Authors' objectives: The KCE was asked to investigate to what extent a selection of innovative oncological drugs that were introduced in the last 15 years have provided added value for the Belgian population and whether the expenditure for this represented efficient use of the available resources.

Authors' results and conclusions: 1. The European Medicines Agency (EMA) typically markets cancer drugs based on surrogate endpoints or single-arm trials, which do not always translate into the substantiated improved outcomes that matter most to patients, i.e. survival and quality of life. The literature reveals that, even after they have entered the market, a significant degree of uncertainty remains for the majority of oncological drugs with regard to their added value for the patient or it turns out that this added value was limited. Coupled with the high prices charged, these uncertainties put policy makers in a difficult position when deciding which treatments should be reimbursed, all the more so as patients' expectations are high. The approval and reimbursement of expensive drugs with little or no benefit leads to inefficient use of the limited resources available and undermines the affordability and quality of our healthcare system. 2. Data from the Belgian Cancer Registry (BCR) on approximately 891 000 tumours from approximately 814 000 patients (diagnosed between 2004-2017) was made available for the purposes of this study. This data was linked to reimbursement data from the Inter-Mutualistic Agency (IMA) for the purposes of mapping expenditure on oncological drugs. The patient survival data was obtained from the Crossroads Bank for Social Security. These different types of observational data were mapped for 12 types of cancer. 3. 40 different types of innovative oncology drugs were selected from these 12 indications, with further information on efficacy and cost-effectiveness then compiled on them. Given the large number of drugs and indications, we primarily based our investigation on findings from systematic reviews of available RCTs (Randomised Controlled Trials) which mapped information on the drugs’ added value. Our observations on (difficulties in calculating) cost-effectiveness are also based on findings from independent researchers published in Health Technology Assessment (HTA) reports. 4. Indications and time periods in which there is a (strong) increase in gross expenditure/average treatment costs, and no clear improvements in overall survival, may raise doubts about the efficacy in relation to survival (and also, therefore, their cost-effectiveness). In line with findings published in international literature, the analysis of the Belgian observational data indicates that automatically associating 'innovative' oncological drugs with a (great) added value in terms of patient survival is unjustified. 5. The independent economic assessments contained in the HTA reports suggest that the greatest uncertainty concerns the estimation of the effect of treatment on overall survival. This was largely due to the following reasons: a lack of studies directly comparing the intervention and the appropriate comparator(s) (head-to-head studies), immaturity of study data, use of surrogate endpoints with an uncertain association with survival and quality of life, and also an occasional crossover of patients which distorts the intention-to-treatanalyses. 6. Unfortunately, the effect on quality of life is also extremely uncertain. This is primarily due to a failure to measure or only partially measure the effect on quality of life during the comprehensive study follow-up (using a generic utility instrument for the economic evaluations) and/or the non-transparent reporting of the results for these estimates. In some cases, the results for this outcome are even deemed confidential, which should never be the case. Furthermore, the impact on quality of life is often not included in the systematic reviews identified. 7. Systematic reviews prove that the relationship between surrogate endpoints (such as progression-free survival – PFS) and critical patient outcomes, i.e. survival and quality of life, is generally weak. This scientific evidence reveals that the automatic use of surrogate endpoints in clinical and economic evaluations is problematic. Specific validation of these surrogate endpoints is required for the indications and drugs for which they are used. 8. RCTs remain the gold standard for estimating the treatment effect of an intervention relative to a comparator. Scientific studies show that observational data often provides inaccurate information for treatment effect estimates. The promotion of non-randomised database analyses as a quick source of "real-world evidence" related to treatment effect is a false solution. 9. In order to provide rapid access to innovative medicines, reimbursement decisions are increasingly making use of managed entry agreements (MEAs) at a time when uncertainties about the added value of these medicines still remain. In principle, this temporary reimbursement should allow time to provide the necessary evidence. Unfortunately, the system offers no incentive for providing reliable evidence in an effective way . 10. The prices agreed in these MEAs also remain secret, which presents an obstacle to setting more acceptable public prices. This makes the entire system opaque. Prices (for the intervention, the comparator and/or the subsequent treatments) are also confidential in almost all economic evaluations. In such cases, cost-effectiveness cannot be calculated independently. 11. There are two key junctures in the life cycle of a medicine that can be used as leverage for obtaining the required evidence; at the point of granting a marketing authorisation (European level) or the moment of awarding a reimbursement (national level).

Authors' recommendations: TO THE EUROPEAN COMMISSION, THE EUROPEAN MEDICINES AGENCY (EMA) & THE NIHDI: We recommend not focusing primarily on early access to “innovative” oncology drugs. The primary concern should be to provide timely access to medicines for which clear and reliable added value for the patient has been demonstrated. Below we give concrete recommendations to achieve this goal. TO THE EUROPEAN COMMISSION, EMA & COMPANIES: We recommend conducting studies already in the pre-marketing phase that are suitable for registration purposes, reimbursement decisions and support for physicians and patients when taking decisions about treatments. Since it is more difficult to provide additional evidence of effectiveness after marketing authorisation was granted, it is crucial to start the necessary studies in a timely manner. In designing these studies, we recommend that there is more focus on including the correct (active) comparator(s), relevant endpoints (including overall survival and quality of life) and adequate follow-up without inappropriate crossover of patients. • A close collaboration of HTA agencies/payers with the support of EMA for this approach to start up practice-relevant studies in the pre-marketing phase must be legally anchored in European law (see recommendation 5). • Given the often uncertain and limited added value of cancer drugs, randomised studies must be prioritised as the most reliable source for estimating the added value of new interventions. Non-randomised observational data should not simply be regarded as a reliable study design for estimating the treatment effect. • These randomised trials should pay due attention to the following: o Including a population that reflects the future target population. o Incorporating the standard treatment as a comparator. Elements of treatment optimisation (e.g. the duration of the treatment) should also already be evaluated in the pre-marketing phase. o Relevant endpoints are quality of life and survival. These should be included in the studies where possible. o Surrogate endpoints can only be useful where they are sufficiently scientifically validated for the specific condition and mechanism of action of the drug. We recommend following the EUnetHTA guideline whereby data on overall survival as well as quality of life should be systematically collected in the metastatic setting (stage IV). o Measuring quality of life, using both disease-specific and generic utility tools (as also recommended by EUnetHTA). Quality of life should also be measured throughout the full follow-up of the study (e.g. also after disease progression). o Strictly avoiding inappropriate crossover of patients in the study. We recommend strict monitoring regarding the timely and complete reporting of all study results. For example, the impact on quality of life must be reported transparently (i.e. results for all treatment arms and all time points when this outcome was measured). The full results of clinical studies should be made public and never be confidential. Like other key endpoints, quality of life and overall survival should be included in the EPAR (European Public Assessment Report). TO THE MINISTER OF PUBLIC HEALTH AND THE EUROPEAN COMMISSION, REGARDING EMA: We recommend that the European Commission adjust the regulatory framework for the EMA, while respecting the difference in competences between the EMA and the national authorities. It should be enshrined that through mandatory early dialogues, the input of the payers and HTA bodies in the member states is taken into account when drafting the protocol of the confirmatory clinical trials. This will help to prevent the studies designed from not providing the information needed to support subsequent reimbursement decisions (see also recommendations 2 and 3). We recommend the European Commission urging the EMA to make more selective use of conditional marketing authorisation if evidence as to the treatment's effect is insufficient. This approval must then be made conditional, with an explicit requirement to collect the required data within a certain period. Conditional approval should be automatically withdrawn if the necessary studies are not initiated/continued/delivered. This should be sufficient incentive to deliver the required data on time. It should be further investigated which criteria can be used for the selective application of the conditional market authorisation and how compliance with the conditions imposed can be monitored. TO THE NATIONAL HEALTH INSURANCE (NIHDI): When reviewing each submitted dossier, we recommend checking that all study results are present (for all studies initiated and for all endpoints incl. quality of life) when reviewing each file. We recommend only accepting the use of surrogate endpoints where they are sufficiently scientifically validated. We recommend using the system of managed entry agreements (MEAs) more selectively and ensuring the necessary data are actually collected. We recommend striving more for an evidence generation system that also helps to resolve the clinical uncertainties. The moment of the reimbursement decision can be used as leverage to achieve this. To make this run efficiently, we recommend the following: • We emphasize that the uncertainties should initially be addressed at European level (see recommendations regarding EMA). Where there is still a major residual uncertainty, it can be determined at a national level which type of study is necessary to answer the outstanding research questions. International cooperation is recommended for this (e.g. in the context of the BeNeLuxA initiative). • We recommend paying attention to the correct study design for collecting further information in order to answer the original research questions. When using observational data from registers, they must be critically examined whether the available data will be able to provide an answer to the open research question. While registry-based RCTs can be a reliable source for identifying treatment effect, this is questionable with non-randomised registry information. We refer again to other requirements for further research (see recommendations regarding EMA). • This question about the correct study design must be asked when concluding the agreement and must be part of this agreement. Failure to start/continue/complete the study on time should automatically lead to termination of the agreement in order to provide the necessary incentives to carry out this study on time. • In order to make randomised research possible in practice, a restriction on the reimbursement to study patients can be considered in a first phase. An intervention by the payer to execute this study can be considered exceptionally (and put into perspective with the expenditure if the intervention were to be reimbursed without any other condition). This does not alter the fact that the companies must make every effort to start the necessary studies before marketing authorisation is granted. Given the lack of transparency and unsustainability of the current confidential price system, we recommend working with other countries to move towards a system with more transparent and acceptable public prices, which would eliminate/reduce the need for confidential agreements with artificially high public prices. An exception to the confidential prices could be that a lower price is agreed in anticipation of more reliable and relevant study results. We recommend making all the assessment files of all reimbursement applications public (including those from before 2019). No results of clinical studies should be treated as confidential. TO THE BELGIAN CANCER REGISTRY (BCR) & NATIONAL HEALTH INSURANCE (NIHDI): We recommend requesting permission to have permanent access to reimbursement data for a longer period (>5 years), possibly until the patient's death. We recommend making it possible to collect more refined data on, among others, sub-populations based on biomarkers (e.g. HER2 overexpression in breast cancer). This can be done by automatically forwarding the test results from the lab systems or by optimising the use of innovative techniques such as Natural Language Processing. It may also be useful to collect biomarker data in view of a relevant historical control group in the rare cases where a randomised trial is really not possible, especially in very small sub-populations. We also recommend systematically collecting data on progression and relapse. We recommend that due consideration be given in every clinical trial to the measurement and timely and complete reporting of quality of life (see also EUnetHTA guidelines in recommendation 3). TO PHYSICIANS, NURSES, PATIENTS, PATIENT REPRESENTATIVES AND INDEPENDENT RESEARCH INSTITUTIONS: We recommend supporting the demand for more reliable and relevant information about the added value of cancer drugs and demanding data that are important for clinical decisions (such as longevity and quality of life). All parties must be aware that rapid access to innovative medicines only makes sense if there is clear added value for the patient. Rapid access without (generating) sufficient evidence of the drug's added value is detrimental to all parties. The industry, physicians and patients must be aware that reimbursement of a contracted product is temporary and may be discontinued, especially if there are uncertainties about clinical efficacy. TO ALL ACTORS IN SOCIETY, INCLUDING PATIENT REPRESENTATIVES, HEALTHCARE PROVIDERS, INDUSTRY, POLICY MAKERS, THE GENERAL PUBLIC, ETC.: A public debate is desirable on various aspects of the reimbursement of medicines, such as identifying the added value of innovative medicines, (rapid) access, affordability, etc.

Project Status: Completed

Year Published: 2021

URL for published report: https://doi.org/10.57598/R343C

Requestor: National Institute for Health and Disability Insurance (NIHDI)

URL for additional information: https://kce.fgov.be/en/do-innovative-medicines-against-cancer-always-have-a-real-added-value

English language abstract: An English language summary is available

Publication Type: Rapid Review

Country: Belgium

DOI: 10.57598/R343C

Organisation Name: Belgian Health Care Knowledge Centre

Contact Address: Administrative Centre Botanique, Doorbuilding (10th floor), Boulevard du Jardin Botanique 55, B-1000 Brussels, Belgium tel: +32 2 287 33 88 fax: +32 2 287 33 85

Contact Name: info@kce.fgov.be

Contact Email: info@kce.fgov.be

Copyright: Belgian Health Care Knowledge Centre (KCE)