AGILE: Seamless Phase I/IIa Platform for the Rapid Evaluation of Candidates for COVID-19 treatment

Conventional evaluation of new medicines is too lengthy (typically 10 years) to meet the urgent need for treating and preventing COVID-19. Our challenge now is to accelerate this, to quickly identify which amongst the large and diverse list of new compounds and repurposed drugs (existing medicines used for treating other diseases) may be life-saving, and transformational.

Large-scale trials are currently evaluating the 'first wave' of repurposed medicines for treating COVID-19. Should these compounds fail to demonstrate benefit (such as has already occurred with candidates such as hydroxychloroquine, lopinavir, tocilizumab, sarilumab), a range of alternative, 'second wave' compounds (with less clinical evidence) have to be examined. Since a majority of experimental treatments which initially seem attractive will eventually prove ineffective, our best chance of finding an effective treatment lies in screening and choosing the most promising candidates as quickly possible. To do this, we need for a 'feeder' programme to advance plausible candidates for clinical evaluation, and to eliminate candidates with little or no prospect of clinical success.

AGILE is a platform for rapid clinical evaluation of potential COVID treatments. By harnessing modern statistical methods within an innovative trial design, we are able to make a seamless transition (for new compounds) from first-in-human use to finding the optimal dose for use in COVID patients. The trial is
i) Pragmatic - assessing outcomes in dozens (rather than hundreds) of patients
i) Adaptive - a small group of participants are initially enrolled, then (depending on safety) followed by further similar groups, with knowledge of safety, optimal dosing and efficacy accruing with each cycle.
ii) Statistically efficient - use of knowledge from 'control' COVID-19 patients within a statistical model allows multiple arms and interventions simultaneously or in sequence to be assessed. Drugs with little prospect of a moderate to significant effect are rapidly eliminated.
iv) Rapid and responsive - similar 'fast-track' programmes for cancer patients are approved by the UK regulator. Working closely with established consortia, AGILE will advance plausible candidates for these consortia to test in a large-scale trials.

Each new compound is included in a separate arm of AGILE. Innovative features of AGILE are the testing (for the first time in humans) in COVID-19 participants, the ability to tailor the study endpoints and participants to the anticipated deployment of the drug in real-life. This means that testing of antiviral drugs (a major focus of AGILE) takes place in community settings outside hospitals, and in people with mild-moderate disease since this is the scenario where these drugs are most likely to be used, to prevent severe disease, and reduce or prevent infections.

AGILE is designed to rapidly identify those compounds which could be game changers in in the battle against COVID19. To achieve this, we propose establishing a UK-wide network of AGILE sites to become a single, national platform for testing new COVID drugs. AGILE has full regulatory approvals, but is currently only operational in one UK site (Liverpool).

AGILE is a 'proof-of-confidence' engine designed to rapidly identify those compounds in phase I which could be game changers in the battle against COVID-19, validated to the point of readiness for onward inclusion into Phase IIb/III trials. Our vision is to build a single UK-wide phase I/IIa trial platform working to evaluate a strong portfolio of credible treatments, advancing only those most plausible compounds onwards. Unique features are the inclusion of COVID-19 patients from the outset (including first-into-human), statistically efficient trial design enabling rapid decisions, and ability to evaluate patients in the community as well as hospitals. Establishing this national early-phase trial platform will also advance preparedness for future epidemics of new viral diseases.

AGILE utilises a multi-candidate, multi-stage, Bayesian adaptive approach, and comprises a Master Protocol with each investigational compound included as a separate trial arms. AGILE is pragmatic (assessing outcomes in small numbers of patients), flexible (adopting primary endpoints appropriate to the action of each compound, with recruitment tuned to its anticipated therapeutic deployment) and seamless (progressing from first-into-human through dose optimisation to early efficacy evaluation). Staggered cohorts (each with 4 treatment, 2 controls) will be sequentially and flexibly added until the optimal dose is established, and thereafter for safety and efficacy evaluation. The adaptive design allows for i) discontinuation of a dose or candidate due to the probability of any promising effect being too low; ii) adding more participants at the same dose, or iii) recommending further evaluation in a late phase trial, through pre-defined stopping criteria for harm or futility. We have built strong links with UK Phase II/III platforms, the Global Therapeutics Accelerator, WHO and the PHE preclinical COVID19 drug screening platform which is in development.