Precision Oncology Trials Need to Reflect the Complexity of Cancer Biology

By the time an oncology program reaches Phase III, many of the most important decisions have already been made. The patient population has been defined, biomarker strategy selected, endpoints have been prioritized, and assumptions have been made about how the therapy should perform in a larger, more complex group of patients.

When a pivotal trial falls short, the result is often described as a drug failure. In some cases, that may be true. But in precision oncology, late-stage setbacks can also point to a different issue. The trial may have carried forward an incomplete understanding of the biology it was designed to test.

A therapy may be rationally designed to target a receptor, pathway or mechanism, but patient outcomes can still vary significantly depending on many factors, including underlying tumor biology that is not yet fully understood. For clinical development teams, the lesson is not that targeted therapies are failing. It is that Phase III success depends on more than identifying a promising target. It depends on whether the biology, biomarker strategy, and patient-selection approach are strong enough to support the trial design.

The target is only the starting point

Tumors are dynamic systems that evolve over time, and different cancer cell populations within the same tumor can respond differently to treatment. Multiple parallel mechanisms may also compensate when one pathway is inhibited, limiting treatment effectiveness even when the scientific rationale appears strong.

This complexity matters because biology, biomarkers, and trial design are deeply connected. If the understanding of the disease is incomplete, that limitation can carry through into biomarker selection, eligibility criteria, and the design of pivotal trials.

A biomarker strategy can only be as strong as the biological assumptions behind it. If development teams focus too narrowly on one pathway, protein, or receptor, they may miss the broader context that determines whether a patient is likely to respond. As a result, biomarkers may become overly simplistic, and trials may not adequately capture the patient populations most likely to benefit.

Biomarker strategy needs to be tested early

A robust biomarker strategy needs to balance clinical applicability, biological relevance, and diagnostic feasibility. This balance is difficult, but essential for trial success.

One major challenge is tumor heterogeneity. Different cells within the same tumor may behave differently, and a therapy may eliminate some cancer cells while others survive and continue to drive disease progression. Many biomarker approaches still rely on bulk tumor analysis, which can overlook meaningful variation among cell populations within the same tumor.

Another common limitation is the size, diversity, and representativeness of the datasets used to develop biomarkers. Data from small or biased patient cohorts can lead to less reliable results, especially when a therapy moves into a broader clinical setting. Larger datasets, real-world data, and more granular technologies such as single-cell analysis can help development teams build a more complete view of the biology they are trying to target and the patient populations they are trying to define.

Phase III assumptions are built before Phase III

The transition from early-stage development to a pivotal trial is one of the most important moments in an oncology program. By Phase III, many key assumptions have already been made: which patients to include, which biomarker to use, which endpoint to prioritize and how broad or narrow the target population should be.

This is why stronger evidence generation earlier in development is so important. Sponsors need to pressure-test whether a biomarker is not only measurable, but meaningful. They need to understand whether response is consistent across patient subgroups, whether prior treatment exposure affects outcomes, and whether disease stage or tumor evolution changes the likelihood of benefit.

This does not mean every trial should become more restrictive. In some cases, therapy may show benefit across a broader patient population. In others, a more selective approach may be necessary. The point is that these decisions should be guided by a more complete understanding of biology, not by convenience or assumptions carried forward from early data.

Patient selection needs more than one dimension

The field is moving beyond single biomarkers toward more complex, multi-dimensional patient selection. Genomic alterations, transcriptional activity, immune context, treatment history, and resistance mechanisms can all influence response in ways that a single marker may not fully capture.

For trial design, this shift will add complexity. It may require more sophisticated diagnostics, better data integration, and closer collaboration with regulators to ensure patient-selection strategies are interpretable and actionable. However, in the long run, this complexity should make trials more informative, not less. Better patient selection can lead to clearer efficacy signals, more efficient development, and a higher likelihood that therapies are tested in the populations most likely to benefit.

Trials must anticipate how tumors evolve

Combination and sequencing strategies will also play an important role in future oncology trial design. They are essential for addressing tumor heterogeneity and overcoming resistance mechanisms. Treatment sequencing is especially important because tumors evolve, and trial designs need to account for how prior treatments, emerging resistance and changing tumor biology may influence response over time.

The next evolution of precision oncology trials will not be defined by abandoning targeted treatment, but by making development strategies more sophisticated. Precision oncology has advanced significantly, but Phase III success depends on more than identifying a target. It depends on understanding the patient, the tumor, and the evolving biology that connects them—and designing trials that are strong enough to reflect that reality.

Anna Pokorska-Bocci, Debiopharm Associate Director, Personalized Medicine