Contemporary dose-escalation methods for early phase studies in the immunotherapeutics era. (November 2021)

• Record Type: Journal Article
• Authors: Araujo, Daniel V.; Oliva, Marc; Li, Kecheng; Fazelzad, Rouhi; Liu, Zhihui Amy; Siu, Lillian L.
• Is Part Of: European journal of cancer. Volume 158(2021)
• Page Start: 85
• Page End: 98
• Publication Date: 2021-11
• Subjects: Dose-escalation methods -- Phase I trials -- Immunotherapeutics -- Model-assisted designs; Cancer -- Periodicals; Neoplasms -- Periodicals; Cancer -- Périodiques; Cancer; Tumors; Electronic journals; Periodicals; Electronic journals; 616.994
• DOI: 10.1016/j.ejca.2021.09.016
• Languages: English
• ISSNs: 0959-8049

Abstract

Abstract: Phase 1 dose-escalation trials are crucial to drug development by providing a framework to assess the toxicity of novel agents in a stepwise and monitored fashion. Despite widely adopted, rule-based dose-escalation methods (such as 3 + 3) are limited in finding the maximum tolerated dose (MTD) and tend to treat a significant number of patients at subtherapeutic doses. Newer methods of dose escalation, such as model-based and model-assisted designs, have emerged and are more accurate in finding MTD. However, these designs have not yet been broadly embraced by investigators. In this review, we summarise the advantages and disadvantages of contemporary dose-escalation methods, with emphasis on model-assisted designs, including time-to-event designs and hybrid methods involving optimal biological dose (OBD). The methods reviewed include mTPI, keyboard, BOIN, and their variations. In addition, the challenges of drug development (and dose-escalation) in the era of immunotherapeutics are discussed, where many of these agents typically have a wide therapeutic window. Fictional examples of how the dose-escalation method chosen can alter the outcomes of a phase 1 study are described, including the number of patients enrolled, the trial's timeframe, and the dose level chosen as MTD. Finally, the recent trends in dose-escalation methods applied in phase 1 trials in the immunotherapeutics era are reviewed. Among 856 phase I trials from 2014 to 2019, a trend towards the increased use of model-based and model-assisted designs over time (OR = 1.24) was detected. However, only 8% of the studies used non-rule-based dose-escalation methods. Increasing familiarity with such dose-escalation methods will likely facilitate their uptake in clinical trials. Highlights: Model-based/assisted methods have not been broadly adopted despite favourable features. Model-assisted methods combine features of both model-based and rule-based designs. Dose-escalation methods can affect the duration and sample size of phase I trials. From 2014 to 2019, only 8% of phase I trials used non-ruled based methods. IO agents should be risk stratified based on therapeutic range and toxicity potential.