Onasemnogene abeparvovec (Itvisma): What the first EU dossier references in an IQWiG assessment reveal

Germany's assessment question had already been included in the Joint Clinical Assessment (JCA) as PICO 12. The final European assessment scope comprised twelve PICOs, with comparative results for seven. PICO 12 lacked suitable comparative data for the German population and the individualised comparator specified by the G-BA.

The Itvisma case shows how information from a European dossier can inform a national assessment. Including a German question in the JCA scope does not, by itself, provide evidence suitable for that question.

Twelve JCA PICOs, but comparative results for only seven

The European JCA of onasemnogene abeparvovec (JCA-MP-2024-12) covered twelve final PICO questions, but comparative evidence was not available across the entire scope. The initial scope contained fourteen PICOs. Two were removed after the proposed indication was restricted to patients aged two years and older. Comparative results were submitted for seven of the remaining twelve PICOs.

The assessment scope included populations distinguished, among other factors, by SMN2 copy number and ambulatory status. Comparators included nusinersen, risdiplam, best supportive care (BSC), intravenous onasemnogene abeparvovec and other individualised treatment options.

This created a broad European assessment framework encompassing several distinct clinical questions. The available studies, however, were not equally suited to answering each one.

JCA scoping and PICO definition explains how these questions are set within the assessment scope. The seven PICOs with results did not represent seven independent clinical comparisons. Several questions drew on the same studies and indirect comparisons.

The JCA report also records difficulties with historical SMA types. The European assessors challenged the retrospective assignment of STEER patients to SMA type 3 solely by age at symptom onset, although none had ever walked independently. Their criticism concerned a subgroup analysis, not the definition of PICO 12. It illustrates why the clinical definition of a patient population matters when assessing the data (JCA report, Section 4.4.1, p. 138).

Distinguishing the assessment scope from the available evidence is essential to interpreting a JCA. A PICO defines a clinical question to be addressed at European level. Whether it can be answered with appropriate data depends on the clinical development programme and the feasibility of relevant comparisons.

The Itvisma evidence did not cover every PICO population

The principal randomised evidence for Itvisma came from STEER, whose study population was narrower than several populations specified in the European assessment scope. STEER investigated intrathecal onasemnogene abeparvovec in non-ambulatory patients with no more than four SMN2 copies.

STEER provided randomised results against a sham procedure, but no direct comparison with nusinersen or risdiplam. Within the JCA, the sham comparison was considered in relation to best supportive care. The manufacturer used indirect comparisons to estimate relative effects against the active treatments.

These analyses drew, among other sources, on CHERISH, SUNFISH Part 2 and further studies in previously treated patients. Statistical adjustments were intended to account for differences between study populations and enable comparisons with the treatments specified in the assessment scope.

The methodological challenges identified in the first JCA reports provide context for these indirect comparisons. The European assessors identified methodological limitations. Differences in patient characteristics, the selection and adjustment of prognostic factors, and reductions in effective sample size constrained the interpretability of the indirect comparisons.

Such limitations can arise when trials with different eligibility criteria are subsequently brought together for comparative analyses. For JCA, the challenge is particularly consequential: evidence must support comparisons against multiple treatments and remain applicable to the population specified in each PICO.

The indirect comparisons, despite their methodological limitations, produced results for certain populations. For PICO 12, however, intervention data were already too limited to conduct any comparative analysis. A different statistical method could not have resolved that lack of data (JCA report, pp. 28–31 and Section 4.3.5).

PICO 12 exposes the limits of national evidence use

The question in the German AMNOG benefit assessment corresponded to PICO 12 of the European JCA dossier, but suitable comparative data were not available for that question.

IQWiG explicitly confirmed this correspondence in its dossier assessment A26-99. The population comprised patients aged two years and older with 5q-associated SMA, at least four SMN2 copies and SMA type 2, 3, 4 or an undefined type.

Germany's Federal Joint Committee (Gemeinsamer Bundesausschuss, G-BA) had specified individualised therapy selected from nusinersen, risdiplam and best supportive care as the appropriate comparator therapy (zweckmäßige Vergleichstherapie, zVT).

The European scope therefore included the German question. However, the evidence did not adequately address both the relevant population and the individualised comparator required for the German benefit assessment.

STEER enrolled patients with no more than four SMN2 copies, whereas PICO 12 covered patients with at least four copies, including those with higher copy numbers. The populations also differed in clinical characteristics and in the comparator strategy required.

Patients with exactly four copies represented a potential overlap, but this did not establish an appropriate comparison for the full German population against the G-BA's specified individualised therapy.

Both the manufacturer and IQWiG found no randomised controlled trial directly comparing onasemnogene abeparvovec with that comparator therapy. Nor did the European dossier provide corresponding comparative results for PICO 12.

IQWiG consequently concluded that an added benefit over the appropriate comparator therapy had not been demonstrated. The final added-benefit decision rests with the G-BA.

From a European perspective, the more informative finding is how the evidence gap arose. Germany's question was included in the JCA scope, but the clinical development evidence did not answer it.

This separates the formal inclusion of national requirements during scoping from their substantive coverage by the clinical development programme. Inclusion of a PICO guarantees neither an appropriate study nor a feasible indirect comparison.

EU dossier references under AMNOG before the JCA report was published

For Itvisma, the manufacturer used the option under Section 4(2a) of the German Ordinance on the Benefit Assessment of Pharmaceuticals (AM-NutzenV) to refer to information in the European JCA dossier. The references concerned Germany's assessment question, which corresponded to PICO 12 of that dossier. The JCA dossier guide explains the European evidence submission.

The European dossier and the final JCA report had different roles. The European HTA Coordination Group endorsed the JCA on 16 July 2026, but the report was not published until 7 September. It was therefore unavailable at the relevant German dossier submission date of 31 July and could not be considered by IQWiG under Section 7(1)(2) AM-NutzenV.

The references to information in the European dossier were permitted and became an additional basis for the German benefit assessment. The dossier contained no comparative data for PICO 12, however, so the option to refer to it could not resolve that evidence gap.

The case illustrates two requirements for using European materials in a national assessment: they must be available at the relevant time and contain data suited to the national question.

What Itvisma means for JCA scoping and evidence planning

The Itvisma case illustrates why JCA preparation cannot end with formally covering as many PICOs as possible. The question is whether the clinical studies and analyses available can actually generate suitable comparative results for the different assessment questions.

The European assessment scope brings together requirements from multiple Member States. It may include populations, comparators and outcomes that were not anticipated in the same combinations when the original clinical programme was designed.

For Itvisma, the boundaries were particularly apparent in the SMN2 copy-number categories and comparator choices. STEER and the supplementary indirect comparisons addressed several European questions. They did not provide suitable comparative evidence for the German population in PICO 12.

Three implications follow for early evidence planning.

First, potentially relevant national populations need to be considered when developing the clinical evidence strategy. Eligibility criteria, genetic characteristics and clinical subgroups can determine whether trial results are usable for a later PICO.

Second, the feasibility of comparisons with relevant treatments needs to be assessed early. The existence of trials involving nusinersen or risdiplam does not in itself support a credible indirect comparison if populations and study conditions are insufficiently comparable.

Third, formal inclusion in the assessment scope must be distinguished from actual evidence coverage. A PICO can be fully specified at European level yet remain without comparative results.

These issues are central to JCA scoping and the assessment scope. Itvisma provides a concrete example from a completed European assessment and a subsequent German benefit assessment.

IQWiG's conclusion does not, in this respect, fundamentally conflict with the European clinical assessment. It demonstrates the national consequence of an evidence gap already visible in the JCA.

For future implementation of EU HTA, the central lesson is that a common assessment scope establishes a European framework for clinical questions. Whether nationally relevant PICOs can be answered with credible comparative evidence is shaped much earlier, through trial design, comparator selection and the planning of comparative analyses.

Sources

1. European Commission: Joint Clinical Assessment Report – Onasemnogene abeparvovec (Itvisma), JCA-MP-2024-12. Published 7 September 2026.
2. Institute for Quality and Efficiency in Health Care (IQWiG): Onasemnogen abeparvovec – dossier assessment A26-99. Version 1.0, 28 October 2026.
3. Federal Joint Committee (G-BA): German benefit assessment procedure for onasemnogene abeparvovec.
4. European Union: Regulation (EU) 2021/2282 on health technology assessment.