Silicone Leakage From Breast Implants and Its Association With Capsular Contracture in 657 Patients

Abstract

Background

Capsular contracture is a common complication following breast reconstruction and augmentation, but there is limited evidence for the association between silicone leaking from the implant and capsular contracture.

Objectives

To determine the association between silicone leaking into the surrounding fibrous capsule and the risk of capsular contracture.

Methods

This multicenter cross-sectional study included patients undergoing revisional surgery for any indication following breast reconstruction or augmentation with silicone implants. Silicone in the fibrous capsule was quantified using histopathological analysis, and capsular contracture was diagnosed using the Baker classification. Multivariable logistic regression was used to assess the association between the amount of silicone leakage and the predicted risk of capsular contracture. The association was tested in a sensitivity analysis excluding potential confounders and misclassifications.

Results

Among 657 included patients (1147 breasts), 272 patients (346 breasts) had either unilateral or bilateral capsular contracture. Capsular contracture was significantly associated with the amount of silicone in the fibrous capsule (P < .001), independent from the time of implantation (P < .001). The excess risk of capsular contracture increased rapidly by 12% (95% CI, 5.0-18), going from 0.00 mL (0-32nd percentile) to 0.26 mL (60th percentile) of silicone in the fibrous capsule. The association was confirmed in the sensitivity analysis of 421 patients (654 breasts) with intact implants.

Conclusions

Even small amounts of silicone leakage are associated with a significantly increased risk of capsular contracture. This highlights the importance of using breast implants with low silicone leakage rates, such as highly cohesive implants, to prevent capsular contracture.

Level of Evidence: 3 (Therapeutic)

Breast implants are widely used for breast reconstruction and augmentation, with ∼2.2 million procedures performed globally each year.^1,2 Unfortunately, many patients develop capsular contracture, affecting 3.6% to 19% of patients after breast augmentation^3-7 and 3.3% to 38% of patients after breast reconstruction^3,5-9 within 10 years of surgery. Capsular contracture is characterized by a thickening and contraction of the fibrous capsule that surrounds the breast implant, leading to hardening and deformation of the breast, and sometimes chronic pain.¹⁰ The current treatment is to perform a partial or total capsulectomy followed by implant replacement.^11,12 However, the risk of capsular contracture recurrence remains high despite surgical treatment.^12,13

Silicone leakage from the breast implant is a suspected risk factor for the development of capsular contracture, but the association has never been tested in a large clinical study.^14,15 Until now, the association between capsular contracture and silicone leakage has mostly been studied in cases of implant rupture.^16,17 However, even intact silicone breast implants have been shown to leak silicone through the implant shell over time.^17-20 It is currently unknown whether silicone leakage from intact implants increases the risk of capsular contracture because the amount of silicone leakage from intact implants is lower and different from what is seen in ruptured implants.²¹ As a result, the general opinion regarding the lifespan of a breast implant is mostly based on rupture rates. According to the American Society of Plastic Surgeons, many patients have their implants replaced after ∼10 years because of an increased risk of implant rupture and capsular contracture.²²

Understanding the effect of silicone leakage is crucial for improving patient outcomes, including guidance on the choice of breast implants, such as saline vs silicone fill or the degree of silicone cohesiveness. Moreover, if silicone leakage into the fibrous capsule increases the risk of capsular contracture, it may be relevant to test strategies for removing silicone from the breast implant capsule, such as a partial or total capsulectomy during implant replacement in the case of implant rupture.

In this cross-sectional study, we aimed to investigate the association between the amount of silicone leakage in the fibrous capsule and the risk of capsular contracture in patients undergoing implant replacement after implant-based breast reconstruction or breast augmentation. First, we hypothesized that increasing the amount of silicone in the breast implant capsule would be associated with a higher risk of capsular contracture independent from the time of implantation. Secondly, we performed a sensitivity analysis of the association with the exclusion of potential confounders and misclassifications.

METHODS

This multicenter cross-sectional study was approved by the Regional Ethics Committee (H-20015276) and the Danish Data Protection Agency (P-2020-176 and P-2020-177) and was conducted according to the principles of the Declaration of Helsinki. Oral and written informed consent was obtained from all patients before inclusion in the study according to Danish legislation. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guidelines were followed in this study.²³

Participants

We recruited patients scheduled for replacement or removal of their silicone breast implants for any indication between September 2019 and July 2023 at 5 public and private plastic surgery departments in Copenhagen. Eligible study participants were patients aged 18 years or older who had previously undergone implant-based breast reconstruction or breast augmentation with silicone implants. We excluded patients who had other implant fillings than silicone, such as saline, and patients with double capsule formation.²⁴ Other exclusion criteria were ongoing breast infection, pregnancy, and breastfeeding within 3 months before surgery. Patient and surgical characteristics were extracted from the patients’ medical records and intraoperative observations.

Exposure

The exposure of interest was the amount of silicone in the breast implant capsule. We obtained a biopsy from the fibrous capsule surrounding the breast implant during the revisional surgery to measure the amount of silicone leakage with histopathological analysis. All biopsies were obtained from the anterior-inferior part of the fibrous capsule adjacent to the inframammary fold. We performed histopathological quantification of silicone on digitized hematoxylin and eosin (H&E)-stained sections in ImageJ (version 1.53k; US National Institutes of Health, Bethesda, MD) using a previously described method measuring the amount of both extracellular silicone in vacuoles and intracellular silicone in foam cells.¹⁹ Observers were blinded for all clinical variables, including whether the patient had capsular contracture. The analysis is described in detail below and depicted in Figure 1.

1. Two regions of interest (ROIs) from the area with the highest and lowest amounts of silicone were selected. Each ROI covered the entire thickness of the breast implant capsule (µm) in a section of 700 µm parallel to the breast implant surface.

2. The area of silicone particles (µm²) in the 2 ROIs was quantified using the threshold function in ImageJ. All artifacts and blood vessels were manually shaded.

3. The proportion of silicone in the breast implant capsule was then calculated, and an average between the 2 ROIs was used:

   $Siliconeratio(%)=AreaROISilicone(μm2)AreaROICapsule(μm2)$
4. To estimate the capsule volume, the radius of the implant was first calculated using the formula for the volume of a hemisphere:

   $Radius(cm)=3×Implantvolume(mL)2×π3$
5. Using the radius, the implant surface area was then calculated:

   $Implantsurfacearea(cm2)=3×π×Radius(cm)$
6. The volume of the capsule was then estimated:

   $Capsulevolume(mL)=Implantsurfacearea(cm2)×Capsulethickness(cm)$
7. Finally, the total amount of silicone in the breast implant capsule was calculated:

   $Totalsiliconeamount(mL)=Capsulevolume(mL)×Siliconeratio(%)$

Outcomes and Definitions

The primary outcome was capsular contracture, diagnosed preoperatively using the Baker classification system (I-IV).¹⁰ Breasts classified as Baker III or Baker IV were defined as having capsular contracture, whereas Baker I and Baker II were considered healthy controls. Time of implantation was defined as the number of years from the insertion of the first breast implant to sample acquisition.

Statistical Analysis

Descriptive statistics were presented as proportions, means with standard deviations, or medians and interquartile ranges (IQRs) depending on data distribution. Patient demographics were stratified based on the amount of silicone leakage into tertiles, presenting potential baseline differences among patients with low, moderate, and high silicone leakage. Baseline differences were tested using Kruskal–Wallis rank sum tests for continuous variables and χ² tests for categorical variables.

In the primary analysis, we investigated the association between the amount of silicone leakage and the risk of capsular contracture. This analysis was conducted using univariable and multivariable logistic regression models, adjusting for the time of implantation. We used likelihood-ratio tests to compare the multivariable models with silicone leakage and time of implantation mapped linearly and with restricted cubic splines and to test for the presence of interaction. The effects of silicone leakage and time of implantation were significantly nonlinear but without evidence of interaction, resulting in a final model using restricted cubic splines with 4 knots. All results were plotted as differences between the baseline risk of capsular contracture with no silicone leakage and increasing amounts of silicone leakage at the median implantation time.

Sensitivity Analysis

We conducted a sensitivity analysis excluding clinical parameters with unclear associations, for example, potential confounders, mediators, and reverse causality, as well as misclassifications to evaluate the robustness of the association between silicone leakage and capsular contracture. Specifically, we excluded cases with implant rupture to minimize the risk of reverse causality, as capsular contracture theoretically could be a cause of implant rupture.¹⁶ Additionally, we excluded breasts with previous implant replacement to improve accuracy on the time of implantation and exclude the influence of previous interventions, such as capsulectomy and previous implant rupture.^12,18 Breasts that had received radiotherapy were also excluded because radiotherapy is a known risk factor for capsular contracture, which may also compromise the implant shell and increase silicone leakage.^25,26

To enhance diagnostic accuracy, we confirmed the clinical capsular contracture status using validated histopathological scoring systems.^27,28 The scoring systems are based on semiquantitative assessments of histopathological parameters significantly associated with capsular contracture, including the thickness of the collagen layer, organization of collagen fibers, calcification, and cellularity in the collagen layer. Breasts graded as Baker III or Baker IV were excluded as false positives if the histopathological score indicated a probability of capsular contracture <50%, and breasts graded as Baker I or Baker II were excluded as false negatives if the histopathological score showed a probability of capsular contracture of >80%. An alpha level of 5% was used to evaluate P-values. All tests and plots were performed in R, version 4.4.1.

RESULTS

The study cohort included 657 patients contributing with biopsies from 1147 breasts (Figure 2). All patients were female with a median age of 47.2 years (IQR 38.5-55.4; range, 19.8-84.9), median BMI of 22.6 kg/m² (IQR 20.5-24.9; range, 15.4-40.2), and a median time of implantation of 14.0 years (IQR 7.1-19.7, 0.1-51.0). Of the 657 patients, 122 patients (19%) had undergone implant-based breast reconstruction, and 535 patients (81%) had undergone breast augmentation. The patients were divided into tertiles based on the amount of silicone in the breast implant capsule: low leakage group, median 0.00 mL (IQR 0.00-0.00), moderate leakage group, median 0.10 mL (IQR 0.04-0.23), and high leakage group, median 2.20 mL (IQR 0.92-5.39). The median time of implantation was 9.0 years (IQR 3.3-14.0) for the low leakage group, 14.0 years (IQR 7.1-18.0) for the moderate leakage group, and 18.0 years (IQR 14.0-22.0) for the high leakage group (P < .001). There were more ruptured implants in the high leakage group compared with the low and moderate leakage groups (low leakage: 3/383 implants, moderate leakage: 23/382 implants, and high leakage: 212/382 implants; P < .001). Moreover, there were more implants with highly cohesive silicone in the low leakage group compared with the moderate and high leakage groups (low leakage: 159/383 implants, moderate leakage: 99/382 implants, and high leakage: 54/382 implants; P < .001). There were no significant differences in the number of breasts with a history of radiotherapy (P = .27) or previous implant replacement (P = .24) within the 3 silicone leakage groups. Details of the patient and implant characteristics are provided in Table 1.

Silicone Leakage and Capsular Contracture

Capsular contracture was diagnosed in 21% (82/383) of the breasts in the low leakage group, 24% (93/382) in the moderate leakage group, and 45% (171/383) in the high leakage group (P < .001). In our unadjusted analysis, we found that the risk of capsular contracture significantly increased with higher amounts of silicone leakage (P < .001). In our multivariable analysis adjusting for time of implantation, we found a similar significant risk of capsular contracture with increasing amounts of silicone (P < .001). Figure 3 shows the adjusted analysis with excess risk of capsular contracture after 14 years of implantation as a function of the amount of silicone in the breast implant capsule. In the low leakage group, there was no silicone in the breast implant capsule and, therefore, no excess risk of capsular contracture because of silicone leakage. In the moderate leakage group, the excess risk of capsular contracture rapidly increased from 0.1% (95% CI, 0.0-0.1) to 12% (95% CI, 5.0-18) when the amount of silicone increased from 0.001 to 0.26 mL corresponding to the 33rd and 60th percentiles. Hereafter, the excess risk of capsular contracture reached a plateau until the 90th percentile (4.13 mL) increasing to 23% (95% CI, 13-34) for the 95th percentile of silicone (9.05 mL; Table 2).

In the multivariable analysis, we found that the time of implantation was a highly significant risk factor for capsular contracture (P < .001) which resulted in an overall and proportional drop in the risk of capsular contracture when adjusting for implantation time (Figure 4).

Sensitivity Analysis

The sensitivity analysis included 421 patients with biopsies from 654 breasts, excluding 238 breasts with implant rupture, 184 breasts with previous implant replacement, and 29 breasts with a history of radiotherapy. Moreover, 80 breasts were excluded because of a risk of misclassification based on the histopathological diagnosis, of which 64 breasts were identified as false positives and 16 breasts were identified as false negatives. The sensitivity analysis confirmed the association between capsular contracture and the amount of silicone leakage independent of the time of implantation in patients with intact implants (P < .001). We found a similar dose–response relationship but with a more moderate increase in the excess risk of capsular contracture going from 0 to 1 mL of silicone in the breast implant capsule. However, the excess risk of capsular contracture in the sensitivity analysis reached an overall higher level with 35% (95% CI, 18-51) at 5 mL of silicone (Figure 4, orange line) compared with 18% (95% CI, 8.8-27) in the adjusted analysis (Figure 4, blue line).

DISCUSSION

The authors of this study fill an important knowledge gap by establishing a significant association between silicone leakage from breast implants and the development of capsular contracture in a large clinical cohort. Our findings indicate that even small amounts of silicone leakage are associated with a significantly increased risk of capsular contracture. Specifically, we observed a rapid increase in risk from the 0 to 33rd percentile with no detectable silicone leakage to the 60th percentile, where there was a 12% increase in the risk of capsular contracture. This was confirmed in a sensitivity analysis, excluding potential confounders such as ruptured implants.

Our study is aligned with previous studies that have shown that silicone particles can migrate into the fibrous capsule surrounding breast implants.^{17,18,20,29-42} However, the clinical significance and a dose–response relationship have not previously been established.^14,43,44 Previous studies have often been based on animal models^43,45,46 or older generations of breast implants that are no longer available on the market.^47-51 Additionally, the current understanding of the role of silicone leakage is largely centered on implant rupture as the critical factor for silicone leakage, assuming that the presence of large amounts of silicone is necessary to drive the development of capsular contracture.¹⁴ Contrary to these assumptions,^52,53 we found that even small amounts of leaked silicone from intact implants were associated with a substantial increase in the risk of capsular contracture. This challenges the view that only ruptured implants pose a risk and highlights the potential risks associated with even every low levels of silicone leakage. A plausible mechanism is that silicone, as an inorganic material, triggers and sustains a chronic inflammatory response in the breast implant capsule, which in turn leads to excessive fibrosis, ultimately resulting in capsular contracture.^54-56 As shown in the sensitivity analysis, the overall excess risk of capsular contracture was higher in patients with intact implants compared with the primary analysis including patients with ruptured implants. This suggests that for similar amounts of silicone, a slow accumulation from intact implants over time has a greater impact on the risk of capsular contracture than a rapid release from ruptured implants.

The authors of previous studies investigating the presence of silicone particles in the breast implant capsule have used various approaches, including manual counting of silicone particles,^20,57 Stimulated Raman Scattering microscopy,^58,59 and Modified Oil Red O staining.^14,18 We chose this semi-automated approach on H&E-stained sections using the threshold function in ImageJ with manual shading of artifacts and blood vessels to allow for analysis of the large sample size in our study.

Our study emphasizes the importance of minimizing silicone leakage as a potential strategy to prevent capsular contracture. We have previously shown that selecting highly cohesive implants can significantly reduce the amount of silicone leakage compared with low-cohesive implants.¹⁹ Saline-filled breast implants could be chosen as an alternative because they have been found not to leak silicone into the breast implant capsule.²⁰ However, they are known to feel less natural and still carry a risk of capsular contracture.⁴⁸ Future studies should investigate whether the effects of silicone leakage on capsular contracture observed in this study are consistent across different patient demographics (eg, breast augmentation vs reconstruction), and how silicone is affected by the type of implant, such as smooth implants, which were underrepresented in our cohort.

Moreover, silicone leakage has been proposed as a trigger for breast implant-associated anaplastic large cell lymphoma (BIA-ALCL) which, similar to capsular contracture, is hypothesized to be caused by an exaggerated immune response to the breast implant.^56,60-62 There were no cases of BIA-ALCL in this study.

Currently, there are no guidelines on how to address the fibrous capsule during implant replacement in breasts without capsular contracture if silicone leakage is evident either from implant rupture or gel bleeding through the intact shell. Our results indicate that it may be important to surgically address the silicone embedded in the fibrous capsule during implant replacement to prevent the silicone that leaked from the old implant from causing capsular contracture around the new implant. This could be achieved by either partial or total capsulectomy, depending on the clinical case. This hypothesis is supported by a study by Swanson who found a significantly higher risk of capsular contracture recurrence following the replacement of ruptured or leaking implants compared with intact nonleaking implants.¹⁶ However, this hypothesis must be tested in future prospective interventional studies that weigh benefits against risks, such as hematoma.

Limitations

This study has several limitations that should be considered. The cohort consisted of patients undergoing implant replacement, which may introduce selection bias as these patients are expected to have more implant-related complications, such as implant rupture, compared with the general population with breast implants. This could potentially lead to an overestimation of the effect of silicone leakage. Unknown confounding factors that influence both silicone leakage and the risk of capsular contracture may also have been present, which we cannot account for in our analysis. Moreover, the accuracy of our silicone quantification was limited by the analysis relying on an extrapolation from a single biopsy, which may not fully represent the total amount of silicone leakage throughout the breast implant capsule. Future studies may consider a multi-biopsy protocol similar to Danino et al;^20,57 however, we tried to account for this potential variation by standardizing the location of the biopsy and by having a very large sample size, which increases the robustness of our findings. Another limitation is the diagnostic accuracy of the Baker classification. The Baker classification system is the most common and broadly accepted method for diagnosing capsular contracture, but it is known to have relatively low reproducibility.⁶³ We accounted for this by including a large number of patients in our cohort and by removing potential misclassifications of the capsular contracture status by only including cases and controls that were confirmed with a validated histopathological scoring system in the sensitivity analysis.²⁷ Moreover, although we find a strong association between silicone leakage and capsular contracture, we cannot infer causality because of the cross-sectional study design. Future prospective studies are necessary to confirm the association and thus strengthen the case for causation. Finally, we did not have access to data regarding the timing of the implant rupture, for example, from ultrasound or mammography which can lead to variation in the time at which patients have been exposed to high levels of silicone.

CONCLUSIONS

In this cross-sectional study of patients with silicone breast implants, we found that the amount of silicone leakage from breast implants to the surrounding fibrous capsule is significantly associated with the risk of capsular contracture. This was independent from other known risk factors, such as time of implantation, implant rupture, and previous implant replacement. These findings highlight the importance of minimizing silicone leakage through the selection of implants with low rates of silicone leakage, such as highly cohesive implants, and in the development of new implants. This is a call for the industry to improve shell barriers or other techniques to decrease gel bleed and researchers to study the long-term performance. In addition, currently, most highly cohesive implants are textured, and because of the risk of BIA-ALCL associated with textured implants, a wider selection of highly cohesive smooth implants is warranted. Furthermore, our findings provide a potential rationale for procedures that remove capsule tissue with embedded silicone such as partial or total capsulectomy as a strategy to decrease the risk of patients later developing capsular contracture, although this hypothesis must be tested in future interventional studies.

Acknowledgments

The authors thank all plastic surgeons and research assistants who helped include patients and obtaining biopsies for the study: Drs Alessandro Venzo, Anders Ulrik, Anna Louise Norling, Anne Katrine Lorentzen, Birgitte Christensen, Bo Jønsson, Bo Sonnich Rasmussen, Camilla Asklund, Caroline Gjørup, Charlotte Caspara Uth, Christian Lang, Christian Sneistrup, David Hebbelstrup, Deborah-Leigh Day, Dogu Aydin, Emil Villiam Holm-Rasmussen, Faye Sarmady, Hans Erik Siersen, Helle Sjøstrand, Hoda Khorasani, Ida Felbo Pold, Jais Oliver Berg, Jakob Astrup, Jakob Felbo Paulsen, Kristina Boldt Strålman, Lea Juul Nielsen, Lena Carstensen, Lilan Elsebet Engel, Louise Bondo, Line Læssøe, Linnea Schmidt, Linnea Langhans, Lisa Toft Jensen, Liselotte Sabroe Ebbesen, Marie Louise von Sperling, Marie Brinch-Møller Weitemeyer, Mette Wolthers, Michael Prangsgaard Møller, Mitra Sepehri, Navid Mohamadpour Toyserkani, Pavia Lumholt, Peter Stemann Andersen, Rami Mossad Ibrahim, Susanne Lambaa, Tina Tos, Trine Foged Henriksen, Ms Leonia Falck Rasmussen, Louise Eggers Rasmussen, Caroline Marie Torp Nygaard, Mikela Kring, Jennifer Jing Liu, Emma Bolette Odgaard, Liv Hart, Sif Birch Mathisen, and Mr Adam Mandrup Timmermann.

Disclosures

The authors declared no potential conflicts of interest with respect to the research, authorship, and publication of this article.

Funding

The work was funded by the Rigshospitalet Research Fund (Copenhagen, Denmark), The Capital Region of Denmark Research Fund (Hillerød, Denmark), the Novo Nordisk Foundation (grant numbers 052198 and 0058322; Bagsvaerd, Denmark), and the Lundbeck Foundation (grant number R413-2022-621; Copenhagen, Denmark). The funders had no role in the study design, data collection and analysis, decision to publish, or preparation of the manuscript. The funding sources were not involved in study design, collection, analysis, or interpretation of data, writing the manuscript, or the decision to submit the manuscript for publication.

REFERENCES

Author notes