FRAMINGHAM (03/22/2004) - Chemogenomic pioneer Iconix Pharmaceuticals Inc. recently got a big boost from major deals with Abbott Laboratories Inc. and Bristol-Myers Squibb Co. (BMS). "These are both very large platform relationships," says Alan Roter, Iconix's vice president of informatics. Financial terms for the arrangement with Abbott were not disclosed, but Iconix stands to earn approximately US$24 million from its multi-year deal with BMS alone.
The deals are partly a sign of Iconix's maturity. The company now has gene-expression and toxicity data for more than 600 compounds in up to eight different kinds of tissues. But these agreements also signal that companies are heeding the U.S. Food and Drug Administration's admonition to pay attention to pharmacogenomics.
Surprise toxicity is a big thorn in pharma's side. "Unless we can reduce attrition, we have no long-term prospects," says Richard T. Robertson, senior vice president of drug safety and pharmaceutical candidate optimization at Bristol-Myers Squibb.
The use of gene signatures for toxicity screening is one of the most sensitive issues on the FDA's fast-moving pharmacogenomics agenda. The agency is adamant it will seek to protect consumers using whatever technology works, even if that means helping bring the field up to date.
Many companies wrestling with gene chip data have expressed reservations about how their results will be interpreted. After all, a "bad" expression signature doesn't always mean the drug is toxic; it could be an artifact of some type.
That's only a problem if, like most companies, you have small data sets. "You need a very large data set to be able to tell if a change is real or not," says Donna Mendrick, vice president of toxicogenomics at Gene Logic.
Recent toxicogenomic survey data from Bioinformatics LLC supports that statement: About a third of respondents said they were going to do more array-based studies, but many also voiced concerns about the reproducibility of their results and "obtaining enough samples for statistical relevance."
That's where companies such as Gene Logic and Iconix come in. By generating massive data sets with strict quality controls, they provide critical baseline data.
"Customers would have to spend a huge amount of money to build a database like ours," Roter says. Companies still do a lot of toxicogenomics internally for secrecy's sake. But with the growing concern about drug failures from toxicity, and the FDA's clear interest in toxicity-determining tests, companies are becoming more interested in services.
"Things really picked up in 2003," Roter says. "Gene expression was moving very slowly until then." Iconix has at least eight other customers besides Abbott and BMS.
Gene Logic, the other leader in this field, says it has about a dozen. "Interest is growing," Mendrick agrees.
Drug safety is so important that pharmaceutical companies are looking for multiple solutions. "Our deal with Iconix is part of an integrated strategy," Robertson says. His group is focused on toxicogenomics and metabonomic solutions right now, but "we won't exclude anything from consideration," he notes.
The use of gene signatures for toxicity screening is one of the most sensitive issues on the FDA's fast-moving pharmacogenomics agenda. The agency is adamant it will seek to protect consumers using whatever technology works.