>> For the assessment of the benefits of drugs for the treatment of rare diseases, for conditional approvals and approvals under exceptional circumstances, the G-BA will in future be given the authority to require pharmaceutical companies to carry out " to demand data collection accompanying the application. What role has application-accompanying data collection already played internationally?
In the UK, the proportion of initial submissions to NICE containing real world evidence (RWE) has increased consistently over the last three years, from 9 % in 2015 to 22 % in 2016 and to 37 % in 2017. And in France, according to IQVIA, around 25 % initial submissions to the HAS already include data from RWE. In general, it can be said that data collection during use is already playing a key role in the recognition of “breakthrough” therapies and “adaptive pathways” and has accelerated the approval process. Under these special approvals, RWE is even accepted as preferred evidence by regulatory authorities to confirm the safety and effectiveness of a drug.
And what does it look like in this country as of today?
In Germany, too, RWE is becoming increasingly important as part of the AMNOG process. To date, the data has been used in particular to describe the unmet medical needs (e.g. rheumatoid arthritis, PAH), to determine the target population and the incidence/prevalence (cancer registry).
Which study/registry designs are used?
Basically, a distinction is made between the following four register forms. On the one hand, this is the register for recording patients with certain diseases (so-called disease registers). Then there are registers to record the supply of patients with certain therapeutic substances or medical devices (so-called product registers). However, there are also registers to ensure the quality of treatment (so-called quality registers) and, last but not least, population-based registers that do not record an outcome (e.g. illness), but rather an “exposure”. Registries are often initiated by scientists, but registries are also increasingly being called for by politicians, such as the cancer or endoprosthesis register. The establishment and expansion of diabetes registers is also currently being discussed as part of the National Diabetes Plan.
And how does the data obtained flow into AMNOG?
In the benefit assessment procedures, for example, macitentan (“Opsumit”) for the treatment of PAH (COMPERA registry), baricitinib (“Olumiant”) for the treatment of moderately to severely active rheumatoid arthritis (RABBIT registry) or crizotinib (“Xalkori”) for the treatment of bronchial carcinoma (CRIP registry), data from patient registries were included. Accompanying data collections have not yet contributed to quantifying the additional benefit. There are no methodological standards for how this evidence can be taken into account when considering accompanying data collections. If one wants to use this data to quantify the additional benefit in the future, then realistic requirements for the evidence and an adaptation of the methods at G-BA and IQWiG are also needed. Against this background, it is to be welcomed that the G-BA has commissioned IQWiG to develop concepts for using the data as part of the benefit assessment.
What level of evidence can actually be generated with this type of data collection and the studies based on it?
In order to demonstrate the effect of a new drug in comparison to a placebo or another effective comparative therapy, a prospective, randomized, controlled, blinded study, the well-known RCT, is usually required. This is and remains the gold standard and is therefore rated with the highest level of evidence by IQWiG and the G-BA. But indication-specific registries can also provide valuable contributions and additions to existing RCT studies.
Which ones, for example?
Particularly when indication-specific registries include a representative sample of GKV patients. Of course, these registers must be carried out methodologically at the highest level and over a longer period of time in order to provide meaningful and complete data. Numerous registers (e.g. RABBIT) fully meet these methodological, technical and structural standards (data standards, IT interfaces, data protection concepts).
As is well known, IQWiG director Prof. Dr. Jürgen diaperer has very little idea about using registers for benefit assessment. He said this in the MVF cover interview (MVF 01/18): “Registers are not suitable for questions of benefit.”
In this case, IQWiG's general criticism of patient registries with regard to various BIAS factors is somewhat exaggerated. Data from patient registers can certainly play a role in deriving an additional benefit. This is particularly true for cases in which an RCT is not feasible. Comparative effectiveness analyzes based on register studies are also quite possible, although the evaluation methodology (propensity score matching) is of crucial importance here. In addition, registry studies can also provide answers to questions that cannot be answered using RCT methodology. This is simply because registers reflect the reality of care and can therefore provide important impetus for improving health care research.
How or how do data obtained through registers differ from the data generated in RCTs?
Registries create evidence that cannot be generated with RCTs, especially with regard to effectiveness under everyday conditions, which health services research has long wanted to represent through the difference between “effectiveness” and “efficacy”. In addition, registries can also provide evidence on the safety of patient groups that do not usually take part in clinical studies: children and seniors, those with severe/mild illnesses and/or patients with comorbidities/multimorbidity or concomitant medications. Registers reflect the reality of care, which can never actually occur in controlled studies due to the high inclusion and exclusion criteria. It's different with registries: Here, high case numbers and long running times even allow rare and/or delayed events to be monitored. In addition, complications and/or drug interactions can be recorded validly and meaningfully.
What do you think of application-related data collection as such?
In my opinion, the regulation in the GSAV is too broad when you look at the current situation. With the planned new regulation, all orphan drugs (including those approved using the standard procedure) will become the focus of possible additional requirements, as will all non-orphans with “conditional approval”. Of the 30 % of the AMNOG medicinal products potentially covered by the GSAV regulation, 8% have so far been limited by the G-BA with registers, i.e. only around 2 % of all medicinal products are currently limited by potential “application data”.
Where specifically?
As a rule, these have so far been orphan drugs for which phase 3 RCTs were not possible. If no phase 3 studies are foreseeable, the G-BA - based on the EMA requirements - already requires registers as a condition in the time limit of the initial decision. If the G-BA is now to receive further rights for accompanying data collection, a binding coordination between the G-BA and the EMA would have to be provided in order to keep the processes synchronized. In addition, the G-BA should be involved in joint consultations between EMA and HTA institutions at European level in order to coordinate possible requirements for approval or additional benefit assessment at an early stage.
Dr. Penske, thank you for the interview. <<
The interview was conducted by MVF editor-in-chief Peter Stegmaier.
Citation note:
