A micro-segmentation of the research enterprise along with drying up drug pipelines and piling up of un-translated basic science discoveries has prompted all the stakeholders of innovation towards development of translational medicine tools for efficacious and expedited entry of products to the market.
Clinical trials are important and have lately come under scrutiny for their efficiency and exploring scope for improvisation in their traditional designs to de-risk the drug discovery enterprise both at the innovators end and for the patients.
Translational medicine tools are being developed for the whole spectrum of the clinical trial process coalescing the phases into an integrated process with greater flexibility and maximum use of knowledge accumulated. The models emerging are adaptive with multi directional knowledge flow over exploratory and confirmatory domains. Some of the key translational tools being included into the process are modelling and simulation, Bayesian probability models, adaptive designs, seamless designs and sample size re-estimations (Orloff et al 2009).
Translational strategies for assessing optimal biological dose, biomarker development and validation, phase 0 trials and novel imaging techniques, adaptive designs for phase II/III protocols, and health technology assessments concurrent to phase III trials help to optimize efficiency of clinical drug development, particularly of targeted therapeutics (Schellens & Beijnen 2009).
Biological, pharmacological and statistical modeling are recommended throughout the process of drug development and help in understanding biological processes, dose selections, and assessment of development strategies respectively. An innovative biomarker strategy is essential for accelerated drug development processes but it is recommended to conduct methodology research in parallel, to demonstrate its correlation with late stage endpoints. Modeling also helps in early development with the use of external information with a scope of data sharing from other development projects or in between companies. In confirmatory phase, simulation assesses the efficiency of study designs, further guiding the development strategy. Bayesian probability models incorporate prior observations for comparison with new experimental values thereby improving safety signal detection in exploratory models (Orloff et al 2009).
An adaptive trial design utilizes provisional data to further modify and improve the study design in a pre-determined way. In the exploratory phase patients may be re-assigned to the most beneficial study arms with a scope of testing many doses at a time for later success in the confirmatory phase. In confirmatory phase it uses the same strategy for efficacy with further scope of sample size re-estimation and change of eligibility criteria. Seamless trial designs attempt to combine objects of separate studies in a single trial like the seamless adaptive Phase II/III clinical trial (Orloff et al 2009).
A variant of classical adaptive designs are the ‘dynamic treatment regimes’ or ‘adaptive treatment strategies’ which allow for dosage level and type to vary on a time scale and according to the subject’s need. Such a reinforcement strategy utilized for management of chronic diseases has also been used for trials such as ‘sequential multiple assignment randomized trials’ (SMART). Such reinforcement learning designs are being used as a tool for discovering optimal treatment regimens for life-threatening diseases like cancer (Zhao et al 2009).
Advanced and flexible trial designs require robust statistical analyses and computations thereby increasing the planning and protocol development time. Challenges are faced in getting these approved by regulatory agencies and ethical review boards. This may lead to the projects being reverted on the traditional path. These novel designs may also require timely accrual rate monitoring, efficient software, communications infrastructure across sites and central un-blinded analysis center along with a flexible drug supply process (Orloff et al 2009).
Challenges could be faced in the implementation of translational medicine tools and techniques across interdisciplinary lines and stakeholders. As a solution to this it is suggested to develop a common vocabulary and an understanding of the benefits of novel trial designs by all the stakeholders. Concurrent development of guidelines with exemplary case studies for the use of novel trial designs is also recommended. At the regulatory interface a procedure for communication needs to be established for discussing constraints and issues which may translate into timely approvals and acceptance (Orloff et al 2009).
References
Orloff J, Douglas F, Pinheiro J, Levinson S, Branson M, Chaturvedi P, Ette E, Gallo P, Hirsch G, Mehta C, Patel N, Sabir S, Springs S, Stanski D, Evers MR, Fleming E, Singh N, Tramontin T, Golub H. The future of drug development: advancing clinical trial design. Nat Rev Drug Discov. 2009 Dec;8(12):949-57. Epub 2009 Oct 9.
Schellens JH, Beijnen JH. Novel clinical trial designs for innovative therapies.Clin Pharmacol Ther. 2009 Feb;85(2):212-6. Epub 2008 Dec 24.
Zhao Y, Kosorok MR, Zeng D. Reinforcement learning design for cancer clinical trials, Stat Med. 2009 Nov 20;28(26):3294-315.