Marcy Oppenheimer, M.D.
In 2005, the Mentor Corporation sought FDA approval to market its silicone gel-filled breast implants. At a hearing before an FDA advisory panel on April 13, the company provided evidence claiming that the implants are safe and effective.
Based on the information provided, FDA scientists indicated that the data failed to show that the implants were safe, and also raised questions about effectiveness.
No Research on Most of their Implants
Mentor is seeking approval of implants of three different shapes. Each is available with two alternative shell surfaces; smooth and textured, for a total of 6 different implant styles. The Core Study, the primary clinical trial that Mentor provided to the FDA in support of FDA approval, includes only one of the implant shapes, and therefore only two of the six implant styles. There are no clinical data on the other two shapes for either type of shell surface.1
Rupture rates refer to the percentage of implants that break each year. Why do rupture rates matter? According to FDA’s guidance to industry:
FDA believes that device rupture is one of the primary safety concerns presented by breast implants. When a silicone gel-filled breast implant ruptures, the patient and the physician may be unaware of it, the body does not have a mechanism for eliminating the silicone gel, and the gel can migrate outside of the capsule into the breast area, the lymph nodes, and distant locations (i.e., extracapsular gel migration).2
FDA thus recommended that manufacturers provide data on the “rate and rate of change of rupture over the expected lifetime of the device.” At the hearing, Mentor’s consultants estimated that the lifetime of their silicone gel breast implants is 25-47 years, which is much longer than previous implants have been found to last. However, the company provided minimal data to support that estimate or to predict rupture rate over that period of time.
Mentor pointed to two studies as proof of a low rupture rate. The Core Study contains data on patients who have had implants for two to three years, but the most accurate measure of rupture was used only during the first or second year.3 This study is obviously inadequate to establish the rupture rate of a device with a lifetime of more than 25 years.
The other study presented by Mentor, the Sharpe/Collis study, was an unpublished, 10-year case series of patients in one plastic surgery practice in the UK. This study had many inadequacies including:
• All patients were treated by one surgeon, who may not be typical of all surgeons.
• Only augmentation patients were included. Reconstruction and revision patients have increased risk of rupture, but were not in the study.
• Patients with capsular contracture or a prior surgical procedure were excluded. These patients have increased risk of rupture.
• All implants were placed above the chest muscle. In the U.S., implants are typically placed below the chest muscle, which increases the risk of rupture.
• Patients with implants removed during the first four years were excluded. These patients may have had ruptured implants not counted in the study.
• Data are subject to selection bias, because patients were limited to voluntary participants from one surgeon’s practice. Patients who were unhappy or angry with their surgeon were unlikely to participate
The FDA concluded that this study is “of limited value in characterizing the rupture rate” of Mentor implants.4 Other rupture data presented by Mentor included publications not specific to their own breast implants.
In summary, Mentor was unable to predict the rupture rate of its silicone breast implants. The company presented two years’ worth of MRI data, with an additional 10 years of scientifically unsound data, to characterize the rupture rate of a device with a lifetime that they predicted would be 25-47 years.
Health Consequences of Implant Rupture
No one knows the exact health consequences of implant rupture, although it is known that silicone leaks from some ruptured implants.5 FDA requested that Mentor elucidate the health consequences of implant rupture,6 but Mentor failed to do so.
FDA concluded that the Core Study data “are of limited value”7 in assessing health consequences. Similarly, the published articles Mentor cites “do not completely address all the health consequences of rupture, and the literature is not specific to Mentor implants.”8
Data from another implant maker, Inamed, raises important concerns about implant safety. Inamed analyzed local complications among a small number of patients with confirmed ruptured implants, compared to patients with confirmed intact implants.9 These data are presented below since Mentor data were so short-term that they were not able to present this type of analysis.
Confirmed Ruptured Implants (17 patients) Confirmed Intact Implants (79 patients)
Lymphadenopathy (swollen lymph nodes)
Seroma (collection of fluid around implant)
It is clear from the table that patients with confirmed ruptures have higher rates of infection, lymphadenopathy, redness, seroma, and skin rash, which are shown shaded above. These symptoms are generally associated with an inflammatory reaction and/or with infection. While these data are based on a small number of patients (only 17 in the confirmed rupture group), they suggest that implant rupture may cause an inflammatory process in the body.
Danish data submitted by the implant manufacturers show that women with ruptured implants are twice as likely to report non-serious pain in the affected breast compared to women with intact implants, and are six times more likely to report breast hardness. In addition, women with extracapsular rupture (where the silicone has leaked outside the immediate area of the implant) are 3 times more likely to report a connective tissue disease, 2 times more likely to report pleuritis, and 1.7 times more likely to report than women with intact implants, although these numbers were not statistically significant.10
Based on a very conservative rupture rate estimate developed by the FDA, 25,000 ruptures would occur among augmentation patients in the U.S. each year if implants were approved for general use. What risks would these women face? No one, including the implant manufacturers or the FDA, can tell them.
Women getting silicone breast implants can expect to have several follow-up surgeries. Mentor’s Core Study indicated that after only two years, 25% of reconstruction patients, 12% of augmentation patients, and 26% of revision patients can expect to undergo a second surgery related to their implants.11 For the augmentation and revision patients, the most common cause of re-operation is capsular contracture, an often-painful hardening of the breast. For the reconstruction patients, the most common cause of re-operation was asymmetry.
Signs and Symptoms of Connective Tissue Disease
Although not provided by the FDA, an analysis presented by Advisory Panel member Dr. Brent Blumenstein indicates an increase in several signs and symptoms of connective tissue disease, such as joint pain and fatigue, for women who have had implants for two years, compared to prior to getting breast implants. The analysis was conducted on all women in the Mentor Core Study, not just those with a ruptured implant. These differences are maintained even when the impact of aging is statistically controlled.
In addition to evaluating the risks of implants, the companies were required to evaluate the benefits as well. It is clear that breast implants increase breast size, but plastic surgeons and implant makers also claim that breast implants help women feel better about themselves and their lives. For that reason, the FDA required implant makers to use objective measures of emotional health, comparing women before they had implants and two years later.
Mentor data show that almost all patients say they are satisfied with their implants after three years. However, their answers to several scientifically developed scales and questionnaires shows that overall, their quality of life remained the same or declined two years after receiving implants. The table below summarizes these results. In the table, a change that was not statistically significant is listed as “no change.”
Assessment Tool Augmentation Reconstruction Revision
Tennessee Self-Concept Scale12
Physical Component Summary of SF-3613
Mental Component Summary of SF-36
Body Esteem Scale14
Rosenberg Self-Esteem Scale15
Functional Living Index: Cancer 16
Better (for delayed post-mastectomy patients)
Better (if a history of cancer)
Besides its own data, Mentor submitted a review of the published literature on the efficacy of implants. FDA reviewers, however, found little support for Mentor’s position that implants improved women’s self-esteem or quality of life. With respect to augmentation patients, the FDA found that “the literature does not provide strong scientific support that breast implants have measurable psychological and psychosocial benefits. . .” For reconstruction patients, the FDA concluded that “[l]iterature that adequately evaluates the short-term or long-term psychological or psychosocial benefits of breast implants . . . was not provided by Mentor.”17
In summary, neither the research conducted by implant makers, nor the published research literature, support the claim that breast implants improve a women’s self-esteem, mental health, or quality of life. This is true whether the patients chose implants after mastectomy, or for augmentation, and whether they were getting breast implants for the first time or to replace problem implants.
Implications of other Research and Federal Activities
Research conducted by an epidemiologist in Canada found that augmentation patients were more likely to be hospitalized, were hospitalized for more days, and used the healthcare system more than similar women who did not have breast implants.18
Although the health effects of breast implants are controversial, the U.S. Department of Justice successfully sued silicone breast implant makers to recover the costs of treating patients with illness caused by implants. Mentor and other implant makers agreed to pay more than $11 million to compensate the federal government for these health costs. The law suit was based on government evidence, not made public, that breast implants caused health problems, which resulted in the government paying medical costs and disability payments.
1 FDA Summary Panel Memorandum, Mentor PMA Review Team, March 2, 2005, p.1.
2 Draft Guidance for Industry and FDA Staff: Saline, Silicone Gel, and Alternative Breast Implants, January 13, 2004, p.27. Available at http://www.fda.gov/cdrh/ode/guidance/1239.pdf.
3 The Core Study contains some three-year data. However, rupture is primarily detected by MRI, and this study called for MRIs in years 1 and 2, but not year 3, so the third-year data is not useful for determining rupture rate. Another weakness of the study is that it excluded patients who had their implants removed. These are patients whose implants may well have ruptured.
4 Mentor Summary Memorandum, p. 80. Available at http://www.fda.gov/ohrms/dockets/ac/05/briefing/2005-4101b1.htm
5 Mentor Summary Memorandum, p. 88.
6 Mentor Summary Memorandum, p.87.
7 Mentor Summary Memorandum, p. 87.
8 Mentor Summary Memorandum, p. 89.
9 FDA Summary Panel Memorandum, Inamed PMA Review Team, March 2, 2005, p. 24.
10 Mentor Summary Memorandum, p. 88.
11 Mentor Summary Memorandum, pp. 52, 53.
12 The Tennessee Self-Concept Scale summarizes an individual’s feeling of self-worth, the degree to which the self-image is realistic, and whether or not that self-image is a deviant one. There are 100 items rated on a 5-point scale ranging from completely false (1) to completely true (5).
13 The SF-36 is a generic measure of 8 health concepts: physical functioning; role-physical; bodily pain; general health; vitality; social functioning; role-emotional; and mental health. The 8 scales can then be collapsed into two summary scales with the first 4 scales comprising the Physical Component Summary (PCS) and the last 4 scales comprising the Mental Component Summary (MCS).
14 The Body Esteem Scale measures female body esteem for a variety of body parts and functions. Assessments include sexual attractiveness, weight concern, and physical condition. There are 35 questions that are rated on a 5-point scale ranging from strong negative feelings to strong positive feelings. Lower scores indicate more strongly negative feelings about a body part or function.
15 The Rosenberg Self-Esteem Scale assesses global and uni-dimensional self-esteem, relating to an individual’s overall feelings of self-worth or self-acceptance. There are 10 items rated on a 4-point scale ranging from strongly agrees to strongly disagrees. A higher score indicates higher self-esteem.
16 The Functional Living Index: Cancer is a subjective instrument designed to assess physical well-being, psychological state, family situational interaction, social ability, and somatic sensation in cancer patients of the reconstruction and revision indication groups.
17 FDA Summary Memorandum, Mentor PMA Review Team, March 2, 2005, p. 70, 73.
18 Tweed A. Health care utilization among women who have undergone breast implant surgery. British Columbia Centre of Excellence for Women’s Health. Available: www.bccewh.bc.ca/PDFs/hcubreastimplants.pdf.