At first glance, drug discovery appears to be a straightforward process of screening compound candidates, optimizing leads, performing preclinical evaluations, and then, with these steps completed, beginning clinical trials. The reality is not so neat. These activities often overlap, merge, and conflict with each other. Parameters such as absorption, distribution, metabolism, excretion (ADME), and toxicology are not simply steps in lead optimization, they are actually an integral part of an ongoing process.
While far from perfect, a wide range of assays have made early ADME useful in providing predictive information. It is in the toxicology stage of ADME/Tox where the pitfalls lie. The downside of unfavorable toxicity is a lot more serious than ADME. Unfortunately, pharma companies have found this out the hard way.
Within the past few years, we have seen the crash of several major drugs in diabetes, CNS, cardiovascular, and weight control even after these drugs reached the market. The toxicity problems with these drugs should have been caught before they reached clinical trials, but they weren’t. This was not because of ineptitude of the companies involved, it was because better tools for determining toxicity were not available. This is why pharmas are desperately seeking approaches that will screen out these compounds during the discovery or development stages. There are many different approaches to improving early tox, but no clear winning strategies. The issues are multifaceted and complex. It will take time, talent, and a lot of money to develop them. This creates opportunities for pharmas and suppliers alike.
Just 20 years ago, toxicology studies were high-information but low-throughput estimates of a candidate drug’s toxicology. Failures were common and expensive. The goal to find better approaches led to the development of cellular assays performed in vitro. Companies like Cellomics developed platforms that would allow investigators to treat live cells with candidate compounds and examine effects of these compounds on cellular structure and cell viability. These assays were faster than in vivo experiments and could be run with much less compound.
Pressure continued for the creation of assays that could be used earlier and at even higher throughput. If high-throughput screening generated several hundred leads per day, high-throughput toxicology was needed to evaluate this output. This was a challenge that was difficult to meet. This is because tox is not a single parameter or even a group of parameters: it is a result or an outcome.
Without a single silver bullet to hit the tox target, the best that can be achieved is a group of alternative approaches that can provide partial prediction of a compound’s toxicity. These approaches are similar to ones already in use in other areas of drug discovery but they have been modified to describe the unique characteristics of a toxic response.
Early tox work can be divided into three general approaches: in silico tools, in vitro assays, and in vivo models. These three approaches to early tox are not mutually exclusive. In silico models are verified with experimental results, and this information is then used to change the model. In vivo techniques can be used with in vitro automated systems. Biological pathways uncovered by one model can be used to identify biomarkers that are utilized with a different technology. This illustrates the nature of early tox testing. One approach seldom provides a complete answer about a compound’s toxicity, but the judicious use of complimentary approaches can provide a wealth of useful information.
Products and Markets
Previous in silico models for early tox promised too much and delivered too little. Today’s best in silico models are based on experimental results and closely connected to specific compounds of interest. Models like Bio-Rad’s KnowItAll, GeneGo’s MetaDrug, and IDBS’ PredictionBase are efficient and easy to use, while Gene Logic, Ingenuity Systems, and Iconix offer comprehensive databases. Sharing of information through government agency initiatives and industry consortia has led to more complete databases.
The market for in silico models and databases for early toxicology totaled over $130 million in 2006, and an 11% growth rate is expected for the next five years.
In vitro testing for early tox is the largest and best established of the three product areas. Cell-based assays using high-content screening are becoming more widely used, providing much more content than traditional biochemical assays. Toxicogenomics is a useful tool, although its importance has been exaggerated. It plays a role but is not a substitute for other methods used in early tox.
Traditional major players like BD Biosciences, Invitrogen, and Promega are leading suppliers of assays. Instrument makers such as Beckman Coulter, Caliper, and GE Healthcare provide useful platforms for cellular assays. The market for in vitro testing for early toxicology totaled nearly $400 million in 2006, and an annual growth rate of 10% is expected for the next five years.
The smallest but fastest growing area is in vivo testing for early tox. Zebrafish are the most promising model organism since they provide whole-animal information from assays that can be run easily in a high-throughput format. Certain, specially designed rodent models are also being used early in the drug development process. Charles River Laboratories is a dominant provider of animal models, but smaller companies like Phylonix and Zygogen have a focus on zebrafish and have demonstrated the versatility of this animal model.
Xceleron is a leader in human microdosing, which is the administration of a subclinical dose of a drug directly to humans. Using an ultrasensitive measurement called AMS, the effects of the drug on humans can be determined. This approach (Phase 0 human trials) is still experimental but is gaining wider use and acceptance. While not strictly an early-tox test, microdosing may have a major influence on this market.
The market for in vivo testing in early toxicology is projected to grow at more than 20% annually starting from a modest $71 million in revenues in 2006. In addition to the markets for early-tox testing tools, there is an important market for outsourcing early-tox testing services. Service suppliers have expanded their range of offerings into the discovery area now that pharmas have become more willing to have these tests run by others.
Large CROs like Covance, Evotec, and Quintiles offer some early toxicology services, although this is a small part of their total business. Small to mid-size service suppliers have found early toxicology to be a profitable niche, and many of these service suppliers sell tools as well. Some start-ups use services as a way to leverage their product sales business.
Leading providers of in silico services include Accelrys and Gene Logic. Albany Molecular and CeeTox are among the many providers of in vitro services, and Charles River and Caliper provide animal testing services as well as products. The market for services in early tox was significant in 2006 and will continue to grow at an annual rate of nearly 15% over the next five years.
Achieving success in developing early tox tests is a major challenge. Current tools are not perfect, but they are getting better. Managing information obtained from tests requires excellent software, since the processes being investigated are quite complex. There is a major difference between ADME and Tox tests. ADME properties are physical measurements and can be determined directly. Toxicity properties are biological and descriptive; they are not as easy to quantify. Many different technology approaches have been taken for early tox. Combinations of different approaches are more useful than any single approach taken alone, and in the marketplace, suppliers and clients will have to work with each other to produce meaningful results.