Evidence generation for innovative products in oncology

With a number of potential breakthrough therapies in the pipeline, there is a great need to discuss the type and amount of evidence needed to get these products on the market. Evidence not only during development, but also thereafter, since a substantial proportion can only be generated post-launch.

During the FIGON Dutch Medicines Days on October 1-2, the RSNN organized two regulatory science sessions around novel cancer treatments and the associated regulatory challenges. During the afternoon session, which was chaired by Dr Marjon Pasmooij (CBG-MEB), four speakers discussed evidence generation for oncology products. They questioned, together with the audience, whether or not randomized controlled trial design (RCT) is still the best trial design for studying ATMPs, and wondered how real-world evidence (RWE) fits in. Moreover, how do we have to deal with the fact that a substantial proportion of the evidence can only be generated post-launch. Again, it became clear that exchanging knowledge and information during the whole development process is essential.

Innovative evidence generation in oncology drug development

The first speaker was Dr Paula van Hennik (CBG-MEB). In regulatory decision making it is all about the assessment of benefit/risk, weighing up the favourable and unfavourable effects, dealing with uncertainties. But then, after marketing authorisation, still more decisions need to be made before a patient is treated with the new drug. Will the cost of the drug be reimbursed? Will it be included in the professional guidelines? Will it really be prescribed following discussion between doctor and patient (informed consent)? These are questions that, when not answered in a timely manner, could lead to potential delays in the introduction of new treatments. Dr van Hennik, therefore, pleaded for a systems approach to speed up development of new oncology drugs (Figure 1).

Figure 1: Scheme of the “total system” and its major elements, presented by Dr Paula van Hennik

Evidence generation in this total system seems best carried out through the double-blind randomised controlled trial (RCT), which is the gold standard for demonstrating the efficacy and safety of new drugs. RCTs have a minimum of bias by design but have known difficulties with respect to lack of heterogeneity and consequently the generalisability of the results. So, they are not perfect and moreover they are being perceived as slow, complex, and expensive. Dr van Hennik discussed some alternative designs, such as uncontrolled studies, modelling & simulation, extrapolation, and the use of real-world evidence, which all have their limitations because of their inability to control bias. She concluded that all these alternative routes provide important evidence, which is not only complementary to the data from RCTs, but also can support the interpretation of data from these trials. However, until now, in her opinion RCTs remain the backbone for evidence generation on treatment effects.

Integrated evidence generation

Figure 2: Pharma’s integrated approach, presented by Alwin Otten

Next, Alwin Otten (Janssen) gave an industry perspective on evidence generation and said that a rethink in pharma is needed. The essential element of evidence generation in pharma is an “integrated approach”, i.e., the integration of different types of evidence at the appropriate time throughout the drug development process. In his view, this “integrated approach” asks for a culture within the company of “how do we bring the right functions, expertise and processes together at the right time?” in the interest of a pro-active (and global) approach to evidence generation and that Regulatory Affairs is in the centre of this (Figure 2).

A major challenge for oncology drugs that many companies face is the requirement of HTA bodies to demonstrate overall survival (OS) benefit. The problem is that HTA methods for assessing new treatments were developed at a time when drugs would deliver at best 2-4 months additional OS and so in the past it was easy for companies to supply HTA bodies with OS information since most patients had died by that time. However, the more recent innovative medicines are much more effective and some of these new drugs do not even reach median progression free survival at the end of the clinical trial: “So, we need to identify alternative strategies, otherwise it will take too long for patients to benefit from these new innovative therapies.” Therefore, pharma companies do not want to wait for OS data before applying for reimbursement. Potential solutions for this, Alwin Otten mentioned, among others, were the acceptance of surrogate endpoints, adaptive pathways and PRIME as early access tools, and the use of real-world evidence (RWE). Within Janssen, they see a full set of opportunities for using RWE during the full product lifecycle, both pre- and post-launch.

The EBMT Registry and CAR-T Cells

Patient registries can be an important source for evidence generation, also for ATMPs. Prof Jürgen Kuball (UMC Utrecht), introduced the audience to the European Bone Marrow Treatment (EBMT) Registry which started in 1970 and is the biggest data source of its kind in Europe. The registry contains data from more than 500 medical centres in more than 50 countries. It receives 30,000 new registrations per year and currently contains data on more than 500,000 Hematopoietic Stem Cell Transplantation procedures. The EBMT registry can register real-life data and even rare and long-term adverse events and make comparisons with clinical results from other (gold standard) therapeutic approaches. A major advantage of the EBMT registry is the ability to follow-up on CAR-T cell treatments. Intensive discussions between the EMA and EMBT registry have taken place on its qualification as a suitable data source for post-authorisation follow-up and risk-benefit evaluations of CAR-T therapy. Assuming that the administration of CAR-T cells will take place within current dedicated centres, the Cellular Therapy form of the EBMT registry should provide a satisfactory coverage of the target population regardless of their transplant activity.

In general, the usefulness of a registry stands or falls with the completeness and quality of the data. In order to reach maximum completeness and power, Prof Jürgen Kuball’s team at EBMT works on continuous simplification of the datasets: as he called it: “Less is more”.

The value of real-world evidence in reimbursement decisions

The last speaker Brenda Leeneman (Erasmus University Rotterdam) introduced the audience to the Dutch Melanoma Treatment Registry (DMTR). This registry was set up in 2012 as a response to the request of the Dutch Minister of Health who made the reimbursement of ipilimumab conditional on: 1: centralisation of care, and 2: the setting-up of a nationwide, population-based registry. It forms a good example of the timely HTA-involvement by ZIN.
The main aims of the DMTR are clinical auditing to improve melanoma care, improving transparency of the quality of melanoma care by registration of quality-indicators, providing insights into real-world outcomes and costs, and creating a platform for research. All patients with unresectable stage IIIC or IV melanoma are registered.

Brenda Leeneman showed the results of a study of the real-world use, survival rates, and safety of ipilimumab. Given the subject of this session the focus was put on the differences between the real-world and the RCT outcomes. She indicated that, although RCTs are crucial for establishing efficacy, RWE is of great added value for enhancing the generalisability of outcomes in clinical practice. Moreover, costs and cost-effectiveness can not only be estimated for one treatment, but also for treatment sequences.

As for remaining challenges in the use of RWE, she mentioned amongst other that RWE outcomes may not always be available at the time of the initial reimbursement decision, how to realise (better) correcting for confounders, and how to deal with missing data. In conclusion she stated that “Real-world data can complement RCT data and may bridge the gap between clinical research and clinical practice”.

Panel discussion

During the panel discussion with all speakers from the afternoon session, the debate on the use of real-world data in relation to RCT data was continued. It was concluded that RWE cannot be a substitute for RCT data. The use of an RCT and/or RWE depends on what you want to know! Moreover, it also depends on the definition of real-world data.

The overall conclusion of the day was that all the different stakeholders within the development process should cooperate. Preferably, in an early stage of the process, one should discuss the type of evidence that needs to be generated. The regulatory process should not obstruct developers and therefore, exchanging knowledge and information during the whole development process is essential. We all want to avoid the situation in which the medicinal product is registered but will not be reimbursed or prescribed.