First-in-human (FIH) studies

Published

March 18, 2026

Primary objective: safety and tolerability. Secondary: characterize PK profile (single and repeated dosing), PD/biomarker readout (to predict therapeutic dose range), and establish a dose range. Need to reach exposures higher than therapeutic to account for potential DDI and organ impairment scenarios later.

Starting dose

Derived from nonclinical data:

  • NOAEL (No Observed Adverse Effect Level) with allometric scaling.
  • MABEL (Minimum Anticipated Biological Effect Level) – preferred when pharmacology-driven toxicity is expected.
  • PKPD, PBPK, and/or QSP modeling to predict human PK and target exposures.

Dose escalation

Typically 2–3 fold between cohorts, adjusted for nonlinearity of PK and steepness of the dose-response curve.

Sentinel dosing

Dose 2 subjects first (1 active, 1 placebo), observe through Cmax, evaluate safety, then dose the remaining cohort.

Single ascending dose (SAD)

  • Subjects receive a single dose of the investigational drug at one dose level, with dose escalation in subsequent cohorts.
  • Typical cohort: 6 active + 2 placebo.
  • Target exposures above the expected therapeutic dose to cover later TQT assessment and DDI/organ impairment-related exposure increases.
  • Sampling: Intensive PK (plasma + urine; >3–5 half-lives to capture full elimination). Can sometimes include triplicate ECGs timed with PK (especially around Cmax), PD biomarkers, safety labs, genotyping, and samples for metabolite profiling.

Multiple ascending dose (MAD)

  • Like SAD, but each subject receives multiple doses at one dose level.
  • Essential for detecting time-dependent PK and accumulation.
  • On some dosing occasions, only pre-dose (trough) samples might be taken.
  • Dose-adjustment rules can be in place to achieve desirable safety/efficacy.
NoteCombined SAD-MAD studies

SAD and MAD studies can be combined under one study, and subjects can participate in both.

Planning considerations

TipFIH planning
  • Validated PK and PD assays must be in place before starting – these can take months to develop and validate.
  • Predefined stopping rules at trial, cohort, and individual subject level (e.g. liver signals, cardiac events, renal markers, severe AEs).
  • Protocols must be flexible enough to add/remove cohorts or adjust dose levels without amendments.