We identified references through searches of PubMed with the terms “endpoints”, “outcomes”, “oncology”, and “trial design”. Only papers published in English were reviewed and there was no defined time period for published articles. Articles were also identified through searches of the authors' own files and cited articles. The final reference list was generated on the basis of originality and relevance to the broad scope of this topic.
ReviewOutcomes and endpoints in cancer trials: bridging the divide
Introduction
The key priorities in the treatment of cancer are to enable affected individuals to live longer or better, and ideally both, than they would without therapy.1 The discussion of outcomes and endpoints in oncology trials needs to take into account cancer type and therapeutic modality as these parameters affect the expectations of both clinicians and patients. Our accompanying Review2 discussed the benefits and challenges of use of overall survival as a trial endpoint including effects of crossover, treatment given after disease progression, and the use of supportive care. Important questions remain unclear, such as who decides what is an appropriate endpoint? How should choice of clinical trial endpoints be explained to patients? Are patient-reported outcomes underappreciated and underused in oncology trials? How should endpoints be implemented to better reflect clinical benefit for patients? And how should these endpoints be integrated into trial design and clinical practice?
Reproducible, valid, and appropriate endpoints are needed to optimise the design and development of clinical trials while ensuring patient safety. Of equal importance is the need to design trials that make the best use of available resources and have clear endpoints of clinical benefit. If intermediate or surrogate endpoints are used, they need to be assessed to ensure they remain a valid alternative. The validity of surrogate endpoints might change with time and their use in trials of novel therapies will inevitably require untestable assumptions. A review of US Food and Drug Administration (FDA) drug approvals spanning 20 years showed that time-to-event endpoints such as time to progression or progression-free survival (PFS) were increasingly used for drug approval—43% of approvals (2006–11) were based on time-to-event endpoints, compared with 33% in the previous 6 years (2000–05) and only 13% in the previous decade (1990–99).3 One approach might be to periodically reassess whether surrogate endpoints remain clinically relevant in view of changing environments and new treatments. In this Review, we discuss important considerations and limitations to achievement of measures of outcome tailored to cancer type, and ultimately, devise a plan to navigate the way forward.
Section snippets
Who decides what is important?
Clinicians, patients, regulatory agencies, and industry are all stakeholders who have an interest in the determination of which endpoints are of clinical relevance. Regulatory agencies carry the burden of responsibility for making the ultimate decision about drug approval. Integral to this decision is the balance between a comprehensive review of safety and efficacy data, and the drive for better therapeutic options for the patient.2, 3 Although the demonstration of a survival benefit is the
How should decisions about endpoints be explained to patients?
Underpinning the controversy about the selection of trial endpoints are life and death questions for patients. Clinical trials are done for reasons that are obvious to investigators, but might be unclear to the general public and often unappealing to patients and even sometimes their treating physicians.
Overall survival might not be a realistically measurable endpoint in settings in which the disease prognosis is good.4, 9 Trials will, therefore, include surrogate endpoints that represent the
Are patient-reported outcome measures underappreciated and underused in oncology?
Patient-reported outcome measures are important in clinical decision making provided they address more than just symptoms. Validated methods to assess quality of life should be used not only to address the providers' questions, but also to ensure the best patient care possible. These measures are increasingly recognised as direct measures of patient benefit and as independent endpoints. The importance of patient-reported outcomes is paramount in the interpretation of gains in survival and to
How should endpoints be implemented to better reflect clinical benefit for patients?
Development of reliable endpoints to allow for the early recognition of clinical benefit to patients is an important goal of clinical trials (figure 1). The issue of using surrogate endpoints is that present measures (eg, RECIST) need time for accurate evaluation and assessment.
PET imaging has been advocated to predict response more accurately and definitively than CT for specific tumour types.28, 29, 30, 31 PET has the ability to differentiate viable tumour tissue from treatment-induced
How should endpoints be integrated into trials and clinical practice?
Drug development has remained substantially unchanged for the past 20 years;1, 43 however, new trial designs challenge this structure.44, 45, 46, 47, 48, 49 Increased research costs have the potential to constrain drug development and affect access for patients to effective therapies.50, 51 Therefore, it is time to challenge the traditional design of clinical trials (Figure 1, Figure 2).
First, the philosophy of phase 3 trials needs to adapt from that of whether the group benefits to who (if
Choosing appropriate endpoints
If the design of trials is evolving, should trial endpoints evolve in parallel? Integration of appropriate endpoints in the design of clinical trials is paramount to the success of a trial. Inappropriate endpoints and poor design might halt the development of a promising drug. For example, gefitinib was first approved for use in advanced non-small-cell lung cancer in the setting of third-line treatment after an initial randomised trial showed that 10% of patients responded to treatment for a
Translating trial results into routine practice
Elderly patients, those with comorbidities or from an ethnic minority, or a socioeconomically disadvantaged population are often under-represented in clinical trials.93, 94 Observational studies after drug approval are necessary to determine whether the clinical benefit identified in trials confers a real benefit in the general population. Side-effects from a drug are often underestimated yet are of key importance to patients.11 Toxic effects from drugs when they are transferred to the
Conclusions and future directions
The aim in designing trials in oncology is to improve patient survival and quality of life, and, ideally, both (panel).1 In our accompanying Review,2 we presented the controversy underpinning current endpoints. In this Review, we provide evidence that the heterogeneity of cancer necessitates individualisation of the design of clinical trials and, therefore, endpoints. This shift in trial design will facilitate meaningful decisions about the clinical implications (ie, benefit to patients) of a
Search and selection criteria
References (100)
- et al.
Endpoints in clinical trials: what do our patients consider important? A survey of the Ovarian Cancer National Alliance (OCNA)
Gynecol Oncol
(2013) - et al.
Quality of life and/or symptom control in randomized clinical trials for patients with advanced cancer
Annal Oncol
(2007) - et al.
2004 consensus statements on the management of ovarian cancer: final document of the 3rd International Gynecologic Cancer Intergroup Ovarian Cancer Consensus Conference (GCIG OCCC 2004)
Ann Oncol
(2005) - et al.
New response evaluation criteria in solid tumours: revised RECIST guideline (version 1.1)
Eur J Cancer
(2009) - et al.
Reporting quality of life in clinical trials: a CONSORT extension
Lancet
(2011) - et al.
Patient-reported outcomes in ovarian cancer clinical trials
Annals Oncol
(2013) - et al.
Early pharmacodynamic assessment using 18F-fluorodeoxyglucose positron-emission tomography on molecular targeted therapy and cytotoxic chemotherapy for clinical outcome prediction
Clin Lung Cancer
(2014) - et al.
Window-of-opportunity trials to evaluate clinical activity of new molecular entities in oncology
Ann Oncol
(2011) - et al.
Dynamic reprogramming of the kinome in response to targeted MEK inhibition in triple-negative breast cancer
Cell
(2012) - et al.
STAMPEDE: systemic therapy for advancing or metastatic prostate cancer—a multi-arm multi-stage randomised controlled trial
Clin Oncol
(2008)