Am J Public Health • Volume 89 • Issue 4

Skeptics may certainly find fault with the third study (the only one to report a significant finding) or with all or any of the statistics described. But few could argue, after examining these studies, that the relative risk for a known and well-defined connective tissue disease is likely greater than 2. Another possibility has been raised, one that calls to mind other difficult-to-study syndromes linked to exposures. In a study addressing Gulf War syndrome, signs and symptoms were often mentioned that proved difficult to describe systematically and therefore difficult to study. What if a suspected silicone exposure syndrome were so mild and transient that it did not warrant a physician visit (as in the Mayo Clinic study). receive a diagnosis (as in the Nurses' Health Study), or require admission to a hospital (as in the Danish and Swedish studies)?

And if such a disorder were, in fact, mild and transient, should it merit the concern that has been shown and the compensation that has been awarded in the silicone implant litigation cases to date? We suggest that neither a well-described disorder with a relative risk of less than 2 nor a transient and mild disorder seems compatible with the number of litigants over silicone implants and the apparent seriousness of their complaints. Some 400,000 women joined in one class action suit for damages, and 170,000 joined in another. Even if there had been 2 million implants undertaken in the United States over the 3 decades in which implant surgery has been practiced (and some estimates put the number closer to 1 million), there is no conceivable way in which a relative risk of 2 or 3 (or even 4) for each of the rare syndromes reported could explain so many exposed women being affected. At most, 2200 out of 2 million unexposed women would be expected to have had any one of the listed forms of connective tissue disorders, most of which are very rare. Doubling the risk among the exposed population yields 4400, and increasing the risk 20-fold produces 44,000. At this rate, there is no way in which 400,000 litigants could all be affected. Extensions of the already-completed studies are ongoing, at least 1 of which is government funded; apparently it is thought in the United States (though not in the United Kingdom or elsewhere) that there is still room for reasonable doubt as to the supposed causal relationships. But if epidemiology is invoked in the interest of public health to prevent the many uses of silicone, the weight of the evidence abstracted here supports the inference that silicone breast implants have not been proved guilty of causing connective tissue disorders.

Hage JJ, Yoe EP, Zevering JP, de Groot PJ

J Hand Surg [Am] 1999 Jan 24:1 73-7

The long-term results of 16 silicone implant arthroplasties following severe destruction of the proximal interphalangeal joint and surrounding soft tissue in 14 patients were evaluated. Two implants were revised (1 arthrodesis and 1 amputation) and are not included in the results. On a visual analog scale of 0 to 100, the subjective evaluation relating to the quality of function and the ability to perform activities of daily living scored 73 on average (range, 0-100; median, 75). The range of active motion compared with the opposite side ranged from 0% to 100% (average, 47%; median, 56%). The lateral deviation, either radial or ulnar, averaged 4 degrees (range, 0 degrees to 9 degrees; median, 4 degrees). Rotational deformity, either radial or ulnar, averaged 8 (range, 0 degrees to 32 degrees; median, 4 degrees). Pinch strength of the operated finger on average was 75% of the opposite nonoperated finger (range, 14% to 164%; median, 70%). Possible statistical correlation between objective and subjective results was evaluated. Subjectively, most patients were satisfied because they were pain free. Objectively, notable rotational deformity, alignment deviation, and loss of pinch strength and range of motion were observed, but these did not correlate well with the subjective assessment. Silicone implant arthroplasty for posttraumatic arthrosis should be used for those cases in which associated adhesions may be corrected by simple tenolysis, allowing for active range of motion shortly following arthroplasty. Even in these cases, the objective results of silicone arthroplasty may not be as good as its subjective outcome.

Author Address

Department of Plastic and Reconstructive Surgery, Academisch Ziekenhuis

Vrije Universiteit, Amsterdam,

The Netherlands.

OBJECTIVE: We previously reported an increased preponderance of a broad range of autoantibodies in symptomatic women with silicone breast implants. The objective of this study was to investigate the frequency of autoantibody production in asymptomatic compared to symptomatic women with silicone implants. One hundred twenty-two asymptomatic women were recruited to our center for autoantibody detection through an advocate dealing with breast implant liabilities. METHODS: Autoantibody detection in 86 asymptomatic women was done blindly on a panel of 15 different antibodies (dsDNA, ssDNA, histones, SSA/Ro, SSB/La, RNP, cardiolipin, phosphatidylserine, pyruvate dehydrogenase, Scl-70, NC-1, silicone, collagen I, II, and IV). Clinical variables, specific questioning about related silicone implant symptoms, and a rheumatological examination were performed blindly by a certified rheumatologist. The findings were recorded and at a later stage compared with positive autoantibody detection. The normal control group consisted of age and sex matched Israeli women without known autoimmune disease. In the positive control group were symptomatic women previously tested for antibody production. The autoantibodies were assessed by ELISA. Values from individual patients were considered positive only when greater than 3 standard deviations above the control mean. RESULTS: The mean ages of 86 asymptomatic and 116 symptomatic women were 46.2+/-11.2 and 45.7+/-8.3 years, respectively. Breast implants were in place for a mean period of 8.2+/-5.0 years in the asymptomatic group and 15.0+/-5.6 years in the symptomatic group. The incidence of increased titers of autoantibodies ranged from 2 to 13% for 13 different autoantibodies among asymptomatic women. Among symptomatic women, 20% harbored 4 autoantibodies and 8% had 6 autoantibodies. The most common antibodies in the asymptomatic group were: dsDNA 8%; ssDNA 9%; SSB/La 13%; silicone 9%; collagen II 9%. No autoantibodies were found for NC-1, Scl-70, or RNP. Among the symptomatic group, the most common autoantibodies were histone ribosomal phosphate, SSA, SSB, Scl-70, cardiolipin, phosphatidylserine, GM2-ganglioside, and NC- 1. Comparison of autoantibody incidence in asymptomatic and symptomatic women with silicone breast implants revealed an increased incidence of anti- SSB/La and anticollagen II in both groups. Polyclonality was more prominent in the group of symptomatic women with silicone breast implants, but also evident in 3 asymptomatic women.

The mean duration of implant in the asymptomatic group was significantly less compared with the symptomatic group (p<0.01). The development of autoantibodies may be related to implant duration.

Author Address

Department of Medicine B, Sheba Medical Center, Tel-Hashomer, Israel.