Safety, Adverse Effects, and Complications of Silicone Implants

Published on February 7, 2012 by

All of the normal risks of surgery and anesthesia can occur with breast augmentation or reconstruction. Infection, bleeding, change in nipple sensation, malposition (hyposensitivity, hypersensitivity), poor healing, anesthetic accidents, and other complications can occur at a rate that is similar to that in any clean surgery in this area on healthy patients. Concerns specific to the breast implant follow.

Repeat surgery

Most women who have implants achieve satisfactory results from one operation, and the implants remain indefinitely without difficulty. As many as 20% of women may need repeat surgery (often more than once). Reoperation may be required for a host of reasons, and for a few women, these devices become high-maintenance items. To place this in perspective, do not expect more from an implant than from any natural body tissue or organ. These devices can be thought of as new body parts and, like our own organs, they may last a lifetime or cause frequent difficulty throughout life.

Capsular contracture

Often referred to as a complication, this phenomenon best is considered an adverse effect. It is the result of the normal process of scar formation resulting from the repair of separation of tissue. One would not consider the skin scar a complication of a cut despite the spectrum of final appearances. Shrinkage or shortening of a scar is a poorly understood phenomenon that varies dramatically among individuals and at different locations and directions on the same person. Contracture around an implant is probably an aborted attempt at extrusion of a foreign body. In truth, the mystery is why contracture does not occur in everyone.

Contracture is the most common adverse effect of breast implants. To achieve a soft, natural-feeling result, the surgical pocket is made somewhat larger than the implant. Normal wound healing forms a scar lining on the pocket surface termed the capsule, which, under ideal circumstances, retains its original dimensions. The oversized pocket permits full flexibility of the implant, often resulting in a breast so soft that the implant is not palpable and closely mimics normal breast mobility and softness.

For reasons that are unclear and appear to be related to a particular woman’s individual biology, the scar envelope sometimes shrinks and squeezes the implant, producing varying degrees of firmness. This commonly is graded on a scale devised by Baker, as follows:

  • Grade I – None: The augmented breast feels as soft as an unoperated breast.
  • Grade II – Minimal: The breast is less soft; the implant can be palpated but is not visible.
  • Grade III – Moderate: The breast is firmer; the implant is felt easily, and its presence is visible.
  • Grade IV – Severe: The implant is firm and often tender, painful, cool, and distorted; its presence is obvious.

Contracture can occur soon after surgery or many years later and may be symmetric, asymmetric, or unilateral. Current theory suggests that low-grade contamination by Staphylococcus epidermidis may be the initiating factor of the contracture, but this is not confirmed.

Capsular contracture is not, in itself, a health risk other than its possible interference with mammography or the risk of surgical correction, if required. However, contractures detract from the quality of the results, with the severity of concern depending upon the individual patient. The best results achieve the ideal of a breast so soft that the implant is undetectable. Yet many women prefer a slightly firm bosom while for others even a severe contracture is only a minor nuisance.

The amount of overlying breast tissue as padding influences both the sense of softness and the appearance. If the tissues are tight, usually the breast has a superior fullness similar to the effect of a push-up bra. The most severe degree of contracture may be unaesthetic or deformed, quite uncomfortable, or chronically painful.

The recent innovation of texturing the implant shell initially showed promise of reducing the incidence of contracture. However, recent data from both US manufacturers have been confusing but suggest that little or no difference exists for saline implants. In contrast, the most recent information concerning gel implants demonstrates statistically less contracture in cosmetic patients (smooth 15%; textured 9%; P
An undesirable adverse effect of texturing is an unpleasant rippling of the breast surface, especially if little overlying tissue padding is present, such as in reconstruction of very small breasts. This may be visible and unsightly or just palpable and annoying, depending on the thickness of normal breast tissue and subcutaneous fat available to mask the irregularities. For most women, firmness is a more acceptable compromise than rippling, especially if it is in the cleavage area.


No evidence exists that the silicone used in breast implants is carcinogenic in humans. More than 7 epidemiologic studies confirm this, at least for the 10- to 30-year periods covered by these reports.[6, 7] Three recent large-scale studies and 2 smaller ones have demonstrated that women with implants may have up to 30% less breast cancer than expected statistically when matched with the general population.

Three animal studies show the same protective effect, and one preliminary report suggests that blood from women with implants kills breast cancer cells in tissue culture. Therefore, all evidence suggests that at least for the greater than 30-year time frame that silicone implants have been available, the risk of humans developing cancer from silicone breast implants is negligible, if not nil. Recent studies also demonstrate that 5-year survival rates are not affected by the presence of these devices.[8]

A recent study from the National Institutes of Health (NIH) demonstrated the development of myelomas in susceptible strains of rats injected with gel in their peritoneal cavities. This was not observed with oil or elastomer. Eighteen cases are now in a newly established registry at NIH. They are clustered in Los Angeles, Arkansas, and Florida (no new cases have been added since the first 18 were collected 4 y ago).

Monoclonal gammopathy of undetermined significance (MGUS) is found in the serum as a monoclonal immunoglobulin G or immunoglobulin A in 1.5% of the otherwise healthy population. This is believed to have some predictive value in determining the risk of myeloma, since 16% of those with elevated levels develop the disease within 30 years (0.8%/y). A small sample of women with implants has demonstrated an elevated MGUS fraction. However, a search through 4 registries containing more than 20,000 women with more than 120,000 years of risk located only 1 case of myeloma. This information should be considered and studied further, but the evidence is too sketchy to generate alarm. A statistical truism is that “an association does not imply a cause-and-effect relationship.”

Polyurethane-coated implants

The polyurethane coat that covered some implants in the United States before 1991 is known to slowly hydrolyze over several years into unknown breakdown products. Of concern is one constituent of the polymer, 2, 4, toluene diamine (TDA). In studies performed in the 1960s, large doses of TDA fed to highly cancer-prone rats produced hepatomas. When viewed in the light of more modern understanding, the validity of these early experiments is questionable.

TDA never has been documented to be a human carcinogen, but because of this animal evidence, current law, known as the Delaney clause, requires that it be banned by the FDA for use in foods, cosmetics, and implantable devices. (Congress has recently amended this clause to make it more pertinent to scientific reality.) However, no evidence exists that TDA is formed in vivo from polyurethane. While TDA is a building block of polyurethane, it is not produced by hydrolysis, since the molecule is cleaved at the urea fraction rather than at the TDA site. An FDA advisory panel hearing held in July 1991 concluded that the probable cancer risk from polyurethane-coated implants is considered negligible (
More recent studies have demonstrated that approximately 80% of women with these devices show traces of TDA in their urine. Traces of TDA also were found in the urine of 11% of the control group, suggesting an environmental source of the chemical. None was found in the blood of patients or controls. Again, the risk of cancer from polyurethane implants was calculated to be less than 1 in 1 million. To date, no cases have been reported in the medical literature of cancer in a woman with polyurethane implants. Because the small but real risks associated with removal are greater than the risk of cancer, the FDA has advised that there is no health related reason that requires removal of these particular devices in asymptomatic women.

Cancer detection

Perhaps the most significant concern regarding breast implants is the possibility of delayed detection of breast cancer. The implant itself is radio-opaque and variably compresses the breast tissues depending upon the particular configuration of a particular woman’s breast architecture and the degree of contracture. A real concern involves the presence of an implant compromising mammography and delaying detection until the mass is large enough to be palpable. Several recent studies have shown that this appears to be only a theoretic risk, since the stage of detection of the breast cancer in women with implants appears to be identical or better than that of the overall population.[9, 10]

Mammographic techniques have improved dramatically in the last few years, enabling the mammographer to minimize the amount of breast that is hidden by an implant. Approximately 9% to more than 20% of breast cancers in all women are invisible on radiographs. Despite the increasing use of mammography, most breast cancers still are discovered by self-examination or physician examination. Many clinicians believe that the presence of an implant can increase the ease of palpation. To date, no cases have been documented in the medical literature in which a diagnosis of breast cancer was delayed by the presence of an implant. This suggests that this is a rare occurrence, at worst.

The Society for Breast Imaging, the American Society of Plastic and Reconstructive Surgeons (ASPRS), and the American Cancer Society agree that a woman with breast implants should be on the same schedule of routine mammography as other women, as follows:

  • Baseline – At the earliest when aged 35 years
  • Biannually – When aged 40-50 years
  • Annually – When older than 50 years

However, a woman with breast implants should avoid screening clinics where only 2 routine views of the breast are taken.[9, 11] Instead, she should be referred to mammographic units accredited by the American College of Radiology, which are familiar with the special displacement (Eklund) views required for proper mammographic evaluation of the implanted breast.[12, 13, 14] Ideally, the woman should try to obtain her studies at the same facility each time so that her films can be observed serially, and she should inform the technician that she has implants. Because extra views are required, the cost of the mammogram for these patients is modestly higher.

Preoperative mammograms in women younger than 35 years should be discouraged. Cancers can be observed best when the breast parenchyma is mostly fat. The young breast has dense stroma, thus mammography contributes little or nothing to diagnosis of any breast disease. In addition, evidence exists that the young breast is much more vulnerable to radiation damage.

For excellent patient education resources, visit eMedicine’s Women’s Health Center. Also, see eMedicine’s patient education article Breast Lumps and Pain.

Rheumatologic disorders

The media and the courtroom have made much about autoimmune disease, human adjuvant disease, or silicone-associated disorder (SAD). Almost every disease and symptom complex from scleroderma to chronic fatigue syndrome to multiple sclerosis and amyotrophic lateral sclerosis has been blamed on the breast implant in various anecdotal reports. More than 200 separate conditions have been listed in plaintiff pleadings as having been caused by these devices. The College of Rheumatology, American Medical Association, and FDA agree that the term human adjuvant disease is inaccurate and inappropriate. These disorders, if they exist, should be labeled simply rheumatologic or immune disorders.

Of those rheumatologic disorders most likely to be caused from exogenous sources, scleroderma (sclerodermalike syndrome is the correct term for the variants that arise from exogenous toxins), lupus, and Sjögren disease are the most probable candidates. Scleroderma is a rare disease with an estimated incidence of approximately 10,000 new cases per year in the United States. A search of the medical literature in 1995 documented approximately 130 cases of scleroderma in women who also happened to have breast enlargement. (This number includes a significant group that was injected with unknown substances such as paraffin or adulterated silicone.)

Since these are rare diseases, they require large-scale epidemiologic studies to determine if a relationship is present. Currently, 31 such studies from 4 countries encompassing a cohort of more than 500,000 women failed to find a statistical relationship between any known or newly recognized disease and silicone.[15, 16, 17, 18, 19, 20, 21, 22, 23, 24]

A panel of experts appointed to review all of the literature for the court with jurisdiction over the implant class action suit summarized their findings as follows:

No association was evident between breast implants and any of the individual connective tissue diseases, all definite connective diseases combined, or the other autoimmune/rheumatic conditions. Sjögren’s syndrome was a possible exception to this statement. This entity requires salivary gland biopsy to meet the published diagnostic criteria. Whether biopsy was actually performed for cases in the studies cited is unknown. The remaining criteria based on dryness of the eyes and mouth with possible immunologic alterations are nonspecific and relatively common in any population group. Thus, the accuracy of diagnosis of Sjögren’s syndrome in the studies incorporated in this meta-analysis is questionable.

Additionally, some claim that silicone causes a totally new syndrome (termed by some silicone-associated disorder), often presenting as a unique variant of fibromyalgia, which may be associated with implants. These all have been based on anecdotal reports, and no epidemiologic studies have been performed to support this claim. On October 22, 1995, the American College of Rheumatology issued the following strongly worded statement critical of claims of such new disorders: “The ACR believes that these studies provide compelling evidence that silicone implants expose patients to no demonstrable additional risk for connective tissue or rheumatic disease.”

Thus, of the many other symptom complexes, illnesses, and disorders claimed by patients, some physicians, the plaintiffs’ bar, and the media to be due to implants, none stand up to scientific scrutiny.

The intensified interest in this issue has generated a great deal of laboratory research in many centers. In some instances, an immune response has been observed in animals, while others have demonstrated positive laboratory findings in a number of women with implants. While animal studies are useful to gain an understanding of processes, extrapolation to humans is often inappropriate. None of the published material to date has presented evidence for a convincing cause-and-effect relationship with any human disease entity. Unfortunately the lay public, the media, and some physicians have equated an immune response, as manifested by laboratory tests, with disease. An immune response is not an immune disease.

As early as 1989, a Plastic Surgery Educational Foundation (PSEF)-sponsored consensus panel on this subject concluded the following:
0.”There is insufficient information available at this time to determine whether silicone in the form of a breast implant can be implicated as a cause of scleroderma-like syndrome or any other autoimmune disease. Judging from the paucity of reported cases in the very large population of implanted women, if a causal association were to be established, the statistical risk would likely be very low. The presence of risk and magnitude thereof can be determined only by appropriate epidemiological research.”
0.”At present, there is no reason to discourage women from considering breast augmentation on the basis of the risk of acquiring or exacerbating a connective tissue disorder. Until the question is answered by further research, it is wise to inform patients that a theoretical risk might exist, especially if they already have a connective tissue disorder, idiopathic Raynaud’s phenomenon, or an affected first-degree relative.”

Therefore, if a relationship is present, it must occur in a small number of women who have a genetic predisposition for the disease for which no predictive test is available. The workshop participants repeatedly emphasized that assuming that the presence of an association between two findings implies a cause-and-effect relationship is flawed logic. This truism is understood poorly by most, thus easily is exploited by those who would profit from vilifying the device. In general, while publicly neutral, privately a strong air of skepticism existed among the participants concerning cause and effect.

In 1991, an FDA panel of experts, having heard all of the then-current evidence that suggested a connection between silicone gel breast implants and rheumatologic disorders, concluded that “the evidence was unconvincing.”

At an NIH-sponsored workshop on silicone implants and atypical rheumatic disease, the evidence for a relationship with known typical disease was again acknowledged as not convincing. The participants affirmed that the diagnosis atypical connective tissue disease was ill-defined, especially for those disorders that rely on subjective symptoms only and are devoid of objective findings or diagnostically specific laboratory findings. The currently classified atypical disorders represent a hodgepodge of subjective symptomatology with or without nonspecific abnormal laboratory tests. They are useful for the purpose of assigning a diagnosis for record keeping, insurance, and patients’ needs for labeling but should not be considered as clearly defined entities.

The absence of a clear definition and classification of these vague ailments makes validating their existence impossible. A substantial portion of the rheumatology community is skeptical that such disorders even exist as defined entities. Thus, pinpointing any specific disorder as caused by breast implants is insupportable. Until consensus is reached and these disorders are defined clearly or discredited (estimated to require 10 y if started today with adequate funding), abuses and misuse of these diagnoses will continue.

Interestingly, manufacturers and physicians are criticized by the plaintiffs’ bar, the media, and unhappy patients for not determining whether implants cause these diseases, when the technology to do so does not exist today, let alone 10-20 years ago.

Finally, several reports state that reversal of the symptomatology can occur with removal of the implants. However, the evidence to date is confusing and suggestive of a placebo effect, the natural ups and downs of the disease, or the results of concomitant treatment.

Contact Us