Drug development in oncology: status quo and outlook with special consideration of surrogate endpoints

In 2011, 30 new agents were approved in the USA and 22 in Europe, including 8 and 7 for the treatment of oncological indications, respectively (1). This trend towards an increasing proportion of oncological drugs, which has already been observed for several years, is continuing in 2012, as current figures from the US Food and Drug Administration (FDA) show: one third – 13 of the 39 newly approved active substances – relate to oncological areas of application (2). Approval in oncology often takes place as part of accelerated approval procedures (“conditional marketing authorization” or “accelerated approval”) and/or as an “orphan drug” for rare diseases. The main reasons given for the significant increase in oncological active substances are: progress in the molecular-biological characterization of tumor growth, development of a large number of new "targeted" active substances, great need for more effective drug therapy options for advanced tumor diseases, demographic development, expiry of patents for blockbusters on Treatment of common diseases and a very lucrative market for oncological drugs.

>> The approval or expansion of the indication(s) for all oncological drugs in Europe today takes place in accordance with Regulation (EC) No. 726/2004 in a centralized procedure by the European Medicines Agency (EMA) - based on phase III clinical studies or II. In a recently revised “Guideline on the evaluation of anticancer medicinal products in man”, which is scheduled to come into force on July 1, 2013, the EMA describes regulatory requirements in detail and aligns them with current developments in drugs in oncology ( e.g. biomarkers, new designs of clinical studies) (3). Recommendations are made for both exploratory phase I/II studies and confirmatory phase III studies, including the characterization of the patient collective and the tumor disease, the type of clinical study and its design, the inclusion and exclusion criteria and the clinically relevant endpoints of the study (3). In addition, methodological considerations are made in appendices, including the use of progression-free survival or disease-free survival (PFS or DFS) in confirmatory phase III studies and the Symptoms reported by patients or health-related quality of life (“patient reported outcome”, PRO, or “health-related quality of life”, HRQoL).
Between January 1990 and November 2002, the FDA approved a total of 71 oncological active ingredients or granted an extension of the indication(s). For more than two thirds of these active ingredients, regular approval was not based on an increase in overall survival (OS), but on endpoints such as objective response rate (ORR), PFS or reduction in tumor-specific symptoms (4) . An analysis published in 2005 of the approvals granted by the EMA for 14 oncological active substances and 27 oncological indications showed that only 7 % of these approvals were based on the primary endpoint and for around 90 % of the approvals, surrogate endpoints such as PFS, time to disease were used as proof of effectiveness progression of the disease and (O)RR were used (5).
Table 1 shows the recently updated evaluation of the drug approvals granted between January 1, 2009 and February 28, 2013 for the treatment of hematological neoplasia and solid tumors by the Drug Commission of the German Medical Association (AkdÄ). The information in Table 1 was extracted from the EMA's publicly available European Public Assessment Report (EPAR). For the treatment of hematological neoplasia and solid tumors, EPARs for 34 active substances in 47 indications were published during the evaluation period. It was mainly an extension of the indication(s) (57 %). A total of 53 clinical studies (“pivotal trials”) were presented, the results of which were supported by other (often uncontrolled) clinical studies. The primary endpoints – OS and PFS – recommended by the EMA and FDA as evidence of efficacy (“benefit”) were examined in a total of 38 clinical studies. Furthermore, PFS (25/53 studies) dominates as the primary endpoint compared to OS (10/53 studies). The previously more commonly used but unsuitable parameter of ORR as a surrogate endpoint in randomized controlled trials (RCT) was only selected as the primary endpoint in 5/53 studies.
This current evaluation by the AkdÄ highlights progress in terms of the quality of approval studies in oncology - especially in the selection of endpoints. However, there are still deficiencies in the approval studies. For example, 89 % approvals were based on the results of only one approval-relevant clinical study, which was more often carried out as an open rather than a double-blind RCT. Despite clear recommendations from the EMA and FDA, data on the PROs or HRQoL (6; 7), which are very relevant for oncological patients, were collected in less than half of the clinical trials. More than two-thirds (70 %) of the pivotal trials recruited fewer than 1,000 patients (Table 1).
Surrogate endpoints – advantages and disadvantages
Good evidence for the efficacy and safety of new oncological agents is based on RCTs and/or meta-analyses or systematic reviews of clinical studies in which meaningful and patient-relevant endpoints such as OS, HRQoL—especially PROs—should be examined (6). However, numerous studies in recent years have shown that as a result of the design and conduct of RCTs in oncology and the interpretation of their results, positive effects such as efficacy are often overestimated and negative effects, especially side effects, are insufficiently recorded (1). In oncology, surrogate endpoints are used as replacements for clinical and patient-relevant endpoints because they allow conclusions to be drawn about new active ingredients more quickly and are sometimes easier to examine. For the sponsors of clinical drug trials, predominantly pharmaceutical companies, this means a shorter study duration, smaller patient numbers, and lower costs of clinical trials, and it also enables faster market access (8). A recently published "metaepidemiological" study analyzed 84 clinical studies with surrogate endpoints and 101 clinical studies with patient-relevant endpoints - including oncological studies - that were published in leading medical journals in 2005 and 2006 (7). A comparison of these studies revealed that clinical studies that use surrogate parameters as primary endpoints report significantly greater treatment effects than those with patient-relevant endpoints (8). However, what is crucial when using surrogate endpoints, especially in oncology, is that they correctly predict positive effects of new oncological active ingredients and that their predictive value with regard to patient-relevant endpoints - especially survival and PROs - has actually been validated using meta-analyses (9–12).
Table 2 compares the endpoints commonly used in oncology in the context of approval studies with regard to their significance for regular or accelerated approvals and their requirements for study design. In addition, advantages and disadvantages of endpoints are presented in keywords in Table 2. The table was taken from a review by FDA scientists (6). Due to the significant increase in accelerated approval procedures or approval of "orphan drugs" in oncology - both in Europe and in the USA - surrogate endpoints have become more important, but are also becoming new in terms of their significance as patient-relevant evidence of benefit or additional benefit approved active ingredients are questioned (10-12). It is undisputed that the primary endpoint OS continues to be considered the gold standard in oncology (6). Meta-analyses of the surrogate endpoints used in oncology studies and studies of the extent of the correlation between surrogate endpoints and OS in different solid tumors have shown that PFS, as the most commonly used surrogate endpoint today, only has a good correlation in a few solid tumors such as advanced colorectal cancer or ovarian cancer on overall survival in RCT shows. PFS can therefore currently only be considered a validated surrogate endpoint for these tumor diseases (10–12). In contrast, for other common solid tumors, such as breast, non-small cell bronchial, renal cell and prostate cancer, there is still a lack of good evidence that surrogate markers used as primary endpoints reliably predict positive effects of new active ingredients relevant to patients in registration studies (10) . The use of composite endpoints, such as the combination of PFS and PROs, is therefore increasingly required in oncology, since with low effect strengths of a new active ingredient in terms of extending the PFS (e.g. by three to six months), there is necessarily a significant reduction Tumor-related symptoms (PROs) that are often crucial for the patient's quality of life must be demonstrated (11).
Statistical modeling of the relationship between PFS and OS suggests that a highly significant extension of PFS with a median extension of survival after tumor progression (“survival post-progression”, SPP) of two months is also very likely (90 %). is associated with a statistically significant prolongation of OS and therefore represents a valid surrogate endpoint. In contrast, the probability of correctly predicting a significant prolongation of OS based on PFS is only 20% at an SPP of 24 months (9). This statement was recently confirmed by analyzes of the relationship between PFS and OS in control and experimental arms of RCTs carried out over the past ten years as part of the FDA approval of new oncological agents (11). They confirm that the use of PFS is an uninformative substitute for the gold standard OS when patients survive for a long time (>24 months) after progression of their solid tumor, but PFS correlates well with OS in relatively short survival after progression. It is important to take into account that the selection and significance of surrogate endpoints is influenced by the type and natural course of the tumor disease, the strength and duration of the effects observed in clinical studies, but also by the availability of effective drug therapy options after the disease has progressed (10).
Suggestions for improvement: What should change before and after approval?
Improving the data situation for assessing the benefits and also the cost-benefit ratio of newly approved drugs is essential for the rational prescription of cost-intensive active ingredients in oncology. To achieve this, standards for preclinical cancer research must be increased (13), the currently (still) valid recommendations in Regulation (EC) No. 726/2004 must be observed more strictly (3), and deficiencies in clinical studies must be further reduced before approval (14). Given the large number of new drugs to be tested and the increasingly common stratifying therapy strategies, the development and use of innovative study designs is necessary (15; 16). Alternative study designs and validated surrogate endpoints are particularly important for studies in rare diseases. This means, however, that after approval - in Europe mostly as "conditional marketing authorization" - the conditions associated with the approval are quickly met and the knowledge that is often still missing for the care of oncological patients under everyday conditions is collected in further studies.
Numerous proposals have been developed to further improve the quality of RCTs in oncology. Key points are:
• adherence to CONSORT recommendations on study design and publication of study results, including adverse events (17; 18);
• increased funding and implementation of phase IV studies or “post-authorization safety and efficacy studies” (PASS/PAES) with the potential to improve medical care practice (19, 20);
• Evidence of differences in endpoints between control and
arm and experimental arm, which are not only statistically significant but also relevant for patients (21);
• the standardization of optimal supportive therapy (“best supportive care”, BSC) (22);
• the detailed description of practical details of the therapy that are essential for implementing the study results into clinical practice: for example, premedication, the use of growth factors or dose adjustments (23).

A detailed discussion of the challenges and possible solutions for ensuring efficient drug supply in oncology can be found, among other things, in a publication by the “Lancet Oncology Commission” (24), in a report that was prepared for the Federal Ministry of Health as part of the National Cancer Plan (25), and in a working paper from an informal discussion group (26). These statements also address other important aspects of post-approval studies, such as the design or improvement of their regulatory and financial framework and the reduction of their administrative effort (25). <<