1. The modern oncology landscape
Oncology clinical trials no longer fit the tidy three-phase framework that shaped twentieth-century drug development. Targeted therapies, immuno-oncology combinations, antibody-drug conjugates and cell therapies have pushed sponsors toward seamless designs, expansion cohorts and tumour-agnostic programs that blur the boundaries between phases.
For biotech sponsors, the implication is practical: the first study you run will likely shape the registration package. Investing early in a defensible design, an explicit dose selection plan and a biomarker hypothesis is no longer optional — it is the difference between a clean filing and an expensive late-stage repeat.
2. Choosing the right trial design
The right oncology trial design depends on the biology of the asset, the target population and the question that has to be answered for the next regulatory interaction. Three families dominate modern practice:
- Basket trials evaluate a single agent across multiple tumour types defined by a shared molecular alteration — useful when biology travels across histologies.
- Umbrella trials evaluate multiple agents within a single tumour type, stratified by biomarker — efficient when a sponsor or consortium can pool arms under one master protocol.
- Platform and seamless designs allow arms, doses or expansion cohorts to open and close based on pre-specified rules — compressing timelines without sacrificing rigour, when governance is in place.
The most common design mistake is choosing a framework for speed alone. A platform trial without a credible statistical analysis plan, an independent oversight model and a clear escalation pathway is a slower trial that looks fast on paper.
3. Biomarker-driven protocols
A biomarker is only useful in a clinical trial if three things are true: the assay is analytically validated, the cut-off is defined before enrolment, and the operational pathway — tissue acquisition, central laboratory turnaround, eligibility confirmation — can hold under real-world site conditions.
Sponsors increasingly need to think in terms of a companion or complementary diagnostic from the first-in-human study onward. Both FDA and EMA expect the diagnostic development plan to travel alongside the drug, not behind it. Late-stage gaps between drug data and diagnostic readiness are a recurring source of approval delay.
4. Endpoint selection
Overall survival remains the regulatory gold standard, but it is rarely achievable in a small biotech's first registration study. Progression-free survival, objective response rate and duration of response continue to support accelerated and conditional approvals when paired with a credible confirmatory plan.
Patient-reported outcomes and time-to-deterioration measures are increasingly part of the conversation — particularly in EMA scientific advice and in payer dialogue. Building those instruments into the pivotal trial, rather than bolting them on later, materially improves the value story at launch.
5. FDA and EMA regulatory strategy
FDA's Project Optimus has shifted expectations on dose selection in oncology. Sponsors can no longer assume that the maximum tolerated dose is the recommended phase 2 dose. Randomised dose comparisons, robust pharmacokinetic and pharmacodynamic data and a written dose optimisation rationale are now standard requirements at the pre-IND and end-of-phase 1 interactions.
EMA scientific advice — and parallel EMA/FDA advice where appropriate — is one of the highest-leverage investments a small sponsor can make. A well-prepared advice meeting can de-risk endpoint choice, comparator selection, statistical approach and the post-authorisation evidence package in a single interaction.
For programs with the potential to address an unmet need, designations matter: Breakthrough Therapy and Fast Track in the US, PRIME in the EU. They are not shortcuts — they are structured access to the agency, and they should be sought when the data genuinely support them.
6. Operational realities
A well-designed oncology protocol still fails if sites cannot execute it. Eligibility criteria that look tight on paper can collapse screen-success rates in practice. Biomarker turnaround times that ignore weekend laboratory schedules delay enrolment. Complex pharmacy preparation requirements limit the site list to a handful of academic centres.
The best protocols are stress-tested against site reality before lock — with investigator input, with the central laboratory and with the CRO that will activate the study. Cost and timeline savings from this step routinely exceed the investment by an order of magnitude.
7. Where to go next
Most oncology programs are decided in the first protocol and the first regulatory interaction. If you are sizing those decisions now, the highest-leverage step is bringing senior clinical, regulatory and operational judgement into the room early — before the protocol is written, not after.
Alpsvia Solutions supports biotech and pharmaceutical sponsors on exactly that work: integrated development plans, study design, regulatory strategy and medical oversight, anchored in oncology.