Cancer drug development is entering a period of rapid transformation, with artificial intelligence accelerating drug discovery, China emerging as a major force in oncology innovation, and regulators worldwide facing increasingly complex questions about how new therapies should be evaluated.
In a recent guest editorial published in The Cancer Letter, Richard L. Schilsky, MD; H. Kim Lyerly, MD; Thomas R. Fleming, PhD; and Renzo Canetta, MD, members of the Accelerating Anticancer Agent Development and Validation (AAADV) Workshop Foundation, examined these shifts and raised what they describe as a defining question for the next decade of cancer drug development: who decides which drug candidates move forward, and why?
Following the publication of the editorial, the authors shared the following commentary with OncoDaily, reflecting on the changing global drug-development landscape, the growing need for cross-sector dialogue, and the questions that will be at the center of the 2026 AAADV Workshop.
The shifting sands of global cancer drug development
By Richard L. Schilsky, MD, H. Kim Lyerly, MD, Thomas R. Fleming, PhD, Renzo Canetta, MD
In the August 14, 2026 issue of The Cancer Letter (volume 52, issue 32), our guest editorial focused on a defining question of the next decade in cancer drug development: who decides which drug candidates move forward, and why.
Major changes have emerged in recent years that are impacting the strategic development of new cancer drugs: the vast number of drug candidates that science can now produce, powered largely by AI; where the drugs originate; how regulators assess the benefit and risk of novel agents; and how health systems decide what innovation is worth. We are managing each issue separately, but they are the same problem.
Anyone who attended a major oncology meeting this year will recognize this sequence. A positive phase 3 trial result is presented and published simultaneously in a major medical journal. The audience vigorously applauds the good news. Then the questions begin, not so much about the biology or the pharmacology that underlies the breakthrough, but about whether the trial’s population resembles the one in the room, whether the control arm is an appropriate standard of care, whether the trial’s primary endpoint truly reflects clinical benefit, and when, if ever, the drug will be available to patients in need around the world.
Major shifts in cancer drug development are now occurring
First, we are likely to see an abundance of novel agents. Computational methods now generate plausible drug candidates faster than the system downstream can evaluate them. It is no longer an issue of the number of structures we can identify, but rather of the expertise of the chemists who can synthesize them, the preclinical models that can triage them, and, above all, the availability of patients and trial sites to test them.
The second shift is the site of origin of these novel medicines. During 2015-2025, U.S.-origin early-stage biopharmaceutical programs grew substantially in absolute terms, but they fell from roughly half of the global total to about a third. In contrast, early-stage oncology development programs beginning in China rose from 381 to 4,152 over the past decade. By 2024, China had overtaken the United States in the absolute number of early-stage oncology programs (1).
The third shift is regulatory divergence. Between 2020 and 2025, the U.S. Food and Drug Administration (FDA) approved 87 novel oncology indications, and China’s National Medical Products Administration (NMPA) approved 94. Of 167 distinct medicines, only 14 were approved by both agencies (2). Some of that gap is expected and appropriate: real differences in disease epidemiology, in unmet need, in population characteristics, and in the standard of care a new agent must beat. Some of it reflects the regions where a sponsor chose to run a trial and what a regulator was subsequently able to conclude about benefit and risk.
While we tend to discuss these as three separate stories, they are tightly interconnected. In fact, each of them moves development decisions earlier, makes them more local, and gives them more downstream consequence than many sponsors can reasonably foresee.
Let’s consider what must be contemplated before a single patient is enrolled in a clinical trial supporting a clinical development strategy: which disease, which population, which comparator, which endpoint, which region, and which regulator will see the package first. Twenty years ago, most of those questions had highly predictable answers. Today each is open, and is usually resolved on grounds of speed, cost, and enrollment feasibility; these are rational grounds, but the decisions that follow may inadvertently confound results which will be interpreted and judged by a diverse group of regulatory agencies.
For example, a comparator in a randomized trial may be chosen because it is standard of care in one country, but the trial results may be submitted to a regulatory body in another country, with a different standard of care. These are design choices whose full cost arrives long after the people who made them have moved on.
Accelerating the discovery of novel compounds will contribute to the increasing complexity of oncology drug development. When candidates are plentiful, and testing capacity is scarce, the decision about which molecules advance becomes the most consequential in the field, and it is made company by company, largely following commercial logic: product novelty, population size, competitive density, likelihood of reimbursement. That calculus is rational for each sponsor and systematically favors a “me too” agent against a validated target in a narrow indication over a novel but not yet fully validated mechanism that might matter far more. Risk-averse sponsors rarely deviate from this path, and no regulator can mandate it.
Cross-sector communication and discussion are more necessary than ever
Some of the issues raised here will eventually be addressed by regulatory guidance, but what would help is earlier and more candid conversation among sponsors, trialists, regulators and patients: alignment on control arms before a protocol is finalized rather than after a filing; multiregional design as the default rather than the remedy; and an honest accounting of what has not worked, which seldom generates a publication and is therefore rarely shared.
Those conversations are often unplanned and unscripted, which is why they are useful but hard to schedule. A biostatistician directly asking a regulator about the suitability of a trial design. A medical director describing a dosing decision that turned out to be too toxic. An investigator learning how the drug application was received in Silver Spring, Amsterdam, and Beijing. A patient advocate saying plainly that a multi-year gap between approval elsewhere and approval at home is not acceptable to the people she represents.
An invitation
The Accelerating Anticancer Agent Development and Validation (AAADV) Workshop has convened these conversations annually since 2004, with the participation of the U.S. National Cancer Institute, FDA, AACR, ASCO, sponsors and patient advocates. In 2024, we reorganized as an independent not-for-profit charitable (501(c)(3)) foundation to broaden access and expand virtual participation, so colleagues outside the United States can take part without crossing an ocean. This year we meet November 5–6, 2026, at the Bethesda North Marriott Hotel & Conference Center in Rockville, Maryland, in person and online.
The program focuses on multiregional clinical trials and global development strategy; benefit-risk assessment, including the sufficiency of a single pivotal trial; dose optimization; the management of immune-related adverse events; whether closely related agents in an already-crowded class are worth developing; and what FDA complete response letters teach us. An evening session is devoted to cancer drug development in China, and I will join Richard Pazdur for a fireside chat on the globalization of cancer drug development.
Details of the program and registration process are at AAADV.org. Hope you can join us.
References:
1. Kang SY, Ji Y. Geographic shifts in early-stage biopharmaceutical innovation. JAMA. 2026;335(15):1355–1356. doi:10.1001/jama.2026.1962
2. Collins G, Pandita D, Andrews H, et al. A cross-national review of novel oncology approvals in the United States and China (2020-2025). Health Aff Sch. 2026;4(7):qxag183. doi:10.1093/haschl/qxag183.
Author affiliations:
Richard L. Schilsky, MD, is chair of the board of directors of the AAADV Workshop Foundation, professor emeritus of medicine at the University of Chicago, and former executive vice president and chief medical officer of the American Society of Clinical Oncology.
H. Kim Lyerly, MD, is a member of the board of directors of the AAADV Workshop Foundation, George Barth Geller Professor of Cancer Research, Professor of Surgery, Duke University School of Medicine.
Thomas R. Fleming, PhD, is a member of the board of directors of the AAADV Workshop Foundation, Professor and former chair of biostatistics, University of Washington.
Renzo Canetta, MD, is a member of the board of directors of the AAADV Workshop Foundation and Chair, AAADV Workshop Program Committee; Former vice president of oncology global clinical research, Bristol Myers Squibb.
Read further on OncoDaily: AAADV 2026: Where the Future of Cancer Drug Development Will Be Debated
