AMNOG 2.0 is the emerging debate about how Germany should balance early access to innovative medicines, robust evidence requirements and the financial sustainability of its statutory health insurance system. Some innovative medicines enter the market with greater evidence uncertainty, additional evidence may only become available after launch, while healthcare expenditure continues to rise.
The debate also matters beyond Germany. The EU Health Technology Assessment (HTA) Regulation has introduced Joint Clinical Assessments (JCAs), creating a common European clinical assessment that feeds into national HTA processes. JCAs have applied to new oncology medicines and advanced therapy medicinal products (ATMPs) since January 2025; orphan medicines will follow in 2028 and all new centrally authorised medicines in 2030. Germany is therefore discussing the future of AMNOG while the European evidence environment is changing. AMNOG 2.0 is not an enacted reform package at this stage, but an ongoing policy debate.
A specific feature of the German system shapes this debate: early benefit assessment does not generally determine whether a newly authorised medicine is available to patients. Newly authorised medicines are generally reimbursable from market entry. The reform question therefore concerns less access itself than how added benefit should be assessed under uncertainty and how that assessment should subsequently translate into the reimbursement amount.
Four questions define the AMNOG 2.0 debate
- How early should patients gain access to innovative medicines?
- How much evidence should be required at market entry to establish added benefit?
- How should additional evidence generated after launch feed back into benefit assessment?
- How can early access to innovation remain financially sustainable?
Orphan medicines, ATMPs and targeted therapies can involve small populations, limited comparative evidence or uncertainty about long-term outcomes at launch. Industry proposals therefore call for greater flexibility in specific treatment situations and use of the best available evidence. At the same time, statutory health insurance expenditure remains under pressure. The resulting policy question is how uncertainty should be managed when early access, evidence requirements and financing pull in different directions.
How much evidence should AMNOG require at market entry?
The first major reform question concerns the evidence threshold for medicines addressing significant medical need when their evidence base remains uncertain at launch. According to the GKV-Spitzenverband, the BMG has discussed special treatment for medicines with regulatory designations including PRIME, Accelerated Assessment, Conditional Marketing Authorisation and authorisation under exceptional circumstances. Under the approach described by the GKV-Spitzenverband, a positive quantified added benefit could be recognised despite greater uncertainty in the underlying evidence.
The problem is particularly visible in small populations and severe diseases. Conventional comparative evidence can be difficult to generate before launch, and long-term outcomes may not yet be observable. Reform proposals therefore include formal recognition of specific treatment situations and greater consideration of the best evidence that can reasonably be generated. The Federal Joint Committee (G-BA) also identifies a potential incentive problem. A special pathway could result in a medicine with limited evidence and a specific regulatory status being assessed under different standards from a medicine with more robust evidence and regular marketing authorisation. The GKV-Spitzenverband similarly emphasises that early benefit assessment serves a different purpose from marketing authorisation and should make evidence uncertainty transparent. The counterargument is that marketing authorisation and German early benefit assessment answer different questions: authorisation concerns the benefit-risk balance, while AMNOG asks whether patient-relevant added benefit is demonstrated against the comparator relevant to German care.
The Institute for Quality and Efficiency in Health Care (IQWiG) goes one step further and questions whether a new assessment pathway is required to deal with best available evidence at all. Different forms of evidence, including single-arm studies with suitable external comparator data, can already be considered within AMNOG. From IQWiG’s perspective, the decisive requirement is that the best available evidence is methodologically and clinically suitable for the specific AMNOG assessment question and allows comparison with the relevant standard of care. For certain special authorisation situations, IQWiG therefore proposes strengthening the obligation to submit suitable best available evidence rather than introducing a fundamentally new assessment pathway.
In an interview with the German healthcare news agency PAG, IQWiG Director Thomas Kaiser further specifies where he sees a need for reform. According to Kaiser, even 15 years after AMNOG was introduced, the necessary comparative evidence is generated in only around half of cases. This affects not only benefit assessment but also the evidence available to inform treatment decisions. He therefore calls for stronger incentives to conduct high-quality comparative studies rather than broadly adapting assessment requirements to gaps in the available evidence.
Source: PAG, OPG Spezial: AMNOG 2.0 – Mit Innovationen Schritt halten, October 2026, interview with Dr Thomas Kaiser, pp. 3–6.
This broadens the reform question: How can AMNOG 2.0 accommodate evidence uncertainty in situations of high medical need without disadvantaging more robust evidence generation – and which changes to the existing assessment framework are actually necessary?
What could change?
A lower or more flexible initial evidence threshold would make management of residual uncertainty after launch more important. The evidence pathway could increasingly look like this: marketing authorisation → initial benefit assessment → additional evidence generation → reassessment. The key question would be which remaining uncertainties can realistically be resolved afterwards.
How should post-launch evidence feed into AMNOG reassessment?
According to the GKV-Spitzenverband’s description of the BMG proposals, regular reassessments are being considered for certain medicines. Data from Germany’s Health Research Data Centre (Forschungsdatenzentrum Gesundheit, FDZ) could potentially contribute in the future. Longer follow-up, new clinical data cuts and suitable real-world data could become part of a subsequent assessment cycle. The discussion about reassessment leads directly to the question of data fitness. The G-BA considers the current FDZ dataset problematic for this type of benefit assessment. Diagnosis and claims data do not automatically capture clinical endpoints, disease stage or other characteristics required for a specific benefit assessment. In rare diseases, sufficiently granular diagnosis codes may also be unavailable. This leads to an important principle for real-world evidence in AMNOG: the availability of a dataset should not determine the assessment question. The decisive issue is whether the data are fit for the specific evidence question.
IQWiG makes a similar methodological point and additionally emphasises the need for prospective planning of real-world data generation. It considers a general focus on FDZ data unsuitable for the proposed reassessment model, while acknowledging that individual FDZ data elements may still contribute to benefit assessment, for example through data linkage in registry-based randomised controlled trials.
Kaiser also identifies a need to reform the statutory framework for Germany’s anwendungsbegleitende Datenerhebung (AbD), a form of post-launch evidence collection that can be required under AMNOG. He criticises the fact that randomised studies cannot be mandated under the current AbD provisions, even as registry-based randomised controlled trials have gained international importance. In his view, the rules can prevent suitable study designs from being required through this mechanism. The debate therefore also concerns which designs the legal framework should permit for mandated post-launch evidence generation.
The central disagreement concerns when reassessment is justified. The GKV-Spitzenverband argues that it should primarily take place when the evidence base has materially changed and a meaningful gain in knowledge can be expected. A more systematic model would require manufacturers to anticipate which evidence could become available two, three or more years after launch and which initial uncertainties it could address. The role of German registries, claims data and FDZ data is covered separately in our analysis of Real-World Evidence in the AMNOG dossier.
How could AMNOG recognise medical need without replacing evidence-based benefit assessment?
The G-BA introduces an alternative approach into the AMNOG 2.0 debate. The regular data-based benefit assessment by the Federal Joint Committee (G-BA) and the Institute for Quality and Efficiency in Health Care (IQWiG) would remain in place for all new medicines. For specific treatment situations, an additional assessment could determine whether there is a “potential for a therapeutically relevant added benefit”.
Under the proposal, such a finding could be considered when the regular evidence does not establish added benefit but there are indications of a meaningful improvement in therapy compared with the appropriate comparator therapy (zweckmäßige Vergleichstherapie, zVT). The supplementary qualitative assessment could consider alternative endpoints, indirect comparisons with greater uncertainty and clinical input from medical experts.
The potential pathway would therefore not automatically translate greater evidence uncertainty into a regular quantified added benefit. Instead, it could introduce an intermediate step:
No data-based added benefit → potential for a therapeutically relevant added benefit → additional evidence → reassessment
For the subsequent reassessment, the G-BA proposes that the evidence base should not be restricted to FDZ data. The pharmaceutical company should be able to determine which data are submitted for reassessment.
This creates a second reform model alongside a more flexible initial evidence threshold: Should AMNOG accept greater uncertainty within the existing added-benefit assessment, or should therapeutic potential be recognised separately until more robust evidence becomes available?
Which endpoints, study designs and evidence standards should apply?
The third reform question concerns how closely German benefit assessment should connect to evidence generated for regulatory approval. According to the GKV-Spitzenverband, the proposals discussed by the BMG would allow endpoints used in the regulatory process to be treated initially as “provisionally patient-relevant” for certain medicines. An endpoint may support a regulatory decision without necessarily meeting German requirements for demonstrating patient-relevant added benefit.
IQWiG specifically challenges the proposed focus on the primary endpoint of the pivotal regulatory trial. Some endpoints that appear in the BMG concept as potentially ‘provisional’ – such as validated functional scores, seizure control or motor function – can already qualify as patient-relevant endpoints in regular AMNOG assessment. At the same time, the primary endpoint of a regulatory trial is not automatically the endpoint that is most relevant to patients for benefit assessment. IQWiG therefore warns that giving the primary endpoint a privileged role could make endpoint selection in clinical development more strategically sensitive.
Industry proposals seek greater recognition of endpoints accepted at European level and key endpoints from pivotal regulatory trials. Other proposals call for more flexibility regarding surrogate endpoints and alternative evidence in specific treatment situations. Greater acceptance of regulatory endpoints could make European and regulatory evidence easier to use within German benefit assessment, while making the subsequent management of uncertainty more important.
Aljona Specht of AstraZeneca calls for disease-related risk factors to be recognised as a separate category within the concept of benefit. In an Ärzte Zeitung interview published on 6 November 2026, Professor Dirk Müller-Wieland cites LDL cholesterol as a causal cardiovascular risk factor relevant to prognosis and argues that such factors should be considered. The special report was supported by AstraZeneca.
A risk factor's prognostic relevance is distinct from validated evidence of a patient-relevant treatment effect. Whether a risk measure reliably captures such an effect depends on the evidence for the specific context; lowering LDL does not automatically demonstrate added benefit in Germany's AMNOG assessment.
Beyond the patient relevance of endpoints, the debate also concerns the development of suitable study designs. IQWiG Director Thomas Kaiser identifies synthetic study arms as a potential future source of evidence. Their use in assessment would, however, depend on empirical research demonstrating that they can produce informative results. This position leaves room for methodological innovation while requiring evidence that new approaches are suitable for comparative assessment.
The reform question therefore goes beyond whether regulatory endpoints should receive greater recognition. It also concerns how patient relevance should be determined and whether that assessment should depend on regulatory status or on an endpoint’s position within the trial design.
How predictable should the German comparator be?
The appropriate comparator therapy (zweckmäßige Vergleichstherapie, zVT) is central to AMNOG because it defines the treatment against which added benefit is assessed. The GKV-Spitzenverband calls for an earlier and more binding determination of the zVT. The vfa similarly calls for greater predictability where clinical trials have been designed in accordance with prior advice from Germany’s Federal Joint Committee (Gemeinsamer Bundesausschuss, G-BA).
Johnson & Johnson also calls for greater predictability in the selection of the appropriate comparator therapy (zweckmäßige Vergleichstherapie, zVT). Urs Voegeli, Managing Director of Janssen-Cilag in Germany, argues in an interview with PAG (OPG Spezial, October 2026, pp. 48–51) that manufacturers need to be able to rely on prior consultations with Germany’s Federal Joint Committee (G-BA) when designing clinical studies. In his view, subsequent changes to the comparator should not undermine evidence generated on that basis. This raises a practical tension between development planning certainty and the need for benefit assessment to reflect changes in the standard of care.
Clinical evidence generation starts years before an AMNOG dossier is submitted. If the German comparator changes after a pivotal trial has been designed or completed, the evidence may become less directly applicable. A more predictable comparator would improve long-term evidence planning, especially as European PICO scoping, JCA requirements and German comparator requirements increasingly need to be considered within one development programme.
Müller-Wieland also questions increasingly granular choices of appropriate comparator therapy (zVT) in Germany. In chronic disease, he argues, the assessment should reflect how risk-stratified treatment strategies are used to improve prognosis in clinical practice. This raises the question of how a comparator suitable for a specific assessment can adequately reflect clinical treatment strategies.
He calls for more precise consultation of medical scientific societies, including on appropriate comparators and the clinical relevance of endpoints. Their clinical expertise can inform the assessment alongside an independent, consistent methodological appraisal of the evidence.
How should AMNOG connect with EU HTA?
Under the EU HTA Regulation, JCAs provide a scientific analysis of the relative clinical effects of a health technology. They support national HTA processes rather than replacing national decisions on added value, reimbursement or pricing. The first JCA report for a medicine was published in June 2026, turning the European framework into an operational assessment process. Germany’s Pharmaceutical and Medical Technology Dialogue places benefit assessment and EU HTA within the same working group.
IQWiG Director Thomas Kaiser sees few fundamental methodological obstacles between the European Joint Clinical Assessment (JCA) and Germany’s AMNOG assessment, as both are grounded in evidence-based medicine. He identifies the treatment of orphan medicines as a major difference: comparative evidence is required from the outset in the JCA, whereas German AMNOG continues to apply a statutory special provision to the early benefit assessment of orphan medicines. Shared methodological principles do not, however, eliminate differences between the questions addressed at European level and those relevant to the German national assessment.
The vfa calls for greater use of European assessment results within AMNOG, closer alignment of analytical requirements, greater consideration of European endpoints and better timing between JCA and the national process. For manufacturers, early alignment of PICO, endpoints, comparator strategy and evidence generation will become even more important. The German concept of the delta dossier describes the additional national evidence and analyses required to translate the European JCA evidence package into a dossier that meets German AMNOG requirements; it is not an official regulatory designation.
Why are orphan medicines and ATMPs central to the debate?
Orphan medicines and ATMPs concentrate many of the tensions behind AMNOG 2.0. Small patient populations, limited comparative evidence, potentially long-lasting treatment effects and difficult trial conditions can make conventional evidence generation challenging. ATMPs have been subject to mandatory JCA since January 2025, while orphan medicinal products will enter the scope in January 2028.
The debate on changing benefit assessment also concerns Germany's AMNOG orphan drug privilege. The Health Financing Commission recommends abolishing it, while IQWiG and KBV propose different approaches to comparative assessment and later full assessments.
Johnson & Johnson points to CAR T-cell therapies as an example of these methodological challenges. In certain aggressive haematological malignancies, small patient populations and difficult trial conditions can limit the feasibility of direct comparisons with established treatments. Urs Voegeli therefore advocates assessment standards that better accommodate such treatment settings. For AMNOG benefit assessment, the key methodological question remains whether alternative forms of evidence can support a sufficiently robust comparison with the relevant standard of care.
Regulatory status and medical need should not automatically be treated as equivalent. The G-BA notes that orphan designation may also be based on an expected “significant benefit” compared with existing treatment options. Orphan designation therefore does not automatically identify a treatment situation in which no satisfactory alternatives are available. This creates an important design question for AMNOG 2.0: Should eligibility for a special assessment pathway depend primarily on regulatory status, or should it be linked more directly to the actual treatment situation and unmet medical need?
IQWiG takes a more fundamental approach to orphan medicines. It proposes that an appropriate comparator therapy should also be routinely defined for orphan medicines and that manufacturers should be required to submit the best available evidence suitable for comparative AMNOG assessment. From IQWiG’s perspective, targeted support for orphan medicines in indications without an active standard therapy could instead be addressed through pricing. This separation would also be closer to the EU HTA framework, where comparative evidence and the relevant standard of care already play a central role.
This creates another reform question: Does high unmet medical need require different evidence standards within benefit assessment, or could it be addressed more directly through reimbursement while maintaining consistent evidence standards?
How can early access and financial sustainability be reconciled?
Behind the methodological debate lies a broader trade-off between rapid access to innovation and the financial sustainability of Germany’s statutory health insurance system. The BMG describes rapid availability of new medicines as an existing strength, while its Pharmaceutical and Medical Technology Dialogue aims to strengthen innovation and security of supply while containing increases in healthcare costs. In the first half of 2026, expenditure growth across statutory health insurance remained above the growth of contribution-based revenues. The potential scope of a special pathway is also part of the debate. According to the G-BA’s analysis, the regulatory criteria under discussion would capture more than 60% of assessments involving new active substances. For 2025, the G-BA reports a range of 71.1% to 89.5%, depending on which regulatory criteria are included. These are calculations presented by the G-BA and depend on the criteria included. This raises a system-level question: Can an assessment pathway still function as an exception for specific treatment situations if it potentially applies to a large proportion of newly assessed medicines?
Johnson & Johnson adds a further dimension to the reform debate: the relationship between the assessment of added benefit and the negotiated reimbursement price. In the PAG interview (OPG Spezial, October 2026, pp. 48–51), Urs Voegeli argues that pricing and expenditure-control measures can limit the economic recognition of therapeutic innovation. He specifically refers to Germany’s Statutory Health Insurance Financial Stabilisation Act (GKV-Finanzstabilisierungsgesetz), price-volume arrangements and so-called Fokuslisten. From the company’s perspective, such measures may weaken incentives to investigate additional indications and affect Germany’s attractiveness as a pharmaceutical market.
This brings two distinct questions into focus: what evidence is sufficient to establish added benefit, and how should an established benefit be reflected in pricing? While the National Association of Statutory Health Insurance Funds (GKV-Spitzenverband) highlights the potential additional expenditure associated with changes to assessment rules, Johnson & Johnson emphasises potential risks to research and investment from restrictive reimbursement conditions. The economic consequences described in the interview represent the company’s position, rather than causal effects established by the source.
IQWiG has also brought cost-benefit assessment into the discussion on how clinical benefit should inform pricing. Under its general mandate, the institute is conducting an illustrative cost-benefit assessment of sacubitril/valsartan to examine how the approach can be applied, presented and interpreted, and whether it could inform price negotiations. Although AMNOG already provides for cost-benefit assessments, IQWiG Director Thomas Kaiser notes that the option has not been used in the past 15 years. It may be particularly informative for medicines with substantial additional costs compared with standard care, including where higher drug expenditure could be offset by downstream savings. This illustrative project does not constitute an adopted reform.
The fundamental question is how much evidence uncertainty a solidarity-based healthcare system should accept in exchange for earlier access to innovative medicines, and how that uncertainty should later be reflected in evidence generation, reassessment and reimbursement.
What AMNOG 2.0 could mean for evidence planning and market access
- Evidence planning could extend further beyond launch if reassessment becomes more systematic.
- Regulatory and HTA evidence could become more closely connected, increasing the importance of coordination between Regulatory Affairs, Clinical Development, HEOR and Market Access.
- European and German evidence strategies could converge further through the sequence: clinical development and PICO planning → JCA → AMNOG → additional evidence generation → potential reassessment.
- The quality of evidence planning itself could become more differentiated. IQWiG proposes distinguishing more clearly between situations in which suitable evidence is absent, suitable evidence shows no added benefit, and evidence is still immature but robust further studies are already planned.
What central question remains open for AMNOG 2.0?
The G-BA and IQWiG statements show that the AMNOG 2.0 debate is no longer only about whether Germany should accept greater evidence uncertainty. It is increasingly about how different forms of evidence uncertainty should be represented within the assessment system.
According to the published statements, the BMG model combines an initial assessment based on the best available evidence with a routine reassessment using additional evidence after three to four years. The G-BA proposes a different approach: the regular data-based benefit assessment would remain in place, while specific treatment situations could additionally receive a finding of ‘potential for a therapeutically relevant added benefit’. IQWiG starts one step earlier and questions whether a new assessment model is required at all. In its view, suitable best available evidence can already be assessed within the existing AMNOG framework, while high-quality evidence planning should be more clearly recognised and incentivised.
This makes a fundamental distinction increasingly important for AMNOG 2.0:
Is suitable evidence missing? Does suitable evidence show no added benefit? Or is the evidence still immature at market entry even though robust additional studies are already planned?
These situations do not necessarily lead to different outcomes under the current benefit assessment framework. This is one area where reform could potentially intervene.
For manufacturers, the quality and feasibility of further evidence generation could therefore become more important alongside the data available at launch. IQWiG’s position points to a specific reform option: stronger incentives for high-quality comparative studies and legal provisions that allow suitable study designs could help close evidence gaps. How uncertainty is addressed would then depend more explicitly on the credibility of plans to generate additional comparative evidence.
AMNOG 2.0 therefore increasingly becomes a question of evidence architecture: How can Germany preserve early access while creating incentives for uncertainty to be reduced through suitable comparative evidence rather than remaining unresolved over the product lifecycle?
Johnson & Johnson’s position extends this methodological debate to the subsequent pricing process. Even if AMNOG 2.0 changes how particular treatment settings are assessed, a separate question remains: how should an established added benefit be reflected in the reimbursement price? The reform debate therefore encompasses both the evidence required to demonstrate added benefit and the economic consequences of that assessment.