Study designs for clinical trials applied to personalised medicine: a scoping review

Abstract

ObjectivePersonalised medicine (PM) allows treating patients based on their individual demographic, genomic or biological characteristics for tailoring the ‘right treatment for the right person at the right time’. Robust methodology is required for PM clinical trials, to correctly identify groups of participants and treatments. As an initial step for the development of new recommendations on trial designs for PM, we aimed to present an overview of the study designs that have been used in this field.DesignScoping review.MethodsWe searched (April 2020) PubMed, Embase and the Cochrane Library for all reports in English, French, German, Italian and Spanish, describing study designs for clinical trials applied to PM. Study selection and data extraction were performed in duplicate resolving disagreements by consensus or by involving a third expert reviewer. We extracted information on the characteristics of trial designs and examples of current applications of these approaches. The extracted information was used to generate a new classification of trial designs for PM.ResultsWe identified 21 trial designs, 10 subtypes and 30 variations of trial designs applied to PM, which we classified into four core categories (namely, Master protocol, Randomise-all, Biomarker strategy and Enrichment). We found 131 clinical trials using these designs, of which the great majority were master protocols (86/131, 65.6%). Most of the trials were phase II studies (75/131, 57.2%) in the field of oncology (113/131, 86.3%). We identified 34 main features of trial designs regarding different aspects (eg, framework, control group, randomisation). The four core categories and 34 features were merged into a double-entry table to create a new classification of trial designs for PM.ConclusionsA variety of trial designs exists and is applied to PM. A new classification of trial designs is proposed to help readers to navigate the complex field of PM clinical trials.

Introducing the Library of Guidance for Health Scientists (LIGHTS)

“…LIGHTS offers guidance for investigators at any stage of a study. In early stages, guidance for clarifying the need for new studies may be particularly useful (eg, for deciding whether a new systematic review is needed), as well as guidance on methodological concepts (eg, Prognosis Research Strategy [PROGRESS] framework for prognostic research), types of adaptive trial designs, guidance for specific design considerations (eg, Difference Elicitation in Trials [DELTA 2 ] guidance for sample size calculation), and reporting guidelines that often provide a well-structured overview of the key steps of a study (eg, a randomized clinical trial or a systematic review). Over the course of a study, other types of guidance may become relevant, including guidance on trial process organization (eg, patient recruitment or data collection), specific analytic issues (eg, handling missing data), presenting results (eg, creating figures), interpreting the strength on evidence (eg, regarding a prognostic factor), and structuring study reports (eg, of a randomized clinical trial or a systematic review).…”

Evaluating Translational Methods for Personalized Medicine—A Scoping Review

“…Compared to systematic reviews that aim to answer specific questions, scoping reviews are used to determine the scope o available evidence in a given field and examine how research is conducted in that field This scoping review is part of the PERMIT project (PERsonalized MedIcine Trials), which aims to map the methods for personalized medicine research and build recommendations for robustness and reproducibility of different stages of the development programs. Although several categorization may be proposed, the PERMIT project considers four main building blocks of the personalized medicine research pipeline: (1) design, building and management of stratification and validation cohorts; (2) application of machine learning methods for patient stratification; (3) use of preclinical methods for translational development, including the use of preclinical models used to assign treatments to patient clusters; and (4) evaluation of treatments in randomized clinical trials [22][23][24]. This scoping review covers the third building block in this framework.…”

Recommendations for robust and reproducible preclinical research in personalised medicine

“…Ideally, robust and reproducible methods should cover all the steps between the generation of the hypothesis (e.g. a given stratum of patients could better respond to a treatment), its validation, and preclinical development, up to the definition of its value in a clinical setting [4][5][6].…”

“…Our results show that, despite the progress in developing innovative and complex preclinical model systems, to date there are fundamental deficits in translational methods, which is one of the obstacles for the implementation of PM. We identified a need for improvements in five critical areas: (1) clinically relevant experimental models; (2) adaptation of standardised protocols, validation procedures, and quality assessment of methods; (3) accurate and transparent reporting; (4) harmonised regulation framework for assessing preclinical evidence; and (5) integration between preclinical and clinical research. These findings are not novel, the issues have been raised by many others before [8][9][10][11][12][13], but the need for adapted guidelines and standards in preclinical research must be addressed to improve translation and enable PM development.…”