Globalization and Clinical Trials: When Does Foreign Data Apply to You?

The STARGLO Problem

In 2025, Genentech submitted data from a trial called STARGLO to the FDA, seeking approval to expand use of their cancer drug Columvi (glofitamab) for patients with relapsed lymphoma. The overall results looked promising: the drug reduced the risk of death by 41% compared to standard treatment.

There was just one catch. Nearly half the patients in STARGLO were enrolled in Korea, Taiwan, and China. Most of the rest came from Europe and Australia. Only 25 patients (9%) were from the United States, far below the typical 20%.

When FDA statisticians broke down the results by region, they found something alarming. In Asian countries, Columvi worked well. In non-Asian countries, the drug showed a 6% higher risk of death compared to standard treatment. The overall benefit was being driven entirely by Asian patients.

The FDA's advisory committee voted 8-1 that STARGLO data were not applicable to U.S. patients. The agency issued a Complete Response Letter, rejecting the application. European regulators, looking at the same data, approved the drug. The same trial, the same results, but different conclusions about who it helps.

STARGLO wasn't an isolated case. The FDA has flagged approximately 25 applications based heavily on single-country or Asia-centric data, mostly in oncology. Several have been rejected or withdrawn. The message is clear: where you conduct a trial matters.

Why Location Might Matter

At first glance, the FDA's position seems strange. Cancer is cancer, right? Cells divide uncontrollably, tumors grow, and drugs either stop them or they don't. Why should it matter whether the patient lives in Shanghai or Seattle?

The reality is more complicated. Multiple factors can cause a drug to work differently in different populations:

These aren't theoretical concerns. There are documented cases where drugs performed differently across populations. The blood thinner warfarin, for instance, requires different dosing in different ethnic groups due to genetic variations in drug metabolism. Some cancer drugs show different efficacy rates in Asian versus Western populations.

The Globalization of Clinical Trials

Clinical trials have become increasingly global over the past two decades. Today, roughly 80% of FDA-approved drugs include data from trials conducted outside the United States. Eastern Europe, Latin America, India, and China have become major hubs for pharmaceutical research.

The reasons are primarily economic. Conducting a clinical trial in the US is extraordinarily expensive. The cost per patient can exceed $40,000, driven by high investigator fees, institutional overhead, and regulatory requirements. The same trial in Eastern Europe might cost $6,500 per patient. In India, even less.

Beyond cost, enrollment is often faster abroad. American patients have many treatment options and are less likely to participate in trials. In countries with fewer alternatives, recruitment is easier. A trial that takes years to enroll in the US might complete in months elsewhere.

For pharmaceutical companies, especially smaller biotechs with limited budgets, the incentives are overwhelming. Why spend $100 million on a US trial when you can get similar data for $20 million abroad?

The Diversity Problem

Here's the paradox: while trials are increasingly global, they're also increasingly homogeneous in misleading ways.

Clinical trials conducted in the US are overwhelmingly white. Despite comprising only 60% of the US population, white patients make up roughly 75% of clinical trial participants. Black, Hispanic, and Asian patients are significantly underrepresented relative to their disease burden.

Meanwhile, single-country trials in China or India may have excellent representation of their local populations but tell us little about whether results apply to the diverse American population.

Congress recognized this problem and, in the 2022 Food and Drug Omnibus Reform Act (FDORA), required the FDA to develop guidance on Diversity Action Plans. Sponsors must now describe how they'll ensure trial populations reflect the intended use population.

For multinational studies, the FDA says enrollment goals should address the overall study, not just the US portion. The goal is trials that are both geographically diverse (conducted in multiple regions) and demographically diverse (reflecting the patients who will actually use the drug).

The Scientific Challenge

Designing a truly global, representative trial is scientifically challenging. You need enough patients from each subgroup to detect differences if they exist. This requires larger (more expensive) trials.

International Council of Harmonisation (ICH) guidelines (E5 and E17) attempt to address this. They provide frameworks for when foreign data can be extrapolated and when bridging studies are needed. The basic principle: you can use foreign data if you can demonstrate that the drug behaves similarly across populations, or if differences are understood and accounted for.

But these guidelines have gaps. They were developed when most drug development happened in the US, Europe, and Japan. The rise of China as a pharmaceutical powerhouse, with drugs developed entirely domestically and then seeking FDA approval, wasn't anticipated.

China's Pharmaceutical Ambitions

China has invested heavily in becoming a global pharmaceutical leader. Chinese biotech companies have developed dozens of innovative drugs, many targeting cancers and other serious diseases. These companies want access to the lucrative US market.

Conducting trials in China makes sense for Chinese companies. They understand the regulatory environment, have established relationships with hospitals, and can recruit patients quickly. Many of their drugs target conditions common in Chinese populations, like gastric cancer and hepatocellular carcinoma.

But the FDA's stance creates a dilemma. Either Chinese companies must conduct expensive, time-consuming trials in the US and other regions, or they risk rejection. Some have chosen to partner with US pharmaceutical companies who can add American patients to ongoing trials. Others have simply accepted that their drugs may never reach the US market.

Critics argue the FDA's position has a protectionist element, making it harder for foreign competitors to enter the US market. Defenders say it's about scientific rigor: if we don't know how a drug performs in diverse populations, we shouldn't assume it works for everyone.

What Good Trials Look Like

The gold standard for regulatory approval is the multiregional clinical trial (MRCT). These trials enroll patients across multiple countries and continents, allowing regulators to see whether treatment effects are consistent across populations.

A well-designed MRCT might enroll patients in:

• North America (US, Canada)
• Western Europe (Germany, France, UK, Spain, Italy)
• Eastern Europe (Poland, Ukraine, Romania)
• Asia-Pacific (Japan, South Korea, Australia, China)
• Latin America (Brazil, Argentina, Mexico)

With patients from diverse genetic backgrounds, healthcare systems, and disease contexts, regulators can be more confident that results apply broadly. If a drug works in German hospitals and Brazilian clinics and Japanese cancer centers, it probably works.

The FDA increasingly wants to see data broken down by region. Did the drug work as well in Asian patients as European patients? Were there concerning differences in safety profiles? Multiregional data allows these questions to be answered.

The Path Forward

Several developments are shaping the future of global clinical trials:

Decentralized trials: The pandemic accelerated adoption of remote monitoring, telemedicine visits, and direct-to-patient drug delivery. These approaches could make trials more accessible globally while reducing reliance on traditional clinical trial sites.

Real-world evidence: Data from electronic health records, insurance claims, and patient registries can supplement clinical trial data. If a drug approved in China is used extensively in Chinese-American populations in the US, that real-world data might support broader conclusions.

Adaptive designs: Modern statistical approaches allow trials to enroll more patients from regions where they're available while still generating valid conclusions about treatment effects.

Regulatory harmonization: Agencies in the US, Europe, Japan, and China are working toward common standards for what constitutes adequate evidence. Greater harmonization could reduce duplicative trials.

What It Means for Patients

For patients, these issues have real consequences. Delays in FDA approval mean delayed access to potentially lifesaving drugs. But approvals based on inadequate data mean potential exposure to drugs that don't work as expected or have unexpected side effects.

The fundamental question is one of epistemology: how do we know what we know? STARGLO showed Columvi worked in Asian patients but potentially harmed non-Asian patients. Without adequate enrollment from diverse regions, we might never have known. How many other drugs have hidden regional differences we haven't detected?

The FDA's answer, increasingly, is: maybe not. The scientific community is still figuring out when foreign data is sufficient and when it isn't. Until then, the most important question a patient can ask about any drug might be: who was in the trial, and were they like me?