PS44: Co-Development of Single or Multi-Target Therapies in Combination: Statistical, Clinical, and Regulatory Issues in Trial Designs

Monotherapy may sometimes be inadequate in treating complex diseases like cancer, cardiovascular, and infectious diseases, requiring use of combination therapy aimed at either single or multiple therapeutic targets. To establish the risk-benefit of a new combination therapy, the traditional approach is to develop a new investigational therapy as an add-on to a previously approved regimen, or to develop a fixed-dose combination of previously approved drugs. This approach of discovery and development of combination therapy by sequentially developing individual drugs as add-ons is a long, arduous process. Fortunately, modern understanding of disease pathophysiology has propelled the biopharmaceutical industry towards the discovery of new molecules directed at specific therapeutic targets. This is exemplified in the chronic weight management space with the recent boom of an estimated 150 anti-obesity molecules (AOMs) in development, including novel GLP-1 receptor agonists (GLP1RA), GIP/ GLP-1/GCG dual and triple receptor agonists that are being combined with other drugs with potentially additional desirable properties such as lessening known side effects for example of nausea and vomiting, providing long-lasting effects (RNA interference through gene silencing), improvement in the quality of weight loss (anti-activin A), and preservation of lean mass (anti-myostatin drugs). With the explosion of new discoveries come the challenges of concurrently developing two or more new therapeutic agents as a combination, where neither entity has been specifically developed as a monotherapy. The 2013 US FDA guidance document for concurrent clinical development of multiple investigational drugs (and/or biologics) refers the industry to the much broader 'combination rule' laid out in Title 21 CFR § 300.50 on fixed-combination prescription drugs. Even though the FDA guidance is not new, the topic of co-development could likely become a hot topic of Statistics research in future. Industry trackers show that the co-development of new molecules is expanding across many therapeutic areas covered by various CDER/FDA medical divisions, e.g., under Office of Cardiology, Hematology, Endocrinology and Nephrology and Office of Immunology and Inflammation. This guidance provides a general roadmap and guiding principles on clinical development strategies to characterize the safety and efficacy of a combination, and to the extent needed or possible, the individual drugs in the combination. Although, a general guidance document is a helpful starting point, it cannot contemplate all scenarios, and the success of a development program hinges on the efficiencies and statistical details of the clinical trials; in other words, the devil is in the details. Along with clinical/clinical pharmacology/regulatory challenges, statistical challenges abound in this area. In this session, there will be three speakers: one from the FDA, one from industry and one from Academia to address a variety of trial design and analysis issues on developing combination drugs: 1. FDA perspective: This presentation educates us about the combination rule in the FDA regulations and highlights the emerging challenges in developing combination products such as the need for superiority testing, factorial designs and multiplicity controls for testing individual components. The talk will aim to provide a practical framework to navigate these challenges and explores use of surrogate endpoints and enrichment strategies. Lessons learnt from real-world cases studies across therapeutic areas on both approvals and setbacks will be provided. 2. Industry perspective: This presentation gives a broad stroke overview of design and analysis challenges from early through late-phase studies. In early phase 1 designs will be explored to determine dose escalation steps and estimate early efficacy for recommended phase 2 doses. In proof-of-concept / dose-ranging phase 2 to provide early evidence on effectiveness of combination and optimize doses for the combination for phase 3, how to design studies to estimate the treatment effect attributable to each drug in the combination. Is a standard factorial design necessary, feasible and scientifically justified to support combination use? Does each individual drug in a combination have activity and can the individual drugs be administered together or do they have to be administered separately. In which scenarios would it be inappropriate to include monotherapy treatment arms of the individual drugs used in combination. Instead of a standard factorial design (AB vs A vs B vs standard of care [SOC] or placebo), can an adaptive factorial design, e.g., early termination of monotherapy arms that are ineffective be used? If the contribution of components has not been previously established prior to a confirmatory phase 3, what statistical rules guide how it should be established. Again, is a factorial design required (e.g., AB vs. A vs. B) and would this make the phase 3 program unnecessarily large 3. Academia perspective: This presentation will tell us about the historical lack of proper design and analysis methods for multi-drug combination studies and then focuses on recent developments. The talk will propose a novel two-stage procedure for obtaining a maximum likelihood estimate of the combination does by utilizing in silico model that builds on experimental data of single drugs and systems biology information. Statistical modeling of joint dose effect and an adaptive Bayesian phase 1 trial design for multidrug combinations will be presented. Lastly, this session will include a cross-disciplinary panel of stakeholders from academia, industry and FDA with expertise in Clinical, Regulatory Affairs and Statistics. Panelists will debate and discuss all of the above challenges in concurrent development of multiple new drugs (or biologics) to be used in combination for an indication. In addition, panelists will be asked to debate on these special topics: 1) co-development of drugs for rare diseases where it may not be possible to do a factorial design comparing combination vs. each monotherapy arm of individual drugs; how should these studies be designed and comparison between which arms is appropriate? 2) what are some efficient pathways for development of existing approved drugs for use in combination for a new indication, and how to identify doses/regimen, and establish safety and efficacy in a new disease.