A review of selected public hearings over several years shows how the focus of the debate has evolved. Earlier proceedings often centred on establishing the clinical efficacy of immunotherapy. Later discussions placed greater emphasis on the treatment setting and comparisons with established options. This is particularly apparent in MSI-H/dMMR colorectal cancer, gastric and gastro-oesophageal junction cancer, endometrial cancer and locally advanced cervical cancer.
Why AMNOG assessments of pembrolizumab (Keytruda) have become more demanding
Pembrolizumab is an anti-PD-1 monoclonal antibody marketed as Keytruda across multiple oncology indications. Under Germany's AMNOG framework, added benefit is assessed for each authorised indication and relevant patient population against the comparator therapy specified for the German assessment (zVT).
As indications expand, the evidence base changes. Pembrolizumab is studied across different treatment lines, combinations and clinical settings, while standards of care continue to evolve. Three recurring questions follow:
- Comparator therapy: Does the trial comparator support an assessment against the zVT specified under AMNOG?
- Patient population: Which biomarker-defined populations, disease stages and prior-treatment groups are adequately represented?
- Endpoints and evidence: What is the clinical significance of observed effects in curative versus palliative settings?
Pembrolizumab in MSI-H/dMMR colorectal cancer: Biomarkers and early progression in KEYNOTE-177
In the 2021 G-BA proceedings on pembrolizumab for metastatic colorectal cancer with high microsatellite instability (MSI-H) or deficient mismatch repair (dMMR), the predictive value of the biomarker was a major topic. MSI-H and dMMR reflect alterations in DNA mismatch repair and help identify patients for whom immune checkpoint inhibition may be relevant.
During the public hearing, clinicians discussed MSI-H as a particularly informative marker. The discussion of KEYNOTE-177 also made clear that biologically informed patient selection does not eliminate uncertainty about individual treatment trajectories.
KEYNOTE-177: Early progression despite biomarker-based selection
KEYNOTE-177 compared pembrolizumab with chemotherapy-based treatment in MSI-H/dMMR metastatic colorectal cancer. Clinical experts discussed patients who experienced early disease progression on pembrolizumab despite having the biomarker profile associated with potential benefit from immunotherapy.
The hearing therefore addressed close monitoring, the timing of treatment response and the possibility of switching to chemotherapy.
KEYNOTE-177 illustrates that a predictive biomarker can define a target population without fully predicting an individual's risk of early progression. An overall population-level effect does not establish which patients will benefit early or may initially require a different treatment strategy.
Pembrolizumab in gastric and GEJ cancer: Oxaliplatin, cisplatin and comparator selection
In the 2024 G-BA proceedings D-1023 and D-1024, the choice of chemotherapy backbone became a key issue. The discussion focused on oxaliplatin and cisplatin in gastric and gastro-oesophageal junction (GEJ) cancer.
Clinical trials are frequently designed years before a national HTA assessment takes place. Standards of care can change during that interval.
The 2024 G-BA hearings: Oxaliplatin as a clinical standard
Clinical experts argued that oxaliplatin had become widely established in routine care. This raised questions about how closely the comparators relevant to the assessment reflected actual clinical practice.
Demonstrating added benefit under AMNOG depends on whether the submitted evidence allows a comparison with the G-BA's specified zVT. Proceedings D-1023 and D-1024 illustrate how a divergence between routine care and the specified comparator can complicate evidence assessment.
For manufacturers, the implication begins at trial design: comparator selection needs to account for possible changes in treatment practice and the comparator likely to be relevant in Germany at the time of assessment.
Pembrolizumab in endometrial cancer: KEYNOTE-868 and a change in the German comparator
The 2025 G-BA proceedings on pembrolizumab in endometrial cancer illustrate the potential consequences of a change in the zVT. During the hearing, KEYNOTE-868 was described as showing statistically significant positive results. At the same time, participants discussed the absence of a suitable comparison against the newly specified German comparator.
The manufacturer referred to rapid changes in the treatment landscape and the difficulty of anticipating them when planning clinical trials.
How a change in zVT affected the discussion of KEYNOTE-868
A change in zVT can make existing trial evidence unsuitable for the comparison required to establish added benefit. Clinical efficacy and added benefit against the specified comparator are distinct assessment questions. A statistically significant trial result therefore does not automatically demonstrate added benefit under AMNOG.
The endometrial cancer proceedings illustrate that positive clinical trial findings may have limited applicability to an AMNOG assessment when they do not support a suitable comparison against the specified zVT.
Manufacturers should therefore consider potential changes in treatment standards and alternative comparator scenarios early in evidence planning.
Pembrolizumab and other checkpoint inhibitors: Established treatment classes change the comparison
The 2025 G-BA hearing on endometrial cancer also addressed the extent to which the clinical effects of different immune checkpoint inhibitors could be distinguished. Clinical experts referred to several randomised trials supporting checkpoint inhibition in the relevant treatment setting.
Growing evidence for a therapeutic class can strengthen confidence in the treatment approach. Under AMNOG, however, the relevant question remains the added benefit of the particular medicine compared with the specified comparator.
Established immunotherapies create new evidence requirements
As immune checkpoint inhibitors become established, the relevant treatment alternatives may change. An indication in which chemotherapy was previously the standard may later include immunotherapy-based options. New studies may consequently need to establish whether a combination provides additional patient-relevant benefit compared with an established active treatment.
The wider adoption of checkpoint inhibitors can make added-benefit demonstrations more demanding when effective immunotherapies are already available as comparators. This does not establish a universal class effect or therapeutic equivalence between checkpoint inhibitors. The indication and comparative evidence remain decisive.
Pembrolizumab in cervical cancer: Overall survival endpoints in a curative setting
The 2025 G-BA proceedings on locally advanced cervical cancer discussed the clinical interpretation of endpoints in a treatment setting with curative intent. Curative treatment aims at long-term disease control or cure, while palliative treatment has different objectives.
Why survival rates at defined time points matter in cervical cancer
The hearing considered whether survival probabilities at defined time points may be particularly informative in curative settings. A landmark analysis can, for example, examine the proportion of patients alive after a specified follow-up period.
Median overall survival and survival probability at a given time point describe different aspects of the survival distribution. The cervical cancer proceedings illustrate why survival endpoints need to be interpreted in the context of curative treatment intent.
What the pembrolizumab proceedings mean for AMNOG evidence strategy
|
Assessment issue |
Pembrolizumab example |
Implication for manufacturers |
|---|---|---|
|
Biomarkers and populations |
KEYNOTE-177, MSI-H/dMMR colorectal cancer |
Consider biomarker selection and clinical heterogeneity |
|
Appropriate comparator therapy |
D-1023/D-1024, gastric and GEJ cancer |
Reassess trial comparators against evolving standards of care |
|
Changes to the zVT |
KEYNOTE-868, endometrial cancer |
Anticipate alternative assessment scenarios and evidence gaps |
|
Endpoints in curative treatment |
Locally advanced cervical cancer |
Interpret survival outcomes within the treatment setting |
The proceedings differ in indication, evidence base and assessment question. Together, they show that established clinical efficacy alone does not demonstrate added benefit in every new treatment setting.
For market access and evidence generation teams, the key considerations are the evolving comparator, the definition of the population relevant to the assessment and the selection of appropriate endpoints.
Conclusion: Pembrolizumab illustrates the evolving demands of AMNOG assessment
The G-BA proceedings on pembrolizumab show how German early benefit assessment evolves alongside oncology treatment. KEYNOTE-177 highlighted biomarker selection and early progression. The gastric and GEJ cancer proceedings focused on comparators. Endometrial cancer illustrated the consequences of a changed zVT, while cervical cancer raised questions about interpreting survival outcomes in curative care.
As pembrolizumab becomes more established, demonstrating added benefit in new indications does not necessarily become easier. Manufacturers need to align clinical evidence early with the comparator, population and endpoint requirements of each AMNOG assessment.
Further reading: Guide to appropriate comparator therapy and AMNOG dossier.