SCIENTISTS have long been seeking to design a cancer vaccine, a magic inoculation that would galvanize the immune system either to destroy a tumor or to prevent the recurrence of cancer once the initial malignancy had been removed.
But they have been hampered by their ignorance about which proteins of a tumor cell, the so-called antigens, would be most likely to provoke a potent immune response and would therefore be useful if inserted into a cancer vaccine.
Scientists have also had great trouble identifying antigens on tumor cells that are not also found on healthy cells, a necessary distinction if they are to design a vaccine directed against malignant tissue alone.
Now immunologists from Belgium report that they have discovered the first set of human tumor antigens meeting the twin criteria.
In laboratory tests, the antigens arouse the activity of white blood cells called killer T-cells, believed to be the arm of the immune system responsible for destroying tumors.
In addition, the new antigen proteins are found exclusively on cancer cells, rather than on normal cells and tumor cells alike, suggesting the antigens are the mutant byproducts of cancerous transformation and are therefore excellent targets for a vaccine.
Dr. Thierry Boon and his colleagues at the Ludwig Institute for Cancer Research in Brussels reported last week that they had found three related antigen proteins on human melanoma cells that provoked strong killer T-cell activity in test-tube experiments.
But when the immunologists examined normal tissue of the skin, lung, liver and elsewhere in the body, the antigens were nowhere to be found.
"I think this is very exciting," said Dr. Alan N. Houghton, chief of clinical immunology at Memorial Sloan-Kettering Cancer Center in Manhattan. "T-cells are probably the most important part of tumor rejection. This provides the basis for the development of a cancer vaccine."
Significantly, one or more of the three tumor antigens were identified in the melanoma tissue of about half the patients they examined, indicating that the antigens might be commonplace mutational events in the evolution of the deadly skin cancer. The antigens also crop up on other types of tumor tissue, although with less regularity.
Other cancer antigens that researchers have identified in the past seemed good only for inciting the body's production of antibodies, which are better at neutralizing bacteria than at destroying tumors.
Boon's results are extremely preliminary and are not likely to be ready for clinical trials within the next several years. But in theory key regions of the tumor antigens may prove useful in a shot that would be given to a patient with newly diagnosed melanoma.
Recognizing the injected antigen slivers as abnormal, killer T-cells would rush in for the attack, dividing into battle-ready clones of cells.
Those amplified legions of killer cells would then assault anything else in the body that happened to bear the same antigens, wiping out pockets of melanoma tissue hidden from the surgeon's scalpel.
Presumably, memory clones of those T-cells would linger in the body for years, helping to prevent a relapse of the cancer.
Boon suggests that more than 25 percent of melanoma patients might be suitable candidates for such a vaccine, but he emphasizes that much work remains to be done before this hope becomes reality.