Surgical procedures involving breast implants, particularly breast augmentation and reconstruction, are commonly performed in the United States and in many other countries, and the procedures are increasing in number.1,2 Despite decades of experience with breast implants, the long-term safety of implants continues to be debated.3,4 When silicone gel–filled breast implants were approved in 2006, the U.S. Food and Drug Administration required the initiation of large postapproval studies to evaluate the incidence of rare adverse events, such as connective tissue diseases, neurologic diseases, cancer, suicide, and a range of other outcomes over 10 years after implantation.5
The Breast Implant Follow-up Study is a large, multicenter, 10-year, observational study being conducted to address the U.S. Food and Drug Administration’s requirement for large, long-term safety studies. The objective of this ongoing U.S. study is to compare the long-term safety of Natrelle round silicone gel–filled breast implants (Allergan plc, Dublin, Ireland) with national norms and outcomes with saline-filled breast implants over the same 10-year period. Natrelle silicone implants are available worldwide in a range of implant options and are approved by the U.S. Food and Drug Administration for primary augmentation, revision-augmentation, primary reconstruction, and revision-reconstruction procedures.6 The 10-year Core Studies7,8 established the safety profile of Natrelle round and form-stable anatomical silicone implants with respect to rates of complications, such as capsular contracture, reoperation, and implant rupture. The Core Studies provided a robust source of data for local complications, but the populations were not large enough to answer questions about rare diseases, such as connective tissue diseases and cancers, many of which occur at rates of fewer than 12 events per 100,000 person years in the general population.9–24 The Breast Implant Follow-Up Study was designed to have sufficient statistical power for detection of these rare diseases.
This article presents safety data from more than 55,000 women in the Breast Implant Follow-Up Study with at least a 5-year follow-up. All subjects underwent augmentation procedures (primary augmentation and revision-augmentation) or reconstruction procedures (primary reconstruction and revision-reconstruction) and received Natrelle round silicone implants or saline implants. Baseline data and surgical characteristics for these subjects have been published.25,26 Rates of targeted adverse outcomes, including cancers, connective tissue diseases, neurologic diseases, and suicides, were compared for Natrelle round silicone implants relative to expected rates (i.e., national norms) and rates with saline implants.
PATIENTS AND METHODS
The Breast Implant Follow-Up Study is an ongoing observational study comparing subjects who elected to receive either Natrelle silicone implants or saline implants. Women who were seeking primary augmentation, revision-augmentation, primary reconstruction, or revision-reconstruction were invited to participate at the time they decided to undergo breast implantation. Data collection began on February 15, 2007, and all subjects were enrolled by March of 2010. This analysis is based on data as of October 1, 2015, at which point all subjects had surpassed the 5-year time point; the maximum follow-up was 8 years. This study was conducted at 1116 U.S. sites.
The study was approved by the institutional review board for each participating study site. It was conducted in accordance with Good Clinical Practice guidelines, and conformed with World Health Organization guidelines. All participants provided written informed consent before enrollment.
Eligible subjects were women aged 22 years or older for primary augmentation and revision-augmentation, and 18 years or older for primary reconstruction or revision-reconstruction. All enrolled subjects were required to be fluent and literate in English or Spanish and to receive one implant or matching implants (both silicone or both saline). Silicone implants were required to be Natrelle round devices. Subjects who currently had saline implants were excluded from the study if they had previously received silicone implants. Subjects were also excluded if they were transgender or if they were deemed by the investigator to be unsuitable for long-term observation.
After surgery, all subjects were asked to complete follow-up questionnaires by means of the Internet, telephone interview, or postal mail once annually for 10 years. These questionnaires collected information on targeted adverse events, and other outcomes (e.g., pregnancy, lactation, and satisfaction) that are reported separately.27 Safety was evaluated through the collection of diagnoses of the following adverse events: connective tissue diseases, neurologic disease, cancer (i.e., breast, lung, brain, and cervical/vulvar), suicide, and self-reported suicide attempts. If a subject reported an adverse event to the plastic surgeon investigator, the physician who made the initial diagnosis (i.e., the diagnosing physician) was contacted to confirm the diagnosis. For subjects with silicone implants, a physical examination was performed by the investigator at years 1 and 4, and will be performed again at year 10, to assess local complications.
For all targeted adverse events, only new postoperative cases were counted. Subjects with a medical history of the adverse event of interest at baseline were excluded from the analysis of that particular adverse event but still included in the analysis of other adverse events. However, subjects with a medical history of any cancer at baseline were excluded from the analysis for all cancers. National norm rates for the rare adverse events were obtained from national data sets.9–24 The process for selecting national norm rates was determined a priori. To avoid potential bias in national norm rates because of confounding factors, the rates were selected to be most representative of the study population (e.g., adjusted for female sex, age, or race when possible). When the national norm was given as a range, the midpoint was used. For rare adverse events occurring at national norm rates between 2.85 per 100,000 person-years and 1.2 per 10,000 person-years, the silicone group was compared with national norms.9–24 The comparison of silicone implants to national norms was conducted using a Poisson distribution.28 To account for differences in the distribution of age and race between the study population and the general female population in the United States, rates were adjusted by the indirect method of standardization. One-tailed tests were performed at the p = 0.05 level of significance; therefore, a two-sided 90 percent confidence interval was used.
For rare and other targeted adverse events, the silicone group was compared with subjects who received saline implants using a Poisson regression model that was fitted using PROC GENMOD in SAS version 9.3 (SAS Institute, Inc., Cary, N.C.), with implant group, indication, race, smoking status, implant location, implant surface type, history of cancer, and history of connective tissue diseases included as categorical variables. Age, implant size, and body mass index were included as continuous variables. A screening univariate analysis was performed, and significant covariates from that analysis were included in the final model. A stepwise selection approach was used to include covariates in the final model, with a value of p = 0.10 used as the selection criterion. The relative risk for silicone versus saline and associated 90 percent confidence intervals were calculated by exponentiating the implant group parameter estimate and 90 percent confidence intervals of the log of the relative risk from the final model. A one-sided p value based on a log likelihood ratio test with 1 degree of freedom was calculated for each targeted adverse event.
Sample Size Considerations
For the comparison of silicone to national norms, power calculations determined that approximately 317,983 person-years of follow-up would be required to yield 80 percent power for a one-sided test to detect a relative risk of 2 for an adverse event with an expected baseline incidence rate as low as 2.85 per 100,000 person-years. Based on an assumed overall loss to follow-up rate of 35 percent (3.5 percent of the original group per year),29 it was determined that 39,390 subjects should be enrolled in the silicone group to obtain approximately 317,983 person-years of follow-up at 10 years. The total follow-up needed for the saline control group was 124,240 person-years (≥80 percent power; one-sided test at the 0.05 level) to detect a relative risk of 2.0 (silicone/saline) for an adverse event with an expected baseline incidence rate of 1.2 per 10,000 person-years. Based on a 35 percent loss to follow-up rate, 15,240 saline control subjects were needed to obtain 124,240 person-years of follow-up at 10 years.
A total of 55,279 subjects constitute the safety population. Of these, 42,873 subjects underwent primary augmentation (29,148 silicone and 13,725 saline), 6837 underwent revision-augmentation (5901 silicone and 936 saline), 4828 underwent primary reconstruction (4648 silicone and 180 saline), and 741 underwent revision-reconstruction (699 silicone and 42 saline). Most of the silicone (91.2 percent) and saline (97.1 percent) implants were smooth surface devices; the other implants were textured devices. Demographic characteristics for the primary augmentation, revision-augmentation, primary reconstruction, and revision-reconstruction groups have been published.25,26 Briefly, primary augmentation and revision-augmentation subjects were predominantly white (63 to 81 percent), and revision-augmentation subjects were a median of 8 years older than those in the primary augmentation group (42 years versus 34 years). Primary reconstruction and revision-reconstruction subjects were predominantly white (77 to 86 percent), and revision-reconstruction subjects were a median of 3 years older than those in the primary reconstruction group (53 years versus 50 years).
All women had surpassed the 5-year time point, with follow-up of 6 to 8 years for the early enrollers in the study. The proportion of subjects who had completed at least one postimplantation questionnaire was 89.5 percent for silicone implants and 78.0 percent for saline implants. The follow-up rate for the 5-year visit was 60.5 percent for subjects with silicone implants and 52.9 percent for subjects with saline implants. The final follow-up rate is not yet known because follow-up through year 10 is ongoing. The total follow-up for all subjects ranged from 157,215 to 189,292 subject-years for women who received a silicone implant and from 61,738 to 64,793 subject-years for women who received a saline implant, depending on the adverse event assessed.
Follow-up was sufficient to detect occurrences and therefore significant differences from national norms for all target adverse events, with the exceptions of relapsing polychondritis and Wegener granulomatosis, both of which had zero occurrences during the study (Table 1).9–15,18,19,21–23,30 The target adverse event of polyarteritis nodosa was also not reported. However, the expected event rate for polyarteritis nodosa (2.35 events per 100,000 person-years)15 exceeds the upper limit of the 90 percent confidence interval of the observed rate (1.6 events per 100,000 person-years), indicating that follow-up was sufficient to detect a significant increase from the national norm for this adverse event.
Rare Targeted Adverse Events: Silicone Groups Compared with National Norms
The rates of rare targeted adverse events confirmed by the diagnosing physician for the subjects receiving silicone implants compared with expected national norm rates are shown for all groups combined in Table 1, for the primary augmentation and revision-augmentation groups in Table 2, 9,11,12,18,19,21–23,30 and for the primary reconstruction and revision-reconstruction groups in Table 3.9–12,19,21,30 None of the rare targeted adverse events, including connective tissue diseases, brain cancer, cervical/vulvar cancer, and neurologic disorders, were observed at significantly higher rates in any of the groups compared with national norms. No cases of breast implant–associated anaplastic large cell lymphoma (ALCL) were reported.
Rare and Other Targeted Adverse Events: Silicone Implants Compared with Saline Implants
Tables 4 through 6 show, respectively, the relative risk of targeted adverse events with silicone versus saline implants for all groups combined, for the primary augmentation and revision-augmentation groups, and for the primary reconstruction and revision-reconstruction groups. (See Table, Supplemental Digital Content 1, which shows the covariates used in the model for adjusted relative risk, http://links.lww.com/PRS/C361.) Silicone implants did not significantly increase the risk of any targeted adverse event compared with saline implants.
Deaths and Suicides
The relative risk and adjusted relative risk of death in the silicone and saline groups for the augmentation and reconstruction indications are shown in Table 7. (See Table, Supplemental Digital Content 1, which shows the covariates used in the model for adjusted relative risk, http://links.lww.com/PRS/C361.) In subjects who underwent primary augmentation (adjusted relative risk, 0.91; 90 percent CI, 0.66 to 1.26) or revision-augmentation (adjusted relative risk, 0.88; 90 percent CI, 0.41 to 1.89), the risk of death was similar with silicone versus saline implants. Similarly, in subjects who underwent primary reconstruction, the risk of death was similar with silicone versus saline implants (adjusted relative risk, 1.22; 90 percent CI, 0.62 to 2.42). Relative risk for the revision-reconstruction group was not determined because no subjects with saline implants died. No implant-related deaths occurred.
The suicide rate among all women with silicone implants (10.6 events per 100,000 person-years; 90 percent CI, 7.0 to 15.4 events per 100,000 person-years) was not significantly higher than the rate in the general female population (Table 1).30 When evaluated for each indication, the suicide rate was not significantly higher than the national norm (Tables 2 and 3). The event rate (per 10,000 person-years) of subject-reported suicide attempts was not significantly higher for women with silicone implants compared with those with saline implants in the primary augmentation group (silicone, 13.0; saline, 16.1; adjusted relative risk, 0.78; 90 percent CI, 0.62 to 0.99) and in the revision-augmentation group (silicone, 11.7; saline, 10.3; adjusted relative risk, 1.27; 90 percent CI, 0.54 to 2.99). The event rate (per 10,000 person-years) of subject-reported suicide attempts was also not significantly higher for women with silicone implants compared with saline implants in the primary reconstruction group (silicone, 10.3; saline, 10.6; adjusted relative risk, 0.72; 90 percent CI, 0.14 to 3.86) and in the revision-reconstruction group (silicone, 11.4; saline, 0.0; adjusted relative risk and 90 percent CI, not applicable).
The Breast Implant Follow-Up Study, the most extensive long-term safety study of silicone implants to date, is providing invaluable information in a large population of women over a period of 10 years after implantation. The multicenter design of this study allows for a broad representation of real-world patients seeking breast implant procedures. Over 5 years of follow-up, the data showed that Natrelle round silicone implants did not significantly increase the risk of any systemic disease over expected rates based on national norms or when compared with saline implant outcomes, regardless of the indication for implantation.
The safety profile of Natrelle round silicone implants with respect to rates of complications, such as capsular contracture and implant rupture, was established in the 10-year Core Study.7 In that study, the Kaplan-Meier risk rates through 10 years varied across indications, ranging from 19 to 29 percent for capsular contracture and 5 to 35 percent for implant rupture.7
The results of this study support the long-term safety of breast implants with respect to systemic disease and are consistent with previous long-term safety studies showing that cancer and connective tissue disease risks are not elevated in women with implants. No evidence of an association between breast implants and an increased risk of any type of cancer was found in a pooled Scandinavian study that followed 3486 Swedish and 2736 Danish women who underwent cosmetic breast implantation (80 percent silicone implants) between 1965 and 1993 for a mean follow-up duration of 16.6 years.31 Possible associations between implants and risk of connective tissue diseases were investigated in a prospective group study that followed 23,847 American women [3950 with implants (70 percent with silicone implants) and 19,897 without implants] for a median of 3.63 years.32 Multivariate analyses adjusted for age, body mass index, smoking, and postmenopausal hormone status showed that breast implants were not associated with large increases in the risk of connective tissue diseases [relative risk of confirmed connective tissue diseases, 1.39 (95 percent CI, 0.82 to 2.35) for women with implants versus those without implants].
A meta-analysis of 32 studies comparing long-term health outcomes in women with and without breast implants4 found possible increased risks for lung cancer, rheumatoid arthritis, Sjögren syndrome, and Raynaud phenomenon. However, the authors reported that the evidence was inconclusive, as most of the studies were not specific to silicone implants and were rarely adequately adjusted for known confounding variables (e.g., smoking status or age). The current results provide more conclusive evidence that women with silicone implants do not have significantly higher risks of lung cancer, rheumatoid arthritis, Sjögren syndrome, Raynaud phenomenon, or any of the rare targeted adverse events.
Breast implant–associated ALCL is a rare T-cell lymphoma that has been associated with silicone implants and with saline implants.33,34 Because the Breast Implant Follow-Up Study began enrolling subjects in 2007, before the publication of the first epidemiologic study of breast implant–associated ALCL in late 200835 and before the profusion of literature on the topic in 2011,36–41 the study was not designed or powered to assess this disease. Nevertheless, no cases of breast implant–associated ALCL were reported in the study. The incidence of breast implant–associated ALCL is difficult to estimate, and incidences suggested in earlier reports35 are now considered underestimates. Realistic estimates of the incidence of breast implant–associated ALCL will be determined as cases of breast implant–associated ALCL come to light42 and as the medical community gains an increased awareness of this potential risk associated with breast implants.
The suicide rate observed in this study was not significantly higher than the suicide rate in the general female population.30 Several studies have reported higher suicide rates among women with implants.43–46 However, the meta-analysis discussed earlier4 found that the evidence from those trials was inconclusive. It has been postulated that the population of women who decide to undergo breast implantation may have underlying differences in demographic, lifestyle, and psychological traits that may put them at risk for suicide.4
This study has certain limitations. First, there is a potential for bias toward underreporting of adverse events by using data confirmed by the diagnosing physician, as subjects may not seek a diagnosis, or the investigator may be unable to contact the diagnosing physician. However, because there is also a potential for bias in relying solely on subject-reported data, it was deemed more accurate to rely on systemic disease reports that were confirmed with the diagnosing physician. Second, there may be some inconsistencies between physicians in diagnostic and reporting practices, especially in a study as large as this one (>1000 sites). The study was statistically powered for 10 years of follow-up, but the data reported herein are based on all subjects passing the 5-year follow-up window. The number of subject-years was not equivalent across different indications. Despite the limitations inherent in observational studies such as this, the results of controlled studies are inherently less biased than retrospective data reviews.
Results from the Breast Implant Follow-Up Study after a minimum of 5 years of follow-up for more than 55,000 subjects provide an unprecedented look at the safety of Natrelle round silicone implants for primary augmentation, revision-augmentation, primary reconstruction, and revision-reconstruction. None of the targeted adverse outcomes, including cancers, connective tissue diseases, neurologic diseases, suicides, and suicide attempts, occurred at significantly higher rates in women with silicone implants compared with national norms and women with saline implants. These findings add to the growing body of evidence confirming the positive long-term safety profile of silicone breast implants.
This study was sponsored by Allergan plc (Dublin, Ireland). Writing and editorial assistance was provided to the authors by Lela Creutz, Ph.D., of Peloton Advantage (Parsippany, N.J.) and was funded by Allergan plc (Dublin, Ireland). All authors meet the International Committee of Medical Journal Editors authorship criteria. Neither honoraria nor payments were made for authorship.
3. Jewell ML. Silicone gel breast implants at 50: The state of the science. Aesthet Surg J. 2012;32:10311034.
4. Balk EM, Earley A, Avendano EA, Raman G. Long-term health outcomes in women with silicone gel breast implants: A systematic review. Ann Intern Med. 2016;164:164175.
6. Natrelle silicone-filled breast implants smooth & biocell texture (directions for use). 2015.Irvine, Calif.: Allergan plc.
7. Spear SL, Murphy DK; Allergan Silicone Breast Implant U.S. Core Clinical Study Group. Natrelle round silicone breast implants: Core Study results at 10 years. Plast Reconstr Surg. 2014;133:13541361.
8. Maxwell GP, Van Natta BW, Bengtson BP, Murphy DK. Ten-year results from the Natrelle 410 anatomical form-stable silicone breast implant core study. Aesthet Surg J. 2015;35:145155.
9. Barnes J, Mayes MD. Epidemiology of systemic sclerosis: Incidence, prevalence, survival, risk factors, malignancy, and environmental triggers. Curr Opin Rheumatol. 2012;24:165170.
10. Bendewald MJ, Wetter DA, Li X, Davis MD. Incidence of dermatomyositis and clinically amyopathic dermatomyositis: A population-based study in Olmsted County, Minnesota. Arch Dermatol. 2010;146:2630.
11. Chakravarty EF, Bush TM, Manzi S, Clarke AE, Ward MM. Prevalence of adult systemic lupus erythematosus in California and Pennsylvania in 2000: Estimates obtained using hospitalization data. Arthritis Rheum. 2007;56:20922094.
12. Chifflot H, Fautrel B, Sordet C, Chatelus E, Sibilia J. Incidence and prevalence of systemic sclerosis: A systematic literature review. Semin Arthritis Rheum. 2008;37:223235.
13. González-Gay MA, García-Porrúa C. Epidemiology of the vasculitides. Rheum Dis Clin North Am. 2001;27:729749.
14. Hazra N, Dregan A, Charlton J, Gulliford MC, D’Cruz DP. Incidence and mortality of relapsing polychondritis in the UK: A population-based cohort study. Rheumatology (Oxford) 2015;54:21812187.
18. Peterson LS, Nelson AM, Su WP, Mason T, O’Fallon WM, Gabriel SE. The epidemiology of morphea (localized scleroderma) in Olmsted County 1960-1993. J Rheumatol. 1997;24:7380.
19. Pillemer SR, Matteson EL, Jacobsson LT, et al. Incidence of physician-diagnosed primary Sjögren syndrome in residents of Olmsted County, Minnesota. Mayo Clin Proc. 2001;76:593599.
20. Uramoto KM, Michet CJ Jr, Thumboo J, Sunku J, O’Fallon WM, Gabriel SE. Trends in the incidence and mortality of systemic lupus erythematosus, 1950-1992. Arthritis Rheum. 1999;42:4650.
21. Wallin MT, Culpepper WJ, Coffman P, et al.; Veterans Affairs Multiple Sclerosis Centres of Excellence Epidemiology Group. The Gulf War era multiple sclerosis cohort: Age and incidence rates by race, sex and service. Brain 2012;135:17781785.
22. National Cancer Institute. Surveillance, Epidemiology and End Results Program. Available at: http://seer.cancer.gov/
. Accessed March 15, 2016.
25. Picha GJ, Singh N, Murphy DK. Natrelle silicone breast implant follow-up study: Demographics, lifestyle, and surgical characteristics of more than 5000 reconstruction subjects. Plast Reconstr Surg Glob Open 2015;3:e489.
26. Singh N, Picha GJ, Murphy DK. Natrelle silicone breast implant follow-up study: Demographics, lifestyle, and surgical characteristics of more than 50,000 augmentation subjects. Plast Reconstr Surg. 2016;137:7081.
27. Alderman A, Pusic A, Murphy DK. Prospective analysis of primary breast augmentation on body image using the BREAST-Q: Results from a nationwide study. Plast Reconstr Surg. 2016;137:954e960e.
28. Breslow NE, Day NE. Statistical Methods in Cancer Research: Volume II. The Design and Analysis of Cohort Studies, 1987.Vol. 82. Lyon, France: International Agency for Research on Cancer.
30. Xu J, Murphy SL, Kochanek KD, Bastian BA. Deaths: Final data for 2013. Natl Vital Stat Rep. 2016;64:1119.
31. Lipworth L, Tarone RE, Friis S, et al. Cancer among Scandinavian women with cosmetic breast implants: A pooled long-term follow-up study. Int J Cancer 2009;124:490493.
32. Lee IM, Cook NR, Shadick NA, Pereira E, Buring JE. Prospective cohort study of breast implants and the risk of connective-tissue diseases. Int J Epidemiol. 2011;40:230238.
33. Brody GS, Deapen D, Taylor CR, et al. Anaplastic large cell lymphoma occurring in women with breast implants: Analysis of 173 cases. Plast Reconstr Surg. 2015;135:695705.
34. Clemens MW, Miranda RN, Butler CE. Breast implant informed consent should include the risk of anaplastic large cell lymphoma. Plast Reconstr Surg. 2016;137:11171122.
35. de Jong D, Vasmel WL, de Boer JP, et al. Anaplastic large-cell lymphoma in women with breast implants. JAMA 2008;300:20302035.
36. Lazzeri D, Agostini T, Bocci G, et al. ALK-1-negative anaplastic large cell lymphoma associated with breast implants: A new clinical entity. Clin Breast Cancer 2011;11:283296.
37. Jewell M, Spear SL, Largent J, Oefelein MG, Adams WP Jr. Anaplastic large T-cell lymphoma and breast implants: A review of the literature. Plast Reconstr Surg. 2011;128:651661.
38. Carty MJ, Pribaz JJ, Antin JH, et al. A patient death attributable to implant-related primary anaplastic large cell lymphoma of the breast. Plast Reconstr Surg. 2011;128:112e118e.
39. Kim B, Roth C, Young VL, et al. Anaplastic large cell lymphoma and breast implants: Results from a structured expert consultation process. Plast Reconstr Surg. 2011;128:629639.
40. Kim B, Roth C, Chung KC, et al. Anaplastic large cell lymphoma and breast implants: A systematic review. Plast Reconstr Surg. 2011;127:21412150.
41. Popplewell L, Thomas SH, Huang Q, Chang KL, Forman SJ. Primary anaplastic large-cell lymphoma associated with breast implants. Leuk Lymphoma 2011;52:14811487.
42. Ramos-Gallardo G, Cuenca-Pardo J, Rodriguez-Olivares E, et al. Breast implant and anaplastic large cell lymphoma meta-analysis. J Invest Surg. 2017;30:5665.
43. Jacobsen PH, Hölmich LR, McLaughlin JK, et al. Mortality and suicide among Danish women with cosmetic breast implants. Arch Intern Med. 2004;164:24502455.
44. Villeneuve PJ, Holowaty EJ, Brisson J, et al. Mortality among Canadian women with cosmetic breast implants. Am J Epidemiol. 2006;164:334341.
45. Lipworth L, Nyren O, Ye W, Fryzek JP, Tarone RE, McLaughlin JK. Excess mortality from suicide and other external causes of death among women with cosmetic breast implants. Ann Plast Surg. 2007;59:119123; discussion 124–125.
46. Rubin JP, Landfair AS, Shestak K, et al. Health characteristics of postmenopausal women with breast implants. Plast Reconstr Surg. 2010;125:799810.
Supplemental Digital Content
Copyright © 2017 by the American Society of Plastic Surgeons