Multinational Trials: Lost in Translation?

Clinical trials have become a multibillion‐dollar business in the United States. Historically, new drugs and devices were evaluated almost exclusively in U.S. academic medical centers. However, over the past decades, clinical trials have increasingly been undertaken outside the United States due largely to increasing costs of U.S.‐based studies, complex regulatory requirements, the lack of timeliness of institutional review boards, lengthy intellectual property discussions with university legal offices, and increased governmental oversight to insure that the Centers for Medicare and Medicaid Services are not billed for study‐related costs. These studies were initially performed in Western Europe or South America, but more recently India, China, and Eastern Europe have become increasingly popular sites for accrual of study patients. It has generally been presumed by both industry and regulatory authorities that studies performed outside of the United States can accurately reflect how drugs or treatment strategies will impact U.S. populations. However, recent evidence suggests that multinational studies may not clearly reflect how a given drug or device will alter the course of disease in a U.S. population.

One recent trial that points out the fundamental risks associated with performing studies outside of the U.S. was the Efficacy of Vasopressin Antagonism in Heart Failure Outcome Study with Tolvaptan (EVEREST) program. ¹ The EVEREST study enrolled more than 4,000 patients from 3 continents and 2 regions in Europe. All patients had been hospitalized for worsening symptoms of heart failure despite optimal medical therapy and were randomized to receive either the vasopressin antagonist Tolvaptin or placebo. Tolvaptan had no effect on long‐term mortality or heart failure‐related morbidity, although it did improve some but not all heart failure signs and symptoms during hospitalization. Inclusion and exclusion criteria had been carefully defined in order to insure that differences in heart failure etiology, heart failure severity, or management practices would not influence the study results.

Nonetheless, after adjustment, South American patients had the highest overall mortality, while Eastern European patients had the lowest cardiovascular death and heart failure hospitalization rate compared with patients in North America. Only 39.9% of patients in South America had coronary artery disease as the etiology of their heart failure compared with 77.4% of patients in North America and 80.5% of patients in Eastern Europe. Patients in North America had the highest rates of comorbidities, including hypertension, hypercholesterolemia, diabetes, chronic kidney disease, severe obstructive lung disease, and peripheral vascular disease, while unadjusted 1‐year mortality was 30.4% in North America compared with only 20.5% in Eastern Europe. The main cause of death in patients in North America was worsening heart failure, while the main cause of death in Eastern Europe was sudden cardiac death.

The marked differences in outcomes across the 3 continents and 2 regions of Europe could be explained by differences in local medical practice, various health care systems, disease etiology, or the use of background heart failure therapies. Although phenotypic differences could certainly account for some of the differences across various continents, genotypic differences could also explain variations in outcomes. Most recently, genotypic differences have been shown to influence the effects of beta blockers in different sets of heart failure patients ² and antithrombotics in patients with coronary artery disease. ³ Thus, it is becoming increasingly clear that while it may be far less expensive to pursue clinical studies in Eastern Europe, India, and China, these studies may not provide data that is representative of how drugs will work in the United States, where the population has admixed genetically for nearly 3 centuries and therefore is unlike the homogenous populations found in Eastern Europe, South America, India, and China.

Ironically, one must wonder whether the sponsors of the EVEREST trial would have published the marked differences in response across the various continents had the study shown that chronic therapy would benefit the overall study population, as the pharmaceutical industry has been reluctant to include genotyping in any multicenter trials or to report intercountry variability. Insuring that clinical trials are performed in a scientifically appropriate manner thus becomes a critical responsibility of the community of clinical and translational scientists (CTS) and requires assiduous efforts to bridge the chasms between T2 and T3. Fulfilling this mission requires a group of community‐based actions: 1) CTS investigators must remain informed of clinical studies that provide important clues regarding population diff erences—a task that this journal will take on; 2) Industry must support genetic studies as part of multinational studies; 3) We must lobby federal regulators to insure that the study populations used in pivotal regulatory trials are representative of U.S. populations; and 4) Academic medical centers must facilitate their ability to eff ectively carry out clinical research.