In Ethiopia there grows a shrub called Maytenus serrata. Its bark andwood yield a substance with striking, but limited, activity againstcancerous cells.

Maytansines, it turns out, also are manufactured by microorganismsbearing the name Nocardia, which look like something between abacterium and a fungus. Some Nocardia species also make the knownanti-cancer drugs, rifamycin and vancomycin.

These are weapons in the arsenal of cancer chemotherapists, butmaytansines, even though they too bind tubulin in tumor cells, neverquite made that grade.

"This cytotoxic drug," recalled medicinal chemist Ravi Chari, "wentthrough Phase II clinical trials at the National Cancer Institute in the1970s. They treated 557 patients, and found that maytansine didn'thave much of a therapeutic window. That is, it's 100-fold morepotent than a typical anti-cancer drug like vincristine or doxorubicin,but lacked specificity between targeting tumors and non-canceroustissues."

Ever since the discovery of monoclonal antibodies in 1978,oncologists have been trying to conjugate these target-seekingproteins onto anti-cancer drugs, and aiming them at specific antigenson tumors.

That still-ongoing effort to enhance specificity by means of theseimmunoconjugates runs into a Catch-22 barrier: What the antibody-antimetabolite package gains on the swings of specificity, it loses onthe round-abouts of potency. That is, in hitting a tumor antigen whilesparing healthy cells bearing the same immunogenic target, itsacrifices its drug's efficacy.

Chari, who is director of chemistry at ImmunoGen Inc., inCambridge, Mass., said this meant "that significant responses wereachieved in animal studies only in high, toxic doses, and clinicalresults were disappointing."

This is the sticking point of most cancer chemotherapy today. Thedrugs can destroy tumors all right, but at doses that threaten todestroy the patient first. These therapeutic compounds act bytargeting rapidly dividing cells _ by definition, growing tumors _while bypassing quiescent healthy tissues.

Healthy High-Dividers Can Prove Deadly

But not all non-cancerous cells are non-dividers. Bone-marrow cells,hair follicles and the lining of the gut also renew themselves rapidly.Whence, patients on chemotherapy lose their hair, suffer unbearablenausea and incur serious anemia. These side-effects limit the amountand frequency of treatment, leaving the tumors to grow back,metastasize and kill.

Chari is senior author of a paper in today's Proceedings of theNational Academy of Sciences (PNAS) titled: "Eradication of largetumor xenografts by targeted delivery of maytansinoids."

He and his co-authors had reasoned, he told BioWorld Today, that ifthe high-potency maytansin drug could only be hooked up to a high-selectivity monoclonal antibody, it might avoid the severe side effectsof current chemotherapy.

Traditionally, most monoclonals in anti-cancer immunoconjugatestrategies have zeroed in on carcinoembryonic antigen (CEA),associated with a wide spectrum of tumors. The ImmunoGen teamsingled out a mouse monoclonal, C242 which is programmedpreferentially to attack glycoprotein antigens on the cells ofcolorectal cancers.

This year, 140,000 new cases of colorectal cancer will be diagnosedin the U.S. alone, and 55,000 Americans will die of the disease.Standard anti-tumor treatments consist of the cut-'em-out, burn-'em-out, poison-'em tactics of surgery, radiation and post-surgicalchemotherapy (mainly with 5-fluorouracil _ 5-FU). Drug treatmentshave a response rate below 25 percent, and unimproved long-termsurvival.

Immunoconjugates remain experimental.

For their preclinical trials described in PNAS today, Chari and hisco-authors took as their "patients" 60-day-old severe combinedimmune deficient mice under whose skin they implanted humancolorectal tumor cells. After letting these cancers grow for a week,they injected eight animals with their monoclonalantibody/maytansine conjugate daily for five days, in amountsequivalent to doses comparable to those of chemotherapy in humans.

This treatment, their paper reported, "completely eliminated anymeasurable tumors within two weeks of the initiation of therapy, andall eight animals were tumor-free for 200 days [duration of theexperiment]," with no serious side effects.

Tumor-bearing control mice, treated with the unconjugated drug andantibody, showed "very little anti-tumor activity," and were moribundat 40 days.

"It's well-known in the field," Chari observed, "that 200 days'remission in a mouse is considered to be a cure. Some people don'teven go beyond 100 days. Two hundred days is like a very oldmouse."

Then the team tested other groups of animals with lower doses oftheir conjugate. "Most cancer drugs today," Chari pointed out, "areused at full maximum tolerated dose [MTD]. The goal is to `push ithigh enough to kill all the cells you can.' We cured animals, injectingless than half of MTD."

Conjugate Spares Only One Cell In A Million

He added: "We can get circulating levels of maytansine in the serumof animals 380-fold above what you need to kill 99.999 percent ofthe tumor cells in humans. That means, we can kill very cell exceptone in a million."

Besides "finding a new treatment for colon cancer," Chari concluded,"we are trying to rejuvenate the immunoconjugate field, by findingsome new idea that the thing works."

ImmunoGen's chairman and CEO, biochemist Mitcell Sayare, pickedup on this theme. "Everyone has asked," he told BioWorld Today,"`Why can't we use monoclonal antibodies to deliver an effectormolecule to kill tumor cells?' With all the discussions of toxins anddrugs and other agents _ ImmunoGen's been working on it since1981 _ finally, we've developed what we think is a conjugate thatactually does what it's supposed to do. But we won't know that untilwe do human studies."

He anticipates such clinical trials "relatively rapidly, within 18months." Meanwhile, efforts center on humanizing the anti-colon-cancer murine monoclonal.

To accomplish this, Chari explained, "we are doing molecularmodeling to find out the structure, and will then do complementarity-determining region grafting _ replacing the mouse region withhuman." This work, he said "is just beginning, and should be finishedin a few months."

His company now is "doing the preclinical work necessary for anFDA submission, which is mainly the Phase I safety and toxicitystudies, and then go into the clinic. Our challenge," Sayare went on,"is that the incremental costs associated with that are beyond whatImmunoGen can afford at the present time. That's one of the reasonswe're looking for a partner."

Meanwhile, he concluded, "We're also looking for fundingelsewhere, even as we speak." n

-- David N. Leff Science Editor

(c) 1997 American Health Consultants. All rights reserved.