Report from the conference on ultrasound-assisted liposuction safety and effects.

The authors report on the Ultrasound-Assisted Liposuction (UAL) Safety and Effects conference held in St. Louis, Missouri, in November 1998. The meeting was convened to discuss how internal UAL works (its physics and mechanisms of action), gain a basic knowledge of the interaction between acoustic energy and tissue, identify safety concerns potentially related to thermal effects and free radical production, and define future research questions. Those attending represented multiple scientific disciplines: plastic surgery, physics, lipid chemistry, cancer biology, and medical biophysics. Participants agreed that scientists do not yet understand the mechanisms of UAL action, although multiple mechanisms are probably involved, such as mechanical forces, cavitation, and thermal effects. Additional research has revealed that long-term complications or negative bioeffects--including DNA damage and oxidation-free radical attack--are probably not serious safety concerns with UAL. Several areas deserving future investigation were proposed.

Microbial growth inside saline-filled breast implants.

In vitro and in vivo experiments were conducted to determine whether intraluminal saline in breast implants can support the growth of common wound-infecting microorganisms over a prolonged period of time. The bacteria tested were Staphylococcus aureus, Staphylococcus epidermidis, Escherichia coli, Corynebacterium jeikeium, Enterobacter cloacae, Klebsiella pneumoniae, and Pseudomonas aeruginosa. Three fungal species also were tested: Aspergillus fumigatus, Paecilomyces variotii, and Candida albicans. In the in vitro study, four organisms survived in flasks of sterile saline for the 2 weeks in which serial cultures were performed: K. pneumoniae, C. albicans, A. fumigatus, and P. variotii. In the in vivo study, 61 white rabbits (122 implants) received both an experimental implant inoculated with one of the test organisms and a control implant containing only sterile saline. They were sacrificed at 1-, 3-, or 6-month scheduled endpoints. None of the control implants containing sterile saline had positive cultures. In contrast, the intraluminal saline was culture positive for 7 of the 10 inoculated organisms after varying lengths of time: S. epidermidis, E. coli, E. cloacae, K. pneumoniae, P. aeruginosa, A. fumigatus, and P. variotii. Samples of capsular tissue also were cultured. Of the 122 capsular tissue specimens, 21 (17 percent) had positive cultures and surrounded both inoculated and sterile implants. In most instances, capsules that were culture positive contained an organism different from the one that had been inoculated in the group. In only 3 cases was the same organism cultured from both the periprosthetic tissue and the intraluminal saline, and these may represent instances of the inoculated organism migrating through the implants filler valves. The data show that several types of bacteria (particularly gram-negative species) and fungi can grow and reproduce in a restricted saline environment for extended periods of time.

Bleed of and biologic response to triglyceride filler used in radiolucent breast implants.

Radiolucent breast implants filled with triglyceride oil have recently entered limited clinical trials. To investigate the questions of oil bleed and the fate of triglycerides that might escape from ruptured breast implants, experiments reported here used peanut oil labeled with radioisotopes so that it could be traced in the urine, feces, and organs of two groups of rabbits. In one experiment, 18 rabbits were implanted with peanut oil-filled implants labeled with tritium to determine whether triglycerides diffuse across silicone elastomer shells. In another experiment, 19 rabbits were injected with 14C-labeled peanut oil to study what might happen to the oil if an implant ruptures. At the end of the follow-up period, we measured radioisotope levels in tissue samples taken from the periprosthetic capsule or injection site of each rabbit, as well as from major organs and the subcutaneous fat on the dorsum opposite the experimental site. One experiment revealed that triglycerides do bleed across the implant shells. Tritium levels were highest in the implant capsule, the omentum, the aorta, and the subcutaneous fat on the nonexperimental side. In the experiment simulating implant rupture, 14C levels were above the background radiation count at the injection site and in the same tissue sites as in the bleed experiment. Both in vivo radiolabeling studies indicate that triglycerides freed from implants by means of bleed or rupture would be absorbed, metabolized, and either excreted or redistributed to the body's normal fat storage sites if they are not needed for energy. In a third in vitro experiment, triglyceride oil specimens were inoculated with various microorganisms associated with wound infections: Staphylococcus aureus, Pseudomonas aeruginosa, Escherichia coli, Staphylococcus epidermidis, and diphtheroids. The data demonstrate that neutral triglycerides used as a breast implant filler do not support growth of common infection-producing bacteria and suggest that triglycerides may have bactericidal properties.

HLA typing in women with breast implants.

Since the 1970s, anecdotal reports have described a relatively small number of women who received silicone gel breast implants and later developed either a recognized rheumatologic disease or unexplained symptoms suggestive of an autoimmune disorder. The study reported here examined whether there is any association between the symptoms seen in implant patients and HLA molecules. One-hundred and ninety-nine subjects were evaluated by HLA typing: symptomatic patients with implants (group I, n = 77), asymptomatic women with implants (group II, n = 37), healthy female volunteers without implants (group III, n = 54), and fibromyalgia patients without implants (group IV, n = 31). A statistically significant 68 percent of group I were positive for HLA-DR53, compared with 35 percent of group II and 52 percent of group III. The fibromyalgia patients were strikingly similar to group I women in terms of HLA-DR molecules, with 65 percent of group IV being positive for DR53. Group I also had a statistically significant increased frequency of HLA-DQ2. Asymptomatic women with implants (group II) had an increased frequency of DR1 and DQ1. In addition, 42 percent of symptomatic patients with implants formed autoantibodies to their own B cells; of these, 81 percent were DR53-positive. Although frequencies of capsular contracture and implant rupture were not significantly different in the two groups with implants, there were statistically significant associations in group I between contractures and ruptures and the presence of DR53 and B-cell autoantibodies. These data suggest that symptomatic patients with implants share important genetic characteristics (primarily HLA-DR53 positivity) that differentiate them from their asymptomatic counterparts. DR53 may be a marker of women who are predisposed by their HLA genotype to develop symptoms following exposure to silicone gel breast implants.