A roadmap to using historical controls in clinical trials – by Drug Information Association Adaptive Design Scientific Working Group (DIA-ADSWG)

Abstract

Historical controls (HCs) can be used for model parameter estimation at the study design phase, adaptation within a study, or supplementation or replacement of a control arm. Currently on the latter, there is no practical roadmap from design to analysis of a clinical trial to address selection and inclusion of HCs, while maintaining scientific validity. This paper provides a comprehensive roadmap for planning, conducting, analyzing and reporting of studies using HCs, mainly when a randomized clinical trial is not possible. We review recent applications of HC in clinical trials, in which either predominantly a large treatment effect overcame concerns about bias, or the trial targeted a life-threatening disease with no treatment options. In contrast, we address how the evidentiary standard of a trial can be strengthened with optimized study designs and analysis strategies, emphasizing rare and pediatric indications. We highlight the importance of simulation and sensitivity analyses for estimating the range of uncertainties in the estimation of treatment effect when traditional randomization is not possible. Overall, the paper provides a roadmap for using HCs.

Cancer clinical trial vs real‐world outcomes for standard of care first‐line treatment in the advanced disease setting

“…quately powered to answer the hypothesis, 35 was underestimated due to historical data being used for trial modelling. 36 A 3-year invasive DFS rate of 89.2% was assumed in the placebo arm; however, the final DFS rate for this cohort was 93.2%.…”

“…Any improvement in outcomes over time as we report needs to be factored into the statistical plan for future studies where the intervention arm of a prior positive study becomes the control arm for the next study. In particular, it is important to ensure that trials are adequately powered to answer the hypothesis, 35 which must be informed by expected outcomes in the control arm. As an extreme example the survival outcomes seen in NO16966 for oxaliplatin‐based treatment plus bevacizumab (median OS 21.3 months), still a standard of care for many patients with metastatic CRC, was already substantially exceeded in the TRACC registry (median OS 26.4 months) and further gains over time should be anticipated.…”

Planning for the Next Pandemic: Ethics and Innovation Today for Improved Clinical Trials Tomorrow

“…In addition, careful thought would have to be given to the rapid improvements in supportive care during a pandemic that could lead to relatively rapid improvements in control arm outcomes, potentially requiring data from early in one platform trial to be considered historical data when shared with another more recent trial, or even within the same trial. Methods for incorporation of historical control data exist (Ghadessi et al 2020). A general methodology would have to be pre-specified and agreed between platform trial designers.…”

“…It is worth considering legislation to compel sponsors to share control vaccine data in the event of a national pandemic emergency. Sharing of control data between studies can accelerate development and reduce the number of patients on control therapy, but a roadmap of planning and methodological issues must be carefully considered (Ghadessi et al 2020). In traditional RCTs, complete elimination of unknown confounders cannot be guaranteed, but randomization and blinding mechanisms reduce the chance of bias.…”

Editorial Commentary: Medial Meniscal Root Repair May Not Be Required During Knee Medial-Compartment Unloading High Tibial Osteotomy

“…Comparisons with historical control groups are problematic and introduce systematic error. 13,14 Selection bias is a common problem, and often, patients have better clinical outcomes. 14 Adaptive designs and Bayesian approaches to analysis are a good alternative but often aim to discount or down weigh historical controls.…”