Unraveling Factors Influencing Early Seroma Formation in Breast Augmentation Surgery

Transcription

1 Breast Surgery Unraveling Factors Influencing Early Seroma Formation in Breast Augmentation Surgery Aesthetic Surgery Journal 2017, Vol 37(3) The American Society for Aesthetic Plastic Surgery, Inc. Reprints and permission: DOI: /asj/sjw196 Marcos Sforza, MD; Rodwan Husein, MD; Connor Atkinson, MD; and Renato Zaccheddu, MD Abstract Background: It is often assumed that seroma formation trails closely behind with incidence rates reported at 1 to 2%. Seroma is highly problematic for both the surgeon and patient and results in both patient anxiety and discomfort, succeeded by frequent outpatient visits, follow-up treatment, increased costs, and potentially hampered aesthetic outcomes. Consequently, it is now more important than ever to study seroma and to assess its pathophysiology and mechanisms of prevention. Objectives: The aim of this study was to isolate and identify risk factors that may be associated with early seroma formation. Methods: The authors reviewed 539 female patients who had undergone bilateral breast augmentation with silicone cohesive gel implants in a period of 12 months. Five possible risk factors were isolated for analysis: patient s age, body mass index (BMI), smoking habit, implant pocket position, and implant size. A total of 15 patients developed early seromas within the one-year postoperative period. Results: Using exact logistic regression with the independent variables treated as binary variables, we found that smoking, BMI, and pocket are associated with increased risk of seroma while we cannot reject the hypothesis that pocket size and age do not affect the development of seroma at 5% significance level. Conclusions: A high BMI, large implant size, submammary pocket, and smoking are factors significantly associated with seroma development whilst age is not. Smoking however was found to be the most detrimental factor as it significantly amplified the effects of other variables. Level of Evidence: 2 Editorial Decision date: September 28, 2016; online publish-ahead-of-print December 7, Breast augmentation surgery is a well-established practice with high success rates and a low risk of complication. Last year over 300,000 women in the United States alone underwent cosmetic surgery to aesthetically enhance the size, form, and feel of their breasts. 1 Despite the high success rates, a number of complications can arise including hematoma, infection, capsular contraction, and seroma to name but a few. 2 Currently, capsular contracture and hematomas are the most frequently occurring complications that lead to revision surgery. However, seroma formation trails closely behind with incidence rates reported at 1 to 2% in some series. 3 Seroma can present both early and late, with the latter being recently linked with malignant growths and prolonged onset. 4 Early seroma formation is defined as periprosthetic fluid accumulation within the first postoperative year, whereas the late form is any moment beyond that time. Prof. Sforza is responsible for the Elective Internship in Plastic Surgery, Dolan Park Hospital, Bromsgrove, UK; and is an Examiner of the Royal College of Surgeons of Edinburgh. Mr. Husein is a Senior House Officer, Royal Salford Hospital, Manchester, UK. Mr. Atkinson is a medical student, Leeds Medical School, Leeds, UK. Dr Zaccheddu is a Plastic Surgeon, Dolan Park Hospital, Bromsgrove, UK. Corresponding Author: Prof. Marcos Sforza, Dolan Park Hospital, Stoney Lane, B60, 1LY, Bromsgrove, England. marcos@marcossforza.com

2 302 Aesthetic Surgery Journal 37(3) Seroma is highly problematic for both the surgeon and patient and results in both patient anxiety and discomfort, succeeded by frequent outpatient visits, follow-up treatment, increased costs, and potentially hampered aesthetic outcomes. Consequently, it is now more important than ever to study seroma and to assess its pathophysiology and mechanisms of prevention. The aim of this study was to isolate and identify risk factors that may be associated with early seroma formation. As matter of reference, whenever the authors refer to seroma formation it is implicit that they are referring to early seromas unless stated otherwise. METHODS In order to assess the incidence of seroma, we performed a retrospective clinical study involving consecutive 539 female patients, undergoing primary breast augmentation surgery with silicone cohesive gel implants between January 2006 and February All surgeries were performed in a private UK hospital, by the two first authors. This clinical audit followed the Declaration of Helsinki guidelines and a written consent for the outlined procedure was obtained from all patients. All patients undergoing primary breast augmentation in that period were included in this study with no exclusion criteria present. Patients were systematically reviewed at 7, 14, and 28 day intervals and whenever necessary for up to 1 year. The smoking habits of all patients were also recorded. For the purpose of analyzing the data, the authors isolated 5 possible variable risk factors: patient s age, body mass index (BMI) (more or less than 30), smoking habit, implants pocket, and size of the implants (more or less than 350 cc). The data were analyzed using exact logistic regression to examine if any of the risk factors can be used to predict the development of seroma. Two independent statisticians reviewed the data. In general, logistic regression can be used when the relationship between the binary/ dichotomous outcome variable and the independent variables can be modeled with the nonlinear logistic function. Nevertheless, regular logistic regression is not suitable for our data set because it applies asymptotic methods that are unreliable with sparse data (ie, small number of patients with seroma). This is why we used exact logistic regression using the Markov Chain Monte Carlo algorithm, which does not depend on asymptotic results as it is based on enumerating the exact distributions of sufficient statistics for parameters of interest in a logistic regression model. 5,6 Our aftercare scheme covers the patient for up to 3 years and entitles them to free revisions. The patients need to return within one year to renew for the next 2 years, otherwise the aftercare is voided. Patients that fail to attend the mandatory appointments as previously stated are persistently contacted and informed about the risks of losing the aftercare. All 539 have returned within the first postoperative year. Surgical Technique The procedures were performed by the same surgical team, with the same brand of textured implants (Eurosilicone, Cédex, France), same equipment (Valleylab Diatherm, Minneapolis, MN), same hospital facilities, and same surgical technique. All patients had the procedure under general anesthesia. During the procedure, the breast implant pocket was created using a Tebbetts monopolar diathermy forceps to minimize surgical trauma and no blunt dissection was performed. 7 All patients received their implants via an inframammary incision, in either the subglandular or subpectoral plane. No antibiotics irrigation neither drains were used. All patients were administered compression socks (Preventex, Leicestershire, UK) and underwent surgery using a prophylactic pneumatic DVT system (Flowtron, Bedfordshire, UK). All participants also received prophylactic enoxaparin. Patients with BMIs of up to 30 were administered 20 mg whilst those with BMIs over 30 were given 40 mg. Enoxaparin was delivered as a single dose at the end of the procedure. All surgical procedures were completed within the hour. All patients had compressive sports bras dressed in the operative room, remaining in situ for 6 weeks in all patients. Hospital stay was 1 night in all cases. All patients received prophylactic antibiotics for 24 hours (Cefprozil 1.5 g IV or Clindamicin 600 mg IV in case of Penicilin or other related allergy) and were discharged with oral antibiotics for 7 days (Flucloxacilin 500 mg four times a day or Clindamicin 300 mg twice daily). Early seroma diagnosis was confirmed clinically. The authors routine is to surgically address their seroma cases and no ultrasound was used. No information regarding seroma s total volume was collected neither bacteriological tests were performed. Patients return to the operating room and have the affected breast implant removed, a sequential cleaning performed according to the authors technique and a brand new implant replaced on the pocket. 8,9 The patients are discharged with compression binders for 6 weeks. There were no recurrent cases. RESULTS Five hundred and thirty-nine female patients were enrolled in the study, and they ranged in age from 18 to 63 years (average, 36.6 years; standard deviation, 8.98). Additional information about patient demographics and complications can be found in Table 1. A total number of 15 unilateral early seroma developed with different pockets within the one-year postoperative window is shown in Table 2. All seromas occurred in the first 6 months after the original surgery and none of the cases referred trauma or report of vigorous physical activity.

3 Sforza et al 303 Using exact logistic regression with the independent variables treated as binary variables, we found that smoking, BMI, and pocket are associated with increased risk of seroma while we cannot reject the hypothesis that pocket size and age do not affect the development of seroma at 5% significance level. The statistical analysis clearly illustrated the effects that can ensue high levels of BMI in the development of seroma. 50.0% of patients with the BMI larger than 30 developed seroma in comparison to 1.89% of their counterparts with BMI lower than 30 (Figure 1). In fact, high BMI increases the odds of having seroma by 15.9 times. The analysis also revealed that the pocket played a significant role is developing seroma and 6.9% of patients with submammary pocket developed seroma in comparison to only 0.8% of patients with submuscular. Correspondingly, pocket increases the odds of developing seroma by 7.5 times. Finally, we found that smoking increases the odds of developing seroma by 19.8 times compared to non-smokers. 14.7% of smokers developed seroma in comparison to only 0.8% of non-smokers (P <.05). A total number of unilateral early seroma developed according to smoking habits is shown in Table 3. On the other hand, while we estimated that, on average, age has small positive effect on seroma development, at a 5% significant level, we cannot reject the hypothesis that it does not effect on seroma development. Similarly, while implant Table 1. Patient Demographics and Complications Demographics Number of patients 539 Age (years) Range Average (standard deviation) 36.6 (8.98) Implant pocket Submuscular 366 (67.9%) Subglandular 173 (32.1%) Implants > 350 cc 124 (23.0%) Smokers 76 (14.12%) BMI > 30 kg/m 2 12 (2.23%) Complications Seroma 15 (2.78%) Wound dehiscence 4 (0.74%) Infections 1 (0.18%) Hematoma 0 (0%) size has larger positive effect on seroma development, we found that this is not significant at a 5% level (Figure 2). Thus, our analysis reveals that smoking, pocket and large BMI are associated with seroma development. Keeping every factor constant, smoking, pocket, and large BMI increase the odds of developing seroma by 19.8, 7.5, and 15.9, respectively (Figure 3). While we had only three patients with all these risk factors, all of them developed seroma, which suggests a call for caution with such patients. Furthermore, having any two of these factors represents a significant risk. The strong associations we found between smoking, pocket type, and BMI with seromas set the ground work for subsequent research and provides us with vital insights about which patients require extra caution regarding possible complications. In addition to identifying the significance of each factor, we also verified that there was an exponential risk as we incorporated the factors one by one up to their totality. Hence, we can firmly conclude that a smoking patient with BMI over 30 will be more likely to develop seroma (Figure 4). Although, as stated above, implant size though has a larger yet insignificant effect on seroma development, smoking increases the odds of seroma with implants bigger than 350 cc (Figure 5). These effects are ceteris paribus effects; regardless of BMI and breast size, smoking increases the odds of getting seroma by 19.8 times, as previously stated. DISCUSSION The pathophysiology of seroma is vague and not well understood but by definition it has been described as a subcutaneous non-infective exudate fluid. Currently, there is no consensus as to the cause of seroma. Several mechanisms have been postulated to account for seroma formation. One initial theory suggests that infection may be the underlying cause of serogeonus fluid formation. While infection can indeed cause periprosthetic fluid formation, several studies, have found significantly negative culture results following fluid cytology study. 10 It has generally been agreed that infective causes cannot be ruled out with regards to seroma development, what includes some of the biofilm theories. Seroma fluid collections have recently been associated with new malignant formations. However, the anaplastic large cell lymphoma is not related with early seromas, but with late ones. This has mainly been described in low-level evidence case reports and it seems unlikely that malignant changes are responsible for all breast augmentation associated seroma. 10 Recent reviews by Jewell et al and Al-Attar et al disregarded both malignant and infective factors and concluded that seroma development following breast augmentation surgery is idiopathic. 11,12 One of the more likely theories stipulate that seroma forms as a result of disruption to

4 304 Aesthetic Surgery Journal 37(3) Table 2. Summary of Seroma Cases According to the Pocket Type Seroma No Table 3. Summary of Seroma Cases According to Smoking Habits Seroma Yes Pocket No Yes Smoking Submammary No Submuscular Yes Figure 2. Graph showing the distribution of seroma cases according to the implant sizes. lymphatic channels, dead space creation, frictional forces, and release of inflammatory mediators.13 The friction theory may also help explain the incidence of late onset seroma, a complication reported more commonly in recent literature mainly with over textured implants.14 Currently there are no studies which directly assess the risk factors associated with early seroma and breast augmentation surgery specifically. Most of the available literature addresses seroma in the context of breast reconstruction surgery, where it is much more problematic. This study aimed to assess the relative correlation of several variables, known to be associated with seroma in other related articles, with the incidence of early seroma. The risk factors assessed were smoking, BMI, age, implant size, and implant position. Smoking Figure 3. Graph illustrating the odds ratio for developing seroma with 95% confidence level. Smoking has long been established to have negative effects on the human body, predisposing individuals to cancer, cardiovascular, cerebral and pulmonary diseases to name but a few. Smokers have also been shown to have worse outcome when undergoing surgery in a variety of fields including plastic and reconstructive surgery. The increased incidence of complications in smokers, particularly seroma, can be attributed to tobacco s general impairing effect on wound healing and vascular repair. Figure 1. Graph showing the distribution of seroma cases according to BMI.

5 Sforza et al 305 Figure 5. Graph illustrating the percentage of patients with seroma for different sizes of breast implants among smokers and non-smokers with interpolation between implant size increments. Colwell et al recently performed a retrospective review of 500 mastectomy with implant reconstruction patients and discovered smoking to be positively correlated to complications such as seroma.15 Similarly, a New York-based study on breast reconstruction, found that the complication rate in smokers to be 27%, as opposed to 13% in non-smokers (P <.001).16 So much so the effect of smoking, that the odds of developing complications were 2.2 greater in smokers (P <.001). However, it must be mentioned that the results from both of these papers discuss total complications, which does not necessarily mean that these findings hold true for seroma incidence exclusively. Interestingly, one study on breast reconstruction, which did analyze the relationship between smoking and seroma, found no significant difference in the variable groups (P =.5793), however overall complications were significantly associated with smoking on the whole.17 A Danish paper investigated the effects of smoking cessation on postoperative complications in 130 patients undergoing breast cancer surgery.18 Participants were required to abstain from smoking for 2 days before and up to 10 days following surgery. The results of this series showed no difference in postoperative complications in both groups, including seroma. The authors believe that a short smoking cessation period is ineffective in reducing the incidence of seroma and other complications. This notion is supported by general surgery articles, which have shown a reduction in postoperative complications and seroma when a 4 to 8 week smoking cessation period was imposed.19 Currently, there is very little definitive data on the relationship between seroma, smoking and breast augmentation specifically. However, our results and those found in breast reconstruction cases show that smoking is generally associated with seroma and increased level of complications. Given the overall serious negative effects of smoking and the positive impact of smoking cessation on seroma incidence, smoking should undoubtedly be considered a risk factor for further studies to validate. Implant Size Another factor analyzed in this paper was implant size. Our results indicated that implanting prostheses with a size of 350 cc or more increases the risk of developing seroma, though not significant at a 5% level. It is not exactly clear why this is so, but we believe that larger implants may increase dead space volume and may also propagate friction and irritation leading to a higher incidence of seroma. Larger implant size are also known to be related to higher incidences of other complications such as capsular contracture and hematoma.20 Lista et al reviewed over 400 breast augmentation patients and found that there was also a significant correlation between larger implant size and postoperative complications (P <.0001).5 It is important to note, that this paper includes seroma as a complication but does not explicitly cite it as a correlative factor. Another study also found greater rates of complication with implants larger than 350 cc (adjusted risk ratio [RR], 2.3; 95% confidence interval [CI], ).21 Interestingly, implant size was found to be the only significant factor associated with complications that, include seroma, when considering other factors such as between age, height, BMI, smoking, or alcohol consumption. The significant impact of implant size has been reiterated further in both mastopexy augmentation and breast reconstruction surgery.22,23 Figure 4. Graph showing the percentage of patients (smokers and non-smokers) with seroma for different levels of BMI.

6 306 Aesthetic Surgery Journal 37(3) Implant Pocket Our findings also indicate that implants placed in the subglandular position significantly increases the odds of developing seroma. A Finnish study on local complications following breast implant surgery also found a higher complication rate with subglandular implants (P =.013). 24 Subglandular placement has also been associated with increased risk of capsular contracture, hematoma, and implant displacement. 20 On the contrary, Henriksen et al found that subglandular implant placement reduced the risk of postoperative complication. 21 This study was powered to primarily address the risks associated with capsular contracture, making it difficult to quantify its direct relevance to seroma risk. Strasser summarizes that subglandular augmentations are associated more so with capsular contracture, implant palpability and rippling, whereas submuscular placements are coupled with malposition and asymmetry. 25 It therefore appears that subglandular implants have more of a mechanical effect and may contribute to a shearing effect and inflammatory process inductive of seroma. The patients before having surgery usually started a course of antibiotics and the authors did not perform bacteriological tests on the seromas fluids. This could have helped to identify any further differences amongst the pockets and should be evaluated in further studies. BMI A larger BMI has long been associated with adverse medical risks and general surgical complications, and the same appears to be true for seroma. From our analysis it was evident that 50% of patients with a BMI greater than 30 developed seroma in comparison to 1.89% of patients with lower BMI. Chen et al concluded similar results, with 64 obese patients developing seroma compared to 14 non-obese patients. 26 The authors calculated obesity to increase the risk of seroma by as much as 10-fold. Similarly, a Chicago-based study investigated the role of obesity in breast reconstruction surgery. 27 The results not only confirmed obesity s role as a major predictor of negative surgical outcomes, but also instilled its relationship with seroma. It was found that obesity also increased the odds of seroma formation by nearly 10%. However, certain studies on oncological breast surgery have failed to demonstrate equal findings. This was highlighted by Gonzalez et al, who found no difference in seroma rates in varying BMI patient subgroups. 28 This study maybe negated in this circumstance as it was conducted on patients undergoing primarily mastectomy rather than augmentation. In addition to which obesity has been generally shown to increase the risk of a whole host of complications including seroma in a spectrum of surgeries within the plastic surgery field. 29 Age Age was another variable studied in our results. Unlike other measured risk factors, older age was not associated with increased chance of seroma. This variability is reflected in the current literature. In plastic surgery as a whole older age groups have not been found to be associated with complications in an array of procedures. An age older than 45 years was found to be significantly associated with postoperative complications in patients undergoing both implant placement and staged tissue expander breast reconstruction. 23 However, some studies in breast surgery argue the contrary. This double standard in results is reflected in the previously mentioned Chen study. The authors found significant correlation between postoperative complications, including seroma, for those aged between 55 to 64, but not those greater than or equal to 65, 45 to 54, or 35 to The volatility of results concerning age are in part due to the greater impedance that other risk factors have on the influence of seroma and operative complications. Despite categorically assessing five potential risk factors of seroma, one of the limitations of this study is that several more variables could have been analyzed. It would have been advantageous to also investigate the relationship of implant type on seroma incidence. This relationship is currently being fiercely debated, with textured implants being shown in some series to be significantly related to seroma formation. 14 Other variables which could also have been included are incision type, drain use, significant weight loss and the presence of co morbidities. Therefore, the authors use the same technique today as they did when started this research. We believe that at this stage there is not enough evidence to change our practice or techniques. However, we strongly believe that we can now provide a better consultation process to patients and we can also be more precise in terms of risk factors and complications to our patients. CONCLUSION A high BMI, large implant size, submammary pocket, and smoking are factors significantly associated with seroma development, while age is not. A BMI over 30 in particular is the strongest indicator for seroma development. Smoking however was found to be the most detrimental factor as it significantly amplified the effects of the other variables. Future studies will be required to validate these results further and to assess the implication of other variables. These findings will hopefully allow clinicians to make more guided preoperative decisions that will in turn aid in reducing the incidence of seroma. Disclosures The authors declared no potential conflicts of interest with respect to the research, authorship, and publication of this article.

7 Sforza et al 307 Funding The authors received no financial support for the research, authorship, and publication of this article. REFERENCES 1. Cosmetic surgery national data bank statistics. Aesthet Surg J. 2016;36(Suppl 1): Bengtson BP. Complications, reoperations, and revisions in breast augmentation. Clin Plast Surg. 2009;36(1):139-56, viii. 3. Lista F, Tutino R, Khan A, Ahmad J. Subglandular breast augmentation with textured, anatomic, cohesive silicone implants: a review of 440 consecutive patients. Plast Reconstr Surg. 2013;132(2): Roden AC, Macon WR, Keeney GL, Myers JL, Feldman AL, Dogan A. Seroma-associated primary anaplastic large-cell lymphoma adjacent to breast implants: an indolent T-cell lymphoproliferative disorder. Mod Pathol. 2008;21(4): Forster JJ, McDonald J, Smith PWF. Markov chain Monte Carlo exact inference for binomial and multinomial logistic regression models. Statistics and Computing. 2003;13(2): Derr RE. Performing Exact Logistic Regression with the SAS System Revised Proceedings of the Twenty-fifth Annual SAS Users Group International Conference, Cary, NC, Tebbetts JB. Dual plane breast augmentation: optimizing implant-soft-tissue relationships in a wide range of breast types. Plast Reconstr Surg. 2001;107(5): Sforza M, Andjelkov K, Zaccheddu R. A successful salvage protocol for breast implants. Plast Reconstr Surg. 2011;128:33e-34e. 9. Sforza M, Andjelkov K, Husein R, Zaccheddu R. Will 1-stage implant salvage after periprosthetic breast infection ever be routine? A 6-year successful experience. Aesthet Surg J. 2014;34(8): de Jong D, Vasmel WL, de Boer JP, et al. Anaplastic largecell lymphoma in women with breast implants. JAMA. 2008;300(17): Spear SL, Rottman SJ, Glicksman C, Brown M, Al-Attar A. Late seromas after breast implants: theory and practice. Plast Reconstr Surg. 2012;130(2): Bengtson B, Brody GS, Brown MH, et al.; Late Periprosthetic Fluid Collection after Breast Implant Working Group. Managing late periprosthetic fluid collections (seroma) in patients with breast implants: a consensus panel recommendation and review of the literature. Plast Reconstr Surg. 2011;128(1): Andrades P, Prado A. Composition of postabdominoplasty seroma. Aesthetic Plast Surg. 2007;31(5): Hall-Findlay EJ. Breast implant complication review: double capsules and late seromas. Plast Reconstr Surg. 2011;127(1): Colwell AS, Tessler O, Lin AM, et al. Breast reconstruction following nipple-sparing mastectomy: predictors of complications, reconstruction outcomes, and 5-year trends. Plast Reconstr Surg. 2014;133(3): McCarthy CM, Mehrara BJ, Riedel E, et al. Predicting complications following expander/implant breast reconstruction: an outcomes analysis based on preoperative clinical risk. Plast Reconstr Surg. 2008;121(6): Nguyen KT, Hanwright PJ, Smetona JT, Hirsch EM, Seth AK, Kim JY. Body mass index as a continuous predictor of outcomes after expander-implant breast reconstruction. Ann Plast Surg. 2014;73(1): Thomsen T, Tønnesen H, Okholm M, et al. Brief smoking cessation intervention in relation to breast cancer surgery: a randomized controlled trial. Nicotine Tob Res. 2010;12(11): Sørensen LT, Hemmingsen U, Jørgensen T. Strategies of smoking cessation intervention before hernia surgery effect on perioperative smoking behavior. Hernia. 2007;11(4): Collins JB, Verheyden CN. Incidence of breast hematoma after placement of breast prostheses. Plast Reconstr Surg. 2012;129(3):413e-420e. 21. Henriksen TF, Fryzek JP, Hölmich LR, et al. Surgical intervention and capsular contracture after breast augmentation: a prospective study of risk factors. Ann Plast Surg. 2005;54(4): Beale EW, Ramanadham S, Harrison B, Rasko Y, Armijo B, Rohrich RJ. Achieving predictability in augmentation mastopexy. Plast Reconstr Surg. 2014;133(3):284e-292e. 23. Roostaeian J, Sanchez I, Vardanian A, et al. Comparison of immediate implant placement versus the staged tissue expander technique in breast reconstruction. Plast Reconstr Surg. 2012;129(6):909e-918e. 24. Kulmala I, McLaughlin JK, Pakkanen M, et al. Local complications after cosmetic breast implant surgery in Finland. Ann Plast Surg. 2004;53(5): Strasser EJ. Results of subglandular versus subpectoral augmentation over time: one surgeon s observations. Aesthet Surg J. 2006;26(1): Chen CL, Shore AD, Johns R, Clark JM, Manahan M, Makary MA. The impact of obesity on breast surgery complications. Plast Reconstr Surg. 2011;128(5):395e- 402e. 27. Hanwright PJ, Davila AA, Hirsch EM, et al. The differential effect of BMI on prosthetic versus autogenous breast reconstruction: a multivariate analysis of 12,986 patients. Breast. 2013;22(5): Gonzalez EA, Saltzstein EC, Riedner CS, Nelson BK. Seroma formation following breast cancer surgery. Breast J. 2003;9(5): Mioton LM, Buck DW 2nd, Rambachan A, Ver Halen J, Dumanian GA, Kim JY. Predictors of readmission after outpatient plastic surgery. Plast Reconstr Surg. 2014;133(1):