Two new studies that suggest mammograms won't save many lives of women under 50 haven't changed the positions of the American Cancer Society and American Medical Association.

The researchers said regular mammograms significantly reduce the risk of dying of breast cancer for women over 50, but offer little lifesaving benefit for women in their 40s. They arrived at their conclusion after reviewing 13 studies.

The cancer society and the AMA, however, continue to recommend that women get a mammogram every year or two starting at age 40.

An AMA spokesman said the 13 studies varied in design and conduct, and that most were done outside the United States and failed to consider any factors other than death rates.

The American Cancer Society said it reviewed the same 13 studies and found no reason to change its belief that women ages 40 to 49 should be screened every year or two.

In 1993, the National Cancer Institute stopped recommending that women in their 40s get regular mammograms, saying there's no evidence the examinations significantly reduce breast cancer deaths in that age group.

Authors of the review of 13 studies couldn't find any significant evidence, either.

"This study's important message for women is that those 50 to 74 who underwent screening mammography had a 26 percent reduction in breast cancer deaths compared with women who did not," said the lead author, Dr. Karla Kerlikowske, an assistant clinical professor of medicine and epidemiology and biostatistics at the University of California at San Francisco.

Routine mammograms yielded an insignificant reduction in breast cancer deaths in women ages 40 to 49 after the same length of follow-up -- seven to nine years, her team found. The reduction was 7 percent, but researchers couldn't be sure the number wasn't due to chance.

The findings were published in Wednesday's issue of The Journal of the American Medical Association. The journal includes a separate cost-benefit analysis of mammography screening, which concludes that standard health insurance should not pay for the procedure for women younger than 50 or older than 69.

The Rand Corp. estimated that a health plan covering 500,000 people would spend $11.9 million over six years to give biennial mammographies to all 50- to 69-year-old women -- and would save 11.9 lives per year.

Screening women ages 40 to 49 every two years would cost an additional $11.7 million over seven years and would save zero to 1.1 lives per year, said researchers led by Dr. Herman Kattlove.

Screening women ages 70 to 74 would cost $1.6 million over three years and would not save a significant number of lives, the researchers said.

Rand, based in Santa Monica, Calif., is an independent think tank.

Dr. Melody Cobleigh, director of the Comprehensive Breast Center at Rush-Presbyterian-St. Luke's Medical Center in Chicago, said she will continue to urge women under 50 to get mammograms because they save lives.

"Over the weekend, I went through the records of patients I'd seen newly diagnosed with breast cancer in 1994," she said Tuesday.

"In our 120 cases, there were 50 women between ages 40 and 50. Thirty-five percent were diagnosed on the basis of screening mammography."

11/94 Mammograms-How often?

Women in their 40s should have regular mammograms, and annual ones offer a better chance of early detection of breast cancer, researchers reported at a radiological meeting Monday.

The two studies, both supporting yearly tests for women in their 40s, follow a decision last year by the National Cancer Institute to withdraw its recommendation that women begin routine mammography in their 40s. The NCI said women should decide for themselves whether to take the exam after consulting with their doctors. The American Medical Association still recommends that women in their 40s have routine mammograms. The AMA also has recommended women start with a baseline mammogram at 35.

The studies were presented at the annual conference of the Radiological Society of North America by Dr. Linda Warren, executive director of the Screening Mammography Program of British Columbia, and Dr. Stephen A. Feig and Dr. Emily F. Conant of the Thomas Jefferson University Hospital in Philadelphia.

Warren, a radiology professor at the University of British Columbia in Vancouver, Canada, studied 150,147 women between July 1988 and September 1993; about 34 percent were under age 50. Overall, she said, 15 percent of the detected cancers were in that age group. And almost 87 percent of the 150 cancers found in women in their 40s were in an early, curable stage -- before the cancer had spread to the lymph nodes, the study found

. Feig, a radiology professor, said recent research indicates tumors in younger, premenopausal women who are still hormonally active tend to grow faster. That, he said, makes early detection among women in their 40s even more critical.

Feig assisted Dr. Emily F. Conant, an assistant radiology professor, on a three-year study at Thomas Jefferson Hospital. Their retrospective analysis of 615 patients 40 and up diagnosed with breast cancer indicated women who had annual mammograms had the best chance of recovery and the most treatment options, said Conant.

Of the 615, 88 women in their 40s had cancers detected solely by mammography; self-examination had not sufficed, Conant said.

12/94 Radiologists ability to read mammograms accurately varies widely.

- Doctors' ability to read mammograms for breast cancer varies dramatically, and some may miss a disturbingly high number of tumors, a study found.

The research suggests that despite all the high-tech diagnostic tools in doctors' hands, using them accurately can be as much an art as it is a science.

Mammograms are one of the country's most common screening tests, recommended annually for all women over 50 to spot cancer before it spreads. The quality of the X-ray machines and film has improved over the past decade, but doctors still must examine the images for signs of cancer.

The latest study, conducted at Yale Medical School, asked 10 community radiologists to look at 150 mammograms, describe what they saw and give their recommendations. All of the doctors came to the same conclusion in just 10 cases.

When radiologists suspect cancer, they suggest that patients receive immediate follow-up, typically a biopsy, an ultrasound scan or more mammograms.

The study found surprising differences in the doctors' ability to see worrisome signs. Overall, they urged further tests to get a definite answer in 87 percent of the women who actually had cancer.

Two of the doctors recommended additional tests in all but 4 percent of the women who had breast cancer. Three others missed 19 percent or more.

"There is a wide range in the accuracy of radiologists," said Dr. Joann G. Elmore, an internist who directed the study. "This may be due to different thresholds of concern" before ordering additional tests of abnormalities.

However, guarding against missing cancer often seemed to come at a high price.

Those who were highly accurate in picking up cancer patients recommended a lot of additional tests in noncancer patients, Elmore noted. "As a woman and physician, I know this can be very stressful," she said.

For instance, one of the doctors who recommended follow-up tests for 96 percent of those with cancer also urged them for 64 percent who were normal. The doctor who did the worst job, missing 26 percent of cancer cases, also suggested follow-up in just 11 percent of those who were free of the disease.

The study was published in Thursday's issue of the New England Journal of Medicine.

In an accompanying editorial, Dr. Daniel B. Kopans, a radiologist at Harvard Medical School, said, "It is disturbing that the interpreters failed to identify some patients with cancer."

Dr. Stephen Feig, a radiologist at Thomas Jefferson University in Philadelphia, said the findings "may not necessarily be a concern" because of shortcomings in the way the study was done.

He said the doctors may have missed tumors because they were denied routinely available data, such as women's previous mammograms, which can be used for comparison, and information on whether they had lumps in their breasts.

Among other findings of the study:

--There were major disagreements among doctors in how to handle 25 percent of the patients. These including such disparities as one doctor recommending a biopsy while another suggests routine follow-up.

--Five months later, the researchers shuffled the mammogram films and asked the radiologists to check them again. The doctors disagreed with their earlier assessments 16 percent of the time.

--Even when doctors concluded that a patient should have a biopsy, they disagreed on whether it should be done on the left breast or the right in one-third of cases.

Elmore said the study should not discourage women from getting mammograms.

"The mammogram is very effective," she said. "It is one of the best screening tools for cancer. It identifies cancer at an earlier stage and helps to reduce mortality."