Decreased Use of Anti-Inflammatory Medications in Autoimmune Connective Tissue Disease Patients Following Breast Implant Removal: A National Analysis

© 2024 HMP Global. All Rights Reserved.

Any views and opinions expressed are those of the author(s) and/or participants and do not necessarily reflect the views, policy, or position of ePlasty or HMP Global, their employees, and affiliates.

Autoimmune connective tissue disorders (CTD) comprise a spectrum of diseases with variable presentation and epidemiology. These conditions include rheumatoid arthritis (RA), systemic lupus erythematous (SLE), Sjögren's syndrome, systemic sclerosis, dermatomyositis, vasculitis, and many other disorders that implicate unregulated inflammation within connective tissue infrastructure.^1,2 Collectively, autoimmune CTD are prevalent and notably affect women much more often than men.³

With a similarly disproportionate impact on women, breast cancer is the most common cause of cancer mortality in women around the world.⁴ As such, consideration of the impact of comprehensive breast cancer care on CTD symptomology in women afflicted by both breast cancer and CTD is essential. Breast implants continue to be the mainstay of post-mastectomy breast reconstruction, comprising 81% of all breast reconstruction procedures in 2019.⁵ However, surgeons and regulatory entities, including the US Food and Drug Administration, have engaged in discourse about the safety of breast implants, citing reports of breast implant illness (BII) entailing systemic symptoms including joint pain, fatigue, and dermatologic findings.⁶ Within this landscape, prior research has reported new autoimmune CTD onset following implant-based reconstruction as well as exacerbation of CTD symptomology in patients with prior CTD history following implant-based reconstruction.^7-11 For example, a landmark 2019 study concluded that patients with silicone breast implants demonstrated CTD rates almost double that of the general population.^12,13

While the symptomatology of BII is poorly understood, removal of breast implants has been shown to reverse BII in some cases.^11,14-17 Likewise, previous research has speculated that removal of implants in patients with autoimmune CTD may ameliorate disease burden.¹⁸ Amidst controversy about the safety and potentially inflammatory nature of breast implants, the present study utilizes a national claims repository to examine anti-inflammatory prescription patterns of patients with known autoimmune CTD diagnoses undergoing implant removal following implant-based breast reconstruction. We aimed to utilize the quantity of anti-inflammatory medications as a proxy for disease burden prior to and following implant explantation to examine the relationship between breast implants and autoimmune CTD symptomatology on a national scale.

Data Source and Study Cohort

Following a previous study involving assessment of breast reconstruction outcomes in patients with autoimmune CTD conducted by a subset of the present paper's authors, an additional retrospective investigation was performed utilizing the ClinformaticsData Mart Database.¹⁹ We used this de-identified claims database to assess demographic and postoperative characteristics of autoimmune CTD patients who underwent implant-based breast reconstruction and subsequent implant removal between January 1, 2003 and March 31, 2021 using Current Procedural Terminology (CPT) codes (Table 1). The Clinformatics Data Mart Database contains inpatient and outpatient claims, as well as demographic and prescription data, comprising 8,140,311,544 claims involving 69,067,157 unique patients. From this dataset, adult female patients continuously enrolled for 6 months or more prior to and after the index procedures were identified.

Table 1. CPT Codes Used to Identify Procedures in the Clinformatics Database

Diagnosis Claims

Insurance claims data were used to identify the year in which the index procedure was performed, patient age at the time of surgery, patient demographics, comorbidities, CTD diagnoses, and postoperative nonsurgical and surgical complications. As in our previous study, CTD diagnoses were identified using International Classification of Diseases, Ninth (ICD-9) and Tenth Revision (ICD-10) codes, including systemic lupus erythematosus, rheumatoid arthritis, systemic sclerosis, Sjögren's syndrome, sarcoidosis, spondyloarthritides, antiphospholipid syndrome, psoriatic arthritis, dermatomyositis, polymyositis, and large, medium, or small vessel vasculitides (Table 2).¹⁹ The Elixhauser Comorbidity Index, using ICD-9 (included only if no existing ICD-10 codes) and ICD-10 codes, evaluated comorbidity levels. As in our previous study, the postoperative complications assessed included nonsurgical complications (hematoma, seroma, wound dehiscence, deep vein thrombosis [DVT] or vascular complication, breast reconstruction deformity, postoperative infection, fat necrosis, tissue necrosis, and nonspecified complications of surgical care) and surgical complications (evacuation of hematoma, evacuation of seroma, other fluid collection or drainage, secondary closure of surgical wound or dehiscence, revision of breast reconstruction, tissue debridement, muscle debridement, or bone debridement) (Tables 1 and 3).^19,20

Table 2. ICD-9 and ICD-10 Codes Used to Identify Connective Tissue Disorders in the Clinformatics Database

Table 3. ICD-9 and ICD-10 Codes Used to Identify Complications in the Clinformatics Database

Prescription Data

Pharmaceutical data identified filled anti-inflammatory prescriptions during the study period using generic names (Table 4). Patients who filled 1 or more anti-inflammatory prescriptions during the preoperative (180 to 8 days before surgery), perioperative (7 days before to 7 days after surgery), and postoperative (8 to 180 days after surgery) periods were discerned and reported by mean (± standard deviation [SD]).

Table 4. Generic Drug Names Used to Identify Anti-Inflammatory Medications in the Clinformatics Database

Outcomes

The following outcomes were investigated in this study population: (1) filling at least one anti-inflammatory prescription in the preoperative period (180 to 8 days before the index surgery), and (2) the number of anti-inflammatory prescriptions filled in the postoperative period (8 to 180 days following the index surgery).

Covariables and Statistical Analysis

Wilcoxon signed-rank testing was used to assess if individual patients' preoperative anti-inflammatory prescription fills matched, exceeded, or fell short of their postoperative anti-inflammatory prescription fills.

Multivariable logistic regression calculated adjusted odds ratios (OR) for at least 1 anti-inflammatory prescription fill in the preoperative period for each independent variable included in the model with the following covariates: (1) age at surgery (modeled continuously); (2) Elixhauser index (0 or 1, 2, 3, 4+); (3) surgery year (2003-2007, 2008-2012, 2013-2016, 2017-2020); (4) education level (high school diploma or less, less than bachelor's degree, bachelor's degree or more); and (5) race and ethnicity (White, Black, Hispanic, Asian).

Multivariable linear regression calculated average differences in the number of anti-inflammatory prescription fills in the postoperative period attributable to each independent variable in the model with the following covariates: (1) age at surgery (modeled continuously); (2) number of preoperative prescription fills (modeled continuously); (3) number of perioperative prescription fills (modeled continuously); (4) the presence or absence of at least 1 complication associated with the index procedure; (5) Elixhauser index (0 or 1, 2, 3, 4+); (6) surgery year (2003-2007, 2008-2012, 2013-2016, 2017-2020); (7) education level (high school diploma or less, less than bachelor's degree, bachelor's degree or more); and (8) race and ethnicity (White, Black, Hispanic, Asian).

Schapiro-Wilk testing was used to determine whether continuous variables were normally distributed. Wilcoxon signed-rank and multivariate regression testing were used for statistical analysis. P values of <.05 were considered statistically significant. All analyses were completed using Stata, version 16.1 (StataCorp LLC).

Utilizing a Wilcoxon signed-rank test, we found that preoperative counts of anti-inflammatory prescription fills significantly exceeded postoperative counts (z = 4.54; P < .001) (Figure). Table 5 describes the demographic characteristics of the study cohort. The mean age of the cohort was 56 (± 12) years, with 85% of patients identified as White. The mean number of fills for the preoperative, perioperative, and postoperative periods were 59 (± 167), 15 (± 37), and 56 (± 178), respectively. Most patients (58%) experienced 1 or more complications. There were high levels of comorbidities, with 84% of patients having an Elixhauser index score of 4 or higher. Most patients (76%) held less than a bachelor's degree, and 19% of patients held a high school diploma or less. Patients most often came from households with a collective income of $100,000 or higher (42%), whereas 19% of patients belonged to a household with an annual income under $40,000.

Figure. Average number of anti-inflammatory prescription fills per day in the preoperative period (-180 to -8 days), the perioperative period (-7 to +7 days), and the postoperative period (+8 to +180 days). A Wilcoxon signed-rank test determined that preoperative counts of anti-inflammatory prescription fills significantly exceeded postoperative counts (z = 4.54; P < .001).

Table 5. Characteristics of the Study Cohort (n = 1105).

Patient factors associated with filling at least one prescription in the preoperative period among the patient cohort are described in Table 6. Older age was associated with lower odds of filling one or more preoperative prescriptions (OR 0.98; P < .001). Relative to those who underwent the index surgery in 2003-2007, those who underwent surgery in 2013-2016 were more likely to fill at least 1 preoperative prescription (OR 1.58; P = .028). Those of higher education levels (bachelor's degree or more) were less likely to fill at least 1 prescription preoperatively than those with a high school diploma or less (OR 0.53; P = .005). Relative to White patients, Asian patients were less likely to fill a preoperative prescription (OR 0.35; P = .040).

Table 6. Patient Factors Associated With Filling 1+ Prescriptions in the Postoperative Period 

Variables associated with average differences in the number of filled postoperative prescriptions are delineated in Table 7. Increased postoperative prescription fillings were associated with greater preoperative and perioperative prescription fills (P < .001). However, among those patients who filled at least 1 prescription preoperatively, higher education levels (bachelor's degree or more) were associated with increased postoperative prescription fills (P = .035).

Table 7. Patient Factors Associated With Quantity of Prescription Fills in the Postoperative Period

Our national claims database study suggests that among patients with an established CTD, removal of breast implants preceded a decrease in postoperative prescription fills of anti-inflammatory medications and thus, by proxy, implant removal may be related to decreased disease burden. Our data elucidates that younger individuals as well as individuals who underwent surgery between 2013 and 2016 were more likely to fill a preoperative prescription. We hypothesize this is due to the increased exposure to social media and the impact advancing technology has on the field of aesthetic reconstructive surgery in recent years.²¹ Likewise, our data shows elevated postoperative prescription fills for individuals with higher education levels possibly due to increased patient comprehension regarding symptom management postoperatively.

Breast Implant Illness in the Context of Connective Tissue Disorders

Symptoms relating to breast implant illness involve multiple body systems, including the musculoskeletal, immune, gastrointestinal, integument, psychological, cardiorespiratory, and nervous systems.^16,17 In the current literature, many of these symptoms arising from different body systems relate to CTD; these include arthralgia, arthritis, myalgia, sleep disturbances, autoantibodies onset, miscarriage, chronic fatigue, persistent infections, Raynaud phenomenon, and symptoms of autoimmune diseases such as scleroderma, Sjögren's syndrome, or rheumatoid arthritis.^10,22

Reports of a relationship between BII and CTD have been documented for many decades, with one 1996 study demonstrating an increased relative risk of CTD in women with breast implants compared with women without breast implants.¹⁵ More recently, this association has sparked renewed interest.^11,13,15,23 As of 2018, Watad et al showed in their cross-sectional study a strong association between breast implants and CTD such as Sjögren's syndrome and systemic sclerosis.²⁴ Although multiple studies have investigated the relationship between BII and CTD, a consensus on the existence and strength of the association between these 2 conditions has not been achieved.^25,26

Alternative views seem to support a more neutral relationship between the 2 variables. Sánchez-Guerrero et al demonstrated no relationship between breast implants and self-reported symptoms of CTD for women with breast implants.²⁷ This study utilized symptoms reported by the American College of Rheumatology for systemic lupus erythematosus, systemic sclerosis, Raynaud phenomenon, photosensitivity arthritis, morning stiffness, xerostomia, dry eyes, and positive rheumatoid-factor tests. Similarly, Singh et al studied 55,279 women with breast implants over a 5-year follow-up period and concluded no increased risk of CTD when compared with national norms.²⁸ While the impact of silicone-based implants on systemic inflammation has not been entirely elucidated, the general symptomatology of BII maintains relevance amidst preexisting CTD.

Literature Regarding Symptoms After Implant Removal

The topic of BII is largely limited to case reports, series, reviews, and retrospective cohort studies, which provide a limited level of clinical evidence regarding the relationship between implants and systemic symptoms. For instance, a consistent theme throughout the literature is that of the subjective symptomatology that women with BII experience. Some of the most commonly cited symptoms include fatigue, myalgia, arthralgia, morning stiffness, sicca, alopecia, and night sweats.^{17,18,22,29-38} Objective clinical signs noted by physicians included ipsilateral axillary/cervical/inguinal lymphadenopathy, joint swelling/tenderness, and trapezius/paraspinal musculature tenderness.^30-32,39

Following explantation, the literature appears to justify both the attenuation and persistence of symptomatology. However, the interpretation of these studies requires an understanding of the heterogeneous presentation of symptomatic patients to isolate confounders. On the one hand, vague systemic symptoms, such as those described above, can show overwhelming resolution with explantation, with persistence of results in 50% of explant patients in some cases.^18,22,29-39 Conversely, for those with a diagnosable autoimmune or rheumatologic condition, such as rheumatoid arthritis, systemic lupus erythematosus, or fibromyalgia, the effects of explantation tend to be transient and are difficult to ascertain given the concurrent usage of immunomodulators.^{18,22,29,31,32,37} Furthermore, this lack of symptom resolution may relate to the persistence of silicone microparticles within the patient or to chronic sensitization by the immune system.^30-32,39

Among patients with no history of CTD who report improved symptoms following explantation, duration of implant exposure may play an important role in assessing prognosis or response. Experiencing BII remains rare in the overall patient cohort undergoing implant-based reconstruction. However, in patients with a longer duration of implant exposure, studies demonstrate not only an increased severity of BII symptoms but a lower likelihood of resolution following explantation.^31,32,36 In effect, these findings make early recognition and explantation paramount in preventing progression and decreasing autoantibody sensitization.

The Clinformatics Data Mart Database enables a powerful retrospective national depiction of CTD presentation amidst breast implant removal. Even so, we have several notable limitations. First, autoimmune CTD are assessed in this study collectively as a binary variable, but these conditions have important discrepancies in presentation, treatment, and outcomes. Future studies should assess CTD with more granularity to investigate the existence of relationships between breast implants and specific CTD conditions. Second, the Clinformatics warehouse includes information on prescriptions filled by patients, which do not necessarily represent the quantity of anti-inflammatory prescriptions consumed by the patient. Additionally, quantity of anti-inflammatory prescriptions is certainly an imperfect proxy for CTD disease burden. Third, there may be inherent heterogeneity regarding medication use and symptom relief following administration. Additionally, important factors relating to social determinants of health—such as insurance status, transportation, and medication costs—may influence a patient's ability to fulfill prescriptions. However, household income, education level, and race were reported.

Our study demonstrates a potential amelioration of autoimmune CTD disease burden following breast implant removal, given a decrease in filled anti-inflammatory prescriptions post-explantation, relative to pre-explantation quantities. While further research with greater granularity and more patient and physician perspectives is required, our findings may support additional pre-reconstruction consideration of the suitability of breast implants in patients with autoimmune CTD. Additionally, this study's results may substantiate expedited explantation in CTD patients with implants who develop new or worsening symptoms related to chronic inflammation.