Cancer Institute Turns To Cell Line Screening

Looking for compounds to kill tumors, Michael Boyd abandons leukemic mice and tries a radically new, unprecedentedly large assay In a sharp break from its past, the National Cancer Institute has abandoned a 30-year-old system for finding compounds that kill tumors and has replaced it with an elaborate but unproven factory-style operation. In the new program, unique in its scope, 20,000 compounds and extracts will be tested against a broad array of human tumor cells each year. The compounds will be submitted by chemical and pharmaceutical companies, while the extracts will be drawn from tropical plants and marine organisms obtained from outside researchers. "What we've created here is a novel and unprecedented resource," says Michael Boyd, who as director of the institute's Developmental Therapeutics Program has assembled the program despite stiff scientific, technical, and institutional resistance. The move is being watched with some skepticism by other cancer researchers. There's wide agreement that the old screen, based on the compounds' effects on leukemic mice, yielded too few new drugs. But critics say there's no guarantee that an assembly line of cell cultures, photometers, and computers will be any more productive. "Only time will tell if it was a good management decision or not," says Raymond Taetle of the University of California, San Diego, a member of the ad hoc peer committee that has kept a close eye on the program. The program, which began last fall and is now in full operation, is expected to cost between $3.5 million and $4 million a year. It employs about 60 people at the NCI's Frederick Cancer Research Facility in Bethesda, Md. If the gamble works, much credit will go to the tenacious and persuasive efforts of Boyd, a 42-year-old chemist and M.D. who was promoted out of the lab six years ago to create the new project and, now that it's finally up and running, is going back to the bench to help it succeed. "All it takes is one powerful drug that would work, really work, against prostate or breast or bowel cancer, and everyone would say it was a great smashing success," says David Korn, the dean of Stanford Medical School and the chairman of the National Cancer Advisory Board. "I don't know what the odds are." The new drug screening project was conceived in October 1983, when Bruce Chabner, director of NCI's Division of Cancer Treatment, promoted Boyd from lab chief to director of the division's Developmental Therapeutics Program with firm instructions to shake things up. "My thinking was that we needed a tough person, a guy willing to take chances, a tenacious fellow who could change the direction of a big ship sailing downstream," says NCI division director Chabner. Boyd says he was "a happily reclusive lab chief, with no aspiration or inclination to a major program management job" overseeing 250 people plus numerous grants and contracts, on a $200 million budget. But he agreed it was time for a change. "I had squawked one too many times" about the program's drug-finding efforts, he says. Over the course of three decades, the NCI had screened hundreds of thousands of compounds, synthetic and natural, in a classic try-everything-on-the-shelf search for ones that would kill cancerous cells. All that NCI scientists and their colleagues in industry and academia had to show for it was 36 licensed drugs. This screen had usually relied, for its first tier, on legions of mice with leukemia. Only those compounds that seemed to retard mouse leukemia were tested further against cancers in humans. It's perhaps no surprise then that chemotherapists gained powerful drugs for fighting leukemia and lymphoma, but little to challenge tumors of the brain, lung, kidney, and other organs. While there was no widespread clamor against the leukemic mouse test, Boyd and Chabner, like many others, suspected the mouse model was overlooking tumor-killing compounds. Boyd knew for certain of one it overlooked. Since his days at Vanderbilt University Medical Center in Nashville in the early 1970s he had been studying moldy sweet potatoes, isolating and then synthesizing a toxin, 4-Ipomeanol, that kills certain lung cells. Its activity is sufficiently specific to lung tumor cells that it has gone into clinical trials. But it has little effect on the conventional mouse antitumor screens. With no new funds available, Boyd and Chabner decided it made more sense to shut down the mouse program entirely and replace it with a new model, based on direct observation of each com-pound's effects on a variety of human tumor cells. Many basic advances in cancer biology have come from cell studies, of course, and cell lines have been used to isolate the active components in natural products extracts. So it was logical, Chabner says, to try to use human tumor cells to find chemicals that killed those tumors. While university and industry labs tend to focus on specific drugs - for instance, the identification of a crucial tumor cell function or structure and an attempt to find a drug that would disrupt it - NCI's goal has always been to screen thousands of compounds at random in the hope of turning up new leads. Boyd's idea involved an unprecedentedly large array of human cell lines for simultaneous tests. "The concept is quite simple, but we've found over the last five years that converting the concept to reality is quite a challenge," Boyd says. "It's been a nightmare, frankly." The mouse testing program seemed to have taken on "a life of its own." The grants and contracts with outside labs took several years to terminate. As for its replacement, "all the experts in the field said this was not possible to do and not worth trying." Taetle adds, "It would be a mistake to say there was ever a big consensus on the best way to proceed." Taetle applauds NCI's efforts to solicit ideas about the proposal from industry and academia. These critics' tough questions forced changes on nearly every point: the goals, the procedures, the data collection, the data analysis. "If you believe there is some benefit in testing a huge array of potential materials from seabeds and forests and mountains, this is as good a way as any," says Korn, whose committee reviews the institute's programs. "But it isn't a tight, scientifically impeccable assay." The problem, in Korn's view, is the difficulty of looking for things that kill cells. "There are huge numbers of trivial things that can kill cells in culture. Therefore, there's a lot of noise in the assay." That's unlike NCI's cell-based screen for new AIDS-fighting compounds, also begun under Boyd, which selects compounds that can keep infected cells alive. "That's a fairly clean assay logic," Korn says. Another problem Boyd's group faced was to select cell lines for the assay. "It was surprisingly difficult to find well-characterized cell lines," Taetle says. Those they considered often lacked good evidence for their organs of origin, for the number of generations since biopsy, and for assurances that the patient had not had chemotherapy. Skeptics wonder whether the cell lines still sufficiently resemble patients' tumors, according to Frederick Valeriote, deputy director of the Division of Hematology and Oncology at Wayne State School of Medicine in Detroit. He asks whether this is the best use of NCI's limited funds, and whether in fact there are many cytotoxic compounds waiting to be discovered. "For a long time there was such a dearth of agents coming through that screen that people got a bit jaded" about the existence of many more of them, he says. Boyd's program screened its first few compounds last summer and, with endorsements from the ad-hoc committee and Korn's committee, reached high gear around year's end. Now that it's up and running, here's how it works: The compounds arrive at Frederick from two places. Pure synthetic compounds, invented by the thousands every year, are submitted by chemical and pharmaceutical firms, either because they want the compound checked or because it looks promising to NCI chemists. In the more complicated half of the program, several botanists and marine biologists working in the field, especially in the tropics, are under contract to submit whatever the local residents recommend for ailments, or any plants and animals that seem to use chemical defenses against predators. Chabner is especially proud of this aspect. "It's sort of a last-ditch effort to screen and characterize products from the rain forest that are disappearing so fast." NCI executed a major plan screening project in the 1960s and 1970s, but it was based on the leukemic mouse test. The compounds and the extracts are tested, at various dilutions, against a battery of 60 different cell lines, which represent seven cancer sites: lung, colon, skin, kidney, ovary, brain, and blood cells. A pink dye adheres to the live cells; a photometer records its intensity. The fainter the pink, the deadlier the compound. The information is recorded electronically, and a computer screen displays any promising results. The rare compound that seems to be selective in its ability to kill or retard cells graduates to the next level of tests, against mice implanted with human tumors. If it passes that test, it goes into the first phases of clinical testing. Now comes a race to see whether the program can prove itself before NCI officials or outside reviewers become impatient and tell Chabner he should cancel it. "We've got to give it 10 years," Chabner says. "Then if there's no positive clinical results I'll be discouraged." Boyd will be doing all he can to see that it does succeed, but from a less-visible position, as chief of a new lab isolating the active components of the natural extracts that pass through the first screen. "I didn't want to leave [the director's job] until I nurtured it to where it could sustain itself," Boyd says. "A few months ago I felt it had reached that point and I could pull the great escape." Chabner reluctantly agreed. So far, most of Boyd's duties are the bureaucratic measures needed to create his new lab. Nevertheless, he says, "just having an extra couple of hours every couple of days to catch up on six lost years of research literature is great fun." David Ansley is a medical writer for the San Jose Mercury-News.

Cancer Institute Turns To Cell Line Screening

Interested in reading more?

Become a Member of