Outcomes trials: the foundation of evidence-based cardiovascular medicine

Friday, 14 November 2014

Outcomes trials are the biggest investment a pharma company will make in cardiovascular (CV) R&D – enrolling tens of thousands of patients and costing up to US$500 million. They are also extremely important, because they are the most effective way to prove that an intervention reduces key medical events. In the context of cardiovascular (CV) medicine, this means major morbidity (e.g. myocardial infarction [MI] or stroke) and death.

For us here at AstraZeneca, high level results from a key outcomes trial – PEGASUS – are expected in Q1 2015. PEGASUS randomised patients with a history of MI and risk factors to either ticagrelor or matched placebo, with a primary endpoint of MI, stroke or CV death(1).

Release of the PEGASUS data will, of course, be exciting for patients and for those of us involved in the trial. However, I want to use this blog post to reflect more broadly on the importance of CV outcomes trials and the work that goes into making them happen.

Outcomes trials – why do them?

Rather than focusing on surrogate endpoints like blood pressure or lipid levels, outcomes trials answer the most important questions about a CV intervention: Does it prevent major morbidity? And does it reduce mortality? Ultimately, these are the questions that all stakeholders – including patients, physicians, payers, regulators and the pharma industry – want to see answered.

CV medicine was the ‘birthplace’ of this type of study and remains the disease area in which most are performed. However, there are other fields in which outcomes trials are also vitally important, for example oncology and respiratory disease.

Why are CV outcomes trials so large?

In the CV arena, most outcomes trials recruit at least 15,000 patients; PEGASUS enrolled > 21,000(1). There are two key reasons for this. First, substantial numbers of patients are needed to achieve a truly representative patient population. Second, the statistical considerations require it: even in high-risk patients, the number of major events over a manageable study period is unlikely to be large. For example, PEGASUS was planned to continue until 1,360 patients had experienced the primary endpoint, which equates to about 6.5% of study subjects(1).

What are the key challenges?

Given their size, there are major scientific and operational challenge in managing an outcomes trial. Substantial numbers of participating centres are needed, and inevitably these will have to come from many different countries. For example, PEGASUS involved around 1,150 sites from > 30 nations across several continents. Apart from meeting the logistical needs and ensuring that regulatory requirements for all locations are met, this broad spread of participating countries has the extra advantage of making the data relevant across the world(1).

Effective and realistic study planning is essential. The centres involved need to be carefully selected, and the inclusion of patients closely monitored. This is to secure that the right type of patients are randomised and to monitor patient safety carefully. An independent Data and Monitoring Board has the responsibility to regularly review efficacy and safety and to recommend changes if necessary.

Collaboration is crucial in managing these trials. For example, in the running of PEGASUS, AstraZeneca is partnering with TIMI, a US academic group that has vast experience with large CV trials – including other collaborations with AstraZeneca, such as the SAVOR and DECLARE studies.

What about data collection?

Managing the study database from a large outcomes trial creates significant logistical complexities. In this regard, it is important that the goals of the study remain focused, and to collect only the most relevant information.

Data collection is a multi-step process that starts with the patient and data entry. It is important to monitor this process to ensure accuracy. Nowadays, this is often supported by remote central monitoring, which for many purposes is more cost effective and no less sensitive than manual checking. Data are transferred electronically to the central database and then further cleaned up. The final, quality controlled data are analysed and may be shared with academic partners when the study is finished, which allows for independent verification.

What is the impact of outcomes trials?

The vast databases that come from these trials are extremely valuable for understanding the effects of an intervention but also the disease. As a result an outcome trial often leads to multiple publications – sometime dozens – and the data generated frequently go beyond the primary objectives of the study. For example, the data can sometimes go beyond answering study-specific questions and also be used for other more general medical questions related to the study population. In addition, outcomes data are often very important in the planning of new studies, thereby leaving a lasting legacy.

At AstraZeneca, we now have substantial experience in planning and managing outcomes trials. They represent a major scientific and logistical challenge, but also provide some of the most important data available to us, and to the wider medical community.


(1) Bonaca et al. (2014). Design and rationale for the Prevention of Cardiovascular Events in Patients With Prior Heart Attack Using Ticagrelor Compared to Placebo on a Background of Aspirin-Thrombolysis in Myocardial Infarction 54 (PEGASUS-TIMI 54) trial. Am Heart J 167(4):437-444.