February 24, 2019


Modified Toxicity Probability Interval Design Version 2

In 2007, Laiya founder developed the original dose-finding TPI design based on toxicity interval, which trail-blazed a new path for phase I clinical trial designs. Over the next 10 years of diligent research, we have developed mTPI (JCO; Ji et al., 2010) and mTPI-2 (Contemp. Clinical Trials.; Guo et al., 2017), and published related papers in top scientific journals. Since mTPI’s debut, it has been adopted by a number of pharmaceutical companies and has passed FDA clinical trial protocol review.

The mTPI-2 design uses interval-based Bayesian designs for dose finding, the decision rules of which correspond to an optimal rule under a formal Bayesian decision theoretic framework. It improves mTPI in terms of safety and has demonstrated superior performance in extensive simulations.


  1. Identifies the MTD with higher power, which improves the overall success rate of a clinical trial.
  2. Flexibility: Applicable to clinical trials of different target toxicity probabilities, which satisfies the need of different oncology drug development.
  3. Better Safety: Preventing patients from being exposed to excessive dose levels.
  4. Simplicity: Decisions table can be generated prior to the clinical trial, making the design more transparent and easier to implement.