Breast Surgery

Aesthetic Surgery Journal Advance Access published May

Preliminary Report

A Preliminary Assessment of the Predictability of Fat Grafting to Correct Silicone Breast Implant-Related Complications

Marcos Sforza, MD; Katarina Andjelkov, MD, PhD;

Renato Zaccheddu, MD; Rodwan Husein; and Connor Atkinson

7,ffil

Aesthetic Surgery Journal 2016, 1-9

© 2016 The American Society for Aesthetic Plastic Surgery, lnc.

Reprints and permission: journals.permissions@oup.com DOI: 10.1093/asj/sjw060 www.aestheticsurgeryjournal.com

Abstract

Background: Fat grafting in breast augmentation surgery is becoming increasingly popular and is al\owing surgeons to fil\ the gaps that implant augmentation alone cannot. However, one current issue surrounding fat grafting is resorption.

Objectives: In this prospective study, the authors present their experience of fat grafting and resorption in 26 patients who had fat transfer to corred deformities or asymmetries following primary breast enlargement surgery.

Methods: The fat utilized was harvested and processed using the Puregraft system. Fat grafting was performed until the problem was visually corrected, fol­ lowed by an additional 300/o overcorrection. Photographs were taken preoperatively and l year postoperatively, and analyzed utilizing cloud-based 3-dimen­ sional imaging software to measure the breast volumes and calculate 1-year resorption rates. Both patients and physicians were asked to rate their satisfaction. Results: The total volume of fat transferred ranged from 160 ce to 360 ce, w h an average of 280 ce per procedure. The absolute volume of tissue resorption shOV\€d considerable variation (median, 42.6 ml; range, 5.5-62 ml). Ho.,vever, there was a very clase correlation between the volume resorbed and the volume implanted: the more volume injected, the more volume absorbed. When expressed as a percentage of implanted tissue, the resorption rate was 27.5 ± O.SO/o (mean ± standard deviation) w h a very narroo range (25.7%-28.90/o). Patients and physicians reported a satisfaction rate of "excellenr in 83.30/o and 750/o of cases, respectively.

Conclusions: Al l year, 25.70/o to 28.90/o of the injected fat had been resorbed in a predictable, reproducible, and effective manner.

Level of Evidence: 4

Accepted far publication February 29, 2016. Therapeutic

Autologous fat tissue is widely utilized today as a soft-tissue filler in both small- and large-volume applications.1 However, a review of the literature indicates that there is considerable variability in tissue retention with autologous fat tissue, leading to unpredictable outcomes.2•3 These differences in re­ tention rates are generally believed to be caused by a combi­ nation of donor-intrinsic factors (eg, age, smoking history, exposure to radiotherapy at the implan! site, etc.), differences in techniques utilized for tissue harvest and implantation, and the presence of components within the graft that can nega­ tively impact retention (eg, lipid from damaged adipocytes, blood cells, and debris). Analysis of the factors affecting reten­ tion has been generally impaired by the paucity of robust,

Dr Sforza is an Examiner for the Royal College of Surgeons of Edinburgh. Dr Andjelkov is a plastic surgeon in prívate practice in Belgrade, Serbia. Dr Zaccheddu is a Plastic Surgeon at Dolan Park Hospital in Bromsgrove, United Kingdom. Mr Husein and Mr Atkinson are Medica! Students, Leeds Medica! School, Leeds, United Kingdom.

Corresponding Author:

Prof Marcos Sforza, Dolan Park Hospital, Stoney Lane, B60 lLY, Bromsgrove, United Kingdom.

E-mail: marcos@marcossforza.com

Presented at: the International Federation for Adipose Therapeutics and Science (IFATS) meeting in October 2012 in Québec City, Québec, Canada (in part); and at the IMCAS meeting inJanuary 2015 in Paris, France, and at The Aesthetic Meeting 2015 in Montréal, Québec, Canada (in ful!).

quantitative data focused on a single clinical application. In this article, we present a prospective study of 26 patients who underwent fat injections to correct deformities or asymmet1ies following breast enlargement surgery with silicone implants, in a total of 50 breasts. In this study, patients received moderate- to large-volume fat grafts that were harvested, pro­ cessed, and i111planted according to standardized 111ethods, and quantitative 3-di111ensional (3D) analysis was applied to obtain reproducible volu111e-retention data. The results dem­ onstrate that this standardized approach provides highly pre­ dictable retention outco111es, wherein the percentage of implanted tissue that is lost falls within a very small range and is independent of the absolute volume of tissue implanted.

METHODS

Over a period of 12 months from January to December 2012, 26 female patients presented to our private practice plastic surge1y clinic with self-described unsatisfactmy results following breast enlargement surgery at various outside plastic surgery clinics. This study had no exclusion criteria and took ali patients consecutively. The corre­ sponding author (M.S.) divided the patients' unsatisfactory results into two general groups: asym111etries (difference in volume, n = 17) and deformities (difference in shape, rip­ pling, capsular contracture, double bubble, n = 9). The corresponding author (M.S.) performed ali of the surgeries. Patients were evaluated and prospectively offered autolo­ gous fat transfers, as opposed to breast implant replace- 111ent, to correct their dissatisfaction. Ali patients were nonsmokers and all patients had a body mass index (BMI) within the normal weight li111its. No patients had material co111orbidities such as diabetes or high blood pressure. No patients had implant replacement surgery or any other as­ sociated surgeries together with the fat grafting procedure. Fat tissue was harvested with a 60 mL Luer Lock syringe (Becton Dickinson, Franklin Lakes, NJ) from the abdominal area utilizing a 2.4 111111 cannula (Sforza Harvester, Tulip Medical Products, San Diego, CA) and processed utilizing the Puregraft system (Puregraft LLC, Solana Beach, CA). After the preparation of the tissue, a single needle-puncture incision was made near the area to be treated. The surgeon applied the fat in a fanlike 111ove111ent with a 20 mL ce Luer Lock syringe attached to a 0.9 111111 cannula (Tulip Injector, Tulip Medica! Products), aiming to apply 1 mL of fat per passage. Fat grafting was perfor111ed until the problem was visually corrected, as determined by the operating surgeon. A narrative video of a similar procedure may be viewed as Supple111entary Material at www.aestheticsurge1yjournal. com.

The volume injected was noted, and then an additional

30% overcorrection was grafted, taking into consideration historically published low absorption rates. The total volume of fat transferred ranged from 160 ce to 360 ce, with

an average of 280 ce per procedure, with a maximu111 volu111e per breast registered in this series of 220 mL.

All surgeries were performed in a specialized plastic surgery hospital with the patients receiving a general anes­ thetic. Ali patients signed an infor111ed consent form for the described procedure. Ali patients had a single dose of l.Sg of cefuroxi111e as a prophylactic antibiotic 1 hour befare the surgery started. Ali patients had co111pression socks (Preventx, Loughborough, United Kingdom) and a prophy­ lactic pneumatic deep vein thro111bosis (DVT) syste111 (Flowtron, Huntleigh Healthcare, Eatontown, NJ) during the procedure. However, no che111ical prophylaxis for DVT was given, according to hospital policy for short procedures (less than 2 hours). Patients were discharged with compres­ sion gar111ents on the harvested area but only soft dressings on the grafted areas to avoid co111pression of the transferred fat.

Photographs were taken preoperatively, and 1 year post­ operatively. These photographs served as a basis for a satisfaction questionnaire in which both patients and inde­ pendent physicians (three board-certified plastic surgeons) were asked to rate their satisfaction. In addition to the satisfaction questionnaire, the photographs were subse­ quently analyzed utilizing cloud-based 3D i111aging soft­ ware (Crisalix, Lausanne, Switzerland). The 3D analysis provided us with the volu111e of the breast both preopera­ tively and postoperatively, allowing us to calculate 1-year

Table 1. lnformation on Patient Demographic and Complications

SD, standard deviation.

Table 2. Tissue Volume lmplanted per Breast

SD, standard deviation.

c:l

.o= oo 'o".

"o.

=ot>

3

Figure 1. The data of the distribution volume of fat grafted.

Figure 3. The percentage of resorption rate demonstrates a very narrow range (25.7% to 28.9%).

absorption rates. It should also be noted that none of the patients in this study received fat grafts in an area with known underlying pathology or any other clinical condition that might render the area to be more hostile to implanted tissue. For example, none of the patients had received pre­ vious radiotherapy or were implanted in an area that had considerable fibrosis and/or physical evidence of impaired vascularity.

RESULTS

The mean age of the patients was 24 years (standard devia­ tion [SD], ± 3.48 years; range, 19-32 years). Additional

Figure 2. There was a clase correlation between the volume resorbed and the volume implanted.

information about patient demographics and complications is in Table 1.

Patient and Physician Satisfaction

All 26 patients completed a 1-year satisfaction questionnaire (Appendix A, available online at www.aestheticsurgeryjournal. com). This questionnaire is the standard patient satisfaction evaluation form that we utilize in our practice, but we ac­ knowledge that the questionnaires were not completed anon­ ymously. All physician satisfaction questionnaires were also completed at 1 year (Appendix B, available online at www. aestheticsurgeryjournal.com). At the 1 year time point, pa­ tients reported a satisfaction rate of "excellent" in 84.6% (n = 22) of cases, "good" in 11.5% (n = 3), and "fair" in 3.8% (n = 1). The 1-year physician satisfaction rated as "ex­ cellent" in 73.1 % of cases (n = 19), "good" in 19.2% (n = 5), and "fair" in 7.7% (n = 2) of cases. During the follow-up period, no patients had evidence of infections, seromas, hematomas, or oil cysts formation after fat grafting. No imaging control on the breasts was performed, because the injections were subcutaneous under a thin skin and the clinical examination of the patients was sufficient to exclude the presence of cysts or lumps in the injected area. In all patients, a successful correction of the presenting problem was achieved without complications, as confirmed in the physicians' questionnaire.

Retention Volume

Data far the volume of fat grafted are shown in Table 2 and Figure 1. As displayed in Figure 1, the volume of fat grafted was not normally distributed; the vast majority of cases

c:l

.o=

oo

'o".

"o.

=ot>

3

Table 3. A ttest Comparing Absorption Between Grafted Volume <150 ml and >150 ml of Fat

Figure 4. The percentage of volume lost was independent of the absolute volume implanted.

(46/50) involved transfer of 2:100 mL of tissue. The median volume grafted per breast was 155 mL.

The absolute volume of tissue resorption showed consider­ able variation (median, 42.6 mL; range, 5.5-62 mL). However, there was a very clase correlation between the volume re­ sorbed and the volume implanted (Figure 2). When expressed as a percentage of implanted tissue, the resorption rate was

27.5 ± 0.8% (mean± SD), with a very narrow range (25.7% to 28.9%) (Figure 3). The percentage of volume lost was inde­ pendent of the absolute volume implanted (Figure 4; slope of the curve = 0.0006; R2 = 0.00115). The same results were ob­ served by comparing the absorption rate between different chosen volumes of grafted tissue. Tables 3 and 4 show the result of a t test comparing the mean absorption rate between grafted volume less than and greater than 150 mL and between grafted volume less than and greater than 120 mL, re­ spectively. The hypothesis was a common mean absorption rate. In neither of these cases were we able to reject the hy­ pothesis of common mean. This is even more clearly observed

Table 4. A ttest Comparing Absorption Between Grafted Volume <120 ml and >120 ml of Fat

Upper

Lawer

lndependent Samples Test

ttest far Equality af Means

95% Confidence lnterval al the Difference

Figure 5. The absorption rate has mini mal variability and in­ creases or decreases independently from the grafted volume.

in Figure 5, where we can verify that the mean absorption rate far different grafted volume levels. Again, we can see that the absorption rate has minimal variability and that it increases or decreases independently from the grafted volume. The same results can be concluded from compa.ring the absorption rate between sections of grafted volume. There were no significant differences on the absorption rates among füe groups, allow­ ing us to say that the predictability of the fat survival was cons­ tant independently of the original deformity. Therefore, the division of the groups became relevant only from a demo­ graphics perspective, allowing us to acknowledge the different patients studied in this article. There was no clinically notice­ able cha.nge on the patients' BMI ra.nge 1 yea.r postoperatively.

DISCUSSION

This study employed a highly sensltlve, objective, and quantitative assessment of volume to examine the rate of tissue resorption following autologous fat grafting. The

c:l

.o= oo 'o".

"o.

=ot>

3

Figure 6. (A) This 25-year-old woman is shown when she presented with asymmetric breasts and (B) 1 year postoperatively. Preoperative and postoperative 3D images were generated with Crisalix.

data demonstrate that a standardized tissue harvest, method of processing, and implantation approach mitigate the vast majority of variability and unpredictability com­ monly ascribed to autologous fat grafting, such that the volume of tissue implanted is the primary determinant of the absolute amount of tissue volume lost. We acknowl­ edge that despite the fact that there were no exclusion crite­ ria to this study, the cohort was very homogeneous and rather young, probably because of the very nature of the original procedure. Some could see this as a favorable factor to improve graft intake. There are no conclusive studies that compare resorption rates according to patient age. Therefore, a statement to advocate this theory will also lack scientific support. The standardized procedure seems to be the best explanation so far.

We acknowledge that the high satisfaction rate encoun­ tered in both the patients' and surgeons' questionnaires have a limitation because they were not filled out anonymously. Nevertheless, the independent surgeons who were evalua­ tors never met the patients, which minimizes such bias.

c:l

.o=

oo

'o".

"o.

=ot>

3

Volume Analysis

Perhaps the most accurate method of measuring fat retention is to excise the fat and weigh the specimen, which is the method utilized in animal studies. Because this is not possi­ ble in human studies, different types of image-based mea­ sures are available. Although the quality of the 1-year clinical

Figure 7. This intraoperative photograph demonstrates that the Puregraft allowed us to remove free lipid, white blood cells, and red blood cells at large.

Figure 8. (A, C) Preoperative photographs of this 31-year-old woman with a previous mastopexy with implants in a subpectoral pocket. She was referred to us from another hospital because of rippling at the upper pole. (B, O) She is shown 1 year after 180 ce of fat was grafted in each breast. (C) Another perspective of the severe rippling that occurred even with a subpectoral pocket.

c:l

.o=

oo

'o".

"o.

=ot>

3

outcomes are clear with routine examination of the patients, the quantitative analysis of retained volume utilizing a vali­ dated 30 imaging software is fundamental to support the statistical significance of our results. The processing of up­ loading photographs by the surgeon is fast (less than 10 minutes) and therefore can readily obtain measurements. This system utilizes a physics-based simulation (Voxel ap­ proach), in which small cubes are utilized to fill the interna! breast to model the interna! breast tissue (Figure 6). Although there are severa! published methods of quantitative volume assessment, this particular approach was chosen because of its sensitivity to real volume change. In a study published by de Heras Ciechomski et al,4 the authors showed that when utilizing this method, there is a mean

reconstruction error between 2 111111 and 4 111111 for both the left and right breasts. However, by taking into account the large 90º angle between the frontal and lateral photographs utilized in the reconstruction, the observed maximum surface error appeared small. Furthermore, the maximum surface error is less than the motion artifacts caused by breathing excursions and the s111all changes caused by patient repositioning when laser or optical sean devices are utilized. Their study verifies the reliability and accura­ cy of the method utilized in this article. 4

We recognize that in utilizing such a method, there are inherent limitations. However, the same limitations apply to any study utilizing acquired images to reconstruct a 30 model. In their facial fat grafting study, Gerth et al5

Figure 9. (A, C) Preoperative photographs of this 29-year-old woman who had previous breast mastopexy implants in a subpec­ toral pocket and four revisional surgeries. She was referred to us from a colleague who believed he had exhausted his options with this patient, who now presented with a severely shaped deformity. The preoperative obligue view (C) demonstrates a severe "double bubble" deformity. (B, O) She is shown 1 year after having 200 ce of fat grafted on the left breast and 100 ce of fat grafted on the right breast. This patient refused that any kind of pocket or implant be changed, and she even refused scar revisions after the previous surgeries. She made it clear that she would only accept a minimally invasive procedure to improve the shape of her breasts. This case demonstrates that fat grafting can be utilized to correct severe deformities caused by a previous breast surgery.

c:l

.o=

oo

'o".

"o.

=ot>

3

acknowledged that the accuracy of the data obtained with the Vectra 30 imaging system (Canfield, Fairfield, NJ) were user dependent and that errors in volume measurement could be introduced.

Sorne studies consider magnetic resonance imaging (MRI) a potentially better method for volumetric measurements.6-8 However, MRI is expensive, time-consuming, and not easily accessible, especially in facilities that specialize in plastic surgery or outpatient clinics.

After the system generates the 3D image, it identifies the landmark limits of the breast and upon request provides the volume of the breast in a very simple manner, similar to other 3D devices such as the previously mentioned Vectra 3D system. Severa! studies already demonstrated that breast augmentation simulations correlate highly with positive surgi­

cal outcomes and increased patient scheduling in specific practices together with higher productive rates.8•9 In our study, 30 simulations were only utilized as a measurement tool, but

their efficacy should lead to other studies in which 3D simula­ tions will also be incorporated in the surgical planning.

The virtual method of analysis utilized in this study elimi­ nated the need for capital investment on hardware, because the technique utilizes a web-based license. Additionally, surgeon-investigators are not limited to a single location ora single point in time to review patient data.

Harvest Methodology

We believe that utilizing a multi-hole cannula, designed specifically for this study, played an important role in ex­ plaining the results. This new cannula is designed with 20 mm x 1 mm holes arranged in a special layout to facilitate bidirectional harvest. The boles are designed with an ele­ vated 1 mm microport in one end, and the ports at the other end are machined with 60-degree cutting edges.

In addition, the asymmetric disposition of the boles in a helical pattern ensure that ali boles see an identical as­ piration pressure. The net effect of the number of holes, in combination with the layout, is to aspirate tissue at the same time, providing more fat harvested with less passages, thus ultimately promoting less trauma to the fat and adja­ cent tissue. By design, the harvested grafted particles were always smaller than 1 mL. This information is still only con­ jectural, because there are no scientific studies to prove this theory. However, it is important to highlight that there are severa! studies demonstrating that larger fat particles tend to have higher absorption rates.11,12

Tissue Processing

This study employed a standardized method for preparing the graft before injection. We utilized the Puregraft system to remove contaminants such as free lipid and red blood cells. In a recent paper by Zhu et al,13 autologous fat grafts prepared with this sterile, single-use approach proved to be far more effective at removing free lipid, white blood cells, and red blood cells when compared with alternative ap­ proaches such as centrifugation. Specifically, the authors reported that free lipid comprised only 0.5% ± 0.1% of the volume of the graft when processed with the Puregraft system, compared with 12.6 ± 1.6% of the volume when the same tissue was processed by centrifugation (Figure 7).

Tissue lnjection

Ali fat was grafted utilizing identical, small injectors (O.9 mm), without any teclmical difficulties. As extensively put forth by Coleman,14 the use of multiple injections with small volumes per passage is fundamental for a lower resorption rate. The reader is also reminded that none of the patients received grafts into hostile recipient areas. This is noteworthy, because it is reasonable to expect a greater

percentage resorption at such sites in the absence of ap­ proaches directed at enhancing retention such as enrich­ ment with stromal vascular fraction cells15 16 (Figures 8 and 9; Supplementary Figures 1-3).

'

Patient Messaging

Results from this investigation can aid in managing patient expectations before performing a grafting procedure. Regardless of the volume injected, a patient's perspective on the retained volume will be masked in part by initial swelling and edema. However, the fact that smaller grafts will have a correspondingly small absolute volume of re­ sorption will be difficult to perceive. At the other end of the spectrum, the larger absolute volume of resorption follow­ ing implantation of a larger graft will be more visually ap­ parent leading to an increased likelihood that the patient will notice the change and have a perception of poor graft retention. Armed with this knowledge, the surgeon can prospectively inform their patients that perception will vary by volume, thus fostering an enhanced comprehension and avoiding misunderstood expectations.

CONCLUSIONS

In summary, careful attention to standardization of tissue collection, processing, and implantation can lead to a high predictability of retention of autologous fat grafts in pa­ tients with a healthy recipient tissue bed. This predictabili­ ty can be expected to improve patient satisfaction.

Supplementary Material

This article contains supplementary material located online at www.aestheticsurgeryjournal.com.

Disclosures

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

Funding

The authors received no financia! support for the research, authorship, and publication of this article.

REFERENCES

l. Kaufman MR, Bradley JP, Dickinson B, et al. Autologous fat transfer national consensus survey: trends in tech­ niques for harvest, preparation, and application, and per­ ception of short- and long-terrn results. Plast Reconstr Surg. 2007;119(1):323-331.

1. Kaufman MR, Miller TA, Huang C, et al. Autologous fat transfer for facial recontouring: is there science behind the art? Plast Reconstr Surg. 2007;119 (7):2287-2296.

2. Chan CW, McCulley SJ, Macmillan RD. Autologous fat transfer--a review of the literature with a focus on breast

   
   

   c:l

   .o= oo 'o".

   "o.

   =ot>

   3

   

   
   

   cancer surgery. J Plast Reconstr Aesthet Surg. 2008;61 (12):1438-1448.

3. de Heras Ciechomski P, Constantinescu M, García J, et al. Development and implementation of a web-enabled 3D consultation too! for breast augmentation surgery based on 3D-image reconstruction of 2D pictures. J Med Intemet Res. 2012;14(1):e21.

4. Gerth DJ, King B, Rabach L, et al. Long-tem1 volumetric re­ tention of autologus fat grafting processed with closed-mem­ brane filtration. Aesth Surg J. 2014;34(7) :985-994.

5. Gosain AK, Klein MH, Sudhakar PV, Prost RW. A volumet­ ric analysis of soft-tissue changes in the aging midface using high-resolution MRI: implications for facial rejuvena­ tion. Pln.st Reconstr Surg. 2005;115(4):1143-1152; discus­ sion 1153-1155.

6. Darcy SJ, Miller TA, Goldberg RA, et al. Magnetic resonance imaging characterization of orbital changes with age and associated contributions to lower eyelid prominence. Plast Reconstr Surg. 2008;122(3) :921-929; discussion 930-931.

7. Creasman CN, Mordaunt D, Liolios T, et al. Four-dimensional breast imaging, part I: introduction of a technology-driven, evidence-based approach to breast aug­ mentation planning. Aesthet Surg J. 2011;3l (8):914-924.

8. Creasman CN, Mordaunt D, Liolios T, et al. Four-dimensional breast imaging, part II: clinical imple­ mentation and validation of a computer irnaging system

   for breast augmentation planning. Aesthet Surg J. 2011;31 (8) :925-938.

9. Wysong A, Joseph T, Kim D, et al. Quantifying soft tissue loss in facial aging: a study in women using magnetic res­ onance imaging. Demwtol Surg. 2013;39(12):1895-1902.

10. Sommer B, Sattler G. Current concepts of fat graft sur­ vival: histology of aspirated adipose tissue and review of the literature. Dennatol Surg. 2000;26(12) :1159-1166.

11. Kato H, Mineda K, Eto H, et al. Degeneration, regenera­ tion, and cicatrization after fat grafting: dynamic total tissue temodeling during the first 3 months. Plast Reconstr Surg. 2014;133 (3) :303e-313e.

12. Zhu M, Cohen SR, Hicok KC, et al. Comparison of three dif­ ferent fat graft preparation methods: gravity separation, cen­ trifugation, and simultaneous washing with filtration in a elosed systern. Plast Reconstr Surg. 2013;131(4):873-880.

13. Coleman SR. Chapter 26: Structural lipoaugmentation. In: Narins RS, ed. Safe Liposuction and Fat Transfer. Boca Raton, FL: CRC Press; 2003:409-424.

14. Zhu M, Zhou Z, Chen Y, et al. Supplementation of fat grafts with adipose-derived regenerative cells improves long-term graft retention. Ann Plast Surg. 2010;64(2) :222-228.

15. Pérez-Cano R, Vranckx JJ, Lasso JM, et al. Prospective tria! of adipose-derived regenerative cell (ADRC)-enriched fat grafting for partía! mastectomy defects: the RESTORE-2 tria!. Eur J Surg Oncol. 2012;38(5):382-389.

:o:,