ESMO 2026: What Will It Take to Deliver the Next Generation of Oncology Trials?

By Oksana Fabri, MD, PhD, Executive Medical Director, Therapeutic Area Medical Lead in Hematology/Oncology

Promising science is no longer the bottleneck. Designing the increasingly complex trials needed to test it is.

At ESMO 2026, many of the most substantive conversations were not about which molecules look promising. They were about whether we can run the trials those molecules need. Targeted agents, ADCs, bispecifics, and cell therapies keep arriving. The trials built to test them are getting harder to deliver: narrower populations, complex biomarkers, demanding schedules, and sites already at capacity.
As an oncology CRO, we see the same pattern from both the sponsor’s and the sites’ perspectives. Most trials that struggle were already in difficulty at protocol sign-off, before the first patient was screened.

What the Evidence From ESMO 2026 Tells Us About Oncology Clinical Trial Design

Endpoints are not always as robust as they look.

A review of 156 high-impact trials with progression-free survival as the primary endpoint found that sensitivity analyses for intercurrent events were mentioned in 10–30% of trials but reported in under 5%. Many therapies are adopted on modest PFS gains, so the question is not only whether a result is positive, but whether it would change practice.

Clear development objectives keep trials interpretable.

When a study is designed to support a specific go/no-go or registration step, endpoints, population, and sample size follow from it. When it tries to answer everything, it usually answers little.

Real-world screening data can overturn design assumptions.

In the EORTC 2129 TREAT ctDNA trial, 56 of 882 patients under ctDNA surveillance turned positive, and those patients more often had extensive nodal involvement. The protocol was then amended to prioritize patients with four or more positive nodes. Input from investigators and sites at concept stage can highlight these issues before the first amendment.

Patients name study burden as a main concern.

A decade of patient-partner reviews of consent documents from 49 EORTC studies found that study procedures and their burden were among the dominant themes, alongside clarity of information and risk. Addressing these issues should go hand in hand with other site-related considerations. Recruitment, diagnostic access, site capacity, and visit schedules have to be solved together, not one at a time.

Earlier dialogue produces more relevant trials.

At a panel during ESMO 2026, speakers called for patient involvement at the design stage and for earlier communication between regulators, investigators, and patient representatives, so that trials address clinically meaningful questions. Nurses, pharmacists, pathologists, and data managers belong in that conversation too.

What This Means For Sponsors

The next generation of oncology trials will hinge on decisions made at the design stage. Feasibility isn’t something to check once the protocol is written. It belongs in the science from the start. Sponsors who test their assumptions early with investigators, sites, and medical experts will be better positioned to bring new therapies to patients sooner and produce data that can drive regulatory decisions and change clinical practice.