Criteria in Phase 1 Trials

Introduction

Phase 1 clinical trials aim to evaluate the safety, tolerability, and pharmacokinetics of investigational products, especially in first-in-human (FIH) settings. One of the most critical aspects of early-phase trial design is the implementation of stopping rules and Dose Limiting Toxicity (DLT) criteria. These safety mechanisms ensure that trials are halted or adjusted when adverse events (AEs) suggest that further exposure at a given dose is unethical or potentially harmful.

What Are Dose Limiting Toxicities (DLTs)?

DLTs are predefined adverse events or laboratory abnormalities that are considered unacceptable due to their severity or duration. They help determine the maximum tolerated dose (MTD) and inform decisions about escalation, de-escalation, or stopping.

DLTs are typically observed during the dose-limiting observation window (usually the first 1–2 weeks post-dosing).

Common DLT Criteria

While DLTs are protocol-specific, they often include:

• Grade 3 or higher non-hematologic AEs (e.g., severe nausea, vomiting, hepatotoxicity)
• Grade 4 hematologic toxicity (e.g., neutropenia, thrombocytopenia)
• Any Grade 3 AE that persists beyond 48–72 hours despite supportive care
• Delayed recovery of organ function beyond protocol-defined timelines
• Death or life-threatening events (Grade 5/4) related to the drug

DLTs must be drug-related (or probably related), as assessed by the investigator and sponsor medical monitor.

DLT Evaluation Period

Typically, a predefined “DLT window” is established—often 28 days in oncology or 7–14 days in healthy volunteer studies. During this period, subjects are monitored intensively for adverse events.

Any event that qualifies as a DLT within this window may affect cohort progression or trial continuation.

What Are Stopping Rules?

Stopping rules are prespecified safety conditions that, if triggered, lead to the temporary or permanent halt of:

• Dose escalation to the next cohort
• Dosing of additional participants in the current cohort
• The entire trial (in case of systemic safety concerns)

Types of Stopping Rules

1. Cohort-Level Stopping Rules

• If ≥2 out of 6 participants at a dose level experience DLTs, escalation is halted.
• If 1 of 3 has a DLT, the cohort may be expanded to 6 for confirmation.

2. Study-Wide Stopping Rules

• One or more deaths possibly related to IP
• Repeated unexpected severe AEs in multiple cohorts
• Significant protocol deviation impacting safety

3. Pharmacokinetic (PK)-Driven Stopping Rules

• If Cmax or AUC exceeds toxic thresholds predicted from animal models
• Non-linear accumulation or altered clearance requiring re-evaluation

4. Laboratory or Biomarker-Based Rules

• AST/ALT > 5× ULN with bilirubin elevation
• QTc interval > 500 ms or increase of > 60 ms from baseline

Designing DLT and Stopping Rules in Protocols

Stopping rules and DLT definitions must be clearly described in the protocol. Components include:

• List of DLT events and grades (based on CTCAE v5.0 or DAIDS grading)
• Assessment window (e.g., Day 1 to Day 14 post-dose)
• Procedures for safety review and escalation decisions
• Roles of investigators, sponsor, and Safety Review Committee

Example from an Oncology Protocol

DLT Criteria	Grading/Threshold
Neutropenia	Grade 4 > 7 days or febrile neutropenia
Thrombocytopenia	Grade 3 with bleeding or Grade 4
Liver enzymes	ALT/AST > 5× ULN or > 3× ULN with symptoms
Non-hematologic toxicity	Any Grade 3+ (excluding alopecia, nausea if resolved quickly)

Role of the Safety Review Committee (SRC)

In many early-phase trials, a Data Monitoring Committee (DMC) or Safety Review Committee (SRC) is established. This independent team:

• Reviews all AE and DLT data after each cohort
• Approves or denies escalation to the next dose level
• Can recommend modifications to dosing strategy or protocol

Real-World Example: How DLT Criteria Stopped a Trial

In a biologic targeting inflammatory pathways, the first two participants in a 6-subject cohort experienced Grade 3 infusion reactions and hypotension. As per the stopping rule of ≥2 DLTs in 6 subjects, escalation was paused. The SRC recommended changing the infusion rate and premedication before resuming. This decision helped prevent a potentially fatal event in future participants.

Adaptive Escalation with Stopping Rules

Modern dose escalation designs (like CRM or BOIN) incorporate stopping rules algorithmically:

• Each DLT updates posterior probability models
• Escalation, de-escalation, or halting decisions are generated in real time
• Probability thresholds (e.g., toxicity > 30%) guide decisions

Regulatory Expectations

• FDA: Expects stopping rules to be pre-specified in the protocol under IND. Any modification post-submission must be justified via protocol amendment.
• EMA: Requires stopping rules and DLT definitions per “Guideline on strategies to identify and mitigate risks for FIH and early clinical trials.”
• CDSCO (India): Recommends using ICH E6 and Schedule Y guidelines for DLT definition and early termination conditions.

Best Practices

• Use CTCAE v5.0 grading to define objective toxicity criteria
• Involve cross-functional teams (clinical, safety, PK) in DLT design
• Define both drug-related and unrelated stopping rules (e.g., device failure, protocol deviation)
• Document all DLTs in real time and ensure SAE reporting compliance
• Conduct simulations to validate stopping rule robustness before trial start

Conclusion

Stopping rules and DLT criteria serve as critical safeguards in Phase 1 clinical trials. They provide the framework for real-time safety assessment and ethical dosing decisions. By carefully defining these mechanisms in protocols and adhering to them during execution, researchers can minimize risk, comply with regulatory standards, and maintain the trust of participants and stakeholders throughout early clinical development.