You may have been there: You walk out of the notary’s office, incorporation papers in hand. A new company’s future dependent upon data produced in vitro, published in a handful of journals. Soon, the excitement of new beginnings gives rise to “translation anxiety”: Will the effects observed in cells be visible in more complex disease models?

Two blocks down the road, another pharmaceutical company files its pre-clinical dossier with the regulators. Next stop: Clinical trials. Will the animal data translate once human volunteers and patients are exposed to the future drug? The rate of failure in Phase 3 clinical trials is still close to 50 percent, and is generally associated with poor translatability between animals and humans (2).

While “de-risking an asset” is often a priority for management, “translatability” haunts the pre-clinical scientists. The choice of efficacy models in early Discovery is dictated by genetic proximity to humans. Later in pre-clinical development, toxicology species will often be selected based on metabolic resemblance with human patients: Whichever species metabolizes the drug candidate most like humans becomes the de facto toxicology species.

The toxicologists’ strategy appears to work: only 24% of cardiovascular drug candidates entering Phase 2 trials failed due to safety concerns since 1990 (1), compared to 50% due to lack of efficacy.

Common approaches to maximizing translation in efficacy assessment use genomics and proteomics to compare the animal and human therapeutic target pathways. When necessary, genetic modification of small animals to replace a human protein with its human isoform is performed, as is the “humanization’ of rodents to present immune responses supported by human cells. The modification of animals to enhance “translation” has been phenomenal over the last 20 years.

The same cannot be said for the diseases induced in these animals: most pre-clinical disease models develop within weeks — months at best — a very acute progression compared to human timelines. How similar is a 28-day rodent disease to a chronic pathology maturing over two decades in humans?

Inflammation, tissue remodeling, fibrosis occur at an accelerated rate in rodents which may be impossible to counter at clinical doses of a drug candidate, leading to incorrect efficacy predictions.

If the translation of safety concerns across species has improved significantly since the introduction of biomarkers (3), prediction of human efficacy remains imperfect, thanks to the sub-chronic nature of animal diseases compared to the longer time-course of their clinical equivalents. Therein lies the challenge of the next decade in drug development.

References:

1- Hwang T.J., Lauffenburger J.L., Franklin J.M., Kesselheim A.S. Temporal trends and factors associated with cardiovascular drug development. J Am Coll Cardiol Basic Trans Sci. 2016, Issue 1: 301–330.

2- Huff, R. The High Price Of Failed Clinical Trials: Time To Rethink The Model. Clinical Leader, October 3, 2016 edition.

3- Wong C.H., Siah K.W. Estimation of clinical trial success rates and related parameters. Biostatistics. 2018;00:1–14.