The i3+3 design for phase I clinical trials

Abstract

Purpose The 3+3 design has been shown to be less likely to achieve the objectives of phase I dose-finding trials when compared with more advanced model-based designs. One major criticism of the 3+3 design is that it is based on simple rules, does not depend on statistical models for inference, and leads to unsafe and unreliable operating characteristics. On the other hand, being rule-based allows 3+3 to be easily understood and implemented in practice, making it the first choice among clinicians. Is it possible to have a rule-based design with great performance?Methods We propose a new rule-based design called i3+3, where the letter "i" represents the word "interval". The i3+3 design is based on simple but more advanced rules that account for the variabilities in the observed data. We compare the operating characteristics for the proposed i3+3 design with other popular phase I designs by simulation. ResultsThe i3+3 design is far superior than the 3+3 design in trial safety and the ability to identify the true MTD. Compared with model-based phase I designs, i3+3 also demonstrates comparable performances. In other words, the i3+3 design possesses both the simplicity and transparency of the rule-based approaches, and the superior operating characteristics seen in model-based approaches. An online R Shiny tool (https://i3design.shinyapps.io/i3plus3/) is provided to illustrate the i3+3 design, although in practice it requires no software to design or conduct a dose-finding trial. ConclusionThe i3+3 design could be a practice-altering method for the clinical community.

The Joint i3+3 (Ji3+3) design for phase I/II adoptive cell therapy clinical trials

“…The P's determine how much deviation is allowed from p T for a dose to be still considered as MTD. The larger the values, the more deviation is allowed, which would result in better operating characteristics since a larger range of doses will be accepted as MTD (Liu et al 2019).…”

“…Traditionally, phase I oncology trials aim to find, among a grid of candidate dose levels, the maximum tolerated dose (MTD) defined as the highest dose level at which the toxicity probability is below or close to a prespecified target probability p T , say p T ¼ 17% or 30% (Love et al 2017;Storer 1989). Both rule-based designs (such as 3 + 3 (Storer 1989), i3 + 3 (Liu et al 2019)) and model-based designs (such as CRM (O'Quigley et al 1990) and mTPI (Guo et al 2017;Ji et al 2010;Ji and Wang 2013)) are commonly used, where the MTD determination solely depends on the toxicity data, i.e. the number of dose-limiting toxicities (DLTs) out of the total number of patients treated.…”

“…We take a different approach. Building upon the i3 + 3 design (Liu et al 2019), we propose the joint i3 + 3 (Ji3 + 3) dosefinding design that 1) is based on a smart algorithm and therefore model free, 2) incorporates both toxicity and efficacy outcomes, and 3) is transparent to physicians and simple to implement. The Ji3 + 3 design is motivated by a phase I/II trial (NCT 03971799) which aims to determine the safety and feasibility of anti-CD33 CAR expressing T cells in children and adolescents/young adults (AYAs) with relapsed/refractory acute myeloid leukemia (AML).…”

PoD-TPI: Probability-of-Decision Toxicity Probability Interval Design to Accelerate Phase I Trials

“…The PoD-TPI design is built upon the mTPI-2 design. Nevertheless, the proposed strategy can be applied to other model-free or model-assisted designs such as 3+3 (Storer, 1989), BOIN (Liu and Yuan, 2015), keyboard (Yan et al, 2017) or i3+3 (Liu et al, 2019). It suffices to change the deterministic decision function A according to the specific design.…”

“…Simpler model-assisted designs like the cumulative cohort design (CCD, Ivanova et al, 2007) and the Bayesian optimal interval design (BOIN, Liu and Yuan, 2015) also uses intervals as the basis for decision making. The latest i3+3 design (Liu et al, 2019) is again based on interval rules, but without a model.…”