The Technology of Change
In the past decade the causes of clinical trial failure have changed dramatically. Back in 1995, 40% of drugs failed in Phase I because of pharmacokinetic issues. Today, due to the success of in vitro and in silico ADME, this figure has been reduced to under 10%. The bad news is that those 30 percentage points of failure have been re-allocated to other, thornier causes of failure, most particularly toxicity, which is showing up not only very late in development but, even worse, after commercialization.
Help, however, is on the way from the same in vitro and in silico technologies that were so successful in addressing the pharmacokinetics problems. New assays, techniques, and models are blossoming, allowing the identification of toxicity not only much earlier and at lower cost than via traditional models, but also more accurately.
Until very recently the pharmaceutical industry has had to rely mostly on preclinical animal models for predicting toxicity. Regulatory agencies continue to rely heavily on such models. The inefficiencies associated with traditional toxicity models are huge, in terms of both out-of-pocket and opportunity costs. These models are time-consuming and expensive to execute, and they are insufficiently predictive of human mechanisms of toxicity.
In the 1980s there were no major withdrawals of pharmaceuticals from the market due to toxicity. In the 2000s, there were over a dozen.
Technological progress is now responsible for faster, cheaper, and more accurate alternatives to animal models. Many new models have recently been developed and validated. Far more are soon to come. Yet, despite strong peer reviews and validation and acceptance in the scientific community, these new models have yet to be widely adopted in the industry.
Many compounds are now entering regulatory preclinical safety testing and clinical trials even though the latest in vitro toxicity testing tools are able to predict that these compounds will fail due to toxicity. If these new cutting-edge technologies were used more widely, the percentage of compounds failing in clinical trials due to toxicity would fall immediately, eliminating huge amounts of wasted investment in drug development, commercialization, litigation, and injury claims.
On the other side of the risk equation, each year many potentially successful compounds are scuttled because they’re toxic to rats and other species used for testing, but those compounds would not necessarily be toxic to humans. For now this is a known unknown. But in a few years we can expect to have technology that will make these evaluations.
As animal toxicity is weeding out 10–20% of all compounds prior to clinical trials, this means that there’s a sizable potential to increase the number of drug approvals from existing pipelines once our technology sufficiently advances. It also means that decades of compounds that failed for reasons of animal toxicity could be revisited, with the potential that a large number of new drugs could come to market as a result of such endeavors.