- © 2013 by American Society of Clinical Oncology
Multidimensional Challenges in Clinical Drug Development, Regulatory Approval, and Marketing
- Corresponding author: Alberto Ocana, MD, PhD, Division of Medical Oncology and Hematology, Princess Margaret Hospital, 610 University Ave, Toronto, Ontario M5G 2 M9, Canada; e-mail: albertoo{at}sescam.jccm.es.
To the Editor:
Cortés et al1 recently published an article entitled “Progress Against Solid Tumors in Danger: The Metastatic Breast Cancer Example.” Although we agree that progress against solids tumors is in danger, we feel that some of the challenges in drug development, regulatory approval, and marketing need elaboration.
First, with respect to selection of an optimal time-to-event end point: It is well accepted that an increase in overall survival (OS) and improvement or maintenance of quality of life are definitive end points. However, because of clinical heterogeneity, it can be difficult to observe statistically significant differences in such end points. Progression-free survival (PFS) has been suggested as a surrogate for definitive end points such as OS. However, in diseases with long survival post progression, gains in PFS do not easily translate into gains in OS. This is seen in both statistical modeling and in clinical trials.2,3 Nonetheless, recent history shows that a strong impact on PFS observed with a given drug will usually translate into an increase in OS, even in diseases with long survival post progression. This is the case with trastuzumab-emtansine as second-line therapy in metastatic breast cancer,4 as well as with the addition of pertuzumab to trastuzumab and docetaxel as first-line therapy in metastatic breast cancer.5 When clinical trial design makes it difficult for improvements in PFS to lead to OS improvement, such as in the setting of frequent cross-over from control to experimental arms, other less optimal methods for assessing clinical benefit should be taken into consideration. The inclusion of composite end points of PFS and patient-reported outcome could be a reasonable approach.2
A second issue is biomarker discovery and improved patient selection. In the era of targeted drugs, it has become clear that agents that are directed against a specific molecular target or those that are guided by a biomarker are associated with greater magnitude of clinical benefit than less specific biologic targeted agents.6 When assessing bevacizumab, it is clear that an unidentified subgroup of women with advanced breast cancer do derive benefit. However, the inability to identify this group, together with concerns regarding the safety and tolerability of this drug, has led to limited endorsement of this treatment by the oncology community (at least in some areas of the world).7 The last few years have seen a reduction in successful conversion from early to late clinical drug development. Although the predominant explanation for this is reduced efficacy or increased toxicity of experimental therapy, a number of drug programs failed because of an inability to identify the potential responsive population (as reviewed in the article by Arrowsmith8).
There is also the issue of real-world benefit and the efficacy-effectiveness gap. Patients who are enrolled in clinical trials are more highly selected and may not be representative of patients treated in general practice. Clinical trials also have numerous exclusion criteria, which include particular comorbidities or the use of certain concomitant medications.9 Consequently, clinical trials alone are likely insufficient to inform of benefit of a particular therapy in general oncologic practice, and confirmatory population-based health outcome studies should be conducted to better define the true effect of new therapies. Stark differences in toxicity have been observed between randomized trials and clinical practice. Such differences may be explained by variations in the use of symptom control measures between academic centers conducting clinical trials and community practice.10,11
With respect to cost-benefit, it is a common human reaction to make an association between the price of products and their quality. Unfortunately, this is not often the case in health care.6 Furthermore, a prominent concern in the current era of global financial instability is that the rate of increase in health care spending is becoming unmanageable. The US Centers for Medicare and Medicaid Services reports that in 1965, health care spending was only 5% of the gross domestic product. By 2020, total health expenditures are estimated to rise to 20% of the gross domestic product, or a fifth of the US economy.12 This is clearly not sustainable, and society will soon need to decide on how to prioritize limited funds to best improve the outcomes of patients.
We certainly agree that the progress in solid tumors is in danger, and to avoid that, an improvement in clinical drug development, regulatory approval, and marketing should be our main goal. Clinicians and researchers will need to overcome the hurdles described if we are to make advances that will lead to real improvements in the outcomes of our patients.
AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST
The author(s) indicated no potential conflicts of interest.
