Antibiotics may soon replace antacids as the best treatment for healing active ulcers and preventing recurrence, but a controversy has arisen over whether this approach is ready for routine use.

Ever since a bacterium found in the stomach and intestine was linked to ulcers, a half-dozen scientific studies have demonstrated the effectiveness of antibiotics in curing ulcer patients. The problem, however, is that the current antibiotic treatment doesn't always work.

For most people, active ulcers can be treated with drugs such as Tagamet that suppress stomach acid production. But once the drug is stopped, the ulcer often comes back.

Antibiotics show promise in permanently healing ulcers. At the same time, some physicians fear that if the antibiotics do not kill all of the bacteria, the germs that survive will grow back and become resistant to subsequent treatments.

"We really haven't worked out a terribly effective and safe way to eradicate the infection," said David Y. Graham, head of gastroenterology at the Baylor College of Medicine in Houston.

"If treatment fails, then the germs often become resistant to antibiotics," said Graham. "We are worried that in one to three years we will have made a whole group of patients who are not treatable."

Barry J. Marshall, a gastroenterologist at the University of Virginia Health Sciences Center in Charlottesville, is more optimistic. In 1983, Marshall and J. Robin Warren, both Australians, discovered the S-shaped bacterium Helicobacter pylori (previously called Campylobacter pylori) in ulcer patients. While some scientists doubt H. pylori is the main cause of ulcers, there is a general consensus that it is associated with the disease. Marshall said that a standardized antibiotic program worked out by his group now cures 80 to 90 percent of his patients.

But he declined to name which antibiotics he uses to kill the bacteria, citing the fact that doctors might try it on ulcer patients without knowing all the details needed for successful treatment.

"Therapy is complicated. It is not ready for everybody's general practitioner yet," Marshall said, "but it could be in another two years."

A study published in the May 26 issue of The Lancet, a British medical journal, shows the problems and the promise of treating ulcers with antibiotics.

Two ulcer experts, E.A.J. Rauws and G.N.H Tytgat at the Academic Medical Centre in Amsterdam, treated 50 patients suffering recurring disease with either colloidal bismuth subcitrate (CBS), a compound similar to Pepto-Bismol, alone or CBS with two antibiotics -- amoxicillin and metronidazole -- what they called the triple treatment. Five patients dropped out because the triple treatment caused too many side effects, including nausea, diarrhea and rash. The treatment, CBS four times a day and antibiotics three times a day, lasted four weeks.

Of the 24 patients who received the triple treatment, the H. pylori was killed in 15, but not in nine -- a 36 percent failure rate. Of the 25 who received CBS alone, bacteria disappeared in only two people. Ulcers did not recur in any of the 17 patients who were freed of H. pylori during the 12 months of follow-up.

Among the 21 Dutch subjects who continued to be infected with H. pylori but were cured of the ulcer by CBS alone, 17 -- or 89 percent -- had a relapse in the first year.

"If you could prevent recurrence, you could help a lot of people," said Jerry Gardner, a gastroenterologist and chief of the digestive diseases branch at the National Institute of Diabetes, Digestive and Kidney Diseases.

Ulcers arise when craters form in the walls of the stomach or the first part of the small intestine called the duodenum. Normally, a layer of mucus protects the linings of the stomach and intestine from the stomach's acid, but once the mucus is lost and the craters form, the acid can cause bleeding and pain.

About 3 million Americans suffered ulcers in 1985, the last year for which figures are available from the National Center for Health Statistics. About half of them are infected with H. pylori, said Baylor's Graham. In addition to the infection, an individual appears to need some genetic predisposition to form ulcers, since many people are infected but never develop ulcers.

Experts believe that about 10 percent of the population will develop an ulcer some time in their lives. Ulcers were responsible for about 46,000 operations and about 6,000 deaths in 1986.

Current treatments, which can cure nearly all ulcers, rely on drugs that either turn off acid production directly, block stimulation of the histamine II receptor in the stomach that turns on acid production, or directly protects the crater from further acid damage until it heals.

"Until 1975, we really didn't have any good therapy for duodenal ulcer disease," Gardner said. "Cimetidine {known by the brand name Tagamet} became available in 1975. It has been given to more than 10 million people. It is a pretty good treatment for duodenal ulcers, but when you stop the therapy, the ulcer comes back."

The Dutch study reported that recurrences happen in 70 to 80 percent of patients after they stop using conventional drugs.

There are a number of difficulties in trying to figure out the best antibiotic treatment. H. pylori is hard to kill because it lives inside the layer of stomach-produced mucus, which keeps away the antibiotic as well as the stomach acid. Some researchers believe successful treatment may require the development of new antibiotics.

Another approach is described in the same issue of Lancet by researchers at the National Medical Laser Center at Middlesex Hospital in London. The treatment, called photodynamic therapy, works by using a chemical to sensitize the bacteria to pulses from a low-power laser. The laser is fired down a fiber optic tube that has been threaded down the throat and into the stomach. When the light hits the chemical that has been absorbed by the bacteria, the chemical becomes toxic and kills the bacteria. The work is very preliminary.

Another problem with finding the best drug treatment, Gardner said, is that half of the ulcer patients in the placebo groups get better all by themselves, making it more difficult to sort out which drug is actually working.

In addition, the Food and Drug Administration has yet to approve CBS, the form of bismuth used in the Dutch study, for use in the United States. Marshall said CBS is a more soluble than the forms of bismuth used in the U.S., such as Pepto-Bismol, and will clear the infection in some patients all by itself. A multicenter study on CBS is just getting under way to gather the data needed for FDA approval.

Until the questions are cleared up, some experts are not recommending a rush to antibiotic treatment for ulcers.

"We are not treating average patients" with antibiotics, Graham said. "We won't just treat anybody with an ulcer." The Baylor doctors only give antibiotic treatment to ulcer patients with a recurring problem severe enough that surgery might be used to cut certain nerves and permanently reduce stomach acid production.

Marshall takes a similar approach. He does not yet recommend that everybody with an ulcer be treated with antibiotics. Instead, he suggests that people with recurrent ulcer problems be tested for H. pylori. If the bacterium is present and current therapy is not working, then patients should consider finding an expert who understands how to use antibiotics properly. If possible, he advocates waiting until the details of the antibiotic treatment can be worked out in the next few years.

Eventually, Marshall said, "ulcers as we know them will be history."