"Costimulation" is the name of the newest game in cancerimmunotherapy. It already works in vivo -- at the mouse level-- as three widely separated research teams are now reporting.

Their strategy was to saddle the surface of each malignant cellwith a genetically engineered protein, B7. This costimulator,they hoped, would turn up the volume of the killer T cellresponse to the cancer's own immunogenically weak, tumor-specific antigen.

In today's issue of Science, tumor immunologist Sarah E.Townsend reports on her team's in vivo experiment in "TumorRejection After Direct Costimulation of CD8+ T Cells by B7-Transfected Melanoma Cells." Townsend is a graduate studentat the University of California's Cancer Research Laboratory inBerkeley. Its director, James P. Allison, co-authored the paper.

They inoculated mice with malignant melanoma cellstransfected with an expression vector containing the cDNAsequence encoding B7. As controls, a second group of mice gotthe vector alone, minus the B7 gene.

By the 38th day, 50 percent of the mice injected withmelanoma cells harboring sham-vectors had "large orulcerating tumors"; by day 53, so did the rest of the placebocohort. But the animals implanted with B7-encoding DNA in thecancer cells were still healthy at day 120, and any tumors theyincurred "had either disappeared or become static."

So far, so good. But would the protection extend as well tosubsequent incursions by melanoma cells without the B7costimulant? It would, they found. Prior exposure to B7-gene-seeded melanoma cell fragments shielded 89 percent of themice from tumor growth for 90 days at least, "whereas tumorsappeared in 80 percent of naive mice by day 18."

"Such a vaccine," wrote Science's John Travis in an editorial,"might prevent metastases in melanoma patients who have hadtheir primary tumors surgically removed."

Peter S. Linsley of the Bristol-Myers Squibb PharmaceuticalResearch Institute in Seattle, described a similar trial of B7costimulation last month in Cell for Dec. 24/31. The Seattleteam's melanoma-laced mice got -- along with their B7 gene --a second DNA sequence. This one encoded a stronglyimmunogenic viral antigen, rather than rely on themalignancy's own wimpy tumor-specific antigen.

Tumors grew in their mice, but disappeared within two orthree weeks. The team then went back and cured establishedmicrometastases.

Molecular immunologist Laurie Glimcher of the Harvard Schoolof Public Health is the third B7 player. Her experiment will bereported in a paper currently under review by a scientificjournal. Glimcher's mice got their B7 gene in a highly malignantmurine sarcoma, which grows in the abdominal cavity's ascitesfluid, like a human ovarian carcinoma.

"I think it's possible," Glimcher told BioWorld, "that patients'tumors can be genetically engineered to express thesecostimulatory molecules, and that such immunization will causedistant micrometastases to regress."

Her co-author, Lee M. Nadler of Boston's Dana-Farber CancerInstitute, is gearing up for just such clinical trials -- "within thefirst half of this year," he told BioWorld. Nadler works inassociation with Repligen Corp. of Cambridge, Mass.

He finds the three teams' B7 mouse experiments "really veryexciting," because "now for the first time, T cell-specificimmunity can be addressed in cancer. Their basic scientificwork gives us a running start from the bench into man. Butwith the caveat that we know before we start that there areinfinite ways this could go wrong."

"There's a giant leap of faith from a mouse into a human, whomay have a big, gigantic, long-standing tumor -- plenty of timefor that person's T cells to become tolerant of the malignancy,"Nadler added. "So I don't know that those T cells in the humancan respond to the B7 transfection."

Nadler is eager to start human B7 trials, but equally skepticalas to their outcome. "We are going into man now aggressively,to do these experiments, but I don't think it makes a lot ofsense that a B7 gene gets popped into every tumor around, andthey vaccinate people. The likelihood of that working just byitself is not very high."

He explains: "One gene, without an understanding of themechanism, does not a therapy make. This requires theantigen, the antigen presentation, the costimulation, and acompetent T cell. All four pieces have to be there, and if they'renot, it ain't gonna work."

Asked if she and Allison were also planning human trials,Berkeley's Townsend told BioWorld, "We're a mouse lab." Shepointed out that B7 costimulation is effective only in tumorsthat express MHC (major histocompatibility complex) antigens."But many MHC tumors can be induced, using gamma-interferon," Townsend said. "So we're trying to establish acollaboration to do co-therapy with gamma-IFN."011593B7

-- David N. Leff Science Editor

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