A team from Imperial College London puts forward an argument for the use of clinical trial simulators in planning future trials in Alzheimer’s disease

The growing global burden of dementia reflects an aging population, where Alzheimer’s disease (AD) is the most common cause of dementia accounting for roughly two thirds of all cases. As yet, there is neither a cure for AD nor any disease modifying treatments and only 5 drugs that provide symptomatic relief. More worryingly, it was reported that 99.6% of the total 413 AD clinical trials that were conducted for potential treatments between 2002 and 2012 had all failed, the reasons for failure remain poorly understood.

Why clinical trials may fail.

A team of epidemiologists (the Immunotherapy Trial Simulation (ITS) Team) led by Professor Sir Roy Anderson and Professor Frank de Wolf within the School of Public Health at Imperial College London have recently published a Commentary entitled ‘Why do so many clinical trials of therapies for Alzheimer’s disease fail?’ in The Lancet (1). The article discusses potential reasons why AD clinical trials are currently failing, which include starting to test therapies too late in disease development, incorrect drug doses, wrong treatment target, and an inadequate understanding of the biology of Alzheimer’s disease. Nevertheless, while it all may be true, the authors propose that there could be a simpler explanation based on the choice of the clinical endpoint for the trials and associated variability in measurement of these endpoints within or between individuals. The endpoints could be scores in cognitive tests, concentrations of specific biomarker proteins in the cerebrospinal fluid or brain scans to measure brain volume changes and protein deposition.

The authors constructed a clinical trial simulator (CTS) that uses computational approaches based on mathematical models of disease induction and progression to explore the potential outcomes of a clinical trial under various trial designs and endpoint choices.

CTSs are powerful tools for increasing understanding of how the pharmacokinetics and pharmacodynamics of a drug influence the choices of the clinical endpoint, sample size, patient recruitment criteria, and clinical trial duration. Indeed, the FDA and EMA regulatory bodies for drug approval have begun to take notice of simulation strategies in an effort to support improved drug development efficiencies. For example, in the midst of an influenza epidemic in 2009, the FDA used a CTS to identify and approve a safe paediatric dose of an experimental drug that had never been studied in children.

For AD, the mathematical model that underpins the simulator can be based on the description of the probability that a patient moves between healthy and diseased states in a defined interval of time in a large sample of patients and takes into account important confounders such as age, sex and presence of any genetic markers associated with AD. The distributions of these large longitudinal studies can be incorporated into the CTS and the effect of a possible therapy believed to slow transitions at a defined efficacy level can be analysed. A striking feature of the AD simulation is the prediction that given the observed variability as derived from observed longitudinal data, it would be difficult to identify a treatment that had even an 80% efficacy in slowing disease progression with a clinical trial that lasted up to 5 years. This would have a profound effect on future AD therapies. The key conclusions from the simulation are two-fold. First, the variability in measurements between individuals implies that it will always be difficult to detect even a good treatment efficacy in a trial of 5 years’ duration. Second, the use of clinical trial simulators can provide powerful insights into the likelihood that a given trial design and endpoint choice are appropriate before an expensive clinical trial is initiated. Such trial simulators however are currently not readily employed by the pharmaceutical industry. The ITS team believes that CTS use, especially when evaluating therapies for diseases with long incubation or development durations that require many years of patient follow-up as in AD, could assist in improving trial outcomes and should be used more often.

The ITS team, in collaboration with the Alzheimer’s Cohort Consortium (ACC), aims to continue development of an AD preventative treatment CTS as an essential tool for improving the design, conduct and efficiency of phase II and phase III AD clinical trials. Further, the team are interested in the impending need for the identification of plasma biomarkers and a biomarker-based diagnostic framework for AD, all of which could contribute to the emergence of an effective treatment option for a disease with a high unmet need.

ITS Team Members in the Department of Infectious Disease Epidemiology, School of Public Health, Imperial College London: Professor Sir Roy Anderson FRS, Professor Frank de Wolf, Dr. Chris Hadjichrysanthou, Dr. Stephanie Evans, Mr. Kevin McRae-McKee, Dr. Mei Mei Wong and Dr Emily McNaughton.

Professor Sir Roy Anderson is a non-executive director of and shareholder in GlaxoSmithKline (GSK), which does not produce therapies for Alzheimer’s disease, and receives research funding from the Janssen Prevention Center, Leiden, the Netherlands; GSK and Janssen did not contribute in any way to the opinions expressed in this Comment.

1. Anderson RM, Hadjichrysanthou C, Evans S, Wong MM. Why do so many clinical trials of therapies for Alzheimer’s disease fail? Lancet. 2017;390(10110):2327-9.