Welcome, this week’s post on the G-BA’s views on endpoints is the second article in the benefit assessments series. You’ll learn about:
As always, I have also added some recommendations for further reading, if you want to find out more.
Next time, we’re doing a quick excursion to two recent IQWiG reports on PROs.
The G-BA assesses the additional benefit of a product based on the data submitted by the manufacturer in the dossier. These can include registrational or pivotal studies, but also additional studies, or other types of evidence.
How does the G-BA determine the level of additional benefit?
The G-BA follows strict criteria to determine whether a product has an additional benefit, i.e., whether it offers a medical benefit over the so-called appropriate comparator therapy (ACT).
These criteria are applied in a hierarchical step-wise manner. Only when the criterion is met, the next one will be reviewed. If the submitted data does not meet one criterion, the G-BA will automatically give the resolution “No additional benefit (proven)”. In that case, the G-BA won’t continue to assess the evidence.
For example, if a study does not compare with the ACT, it is irrelevant whether the product showed a statistically significant reduction in mortality.
In this post, we focus on the last steps in this hierarchy, the endpoints.
Endpoints defined in IQWiG methods
The IQWiG methods (current version 6.1, in German) describe the details of how the IQWiG and G-BA view and consider endpoints.
Overall, the G-BA and IQWiG review whether an intervention shows a positive effect, a benefit, or a negative effect, i.e. harm, for a patient-relevant endpoint. The data need to show a causal relationship between the effects and the treatment, i.e. the observed effects can be attributed solely to the intervention.
Patient relevance is crucial for the acceptance of an endpoint. Patient-relevant endpoints reflect how a patient feels, how they perceive their functions and activities, or whether they survive. Here, the G-BA and IQWiG take both the intended and the unintended effects into account.
Outcome categories
The G-BA and IQWiG differentiate between the following 3 types of outcomes:
- Mortality
- Morbidity (symptoms and side effects)
- HRQoL
In addition, the G-BA and IQWiG are open to considering the “treatment effort” related to the intervention and disease, as well as patient satisfaction in terms of health-related aspects. However, these two additional endpoints cannot be the sole basis for assigning an additional benefit.
Mortality is a key endpoint as it demonstrates whether there is an increase in a patient’s lifespan.
The disease-related morbidity endpoints should capture whether there is an improvement in the overall health status and a reduction in the duration of the disease. The therapy-related morbidity endpoints show if there is a reduction in side effects.
Lastly, HRQoL endpoints can demonstrate whether the intervention offers an improvement in the quality of life. To assess HRQoL, the G-BA and IQWiG accept only those instruments that are suitable and evaluated for use in clinical studies.
Diagnostic measures are not per se patient-relevant. They can only have an indirect benefit if used as a companion diagnostic to a therapeutic intervention. There needs to be a proven benefit of the intervention depending on the test result.
Additional benefit rating by G-BA based on endpoints
The G-BA and IQWiG assess the certainty and extent of additional benefit for each endpoint individually. Then, they weigh these findings against each other to determine the overall additional benefit of the intervention.
The table below provides a rule of thumb for determining the extent of additional benefit based on the effect seen in the four overall outcome categories.
Surrogate endpoints considered by the G-BA
Surrogate endpoints can be simpler and faster. They can themselves be patient-relevant, but clinical trials often use them instead of patient-relevant endpoints. However, the IQWiG regards most surrogate endpoints as not reliable and misleading. Therefore, surrogate endpoints, that are not themselves patient-relevant, will only be accepted when they have been sufficiently validated
- using appropriate statistical methods,
- in the target patient population,
- for the same or comparable interventions (same mechanism of action).
The surrogate endpoint will be considered valid when the effect on the patient-relevant endpoint can be sufficiently explained by the effect on the surrogate endpoint.
Surrogate endpoints can be a hurdle at the German early benefit assessments. They are often included in registrational trials. Regulators generally accept these to assess the overall risk/benefit profile of an intervention. However, the G-BA and IQWiG focus on the patient-relevant benefit of the intervention, which can create a misalignment between regulators and the German HTA bodies.
Validation of surrogate endpoints
There is no standard method or best practice to validate a surrogate endpoint. However, the G-BA and IQWiG prefer the use of correlation-based methods. These methods often need a meta-analysis of several RCTs, which measure both the effects of the surrogate as well as the patient-relevant endpoint. This meta-analysis needs to show a high level of correlation both at the study and individual level. In addition, the validation studies need to have a high certainty. Alternatively, the manufacturer can compare the effect on the surrogate endpoint from the studies with the so-called surrogate-threshold-effect (STE). Here, the extent of the statistically significant effect on the surrogate endpoint determines whether it is accepted.
Review of early benefit assessments of analogues
The acceptance of endpoints is often disease area-specific. Thus, in order to anticipate whether the G-BA or IQWiG will accept a certain endpoint, I recommend reviewing the early benefit assessments of analogues products in the same therapeutic area. The commentary quite clearly states which endpoints the G-BA or IQWiG accepted and why they rejected others.
It is important to pay close attention to the operationalization of the endpoint, i.e. how exactly the trial defined and measured the endpoint. For example, two trials might use the same HRQoL instrument, but the G-BA might only accept it in one assessment, since the other trial used the instrument at the wrong times or used a non-validated MID.
Further reading related to HTA in Germany
- Regular and abbreviated early benefit assessments – Is this the end of orphan drug privileges?
- HTA guide
- EUnetHTA21 starts preparation for joint HTA from 2024
- Digital treatments and medical applications (DiGA) – Assessment process in Germany
- G-BA assessment of orphan drug gene therapies, Zolgensma and Libmeldy
- Why did the G-BA assess Trimbow now?
- IQWiG methods version 6.0, English
Newsletter
If you liked this post and don’t want to miss the next one, sign up for my free weekly newsletter!
3 thoughts on “Which endpoints are important to the G-BA and IQWiG?”
Comments are closed.