>> Mr. Dr. Bruns, why is it that the scientific evidence that is now available in oncology is arriving late in care?
Bruns: All health care – not just in oncology – is in the hands of self-administration and is characterized by different billing systems for inpatient and outpatient care. Each of these isolated areas has its own internal set of rules.

Where is the cross-sector care?
Bruns: It doesn't exist as a rule yet. The National Cancer Plan and the National Decade Against Cancer can only be something of a bypass. These approaches can work in parallel for a while and, little by little, try to intervene in the supplying systems and bring them together.

It is working?
Bruns: Approaches can be seen: for example, through the obligation to establish clinical cancer registries in all federal states. Something is also happening in terms of creating a more sustainable financing basis for outpatient cancer counseling. In my opinion, instead of proclaiming a new decade of research against cancer, it would be important to finally ensure that the translation from research into care is reflected in social law in order to close the gap between existing evidence and actual care.

A tried and tested approach would be to form centers.
Bruns: According to the current legal situation, innovations always come into widespread use straight away or not at all. When it comes to new procedures, it should be considered whether it would not be better if they could not initially be used by all service providers, but only by those who have the appropriate skills and, above all, who are willing to do so, in parallel with the high-quality work they provide to collect supply data. With innovations, which are usually expensive in oncology, it would therefore make a lot of sense to establish knowledge-generating care, which was already described in 2017 by the cross-party AG Future Oncology and is currently being demanded by us and the DGHO. Because you can only learn from data.

However, studies and real care are very different.
Bruns: The basic idea of any clinical study is always to generate data. It is an ideal idea that the entire care landscape would suddenly be dedicated to data collection with a study nature when introducing innovations. This won't work that easily. As an interim solution, a selective selection of service providers who carry out and document the treatment in selected structures would be required. The transition to the next stage could then be developed relatively quickly on the basis of selective contracts.

If industry plays a role in this, there is always the fear that data should only be collected to make research and development cheaper or to create a new business model. What would be your petition, Dr. Lettrich?
Lattrich: If we keep in mind that only a fraction of the existing data is currently used for research and care, it should be clear to everyone involved in the system that we must come to a common solution in the interest of society as a whole and put particular interests aside. And of course we – as a research-based healthcare company and provider of innovative therapies – want to help shape this solution process and contribute our experience. If this ultimately leads to research and development not only becoming more precise and faster, but also more cost-efficient, this is certainly a welcome effect.

What could a transparent data system look like?
Lattrich: I think, as a first step, we in Germany first have to agree on which data should be collected and how we can achieve a uniform processing that would make meaningful use of the data possible in the first place. Because part of our reality is that the management of health data in Germany is still highly fragmented.

How is this done at the cancer registry, Mr. Bruns?
Bruns: The cancer registries collect defined data sets and are basically already asked to make them available for research. The basic idea of the accompanying data collection after approval as laid down in the GSAV is, in principle, exactly what knowledge-generating care means: namely, establishing certain observation points with approval and then observing them over a sufficient time horizon.
However, the G-BA should be able to demand the accompanying data collection from the pharmaceutical companies.
Bruns: It's crazy that this was made the responsibility of the companies, but that was almost to be expected.

But generally…
Bruns: ...that makes sense if the newly acquired data is integrated into the existing clinical cancer registries. Then, if necessary, you can match it with other data - anonymized of course - and possibly come to new insights. Without this match, you will only have singular data fragments for individual medications with specific but different observation points, with the anamnesis plus the other collected data missing.

Who should be responsible for matching this data?
Bruns: In any case, this should not be the task of IQWiG or the G-BA, but rather that of the clinical cancer registries. To do this, however, a self-administratively neutral, higher-level unit would have to be created, which would still have to be founded.

Would this be an effective model for Roche?
Lattrich: I believe that a purely national perspective does not go far enough here. Let's take rare diseases, for example cancers with extremely rare driver mutations. Here we are reaching natural limits with purely national data collection due to the low number of patients. It is all the more important that the standards we define for Germany harmonize with existing international standards.

Roche would generate a public use file that could be used openly for research?
Lattrich: In this context, we are currently gaining valuable experience with our strategic partner Flatiron Health. In cooperation with clinics and research institutions in the USA, the company has quickly established a platform that collects, networks and analyzes data from oncological care in a standardized manner. We are now talking about data from more than two million cancer patients. And this data can be used by regulatory authorities, clinics, science and other research companies. Of course, the general conditions in the USA – and certainly the US mindset – cannot be transferred one-to-one to Germany. But the example shows that solutions have long been possible - and the industry can make a valuable contribution here.

Bruns: In fact, Flatiron is basically leading the way by creating new evidence for science at the patient level. The disadvantage is that this creates a data monopoly and possibly a mental block when dealing with the data. The combination of digital patient data with a register-based evidence file would fundamentally generate a high benefit for the patient and for the decision-maker in the treatment process.

Lattrich: That's right, the potential for care and research is enormous - it is all the more important that mental barriers are overcome and the common goal takes center stage. Of course, companies like Flatiron Health and Foundation Medicine are now part of the Roche Group - but above all, they have developed solutions that can already offer added value for the treatment of cancer patients. And it is certainly clear that the number of providers of digital solutions will continue to increase in the future.

Different providers of corresponding solutions would be possible?
Bruns: Basically, you will only get sufficient data diversity through a large number of service providers. There have to be rules that you have to adhere to as a basic requirement for participation in the system. It is important that these are the same rules for everyone. Because when in doubt, all sides have their own “hidden agenda”.

What can be done to ensure that a variety of service providers works while adhering to the same rules?
Bruns: Politicians will certainly promote or at least accept the issue of service provider diversity because they don't want to create monopolies. That's why clear rules are needed that are formulated without self-interest and that reach as broad a consensus as possible.

But it's about time: According to current figures from the DGHO, the prevalence and incidence of oncological cases will increase dramatically in the coming years due to the demographic increase of older people in our society alone. What could the supply of the future look like?
Lattrich: For us there is no question that we will only exploit the potential inherent in digitalization and general scientific progress in the fight against cancer if we establish research-based care on a broad scale. We must capture the knowledge that is generated in huge quantities every day in routine care and make it usable for both therapeutic decisions and the research and development of future therapies.

Bruns: Right, and for that we need data first and foremost. If we don't document what care looks like now, we won't know what care will look like tomorrow. This was the driving point behind the introduction of clinical cancer registries.

How will the future of cancer care continue to change from an industry perspective?
Lattrich: We are already seeing that therapies are being developed more and more specifically for smaller and smaller patient groups. This trend toward personalization of cancer therapies will continue to advance. And that goes hand in hand with a fundamental rethinking of oncology: we no longer see cancer as just a disease of organs, but above all as a disease of genes. Cross-entity treatment strategies that specifically target individual driver mutations, regardless of the organs affected, will continue to increase in the future and pose completely new challenges for research and development, and in a second step also for care.

For example?
Lattrich: Let’s take our test substance entrectinib as an example. This is aimed specifically at individual driver mutations that, on the one hand, play a role in various types of cancer, but on the other hand are extremely rare. The challenge is to find the patients who are suitable for treatment with entrectinib. For research, this means, for example, that we cannot set up extensive phase III study programs here - we would need decades to recruit the appropriate number of patients. At the same time, it would not be ethically justifiable to treat patients who are candidates for this highly specific therapy with chemotherapy or even a placebo in a comparison arm. This also requires a rethink: More flexible and modern study designs must be recognized, as must, for example, the use of real world data to generate virtual comparison arms.

The approval authorities have already responded with “conditional approvals” for diseases with a high “medical need”. How do you feel about this, Dr. Bruns?
Bruns: The current commitment is that the health systems allow market access without any major additional requirements. However, the systems could also require that medications with a “conditional approval” either only be used in centers and/or that corresponding data be collected. The issue of data transparency then arises with the decision to approve Phase II, because the resulting deal essentially creates the data that is actually to be collected in Phase III together. This means: Anyone who wants to receive approval in the future based on “conditional approval” must be prepared to transparently share the data collected afterwards with science.

As an act of shared responsibility for better care?
Lattrich: That's right, when it comes to generating data, it's obviously not about someone having the data and sitting on it. In order for the patient to benefit from it, the data must be made available to the various participants in the healthcare system.

And suddenly the demand for complete data transparency arises.
Bruns: The industry is always relatively quickly exposed to accusations that it wants to keep data opaque. However, that is not true. As long as data is collected as part of development, there are legitimate patent interests. At this point two completely different systems meet. On the one hand, there is the supply and self-administration side, which works on a non-profit basis and is funded by contributions. On the other hand, the pharmaceutical companies, which are committed to shareholder value and their corporate strategy, usually have to give in due to their weaker position as a demanding and supplying industry. For me, the demand for total data transparency is idiotic. Anyone who wants private research and development must also be guided by the principles of the market economy. If one day the industry really has to disclose all of its study data after phase I, II or even III, there will be no more innovations.

Back to accelerated approval: What if the associated hope for patient benefit is not confirmed in practice?
Bruns: If this hope does not materialize, the ripcord must be pulled just as quickly to prevent harm to the patient. However, this requires transparent data collection from the real supply, which is why the Bundestag correctly introduced the accompanying information after approval with the GSAV.

Lattrich: Our goal is that every patient receives exactly the therapy that optimally supports them in the fight against their individual illness. Conversely, of course, this also means: We have no interest in our medicines being used on patients who do not benefit from them. In this context, however, it is important that the evidence from the data from practice is then recognized accordingly.

If this data is not backed up with a certain level of legality, IQWiG will probably not recognize this register data.
Bruns: That would of course be fatal. If the legislature considers this data to be relevant, it must also be recognized, otherwise the effort would be absolute nonsense. But it probably won't come to that, because if the G-BA demands accompanying data collection from the pharmaceutical company, it must also commit to classifying it as relevant.

Is this the beginning of the “brave new world” of real world evidence?
Bruns: We will certainly only be happier if they are collected in a structured manner.

Lattrich: If we succeed together in collecting the data in a structured manner and thus making it usable in the first place, then we will certainly be able to accelerate progress in oncology again.

Then let's look into the future: Where will we be in cancer treatment in ten years? Is Federal Health Minister Spahn right that cancer can be defeated in just a decade?
Bruns: I don't believe it. That's ambitious, you have to give him that, but the timeline isn't right. Unless the topic is addressed in a much more comprehensive way than before.

Lattrich: Medical progress in oncology is enormous – and modern therapy concepts have made a decisive contribution to this. Let's look back 25 years: At that time, for the vast majority of patients, there was basically only chemotherapy, which was administered according to the watering can principle. Today, highly specific therapies that target the tumor are firmly established in many types of cancer. But it is also clear that we will only defeat cancer with a holistic approach - and this also includes the question: How can we prevent cancer from developing in the first place?

However, a relevant proportion of cancer cases, around 30 to 40 percent, are lifestyle-related.
Bruns: This is about interaction and the accumulation of risks. This is also a data issue. When we talk about the future of oncology, it's easy to say that the future of oncology will be increasingly data-driven.

But everything must not only be documented, but also be able to be scientifically evaluated. It must be possible to generate knowledge from any documentation, including from the new accompanying data collection, in order to gradually reduce the existing knowledge gap between scientific evidence and the reality of care.
Lattrich: Right. We have to close the gap between what is already theoretically possible today and widespread care - and for this, standardized collection of data from routine practice is simply essential.

Bruns: That's why we need standardized evidence and patient files. And that's why you have to define in advance exactly what you want to know and what benefit the knowledge to be aggregated will have for the patient, the doctor and society. Then there will also be sufficient compliance not only from patients, but also from the doctors and medical professionals who have to carry out the documentation.

This also means a bit of a cultural change, as at the moment documentation is actually only documented for billing purposes or, in case of doubt, for backup purposes.
Bruns: That's right, in many clinics today the idea of being able to store the collected data somewhere in the basement is actually still documented, but not of making the data available to the healthcare system and science. The latter would be the more sensible approach because the data continuum would then grow with each patient treated and the knowledge acquired would not be lost when the treatment ends, the patient changes sector or dies. This includes not only treatment data, but also data from biobanks of living or deceased patients. The name of the patient is completely unimportant in order to take the wind out of the sails of data protection officers. This is about tumors with specific sequences that need to be submitted to meta-analyses together with patient histories.

That would be the learning or research supply.
Bruns: Right. Of course, we are making the world of medicine more uncomfortable than before. That's why you have to think about solutions and ways to make documentation work easier at the same time. This also includes appropriate financing, perhaps even the training of assistants who could carry out data collection much more efficiently.

Lattrich: In this context, we certainly have to convey more strongly that the doctor has a clear advantage through data collection. When we talk about existing assistance systems, it's never about replacing the doctor - it's about relieving the doctor so that he can have more time for treating his patients or doing research.

But assistance systems are one thing...
Lattrich: … and decision-making systems are another thing. If such systems run in the background, they can check plausibility in the documentation and suggest improvements. But once again: Corresponding systems can never replace the doctor and the doctor's decision, but they can help to improve treatment and documentation in the interests of the patient.
The doctor may even be happy to document it if he can derive direct benefit from it.
Bruns: Ultimately, the most important analogous advance in recent years has been the introduction of tumor boards, which aggregate the knowledge of several minds to make a decision. Expanding this knowledge to include data from real care is a logical consequence. Unfortunately, too little is known about how companies like Flatiron or Foundation Medicine work.

Lattrich: The commonality between Flatiron Health and Foundation
Medicine is that they use digital technologies to network and analyze data and ultimately make it usable for research and patient care. Of course, these companies bring changes to our system - and it comes as no surprise to us that these changes are sometimes met with skepticism. But I am firmly convinced that the services and solutions offered by Flatiron Health and Foundation Medicine will be a natural part of routine care in just a few years, as they can ultimately support the doctor in finding the optimal therapy for his patient to select.

Dr. Bruns and Dr. Lattrich, thank you very much for the interview. <<

The interview was conducted by MVF editor-in-chief Peter Stegmaier, and participation in the preparation of the questions: Prof. Dr. Reinhold Roski.

Citation note: