A Phase 1 dose-escalation trial is built primarily to characterize safety, tolerability, pharmacokinetics, and a dose range for further study. It can reveal early evidence of biological or clinical activity. It is rarely designed to produce a reliable comparative efficacy estimate.
The most useful reading question is not “Did anyone respond?” It is “How did the design move from one dose to the next, and what evidence supports the recommended dose?”
Map the trial into parts
Many first-in-human programs have at least two sections:
- dose escalation: small sequential cohorts receive increasing doses under prespecified rules;
- dose expansion: additional participants receive one or more selected doses to characterize safety, pharmacology, and activity in defined populations.
Some trials add single-ascending-dose and multiple-ascending-dose portions, food-effect cohorts, or formulation comparisons. Results should remain connected to the part that generated them.
An expansion cohort is larger than an escalation cohort, but it is not automatically randomized or confirmatory.
Dose-limiting toxicity is a protocol definition
A dose-limiting toxicity, or DLT, is not simply any serious adverse event. The protocol defines which toxicities count, their severity, attribution, and the observation window.
Read the definition closely:
- Which grading system was used?
- Must the event be considered treatment-related?
- Do laboratory abnormalities count without symptoms?
- How are delays, missed doses, or supportive care handled?
- What happens when a participant is not evaluable for the DLT window?
A short window can miss cumulative or delayed toxicity. Later adverse events remain important even if they do not enter the formal escalation decision.
The maximum tolerated dose is an estimate
In a traditional 3+3 design, cohorts often begin with three evaluable participants. Escalation or expansion depends on the number experiencing DLTs. The maximum tolerated dose is then chosen around the highest tested dose meeting the rule.
This design is simple but statistically imprecise. Few participants receive each dose, the selected dose can vary with event order, and the method uses limited information from neighboring cohorts.
Model-assisted and Bayesian designs can estimate a target toxicity probability more efficiently, but they depend on stated assumptions and correct implementation. Whatever the design, “MTD” should not be read as a sharp biological boundary between safe and unsafe doses.
MTD and recommended Phase 2 dose can differ
The highest tolerable dose may not be the most informative or sustainable dose. A recommended Phase 2 dose can incorporate:
- acute and chronic tolerability;
- dose interruptions and reductions;
- pharmacokinetic exposure;
- target engagement or pharmacodynamics;
- preliminary activity;
- formulation practicality;
- variability across participants;
- exposure-response modeling.
Modern targeted therapies may reach a biologically active plateau before an MTD is observed. A report should explain the totality of evidence behind dose selection, not state only that a committee chose it.
Pharmacokinetics connect administered dose to exposure
Key parameters can include maximum concentration, area under the concentration-time curve, time to maximum concentration, half-life, clearance, and accumulation after repeat dosing.
Look for:
- individual as well as summary data;
- dose proportionality;
- variability and outliers;
- accumulation over cycles;
- antidrug antibodies for biologics;
- effects of food, organ function, or concomitant drugs;
- whether exposure overlaps with toxicities or pharmacodynamic changes.
A long half-life can complicate escalation because participants at one dose may still be accumulating exposure when the next cohort begins.
Sentinel dosing and staggered enrollment manage uncertainty
First-in-human protocols may dose one participant before the rest of a cohort or stagger dosing so early safety information is available before broader exposure. Starting dose and escalation increments should be justified using nonclinical pharmacology and toxicology, prior related products, predicted exposure, and uncertainty factors.
Aggressive percentage increases at low doses can be reasonable when exposure is far below a biologically active range; smaller increments are usually expected as relevant exposure or toxicity is approached.
Early response rates have unstable denominators
Phase 1 efficacy summaries can use denominators such as all enrolled participants, all treated participants, participants with measurable disease, or only evaluable participants with follow-up. Excluding early discontinuations can inflate apparent response.
Responses should be described with exact numerators, confidence intervals, duration, assessment schedule, and prior therapy. Multiple tumor types or disease subgroups should not be pooled when biology and baseline prognosis differ.
Uncontrolled response does not show improvement over standard care. Exceptional durable responses can justify further study, but they remain signals rather than comparative proof.
Expansion cohorts create multiple analytical opportunities
As a program adds doses, biomarkers, combinations, and disease groups, the number of possible summaries grows. Protocol amendments may be scientifically appropriate, but readers need dates and prespecified objectives to distinguish planned evaluation from data-driven selection.
Check whether the report includes every opened cohort and whether recruitment stopped for lack of activity. Selective publication of promising cohorts distorts the program-level picture.
A practical extraction template
For each dose level, record:
- number treated and DLT-evaluable;
- dose and schedule;
- DLT count, type, and timing;
- all grade 3 or higher treatment-related events;
- discontinuations, delays, and reductions;
- pharmacokinetic exposure and variability;
- pharmacodynamic evidence;
- response denominator and follow-up;
- dose-selection decision;
- deviations from the escalation algorithm.
This table often reveals more than the narrative abstract.
Bottom line
A Phase 1 escalation study is a structured learning process under uncertainty. Its strongest conclusions concern observed toxicities, exposure, pharmacology, and the rationale for further dosing. Early activity can guide development, but small selected cohorts and flexible expansion mean it should not be promoted as confirmatory efficacy evidence.
Sources
- U.S. Food and Drug Administration. Estimating the Maximum Safe Starting Dose in Initial Clinical Trials for Therapeutics in Adult Healthy Volunteers.
- U.S. Food and Drug Administration. Expansion Cohorts: Use in First-In-Human Clinical Trials to Expedite Development.
- International Council for Harmonisation. E8(R1): General Considerations for Clinical Studies.
- International Council for Harmonisation. M3(R2): Nonclinical Safety Studies for Human Clinical Trials.
- U.S. Food and Drug Administration. Optimizing the Dosage of Human Prescription Drugs and Biological Products.
Continue reading
Editorial responsibility and corrections
Published by Serum Research. Sources checked October 1, 2026. This is educational material; no independent scientific, clinical, statistical, analytical, or regulatory review is claimed. Read how we review evidence or send a correction.
For Research Use Only. Not for Human Consumption. This page does not provide instructions for medical use or establish the identity, quality, safety, or equivalence of a catalog material.