National Cancer Institute scientists believe they have isolated a human leukemia virus.

The discovery, which they call "exciting" and other authorities call "highly significant," could trigger new interest in searching for viruses as causes of human cancer. Many scientists have turned away from the search for human cancer viruses, and NCI, part of the government's National Institutes of Health, has been criticized for spending too much money on virus research.

The find could also lead to new ways of treatment and even prevention, though it is far too early to say what may be possible.

Dr. Robert Gallo, chief of NCI's Laboratory of Tumor Cell Biology, and his coworkers have found the virus in four patients as well as antibodies to the virus--solid evidence of the body's effort to resist it--in 12 more.

Japanese doctors at the University of Kyoto, collaborating with Gallo, reported last week that they, too, had found the virus in some patients.

Viruses cause cancers without a doubt in several animal species. And they cause leukemia--blood cell cancer--in cows, cats, mice and chickens, as well as gibbon apes, close relatives of humans.

But a decades-long search for human cancer viruses has had thin results. Viruses have been at least "associated" with three human cancers, meaning they are somehow involved in the cause with other factors. The three are liver cancer, cancer of the cervix and one kind of lymphoma, found mainly in Africa.

Gallo's discovery "gives a strong stimulus to reopen investigations into viral involvement" in human leukemias, Dr. William Jarrett of the University of Glasgow has told Oncology Times, a publication for cancer researchers.

Gallo's main colleagues in this work have been Bernard Poiesz, Marvin Reitz Jr., Francis Ruscetti, V.S. Kalyanaraman, Marjorie Guroff and M. Sarangadharan.

"What we have found," Gallow said in an interview, "is a new retrovirus or RNA virus." RNA, or ribonucleic acid, and DNA, deoxyribonucleic acid, are nature's two main kinds of genetic material.

Retroviruses cause the known animal leukemias. The one that causes cattle leukemia, a disease highly destructive to livestock, is at least distantly related to the virus NCI has just found.

Retroviruses are also of high interest today.

It is known that they sometimes carry so-called transforming genes into cells. These may promote normal growth or, sometimes, lead to cancer--either by inserting their new information wrongly or by turning on another gene that makes cells grow inappropriately.

Until now, said Dr. David Baltimore, a Massachusetts Institute of Technology Nobel Prize-winner, the lack of any known human retrovirus has been a major gap, "so even if this virus is just a human virus and causes no known disease, it is important."

Gallo said, "We can't yet say this virus causes a leukemia." But he also said, "We clearly are getting an association of this virus with a form of leukemia," and though "we can't say yet that it is an etiologic originating agent, we suspect it."

He has at least provisionally named the virus "HTLV" for "human T-cell leukemia-lymphoma virus"--"T-cell" because this is the kind of white blood cell that proliferates in this kind of leukemia.

One of the main problems, always, in studying leukemias and lymphomas--lymph gland cancers, closely related to the leukemias--has been getting T-cells to grow in the laboratory.

Six years ago, Gallo, Doris Morgan and Ruscetti found a protein they call TCGF for "T-cell growth factor." This has proved to be the magic key to growing T-cells and leukemia and lymphoma cells well and reliably.

It is this technique that has enabled the NCI group to isolate and grow HTLV virus from four patients with an unusual "and very aggressive" leukemia or lymphoma that often closely resembles a disease called "Sezary leukemia."

Because isolating and growing a virus from any one patient is still long and laborious work, the scientists have screened blood samples from many other patients and found 12 with HTLV antibodies. They have shown that such antibodies are not in the blood of 200 healthy persons.

Antibodies, proteins the body makes to fight aggressors, "are flags that tell us where the virus has been," Gallo explained.

Now, he said, there must be much more study, including epidemiological surveys to see how widespread the virus may be. This will include collaboration with Japanese doctors, since a closely related or perhaps the same form of leukemia is found in a kind of cluster in southwest Japan, and blood samples from Japanese patients have revealed telltale antibodies to the NCI virus.

The Gallo evidence, said Dr. David Katz of the Scripps Clinic and Research Foundation is "quite impressive and convincing." But, said Dr. Max Essex of Harvard, proving that this or almost any new-found virus causes human cancer will be a slow process at best.

Gallo in 1975 felt sure he had isolated a leukemia virus, but was unable to prove it. What he feels "99 percent sure of today," he said, is that he just "couldn't keep the virus growing then" and it was replaced in his lab cultures by a contaminant, a primate leukemia virus.

How do people get the new-found virus, if it is indeed a leukemia virus?

"We don't know," he said. "We don't think it, or any cancer, is infectious in the common way, like a cold. There may be infection at a very low level, probably only in genetically susceptible individuals.

"We can speculate, however, that this virus sometimes reaches humans from someplace outside them. You're not born with it in your genes. Then, in some susceptible people, it replicates." That is, it grows.

And in persons whose cells permit such growth it may unlock a gene that starts another step, or steps, that may lead to a cancer in a process still only dimly understood.