Blog | Preclinical and Clinical Laboratory Services

Drug Discovery Success Rates | Role of Preclinical Study Design.

Written by MD Biosciences | Apr 22, 2015 5:16:55 AM

In 2014, an article was published in Nature analyzing the clinical development success rates for investigational drugs. It's no surprise that the success rates are still somewhat dismal with 1 in 10 drugs that enter clinical phases pushing through to FDA approval. The article breaks down the success rate in each phase for differing classes of drugs as well as various therapeutic indications. NMEs were found to have the lowest success rates in every phase of development (7.5%) whereas biologics had nearly two times the success rate (14.6%).

While the article continues with plenty of statistics for all to devour, the more interesting part of the article looks at the root cause of drug development failures (described as the discontinuing of a drug development program). The 4 main categories for why drug dievelopment programs are stopped at such late phases are efficacy, safety, commercial and a broad category of unknowns. The authors found that over half of the suspensions were contributed to some measure of efficacy, although it is unknown if that is due to poor study design or lack of biological activity.  While there are plenty of improvements that can be made at the clinical trial stages, backing up further into early stages such as the basic research and preclinical stages can play a role in increasing the success rates. 

More predictive animals models, earlier toxicology, biomarkers and targeted delivery are all points mentioned for improving the preclinical stage for predicting successes at clinical phases.

MD Biosciences recognizes this and is constantly adding relevance to its models for correlating or closing the gap between preclinical and clinical. We do this in a couple of ways to :

Phenotypic screening: our scientists have developed an in vivo matrix based screening program that provides valuable phenotypic measures with biomarkers and predictive toxicology data early on. Leads can be selected and moved forward based on this combination of data that is determined in a true in vivo environment (not a cell based one that mimicks only to not hold up once brought into in vivo). Cell based screens may not predict the activity that observed in vivo as complex pathways and mechanisms are not involved as they are in whole animal systems.

Large animal, pig models. We are constantly developing preclinical models in larger species such as the pig/swine in order to provide more translatable results. While there is tremendous value in using the rodent models (particularly in early stages), the relevance of the pig to human is greater than rodents thereby strengthening the data. Studies available in the pig currently are post-operative pain, neuropathic pain, nerve block, wound healing and arthritis. The pig is the ideal species for the fact that so many systems are similar between the human and pig. 

Our scientists work with many companies on long term collaborations to bring drug discovery pipelines from screening stages to either licensable or clinical phase stages using a combination of the approaches listed above during preclinical study design phases. While this post here is in no way meant to be in depth of all the strategies or contributing factors to clinical phase failures, it is meant to highlight some of the ways that can be incoroporated into preclinical phases to facilitate success later on. Please contact a scientist to discuss your program specifically.