By David N. Leff
Imagine a clinical trial with 60 cancer patients, each treated with 60,000 different experimental agents -- one at a time.
That is the scene conjured up by physician and biophysicist John Weinstein, a senior researcher in the National Cancer Institute's Laboratory of Molecular Pharmacology.
Weinstein is first author of an article in the current issue of Science, dated Jan. 17, 1997. Its title: "An information-intensive approach to the molecular pharmacology of cancer."
That approach, he told BioWorld Today, exploits two resources:
* "Sixty cancer cell lines, derived from a wide variety of tumor types -- lung, colon, breast, for example -- against which, since 1990, we have screened more than 60,000 natural and synthetic products, from academic and industrial sources."
* "A computer algorithm named DISCOVERY, which," Weinstein explained, "integrates different types of information about the compounds, and displays them in ways suited to human pattern recognition.
"In some sense," he observed, "this is akin to the multiple biochemical assays at the molecular level, but done in whole cells."
Weinstein pointed out, "One of the special aspects of this program is that the cell lines are being characterized with respect to a large variety of the interesting sorts of cellular factors, ranging from oncogenes to tumor suppressors to transport molecules.
"We are searching for targets that have not previously been thought of as targets."
"Our original hypothesis," he continued, "was that if one found agents selectively active against, let's say, breast or colon cancer, that would tend to carry forward into the clinic. And that hypothesis remains to be validated, as these agents progress in clinical trials."
So far, the program has begun Phase I studies of five putative anticancer compounds assessed and analyzed by the 60-cell-line resource. "They were selected," Weinstein pointed out, "partly on the basis that the information-based technology showed them to be unlike other agents already in the clinic."
Physician and pharmacologist Edward Sausville, a co-author of the Science paper, manages NCI's developmental therapeutics program. "We select such compounds for evaluation in the screening system," Sausville said, "based on their uniqueness of structure and activity pattern. That they are not analogs of currently available compounds."
He described to BioWorld Today the Phase I studies of these five agents in a wide variety of cancers -- renal, prostate, lymphoma, colon -- the full range of tumor types:.
* "A flavopiridol came to us originally from Hoechst -- now Hoechst-Marion-Roussel -- in 1991 or so," Sausville began. "Our screening studies revealed it to have high potency as a cyclin-kinase inhibitor, with anti-tumor activity in animals."
* "Kyowa Hakko Kogyo Co., Ltd., Tokyo," he continued, "provided two compounds, a quinocarmycin analog and an agent code-named UCNO 1. Both showed differential evidence of growth inhibitory effect, and some in vivo activity in animals." The former product, he observed, "is probably a DNA-interactive agent."
* A cyclic depsipeptide from Fujisawa Research., Osaka, "is very potently active in in vivo models, and has a novel structure."
* A spicamycin analog, supplied by Kirin Brewery Co. Ltd., Tokyo, is under Phase I study at Boston's Dana Farber Cancer Institute. "We have great hope for this agent in relation to certain colon tumors," Sausville said.
He added, "We're very excited at the possibility of this screening approach being married to the important breakthroughs in the biotechnology industry. If compounds emerge from our screening system that don't have targets clearly demarcated, having the ability to potentially engage in a dialogue with the biotech industry, to learn more about those compounds, would be something that we'd be very much interested in.
"A second use," Sausville concluded, "would be to actually use this screen for novel molecules coming from the biotechnology industry, in an effort to discern patterns of activity that might increase enthusiasm for their development." *