EU Joint Clinical Assessment after the first reports: How should companies interpret the findings?
At this stage, it remains too early to assume that these cases are representative of future JCAs. All three completed reports concern orphan medicines in rare cancers, and cover different technologies, development programmes and evidence settings. They are nevertheless useful because they demonstrate a range of findings that can emerge at the PICO level and raise important issues for national HTA, pricing and reimbursement.
A practical way to review progress is therefore to consider the case for each PICO rather than trying to assign one overall interpretation to each medicine. Some findings may support a relatively straightforward national assessment. Others may require additional analysis, context or a clearer explanation of why uncertainty should be given more or less weight in a particular national decision.
Three assessments in different evidence settings
The lurbinectedin (ZEPZELCA) assessment, published just this week, was arguably the most straightforward of the three assessments to date. It considered lurbinectedin in combination with atezolizumab as maintenance treatment for adults with extensive-stage small-cell lung cancer whose disease had not progressed following first-line induction therapy. The assessment contained one principal PICO and was supported by a direct randomised comparison with atezolizumab alone. [2]
The tarlatamab (IMDELLTRA) assessment, also published just this week, was broader. It also concerned extensive-stage small-cell lung cancer but this time covered seven PICOs. The evidence package included direct comparisons for three PICOS, network meta-analysis for 2 PICOs, an unanchored comparison (MAIC) for one PICO, and an anchored indirect treatment comparison (Bucher) for the remaining PICO. The evidential position therefore varied within the report according to the relevant population, comparator and analytical method. [3]
Tovorafenib (OJEMDA), the first JCA to be published, was assessed for paediatric low-grade glioma, and presented an altogether different challenge. The development programme was centred on a single-arm, open-label Phase II study. The assessment scope contained eight PICOs, but comparative evidence was submitted for only part of the scope, with an unanchored matching-adjusted indirect comparison used for one central PICO. Other PICOs were supported only by descriptive data or were not included in the relative assessment because suitable comparative information was unavailable. [4]
These differences are relevant when considering what the reports may tell us. A focused PICO supported by a relevant randomised trial is more likely to produce an interpretable estimate of relative effect than a PICO relying on disconnected studies, small subgroups or an external comparison. The first three reports therefore illustrate different evidence situations to guide what might be expected in different situations.
Classifying JCA outcomes
A practical way of assessing the JCA findings is to classify them according to the issues that are likely to surface at the national level. This helps distinguish between evidence bases requiring different priorities for national strategies and supporting analyses. We outline some practical ways to organise these:
- Evidence demonstrates relative benefit: Here, comparative evidence shows a credible improvement in relevant outcomes for a specific population and comparator. The national task is then to focus on translating that finding into the local treatment pathway, economic model and reimbursement case. The main questions are likely to concern clinical relevance, applicability and value rather than whether a relative effect has been shown at all.
- Evidence suggests directionally favourable outcomes, subject to uncertainty: Here, the evidence tends to favour the technology, but the result is weakened by imprecision, risk of bias, indirectness, small sample size, incomplete adjustment or other methodological limitations. The national strategy priority is likely to focus on explaining the source and consequences of the uncertainty, identifying whether further evidence is available that might otherwise be out of scope in the JCA, and showing whether the uncertainty is likely to alter the direction or magnitude of the effect.
- No evidence of relative improvement in outcomes: Comparative evidence may be available, but it does not show a clear advantage for the relevant outcome, comparator or population. This may lead companies when preparing national submissions to place greater emphasis on other parts of the evidence package, such as convenience, treatment burden, economic impact, subgroup relevance or outcomes not fully captured in the JCA relative-effectiveness analysis. The scope for doing so will depend on how national bodies weigh these factors and whether they consider them sufficiently supported.
- Not assessable via JCA methodology: In this case, the required comparison cannot be evaluated because evidence, analyses or supporting documentation are absent, incomplete or unsuitable for interpretation. This is different from showing no added effect. It indicates that the JCA could not reach a reliable comparative conclusion. At national level, this may require additional analysis, new evidence, a clear explanation of why the gap exists, or an argument that the comparison is not central to local decision-making.
These categories are likely to be more useful when applied at PICO level rather than to the product as a whole. The same medicine may have direct and persuasive evidence against one comparator, uncertain evidence against another and no assessable evidence in a relevant subpopulation. Tarlatamab is a good example of how the evidential position can vary within one report, depending on the comparator and the method used.
For companies, this suggests that the main output of a JCA should be read as a set of PICO-specific evidence positions. Each position is likely to require a different response in subsequent national submissions.
Discontinued assessments and the question of assessability
The two discontinued JCAs provide an additional learning. They show that in addition to the underlying evidence base, whether the dossier presents that evidence in a complete, transparent and assessable form is also critical.
For Tacquell, the assessors identified missing information retrieval for several PICOs, insufficient description and justification of analytical methods, missing underlying documentation, absent or incomplete sensitivity analyses, and inadequate presentation of relative-effectiveness and safety results. The requests also covered relative effect measures, methodological justification and documentation needed to assess the certainty of external comparisons. [5]
For Catequentinib, the missing elements included transparent identification and synthesis of the available evidence, correct presentation of the requested PICOs, explanations for omitted results, feasibility assessments for indirect comparisons, justification for departures from methodological guidance, and more complete reporting of study searches, selection and characteristics. [6]
It is therefore important to distinguish between a limited evidence base and an incomplete submission. Evidence may be limited because the population is small, because randomisation is difficult, or because practice differs between countries. These limitations may be inherent to the disease area or development programme. A dossier may also be incomplete because searches, methods, analyses or supporting documentation have not been adequately provided. That is a different issue and is more likely to be viewed as avoidable.
The practical implication is that companies need to provide a sufficiently complete account of what evidence exists, how it was identified, how it was analysed and where the limitations remain. A sparse evidence base can still be assessed if it is presented transparently. An incomplete evidence package may prevent the assessment from proceeding at all.
Designing the core evidence and anticipating the peripheral PICOs
For companies in late-stage development, it is not realistic to design a pivotal trial solely around the full anticipated JCA scope. Trial design needs to balance patient needs, different jurisdictions, regulatory requirements, the scientific and medical concerns, international standards of care as well as the needs of HTA bodies and payers.
The priority should generally be to generate the strongest feasible evidence for the core population in which the company expects the medicine to be used and reimbursed. This includes selecting a relevant comparator, measuring important clinical and patient outcomes, and ensuring that the trial can answer the main relative-effectiveness question.
Even if health technology developers knew what the scope would be at the time of designing the pivotal trial, it is unlikely that every JCA PICO could be addressed through that one trial alone. Some PICOs will reflect narrower subpopulations, different treatment sequences or comparators that are relevant in only a limited number of Member States. These more peripheral questions should still be anticipated, but they may need to be addressed through an evidence-synthesis strategy rather than through the pivotal trial itself.
Unanchored treatment comparisons
The first reports provide useful examples of the difficulties associated with unanchored indirect treatment comparisons. These methods are used where direct comparative evidence and a connected evidence network are unavailable. They may be necessary, particularly in small populations, but they require assumptions that are difficult to verify.
The tovorafenib report states that unanchored indirect comparisons have an “inherently higher risk of bias” than comparisons informed by randomised evidence. It also notes that validity depends on adequately adjusting for prognostic variables and effect modifiers, with the assumption that all relevant factors have been identified and measured being particularly demanding in practice. The report identified additional uncertainty because the information needed to assess some assumptions was incomplete. And it also noted uncertainty associated with small sample sizes and the reduced effective sample size after matching. These limitations do not mean that the analysis provides no information. They do mean that a favourable estimate may remain difficult to interpret as a reliable measure of the relative treatment effect.
However, the key issue from all of this, is whether the required assumptions can be supported well enough for the result to influence national decision-making. Where an unanchored comparison is likely to be central, the availability of relevant individual patient data, the identification of effect modifiers, the comparability of outcome definitions and the degree of population overlap need particular attention.
Orphan medicines and ATMPs are not a single evidence category
All three completed JCAs concern orphan medicines, yet they produced very different assessment situations. This suggests that orphan status itself does not determine whether the JCA will be difficult. Nor should an ATMP automatically be assumed to have an evidence base that is unsuitable for relative assessment.
The more relevant issue is whether good comparative data are available. Some orphan medicines may be supported by a suitable randomised trial and a relatively focused PICO, as illustrated by lurbinectedin. Others may depend on single-arm evidence, small subgroups or external comparisons, as in parts of the tovorafenib assessment.
The likelihood of falling into the “directionally favourable but uncertain” or “not assessable” categories may therefore be greater in rare diseases and some advanced therapies, but it is not inevitable. The difficulty arises where small populations, heterogeneous disease, a changing treatment pathway or the absence of an established standard of care prevent a reliable relative comparison.
There is also a reasonable question about how well a method designed to estimate relative effects can accommodate settings in which comparative evidence is unlikely to become available. The JCA methodology assesses the submitted comparative evidence and its uncertainty; high unmet need, conditional regulatory approval or the rarity of the condition does not itself remove the requirement for an interpretable relative assessment. The assessment framework itself may therefore have limited scope to say something favourable about a medicine where the clinical need is clear but comparative data are not capable of supporting a reliable relative-effect estimate.
This is not necessarily an argument for relaxing evidential standards. It does, however, suggest a limitation in what the JCA can conclude in some high-unmet-need settings. National HTA bodies will still need to distinguish between lack of benefit, lack of evidence and uncertainty that is difficult to avoid because of the disease context.
Stakeholder input and patient-relevant outcomes
The reports document input from patients, carers and clinical experts, but the tovorafenib assessment shows some of the constraints on how that input can influence the relative assessment.
The assessment scope requested health-related quality of life, neurological and cognitive symptoms, vision, balance, hearing, motor function, seizures and fatigue. However, comparative quality-of-life results were not available for the main assessed population. During review of the draft report, the patient or carer contributor described the lack of quality-of-life data as a major evidence gap, particularly in a chronic paediatric condition where functional and cognitive outcomes are important. The contributor also raised the absence of discussion of growth, development and long-term treatment considerations.
This highlights a potential limitation of the process. The JCA can identify outcomes that matter to patients, but it may remain largely silent on them where a sufficiently robust relative assessment cannot be performed. The result is methodologically understandable, but it may leave some of the most important aspects of treatment experience outside the substantive findings.
This is important because it shows where the formal assessment provides only a partial account of treatment value. Those gaps may need to be taken forward more explicitly in national appraisal, particularly where quality of life, functioning or long-term treatment burden are central to the condition.
PICO burden and fragmented evidence positions
The number of PICOs remains an operational concern. The three completed reports contained one, seven and eight PICOs respectively, which are in line with the average numbers of PICOs expected from earlier modelling prior to the JCA going live. Nevertheless, they still create substantial analytical and reporting requirements. [7]
The burden is also qualitative rather than simply numerical. Each PICO may have its own studies, assumptions, comparators, subgroups, analytical methods and degree of uncertainty. A single report can therefore contain several different evidence positions for the same medicine. Companies need to understand which of these positions are central to the intended use of the product and which are likely to be important only in particular national settings.
The national consequences remain uncertain
The main unresolved issue is how national HTA bodies will use the reports. JCAs assess relative clinical and safety effects along with their uncertainty; they do not determine price, reimbursement or overall value. National authorities remain responsible for appraisal, economic evaluation and final decision-making, while giving due consideration to the JCA.
At the time of writing, there are no examples of HTA outcomes for these products, and to what extent the JCA will have influenced decision making. A directionally favourable but uncertain result may be treated differently from a finding of no demonstrated improvement. A PICO that is not assessable may be central in one country but peripheral in another. National bodies may also place different weight on unmet need, economic evidence, social impact, treatment pathways and evidence generated after the JCA cut-off.
For companies, the practical task is to understand the PICO-specific position established by the JCA and prepare the appropriate next step. Demonstrated relative benefit needs to be translated into national clinical and economic value. A favourable but uncertain finding requires a clear account of the uncertainty and its decision consequences. No demonstrated improvement may require a broader, well-supported value case. A non-assessable finding may require additional evidence, or an explanation of why the relevant comparison should receive limited weight locally.
The first reports do not yet allow reliable conclusions about national reimbursement outcomes. They do, however, provide an initial view of how different evidence structures are handled and where companies are likely to face further work after the European assessment has been completed.
Contact:
If you are interested in discussing any of the issues above for your company/drug development program, please contact me through my email address dniven@nivenbiopharma.com . Feel free to also visit my website at www.nivenbiopharma.com for more information.
I have no conflicts of interest in the production of this article.
Sources:
[1] List of ongoing Joint Clinical Assessments, European Commission, 8 July 2026
[2] Joint Clinical Assessment Report Lurbinectedin, HTACG, July 2026
[3] Joint Clinical Assessment Report Tarlatamab, HTACG, July 2026
[4] Joint Clinical Assessment Report Tovorafenib, HTACG, June 2026
[5] Information, Data, Analyses and Other Evidence Missing from the Dossier of the Joint Clinical Assessment of the Medicinal Product Tacquell European Commission, February 2026
[6] Information, Data, Analyses and Other Evidence Missing from the Dossier of the Joint Clinical Assessment of the Medicinal Product Catequentinib European Commission, February 2026
[7] PICO Exercises, European Commission, Feb 2025
