The approach is still experimental, but holds promise for millions of women left with cratered areas and breasts that look very different from each other after cancer surgery. It also might be a way to augment healthy breasts without using artificial implants.
So far, it has only been tested on about two dozen women in a study in Japan. But doctors in the United States say it has great potential.
"This is a pretty exciting topic right now in plastic surgery," said Dr. Karol Gutowski of the University of Wisconsin-Madison. "There are people all over the country working on this."
The Japanese study was reported Saturday at the San Antonio Breast Cancer Symposium. The company that developed the treatment, Cytori Therapeutics, plans larger studies in Europe and Japan next year.
More than 100,000 women have lumps removed each year in the United States. These operations, lumpectomies, often are done instead of mastectomies, which take the whole breast. But they often leave deformities because as much as a third of a woman's breast may be removed.
"It's almost a euphemism" to call it a lumpectomy, said Dr. Sydney Coleman, a plastic surgeon at New York University who has consulted for Cytori and is interested in the stem cell approach.
The defect "initially may not be as noticeable" but it often gets worse, especially if the woman also has radiation treatment, said Dr. Sameer Patel, a reconstructive surgeon at Fox Chase Cancer Center in Philadelphia.
"There's a growing push to try to involve the plastic surgeon particularly for this reason - to try to avoid a defect," but once one develops, options to repair it are limited, Patel said.
The implants sold today are for reconstructing breasts after mastectomies. They aren't designed to fix odd-shaped deformities from lumpectomies or radiation.
"Each one is so different, there's no little thing you can just pop in there," Gutowski explained.
Doctors can try making the other breast smaller so they match, transplanting a back muscle to boost the flawed breast, or rearranging tissue to more evenly distribute what's left. But these involve surgery and leave scars.
Mini implants of fat tissue have been tried, but they often get resorbed by the body or die and turn hard and lumpy. The recent discovery that fat cells are rich in stem cells - master cells that can replenish themselves and form other tissues in the body - renewed interest in their use.
In the Japanese study, doctors liposuctioned fat from 21 breast cancer patients' bellies, hips or thighs. Half was reserved as the main implant material; the rest was processed to extract stem cells and combined with the reserved fat. This was injected in three places around a breast defect.
Doctors think the stem cells will keep the tissue from dying and form lasting mini implants.
Eight months after treatment, "about 80 percent of the patients are satisfied" with the results, said the lead researcher, Dr. Keizo Sugimachi of Kyushu University in Fukuoka, Japan.
There was a statistically significant improvement in breast tissue thickness at one and six months after treatment.
Doctors with no role in the research say longer study is needed to see if these results last.
The treatment is expected to cost $3,000 (euro2,067) to $5,000 (euro3,446), said Cytori's president, Dr. Mark Hedrick. The company sees potential for cosmetic breast augmentation of healthy breasts, but for now "our plan is to focus on an unmet medical need" in cancer patients, he said.
The American Society of Plastic Surgeons says doctors must be cautious about using fat cells for cosmetic purposes until more is known. Gutowski heads a task force the society formed to study the science. Coleman is a member.
"It's got great potential not only for breast but other cosmetic and reconstructive purposes," like filling in facial defects from cancer or trauma, Gutowski said. "Imagine the aging face."
Better cosmetic treatments may encourage more women to choose lumpectomies. Some have chosen mastectomies because they are concerned about being left with a defect, especially younger women.