinuations do not affect the EMA’s assessment or its timeline.[1]

The consequence is that no endorsed JCA report is produced for national HTA bodies to use. National assessments can still proceed, but developers lose the chance to enter reimbursement discussions across Member States with a shared EU clinical assessment already behind them.[2]

The JCA process stopped because the comparative evidence could not be fully assessed.

That distinction matters. Before comparative evidence can be persuasive, it must be assessable. An assessor must be able to find the evidence trail, understand why the studies can be compared and verify how the results were produced.

The three dossiers reached the same procedural outcome for different reasons. Together, they show that a JCA evidence package needs coverage, credibility and traceability.

A JCA can stop before the findings are assessed

The EU JCA Regulation sets a clear threshold for the dossier.

It must contain the complete and current evidence needed for the assessment scope. The methods must be appropriate for every research question. The dossier must also be structured transparently and include the underlying documents needed to verify its analyses.[2]

If the dossier does not meet these requirements, the European Commission asks for the missing information. If the updated dossier still falls short, the JCA must be discontinued.[2]

This happens before the evidence can do its intended job. Assessors cannot judge a treatment comparison if they cannot establish which studies were found, which PICO the analysis answers, whether the studies are sufficiently comparable or how the result was calculated.

PICO stands for population, intervention, comparator and outcomes. Each PICO is a specific question about which patients receive the medicine, what it is compared with and which outcomes matter.

Catequentinib did not show a complete route through every PICO

The catequentinib dossier covered two proposed indications: synovial sarcoma and leiomyosarcoma.[3]

After the Commission’s second request, the updated dossier still did not provide a transparent and comprehensive account of all the relevant evidence. Search methods, study selection and reasons for excluding evidence were incomplete or unclear across several sources.[3]

The assessment scope and its PICOs were not reproduced as requested, and missing results were not adequately explained.[3]

For synovial sarcoma, no relative effectiveness or safety evidence or analyses were provided for the PICOs in the scope. For leiomyosarcoma, some evidence was present, but important study descriptions, certainty assessments and sensitivity-analysis methods remained missing.[3]

The Commission asked why feasibility assessments for possible indirect comparisons had not been conducted and for that decision to be justified.[3]

This does not establish that a credible indirect comparison was possible. It means the assessors did not receive the explanation and supporting evidence needed to evaluate why that route had not been pursued.

The problem began before the statistical model. The full evidence landscape had not been made visible.

Tacquell submitted comparisons that could not be fully checked

Tacquell showed a different problem.

Information retrieval was missing for five PICOs. For other PICOs, comparative analyses had been submitted, including an unanchored external comparison.[4]

An unanchored comparison uses evidence from separate studies without a shared randomised control group connecting them. This makes differences between the studies particularly important.

The Commission identified missing methods, relative-effect estimates, Kaplan-Meier curves, safety results, information about missing data and sensitivity analyses for key PICOs. It also found that there was not enough supporting material to assess certainty or inspect parts of the risk-of-bias assessment.[4]

The program code for the unanchored comparison was missing.[4]

An analysis can exist and still be unassessable. A final estimate does not show which data were used, which assumptions were made, how uncertainty was handled or whether another analyst could reproduce it.

Sintilimab relied on studies whose comparability could not be assessed

Every sintilimab PICO relied exclusively on indirect comparisons.[5]

That made the studies behind those comparisons central to the whole dossier. Yet the submitted information did not adequately describe the studies, interventions, treatment courses, outcome definitions, follow-up or imaging schedules.[5]

Important study-level results feeding the comparisons were also missing. This included information about censored patients in the ORIENT-11 Kaplan-Meier curves. A patient is censored when their exact event time is unknown, often because they had not experienced the event by their last follow-up or were no longer being followed.[5,6]

Search reporting was incomplete too. A ClinicalTrials.gov search was missing, searches of national HTA websites were incomplete, and some search strategies and electronic search records had not been supplied.[5]

Assessors therefore could not judge whether the contributing studies were sufficiently comparable. This includes study design and patient characteristics that affect outcomes regardless of treatment or change the relative treatment effect. The technical term for this is exchangeability.

Because all the PICOs depended on indirect comparisons, the Commission concluded that sintilimab could not be assessed against any comparator.[5]

The Commission noted that much of the missing study information was available in the ORIENT-11 clinical study report or comparator publications.[5]

Evidence can exist and still be functionally missing from a JCA if assessors cannot find it where they need it, connect it to the analysis and understand how it supports the result.

Every PICO needs an evidence route or a documented gap

The JCA dossier guidance anticipates that some PICOs may not have a comparative result. It asks developers to identify those PICOs and explain why the result is unavailable.[7]

That means every PICO needs a documented route:

• a direct head-to-head comparison;
• an anchored indirect comparison through a shared control;
• a potentially feasible unanchored or external comparison;
• descriptive evidence that cannot support a relative-effect estimate; or
• no credible comparison with the available evidence.

The last category can be the correct conclusion. Our review of the Itvisma JCA showed five PICOs for which the assessors accepted that the available evidence could not support a credible comparison.

But a documented search that finds no eligible evidence is different from not reporting the search. One demonstrates the gap. The other prevents assessors from knowing whether the gap is real.

The evidence search is part of the comparison

Indirect comparisons are often treated as a statistical task that begins once the comparator studies have been selected.

The JCA requirements show why that is too late.

Comparator studies must be identified systematically. Plausible studies connecting the evidence network should be considered. Their populations, treatments, outcomes, follow-up and potential treatment-effect modifiers must then be compared.[7,8]

A treatment-effect modifier is a patient or study characteristic that changes how well one treatment works relative to another.

This work determines whether an indirect comparison is possible and which method the evidence can support. It also reveals where the evidence bridge breaks.

We explain the methods themselves in our field guide to indirect treatment comparisons.

Advanced methods cannot create missing evidence

Methods such as matching-adjusted indirect comparison can adjust for some measured differences between studies. They cannot adjust for a variable that was not measured or reported. They cannot create overlap between patient populations or make incompatible outcomes equivalent.[9,10]

This is particularly important for unanchored comparisons. Without a shared randomised control group, the analysis must account for all important differences that affect prognosis or modify the treatment effect. In practice, some of these differences may be unknown or unavailable.[9,10]

A more sophisticated model cannot repair an incomplete search, missing study characteristics or absent patient variables.

The method comes after feasibility. First establish what data exist and whether the studies can credibly answer the same question.

Code and supporting files are part of the evidence

The supporting appendices are easy to treat as an administrative task. They are not.

The JCA dossier must provide enough material to verify the work. Depending on the analysis, that may include:

• complete search strategies and electronic search records;
• study-selection logs and reasons for exclusion;
• protocols and statistical analysis plans;
• clinical study reports and study-level input data;
• risk-of-bias assessments and their supporting material;
• software details, program code and output; and
• diagnostics and sensitivity analyses.[2,7,8]

The Annex 1 tables specify detailed inputs for direct and indirect comparisons. For time-to-event outcomes, these include sample size, events, censored patients, median estimates, hazard ratios and their confidence intervals where relevant.[6]

These files allow assessors to trace a result back through the model, the studies and the original search.

Traceability is part of the evidence package.

The second request gives 15 days, not a second evidence programme

The standard deadline for the initial dossier is generally 100 days after the Commission’s first request.[8,11]

After a second request, the default response period is 15 days. It is shorter for accelerated marketing authorisation procedures, certain variations and cases where only minor information is missing.[8,11]

That may be enough to correct an omission when the work already exists and is organised.

It is not a sensible point to begin systematic searches, rebuild evidence networks, resolve incompatible data, design several new analyses and assemble reproducible code.

The second request should be treated as a correction window, not an evidence strategy.

Build the dossier around coverage, credibility and traceability

These cases suggest a practical readiness check before submission.

Coverage: Has every required source been searched? Can every PICO be linked to direct evidence, an indirect evidence route or a clearly justified gap?

Credibility: Can the studies support the proposed comparison? Are the method, assumptions, diagnostics, uncertainty and sensitivity analyses defensible?

Traceability: Can an assessor follow the evidence from the search to the study, PICO, model and result? Are the underlying documents, code and outputs available?

This should be designed as one evidence chain. It should not be divided into a literature review, an indirect comparison and a dossier-writing exercise that meet for the first time near submission.

Evidax TRACE maps the published evidence landscape, comparator studies, outcomes and potential treatment-effect modifiers with expert oversight and a complete audit trail. Evidax Comparison Blueprints then establish which comparisons the available evidence can support, which method is defensible and what assumptions, diagnostics and supporting materials the analysis will require.

The purpose is not to force an analysis into every PICO. It is to know which questions can be answered, which cannot and what needs to change while there is still time to act.

Take home

• A JCA can be discontinued before its comparative findings can be fully assessed.
• Every PICO needs a documented evidence route or a transparent explanation of the gap.
• An indirect comparison needs assessable studies, defensible assumptions and clear uncertainty analyses.
• Search records, study inputs, code and supporting documents are part of the evidence package.
• Complex methods cannot create missing studies, patient variables or population overlap.
• The second request is a short correction window. Assessment readiness must begin before the formal dossier period.
• Discontinuation is procedural. It is not evidence that the medicine is ineffective or a refusal of reimbursement.

References

1. European Commission. Updated list of joint clinical assessments. News announcement published 10 September 2026; workbook updated 21 September 2026; discontinued-JCA data extracted 17 September 2026. Available from: https://health.ec.europa.eu/latest-updates/updated-list-joint-clinical-assessments-2026-09-10_en
2. European Parliament and Council of the European Union. Regulation (EU) 2021/2282 of 15 December 2021 on health technology assessment and amending Directive 2011/24/EU. Official Journal of the European Union. 2021;L458:1-32. https://eur-lex.europa.eu/legal-content/EN/TXT/?uri=CELEX:32021R2282
3. European Commission. Information, data, analyses and other evidence missing from the updated dossier of the joint clinical assessment of the medicinal product catequentinib. JCA-MP-2024-03. 8 May 2026. https://health.ec.europa.eu/document/accbe9bd-f522-4248-8692-bca8ed38ff4d_en
4. European Commission. Information, data, analyses and other evidence missing from the dossier of the joint clinical assessment of the medicinal product Tacquell. JCA-MP-2025-02. 13 February 2026. https://health.ec.europa.eu/document/3331e7f8-6dba-493b-b3ab-d21a692f5034_en
5. European Commission. Information, data, analyses and other evidence missing from the updated dossier of the joint clinical assessment of the medicinal product sintilimab. JCA-MP-2025-08. 15 September 2026. https://health.ec.europa.eu/document/2b1ab18c-d3a2-4b7d-a929-5f75e8ba9021_en
6. Member State Coordination Group on Health Technology Assessment. Guidance on filling in the joint clinical assessment dossier template: Annex 1, table template collection. Adopted 28 November 2024. Available from: https://health.ec.europa.eu/publications/guidance-filling-joint-clinical-assessment-jca-dossier-template-medicinal-products_en
7. Member State Coordination Group on Health Technology Assessment. Guidance on filling in the joint clinical assessment dossier template: medicinal products. Version 1.0. Adopted 28 November 2024. https://health.ec.europa.eu/publications/guidance-filling-joint-clinical-assessment-jca-dossier-template-medicinal-products_en
8. European Commission. Commission Implementing Regulation (EU) 2024/1381 of 23 May 2024 laying down procedural rules for joint clinical assessments of medicinal products for human use at Union level and templates for those assessments. Official Journal of the European Union. 2024. https://eur-lex.europa.eu/eli/reg_impl/2024/1381/oj
9. Member State Coordination Group on Health Technology Assessment. Methodological Guideline for Quantitative Evidence Synthesis: Direct and Indirect Comparisons. Adopted 8 March 2024. https://health.ec.europa.eu/publications/methodological-guideline-quantitative-evidence-synthesis-direct-and-indirect-comparisons_en
10. Member State Coordination Group on Health Technology Assessment. Practical Guideline for Quantitative Evidence Synthesis: Direct and Indirect Comparisons. Adopted 8 March 2024. https://health.ec.europa.eu/publications/practical-guideline-quantitative-evidence-synthesis-direct-and-indirect-comparisons_en
11. Member State Coordination Group on Health Technology Assessment. Procedural guidance for JCA medicinal products. Version 1.0. Adopted 28 November 2024. https://health.ec.europa.eu/publications/procedural-guidance-jca-medicinal-products_en