Surgical removal of the blackish-bluish skin lesion gives an early-onset melanoma patient an 85 percent chance of living for five years_ the usual survival time that predicates success of cancer therapy ingeneral.

But if the melanoma metastasizes to the brain, those five-year oddsgo down to somewhere between five percent and zero.

For melanoma tumors that have spread to other organs in the body,but not yet to the central nervous system (CNS), standard treatment issurgical ablation followed by radiation and/or chemotherapy (nowincluding interferon) to further reduce the malignancy.

Then, still-experimental immunotherapy seeks to unleash the body'sown immune defenses to wipe out what cancerous cells remain.

When it comes to the brain, even immunotherapy is still a forlornhope; the CNS doesn't seem to play by the same rules as the rest ofthe body's immune system.

That system relies mainly on three echelons of lymphocytes _ CD4+T cells, CD8+ T cells and natural killer cells _ to hunt down and killoff metastatic melanoma tumor cells.

"In a sense," observed pediatric oncologist David Ashley, of DukeUniversity, "the brain itself is working against immunotherapyapplied to brain tumors." Its first-line self-defense against tumor-attacking lymphocytes is the blood-brain barrier (BBB). This almostimpermeable membrane envelops the entire CNS. It admits small-molecule soluble nutrients and gases, but excludes most cells andlike-sized contaminants whose admittance might cause inflammationor worse.

That's one reason why recent attempts to challenge metastases ofmelanoma in the body with cells carrying genes that express T-cell-boosting cytokines have fallen short at the off-limits entrance to theCNS.

Now a task force of pathologists, oncologists and neurosurgeons atDuke University Medical Center in Durham, N.C., has constructedand tested a panel of vaccines based on a strategy that parachutesthose cytokine-stimulated lymphocytes behind the lines of the BBB.

Their report in the current Proceedings of the National Academy ofSciences (PNAS), dated Sept. 17, 1996, tells how: "Subcutaneousvaccination with irradiated, cytokine-producing tumor cellsstimulates CD8+ cell-mediated immunity against tumors located inthe `immunologically privileged' central nervous system."

Systematically, one after another, the team tried the effects of half adozen cytokines _ interleukins 2, 3 4 and 6, interferon-g andgranulocyte-macrophage colony-stimulating factor (GM-CSF) _ inpreventing or curing melanoma in mice.

Using a retroviral vector, they transfected six cultures of malignantmurine melanoma cells in vitro with the genes expressing eachcytokine. Then they injected the putative vaccines subcutaneouslyinto the bloodstreams of mice.

"The same technology," Ashley, a co-author of the PNAS paper, toldBioWorld Today, "has already been applied here at Duke in humanPhase I/II trials of renal cell carcinoma and breast cancer, andstimulated immune responses."

In their blood-borne delivery system, the transfected tumor cells dulydelivered their cytokine-expressed DNA payloads, which crossed theBBB. Of the six vaccines one, the GM-CSF, got high marks for bothprophylactic and therapeutic efficacy.

Thus, in a typical experiment, one cohort of mice received asubcutaneous shot of 100,000 transfected melanoma cells, while acontrol contingent got the same dose of untransfected cells. Oneweek later, both groups were challenged with 200 raw tumor cellsdelivered directly to the brain.

Within 25 days on average, all 33 control animals were dead of theircancers. In contrast, 49 active-ingredient mice protected with GM-CSF, IL-6, or IL-3 survived twice as long, 49 days. What's more,eight of the 23 animals vaccinated with GM-CSF lived more than 100days. At that point they were re-challenged with another 200 tumorcells to the brain; seven survived for more than 60 additional dayswith no evidence of cancer.

Results with the other cytokines were mixed. IL-4 and IFN-gperformed no better than controls. As for IL-2, it turned out to beapparently toxic, diminishing median survival times to 16.5 days.

So much for prevention. As for actual therapeutic treatment,unvaccinated mice received 200 melanoma cells to the brain, whichstarted an established tumor within three days. Then, 20 of them got asingle shot of one million GM-CSF-producing cells, and survived 30days, a week longer than controls. Three of the 20 lived around 100days, and then survived a second CNS challenge for at least 60 moredays.

Neurosurgeon John Sampson is the PNAS paper's first author. Hetold BioWorld Today: "We watched the animals for a total of 250days, at which time we killed them and looked at their brainshistologically. There was no tumor evident. So, as much as you canbe certain in animal experiments, these mice appeared to be cured ofthe tumor."

Besides demonstrating the prowess of GM-CSF, "The main findingof our paper," Ashley said, "is that this technique, which has beenused to treat peripheral tumors such as renal cell carcinoma, isapplicable to brain tumors."

The Duke group, he continued "is now working on a true primaryhuman brain tumor in animals. That is, our current work is extendingthese studies to looking at an anaplastic astrocytoma, which is aprimary brain tumor." "The same vaccines are effective in theprimary brain tumor model as in the metastatic ones," Sampson said.

Besides melanoma and astrocytoma, their "number one brain tumorcandidates for this immunotherapy strategy," he said, "would includemetastatic breast cancer and primary glioblastoma multiforme. But,"he added, "that's down the line."

About a year down the line. Within 12 months, the group hopes toenroll its first human patient.

"The scenario," Ashley suggested, "would be: Patient comes in with anew brain tumor. Has surgery; we get some of the tissue, and use ourretroviral vector to transfect the cells with whatever cytokine is themost effective in the human system. We then use those cells as asubcutaneous immunization, which is _ hopefully _ effective.

"Besides eradicating pre-existing malignancy," he continued, asecond scenario might be: "Patient comes in, has surgery, radiation,chemotherapy, and then a vaccine to prevent the tumor from comingback.

"Such prevention of recurrence," Ashley observed, "may beimportant."

He said, "The first human trial within a year is relatively realistic."

Between now and then, he concluded, "What we want to do over thenext 12 months is a series of toxicity studies in animals to make surethat no cross-reactive antigen in the vaccine cells could cause anautoimmune disease. This is sort of frontier stuff." n

-- David N. Leff Science Editor

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