Generating comparative evidence on new drugs and devices before approval

Fewer than half of new drugs have data on their comparative benefits and harms against existing treatment options at the time of regulatory approval in Europe and the USA. Even when active-comparator trials exist, they might not produce meaningful data to inform decisions in clinical practice and health policy. The uncertainty associated with the paucity of well designed active-comparator trials has been compounded by legal and regulatory changes in Europe and the USA that have created a complex mix of expedited programmes aimed at facilitating faster access to new drugs. Comparative evidence generation is even sparser for medical devices. Some have argued that the current process for regulatory approval needs to generate more evidence that is useful for patients, clinicians, and payers in health-care systems. We propose a set of five key principles relevant to the European Medicines Agency, European medical device regulatory agencies, US Food and Drug Administration, as well as payers, that we believe will provide the necessary incentives for pharmaceutical and device companies to generate comparative data on drugs and devices and assure timely availability of evidence that is useful for decision making. First, labelling should routinely inform patients and clinicians whether comparative data exist on new products. Second, regulators should be more selective in their use of programmes that facilitate drug and device approvals on the basis of incomplete benefit and harm data. Third, regulators should encourage the conduct of randomised trials with active comparators. Fourth, regulators should use prospectively designed network meta-analyses based on existing and future randomised trials. Last, payers should use their policy levers and negotiating power to incentivise the generation of comparative evidence on new and existing drugs and devices, for example, by explicitly considering proven added benefit in pricing and payment decisions.

Introduction

A record-breaking number of new drugs and devices have entered the market in the last 10 years. In 2018, the US Food and Drug Administration (FDA) granted approval to 59 drugs and 106 devices (compared to an average of 28 drug approvals per year during the preceding decade), and the European Medicines Agency (EMA) approved 42 new drugs. In addition to new drugs for established therapeutic areas with large numbers of existing treatment options (eg, antidepressants for depression,¹ statins for coronary heart disease,² and HbA1c-lowering therapies for diabetes³), the research and development pipelines of pharmaceutical and device companies have delivered new therapies for rare diseases.⁴ For example, several new agents are now available for the treatment of multiple myeloma,⁵ chronic myeloid leukemia,⁶ Gaucher disease,⁷ and pulmonary arterial hypertension.⁸ This is good news for patients, since some of these novel therapies have turned out to be beneficial.⁹ For example, drugs like imatinib for chronic myeloid leukaemia and sofosbuvir for hepatitis C have transformed clinical outcomes, improving and extending the lives of patients with these serious and life-threatening conditions.10, 11 However, other new drugs like the HbA1c-lowering rosiglitazone have turned out to have differing benefit–risk profiles than expected in certain populations and subsequently been removed from some markets.12, 13, 14 Also, there have been several major safety crises related to high-risk medical devices, resulting in their market withdrawal, such as pelvic mesh,¹⁵ and metal contraceptive implants.¹⁶

The market entry of larger numbers of new drugs and devices might also paradoxically complicate treatment decisions if little or no data exist on the comparative benefits and harms of new versus existing alternatives. What the treatment of choice for a patient with condition x should be, is a key question for clinical practice.¹⁷ Without data on comparative benefits and harms, it might be difficult for patients and clinicians to identify the appropriate therapy.

In this Series, we describe and highlight some fundamental principles related to developing comparative data on drugs and devices, particularly if multiple options exist to treat the same condition. Our primary focus is on the FDA and EMA, which serve as gatekeepers to the largest pharmaceutical markets worldwide that collectively account for over 60% of total sales. In the USA, the FDA is also responsible for medical device regulation; in the EU, notified bodies designated by national authorities are responsible for conformity assessments of devices (table 1).¹⁸ FDA and EMA are tasked with the goals of granting expeditious access to promising new treatments while also requiring adequate data before approval to protect patients from ineffective and potentially harmful products. Regulatory agencies’ evidence standards for approval shape the quantity and quality of clinical studies generated on new drugs (and also devices in the USA).

In this first paper in the Series, we examine the availability of comparative effectiveness data, and outline how the current regulatory approaches to approving new medicines and devices address the evidence needs of patients, clinicians, and other decision makers in health systems. Recent policy changes aimed at speeding up the development, review, and approval of new products have complicated system-wide efforts to generate comparative data on drugs and devices before and after approval. We therefore propose strategies to improve the future availability of comparative data at the time of market entry. The second paper in the Series focuses on the generation of comparative evidence in the post-marketing period for drugs and devices but also interventions for which often no commercial developer and no dedicated regulatory system exist—eg, surgical interventions. The Comment accompanying the Series analyses the ethical tensions in comparative effectiveness research.