Project Optimus and What It Means for Small Biotechs

For decades, oncology dose selection followed a simple logic: escalate until you hit toxicity, call the highest tolerable dose the answer, and move on to efficacy. That approach made sense for cytotoxic chemotherapy, where more drug generally meant more tumor kill, and where patients were treated for a limited number of cycles.

It fits poorly with modern oncology. Targeted therapies, antibody-drug conjugates, bispecifics, and immunotherapies often plateau in efficacy well below the maximum tolerated dose (MTD), and patients stay on them for months or years. A dose that looks tolerable in a six-week dose-limiting toxicity window can be miserable — or unsustainable — at month eight.

FDA's Oncology Center of Excellence launched Project Optimus in 2021 to address this gap. For large pharma, it is a planning adjustment. For a small biotech with one lead asset and eighteen months of runway, it can reshape the entire development plan.

What Project Optimus Is (and Isn't)

Project Optimus is an FDA initiative to change how oncology doses are selected and justified. Its central idea is that the goal of early development is to find the dose that offers the best balance of efficacy and tolerability — not simply the highest dose patients can survive.

The initiative was formalized in FDA's guidance on optimizing dosage for oncology drugs, issued in draft in early 2023 and finalized in 2024. (Confirm the current version and any updates on FDA's website before building plans around specifics, since expectations continue to evolve through review-division feedback and precedent.)

What Optimus is not is a rigid checklist. There is no single mandated design. FDA's position is that the right approach depends on the drug's mechanism, therapeutic index, and available data — which is exactly why it creates uncertainty for sponsors without in-house regulatory strategy experience.

Why the Problem Was Real

It is tempting to read Optimus as regulatory overreach. The underlying evidence is hard to dismiss. A substantial share of newer oncology drugs, particularly small-molecule targeted therapies, required post-approval dose changes or additional dose-related studies because the approved dose caused more toxicity than necessary. Patients on these drugs frequently needed dose reductions in real-world use.

Why Small Biotechs Feel This the Hardest

Large sponsors can absorb Optimus. Small biotechs face a different equation:

None of this means small companies cannot succeed under Optimus. It means they have less margin for late surprises — so planning earlier matters more.

How to Build an Optimus-Ready Program

Start dose optimization in Phase 1, not after it

The most common strategic error is treating dose escalation and dose optimization as sequential steps. Under Optimus, they overlap. Design your first-in-human study with optimization in mind:

• Use model-assisted or model-based escalation (such as BOIN or CRM) rather than 3+3, so you generate richer dose-response information with fewer patients at sub-therapeutic levels
• Collect PK and PD data systematically from the start, with sampling schedules that support later modeling
• Build biomarker assessments into the protocol, particularly any pharmacodynamic marker that shows target engagement
• Plan expansion cohorts at more than one dose level so you are not forced to choose a single dose before you have comparative data

Define your "candidate doses" early and justify them

Going into a randomized dose comparison, you should be able to explain why each dose is in the study. That rationale typically draws on nonclinical pharmacology, Phase 1 safety, PK exposure, and any early efficacy signals. If you cannot articulate why you are testing dose A versus dose B, FDA reviewers will ask — and a weak answer costs time.

Consider where the randomized comparison should live

There are several structures sponsors use, each with tradeoffs:

There is no universally right answer. The best choice depends on the therapeutic index, patient availability, competitive landscape, and financing horizon. This is a decision worth pressure-testing with FDA rather than assuming.

Use quantitative tools to stretch small datasets

Exposure-response analysis and pharmacometric modeling allow you to extract more insight from limited patient numbers. For a small company, engaging a clinical pharmacology or pharmacometrics partner early is often a higher-return investment than adding sites.

Track tolerability beyond cycle one

Traditional dose-limiting toxicity windows capture acute events but miss cumulative or late toxicity. Under Optimus, collect and analyze dose reductions, interruptions, discontinuations over time, patient-reported outcomes on tolerability and quality of life, time on treatment and relative dose intensity, and late-onset adverse events. These data both strengthen your dose justification and protect you from an unpleasant surprise when FDA asks why so many patients in your dataset needed dose reductions.

Engage FDA early and specifically

Optimus is a case where early regulatory dialogue pays off disproportionately. A Type B or Type C meeting to discuss your dose-selection strategy before you launch a registrational trial is far cheaper than discovering a gap during review. Come prepared with a clear summary of the data supporting each candidate dose, your proposed randomized comparison design including endpoints and decision criteria, your rationale for how the results will inform the registrational dose — and specific questions, not open-ended requests for approval.

Operational Realities Sponsors Underestimate

Design is only half the challenge. Executing a multi-dose program creates operational burdens that catch small teams off guard:

• Drug supply and packaging: Multiple dose levels mean more complex manufacturing, labeling, and inventory planning
• Randomization and blinding: Deciding whether dose arms can be blinded, and how to manage unblinding risk when adverse event profiles differ between doses
• Site training and consistency: Sites need clear guidance on dose modification rules so interruptions and reductions are applied consistently
• Data infrastructure: PK sampling, biomarker collection, and PRO capture add data streams that must be integrated cleanly
• CRO capability: Not every CRO has deep experience with dose-optimization designs or pharmacometrics integration. Confirm this during vendor selection.

Cost and Timeline: A Realistic Perspective

Optimus does make oncology development more expensive and slower in many programs. But the comparison that matters is not "Optimus versus no Optimus." It is dose optimization done deliberately up front versus dose optimization forced on you later. A postmarketing requirement, a delayed approval, or a poorly tolerated label dose that hurts commercial adoption can each cost far more than a well-designed early-phase comparison.

A Practical Checklist for Small Biotechs

Before advancing beyond Phase 1, pressure-test your plan against these questions: