Abstract
Autologous buttock augmentation with fat grafting has emerged as one of the preeminent modalities for body contouring employed by plastic surgeons today. Since 2009, we have used the roller pump injection technique.
This procedure can be performed safely without specialized equipment and eliminates the tasks of manual graft preparation and injection. We describe our technique and standard safety measures. The anatomy of and complications associated with liposuction and fat grafting were recorded and reviewed.
Retrospective chart review of 916 patients who underwent autologous buttock augmentation by this method from February 2009 to November 2016 was performed. All procedures were performed under general anesthesia at the same surgical center. Liposuction was performed and using a roller pump, the fat layer was propelled through an open-ended cannula into the recipient site.
Mean volume of fat removed and fat grafted in each patient was 3156 mL and 1807 mL per patient, respectively. There were complications in 13 patients for a rate of 1.4% with 10 (1.1%) related to fat grafting. Fourteen patients (1.5%) had subsequent procedures for volume and four patients (0.44%) for asymmetry. There were no venous thrombolic events, fat embolic events, or deaths.
The roller pump injection technique for buttock augmentation with fat grafting is safe and efficient. This technique minimizes preparatory effort and does not require additional equipment. We were unable to identify variables associated with complication risk due to the power of this study and the low percentage of complications.
Autologous buttock augmentation with fat grafting has emerged as one of the preeminent modalities for body contouring employed by plastic surgeons today. Although augmentation of the buttocks with fat has been performed for more than 30 years,1 the procedure has recently exploded in popularity. Just last year, 19,019 fat grafting procedures to the buttocks were performed in the United States, an increase of 2.9% from the year prior.2 The purpose of this procedure is to achieve a redistribution of volume to shift the proportions of waist-to-hips and modify the posterior projection, upper, and lateral fullness of the buttocks.3-5 Depending on patient body habitus and goals, this is achieved by some combination of lumbosacral, flank, hip, inner and outer thigh liposuction, and fat grafting throughout the superior buttock, lateral gluteal depression, trochanter, ischium, lateral and posterior thighs.3,4,6-9 Some respondents from a population analysis favor a shift from the traditional waist-to-hip ratio of 0.7 to 0.6 with an emphasis on the greatest point of projection lying at the midpoint of the buttocks.10 Patient satisfaction after this procedure is reported to be 86.3% to 97.1% with their buttocks11-13 and 100% with their waistline contouring.12
There are many different practices to harvest, prepare, and inject the fat graft and there is no consensus on the optimal fashion.11,14-21 Two recent meta-analyses reviewing 22 articles with 3567 total patients from 1992 to 201522 and 19 articles with 4105 total patients from 1996 to 201523 that underwent autologous gluteal augmentation by fat grafting report an overall complication rate of 9.9% and 7%, respectively. This is much lower than the multicenter reported complication rate of 38.1% for silicone buttock implants.24 Complications may be either regional or systemic, and include infection, contour deformity, oil cyst formation, hematoma, acute blood loss anemia, hypovolemia, deep venous thrombosis (DVT), pulmonary embolism (PE), fat embolism, and death.3,6,22,25,26 It is also reported by Cardenas-Camarena from their series of 789 patients, that with increased experience the complication rate decreased despite increasing volumes of fat injected.6 To satisfy the growing demand by patients for the procedure and for higher grafting volumes per patient,6,13 we have devised a technique that allows for safe, efficient buttock augmentation for use across the spectrum of patients’ goals.
Our technique was introduced by the senior author (H.A.M.) in 2010,27 and we have been performing this procedure facilitated by a novel roller pump injection technique since 2009 on more than 900 patients with an overall complication rate of 1.75% and a donor/recipient specific complication rate of 1.09%. The procedure can be performed without additional equipment and minimal operative time is added after the graft harvest and the waistline contouring by liposculpture is completed. With this technique, much of the laborious tasks of manual graft preparation and injection are eliminated and the graft is injected into the recipient site with a “no-touch” technique. We will describe our technique, the equipment we use to perform it, and demonstrate the safety and efficiency of the procedure.
METHODS
A retrospective chart review of 916 consecutive patients who underwent autologous buttock augmentation with fat grafting with the roller pump injection technique over a seven-year period from February 2009 to November 2016 was performed. Our study involved the collection and retrospective review of existing data, without influence on patient care or outcomes. Protected Health Information is stored in a secure electronic medical record. In accordance with the regulations of the US Department of Human Health Services28 and following the ethical standards set forth by the World Medical Association,29 this study is considered exempt from Institutional Review Board approval. Patients have scheduled follow ups for intervals throughout the first year and many remain with our practice for additional procedures and treatments in our spa for considerably longer.
All procedures were performed under general anesthesia by six surgeons at the same outpatient surgical center. Patients with indications for medical clearance determined by the algorithm of the American Society of Anesthesiologists for ambulatory patients were cleared preoperatively by their internist or cardiologist. Patient demographics and comorbidities, smoking status, use of oral contraceptives and hormone replacement therapies were reviewed. All patients who had this procedure performed by this method during that time period were included. Smokers were not excluded nor were special precautions made on their behalf. Our routine is to have patients discontinue oral contraceptives and hormone replacement therapies two weeks prior to surgery. Patients undergoing high risk procedures consisting of high volume liposuction (>5 L) or multiple concurrent procedures were started on therapeutic anticoagulation on the first postoperative day per our previously published routine.30 Those patients who had greater than 5 L of lipoaspirate after fat stratification or multiple concurrent procedures were kept overnight for monitoring. Our surgicenter limits large volume liposuction to a maximum of 10 liters. Concurrent procedures were any procedure outside of suction-assisted lipectomy and fat grafting and included those throughout the spectrum of aesthetic plastic surgery.
Surgical Technique
Patients are marked while standing in the preoperative area and prior to being escorted to the operative room where a standing, circumferential prep is performed with chlorhexidine scrub. The patients are then placed on the table with sterile towels and drapes beneath and above them. This is to allow us to change the patient’s position intraoperatively without having to repeat the prep and draping. All patients undergo general anesthesia during the operation. Depending on surgeon’s preference, a superwet or tumescent infiltration technique with 75 mL of 1% lidocaine, 3 mL of 1:1000 epinephrine, and 30 mg of triamcinolone mixed in 3 L Lactated Ringer’s solution is used in the marked regions for liposuction. No tumescent solution is used in the recipient site. The infiltration is performed using a roller pump (Medco Manufacturing, Spring, TX) and a 4-mm basket cannula and time is elapsed until there is noticeable skin pallor. Liposuction powered by the Medco Aspirator IV (Medco, Manufacturing) is then performed with or without the assistance of the MicroAire device (MicroAire Surgical Instruments, Charlottesville, VA) for contouring and fat harvest with either a 4- or 5-mm basket cannula. The lipoaspirate is collected in a 3000 mL autoclavable suction canister (Medela, McHenry, IL) and then permitted to stand and stratify by density. This process is passive and allowed to occur for at least 15 minutes and for up to as long as an hour while the liposculpture is being completed. The graft is prepared by extracting and discarding the layer of liquid that settles beneath the fat layer by direct suction using a small bore tumescent cannula. The small pores in this cannula allow for the extraction of tumescent fluid by suction without loss of fat aspirate. Approximately 10% or less of the fluid is left dispersed throughout the lipoaspirate. Early experience used additional filtration techniques to remove a greater percent of residual fluid but resulted in a less consistent flow. Permitting the retention of this small amount of fluid allows for the optimal propulsion of the fat graft through the roller pump and is resorbed by the body postoperatively. Antibiotic solution is then stirred in to the lipoaspirate. The routine additive is 1 gram of cephalexin per 3 liter canister. Gentamycin may be substituted for patients with penicillin allergy.
A section of infusion tubing compatible with the roller pump is spliced between two segments of liposuction tubing and secured with sterile-processed Hyper Tough 8-inch cable ties (Wal-Mart Stores, Inc., Bentonville, AR). One end of the liposuction tubing is placed within the collection canister and a 4- or 5-mm basket cannula is secured on the other end with a zip-tie. We do not use a 6-mm cannula in fat grafting cases as we have found that fat harvested with this caliber cannula will occasionally create a blockage in the system. The preparation of this equipment takes under a minute and may be prepared in the sterile field before the operation to save time (a video demonstrating the assembly is available online as Supplementary Material at www.aestheticsurgeryjournal.com). The interspersed segment of infusion tubing is passed off of the sterile field and placed into the roller pump machine (Figure 1). By foot pedal activation, lipoaspirate is the propelled by the roller pump from the collection canister back through the combination tubing and out of the basket-tipped cannula. The rate of injection is adjusted to about 3 mL/sec and can be measured by filling a 10 cc syringe in a 3 second period. With the equipment prepared ahead of time, the time from completion of liposuction until beginning the fat injection is under a minute.
Rendition of assembly of components for roller pump injection technique with regards to sterile and nonsterile components.
The fat graft injection may be performed in each buttock while the patient is in the lateral decubitus position immediately following the liposculpture of the ipsilateral hip, flank, and bra-line and then proceed to turn the patient to the other side and repeat. Additionally, it may be performed in both buttocks simultaneously while the patient is pronated after completing the bilateral liposculpture. Some surgeons will reuse incisions from the liposculpting and others will make a new stab incision about the midpoint along the superior gluteal contour and another out by the posterolateral thigh. The injection is performed from a superior to inferior direction employing a fanning pattern at varying depths to distribute the fat graft in layers between the muscle fascia and the dermis. Some expansion of the recipient bed is expected from the continuous tunneling of a 4- or 5-mm basket cannula during the fat injection but we do not otherwise perform any fashion of buttock site expansion techniques to increase receptive capacitance.
The vast majority of patients have an equal amount of graft injected into each buttock, but some patients have asymmetry or unilateral contour deformities that require additional fat to one side. Areas determined during the preoperative markings to be hollow and require contour correction have additional volume placed there until this region is flush with the surrounding tissue. This includes areas in the trochanteric region. Selected patients with narrow thighs who desire an augmented lateral buttock projection may require a tapering technique in which grafting is performed from the hip down to the mid-thigh or even the knee from a wide-to-narrow, thicker-to-thinner wedge-shaped fashion. However, lateral thigh fat grafting is not routinely performed. If performed nearly continuously, the rate of 3 mL/sec allows for a unilateral augmentation of 900 mL to be completed in five minutes. This rate of injection may be adjusted according to the surgeon’s level of comfort. A review of buttock fat grafting procedures during the month of March 2017 shows a range of volumes from 500 to 1700 mL per side performed over a range of 1 minute 53 seconds up to 6 minutes 15 seconds of pump activation time per side. The mean rate during these time trials was 4.53 mL/sec, not accounting for stoppages while the stopwatch continued.
The endpoint is determined by the patient’s goals and amount of fat available to be grafted. For those who wish for simple contour enhancement, grafting is ceased at the time which the surgeon has corrected all preoperatively determined areas of deficiency and has slightly overcorrected areas of desired prominence in hip, buttock, and thigh areas. For those who wish for maximal volume and projection, augmentation is performed until maximal skin capacitance. This is determined by a dimpled appearance and a firm, difficult to compress feeling of high skin turgor. Removal of the cannula from the port may allow for the efflux of some of the graft. Manual pressure for a few moments will occlude the port even after the pressure is removed. At the immediate completion of maximum volume grafting, the skin may have a dimpled appearance from subcutaneous deposits of fat which will resolve as the graft volume equilibrates to lower pressure areas over the first few minutes.
Secondary liposuction may be performed to refine the buttock and hip contour or adjust the volume and location of the fat grafting. Firm massage may also be used to modify contour where fat has been deposited. We do not close our stab incisions from liposculpture but some surgeons close the ports used for fat grafting. At the end of surgery, patients are dressed with compression garments selected depending on the regions where liposuction performed. We do not use specialized garments with cutouts or pressure relief for the buttocks nor do we implement any pressure precaution or instruction for positions to avoid.
RESULTS
Over a period of more than seven years, beginning in February 2009 until November 2016, 916 patients underwent liposculpting of the trunk with fat grafting of the buttocks facilitated by the roller pump injection method. The mean follow-up time for all patients was 325 days (range, 3 days-8 years and 5 days). The patients were 97% female (n = 895) with a mean age of 38.9 years (range, 18-76 years) and mean BMI of 26.3 kg/m2 (range, 17.2-56.5 kg/m2). Among the population 9.3% was actively smoking at the time of operation (n = 85) and 26 patients were known to be diabetic (2.8%). There were 106 patients taking oral contraceptive pills or other forms of hormone replacement up until two weeks prior to operation (Table 1).
Patient Population
| Description of population | No. of patients | Percentage of population |
|---|---|---|
| Female | 895 | 97% |
| Tobacco users | 85 | 9.3% |
| Hormone therapy/oral contraception use | 106 | 11.6% |
| Diabetes mellitus | 26 | 2.8% |
| Previous liposuction | 178 | 19.4% |
| Undergoing concurrent procedures | 502 | 54.8% |
| High volume liposuction | 118 | 12.9% |
| Description of population | No. of patients | Percentage of population |
|---|---|---|
| Female | 895 | 97% |
| Tobacco users | 85 | 9.3% |
| Hormone therapy/oral contraception use | 106 | 11.6% |
| Diabetes mellitus | 26 | 2.8% |
| Previous liposuction | 178 | 19.4% |
| Undergoing concurrent procedures | 502 | 54.8% |
| High volume liposuction | 118 | 12.9% |
Patient Population
| Description of population | No. of patients | Percentage of population |
|---|---|---|
| Female | 895 | 97% |
| Tobacco users | 85 | 9.3% |
| Hormone therapy/oral contraception use | 106 | 11.6% |
| Diabetes mellitus | 26 | 2.8% |
| Previous liposuction | 178 | 19.4% |
| Undergoing concurrent procedures | 502 | 54.8% |
| High volume liposuction | 118 | 12.9% |
| Description of population | No. of patients | Percentage of population |
|---|---|---|
| Female | 895 | 97% |
| Tobacco users | 85 | 9.3% |
| Hormone therapy/oral contraception use | 106 | 11.6% |
| Diabetes mellitus | 26 | 2.8% |
| Previous liposuction | 178 | 19.4% |
| Undergoing concurrent procedures | 502 | 54.8% |
| High volume liposuction | 118 | 12.9% |
The mean volume of separated fat removed from each patient was 3156 mL and the mean volume of fat grafted was 1807 mL total per patient. Among the patients, 12.9% (n = 118) underwent high volume liposuction (>5 L) and 54.8% (n = 502) underwent at least one other concurrent procedure. Over the course of our experience, the volume of fat grafting had increased overall (Figure 2). In 2009, the average amount of total fat grafting per patient was 1290 mL (645 mL per side). In 2016, the average amount of total fat grafted per patient is now 2004 mL, just over one liter (1002 cc) per buttock (Table 2). The maximum amount of fat grafting we have performed to date in a single patient was 4800 mL, or 2400 mL per buttock. This patient did not have any complications. We did not find any trends in the volume of liposuction performed per patient over this period nor did we identify a correlation between volume of liposuction and volume of fat grafting. Higher BMI does permit for larger fat grafting volumes (Figure 3).
Bar graph showing the mean total lipoaspirate and fat grafting volumes per year from 2009 to 2016. The blue bars represent mean lipoaspirate volumes and the yellow bars represent mean total fat grafting volume for each year of our seven year experience. The solid line represents the overall increase in total fat volume grafted per patient per year from 1290 mL to 2004 mL per patient.
Case Data by Year
| Year | No. of cases | Mean lipoaspirate volume (mL) | Mean total fat graft volume (mL) |
|---|---|---|---|
| 2009 | 58 | 3533 (range, 1300-10350) | 1291 (range, 200-2200) |
| 2010 | 68 | 3074 (range, 500-9500) | 1497 (range, 260-2400) |
| 2011 | 85 | 3357 (range, 300-10400) | 1655 (range, 200-3000) |
| 2012 | 98 | 3071 (range, 400-10900) | 1942 (range, 182-3400) |
| 2013 | 108 | 3293 (range, 600-10000) | 1873 (range, 190-3400) |
| 2014 | 153 | 3017 (range, 200-11500) | 1785 (range, 100-3200) |
| 2015 | 168 | 2792 (range, 400-8400) | 1890 (range, 200-4800) |
| 2016 (through Nov) | 178 | 3408 (range, 280-9600) | 2005 (range, 280-4400) |
| Year | No. of cases | Mean lipoaspirate volume (mL) | Mean total fat graft volume (mL) |
|---|---|---|---|
| 2009 | 58 | 3533 (range, 1300-10350) | 1291 (range, 200-2200) |
| 2010 | 68 | 3074 (range, 500-9500) | 1497 (range, 260-2400) |
| 2011 | 85 | 3357 (range, 300-10400) | 1655 (range, 200-3000) |
| 2012 | 98 | 3071 (range, 400-10900) | 1942 (range, 182-3400) |
| 2013 | 108 | 3293 (range, 600-10000) | 1873 (range, 190-3400) |
| 2014 | 153 | 3017 (range, 200-11500) | 1785 (range, 100-3200) |
| 2015 | 168 | 2792 (range, 400-8400) | 1890 (range, 200-4800) |
| 2016 (through Nov) | 178 | 3408 (range, 280-9600) | 2005 (range, 280-4400) |
Case Data by Year
| Year | No. of cases | Mean lipoaspirate volume (mL) | Mean total fat graft volume (mL) |
|---|---|---|---|
| 2009 | 58 | 3533 (range, 1300-10350) | 1291 (range, 200-2200) |
| 2010 | 68 | 3074 (range, 500-9500) | 1497 (range, 260-2400) |
| 2011 | 85 | 3357 (range, 300-10400) | 1655 (range, 200-3000) |
| 2012 | 98 | 3071 (range, 400-10900) | 1942 (range, 182-3400) |
| 2013 | 108 | 3293 (range, 600-10000) | 1873 (range, 190-3400) |
| 2014 | 153 | 3017 (range, 200-11500) | 1785 (range, 100-3200) |
| 2015 | 168 | 2792 (range, 400-8400) | 1890 (range, 200-4800) |
| 2016 (through Nov) | 178 | 3408 (range, 280-9600) | 2005 (range, 280-4400) |
| Year | No. of cases | Mean lipoaspirate volume (mL) | Mean total fat graft volume (mL) |
|---|---|---|---|
| 2009 | 58 | 3533 (range, 1300-10350) | 1291 (range, 200-2200) |
| 2010 | 68 | 3074 (range, 500-9500) | 1497 (range, 260-2400) |
| 2011 | 85 | 3357 (range, 300-10400) | 1655 (range, 200-3000) |
| 2012 | 98 | 3071 (range, 400-10900) | 1942 (range, 182-3400) |
| 2013 | 108 | 3293 (range, 600-10000) | 1873 (range, 190-3400) |
| 2014 | 153 | 3017 (range, 200-11500) | 1785 (range, 100-3200) |
| 2015 | 168 | 2792 (range, 400-8400) | 1890 (range, 200-4800) |
| 2016 (through Nov) | 178 | 3408 (range, 280-9600) | 2005 (range, 280-4400) |
Graph demonstrating the BMI of each patient with relationship to the total fat grafting performed. The solid line demonstrates a trend for patients with higher BMI to have had higher total volumes of fat grafting.
There were 16 total complications in 13 different patients for a complication rate of 1.4%. There were 10 complications limited to either donor or recipient sites of fat grafting for a complication rate of 1.1%. Five patients required intravenous antibiotics for cellulitis. Two of these patients had a cyst with cellulitis and were found to have purulence when drained. One patient had an oil cyst that aspirated 155 ml at 9 months postoperative and will require contour correction. There were no systemic infections, venous thrombolic events, fat embolic events, or deaths.
None of our complications occurred in patients who had less than 1000 mL of fat grafting. Twelve of 13 patients with complications had between 1000 mL and 2000 mL total of fat grafting (500 to 1000 mL per buttock). These patients had BMI range of 19.97 to 29.55. There was only one patient with complications that had more than 2000 mL of fat grafting despite there being 292 patients (31.8%) in this group. This patient had 3200 mL total of fat grafting (1600 mL per buttock), a BMI of 36.8, and had three complications related to abdominoplasty flap loss rather than to donor or recipient site of buttock fat grafting. There was no correlation between high volume liposuction, volume of fat grafting, or multiple concurrent procedures and overall risk of complications (Table 3).
Patient and Complication Breakdown by Volume of Fat Grafted
| Total volume of fat graft (mL) | No. of patients | Percentage of population | Complications |
|---|---|---|---|
| 1-999 | 137 | 15% | 0 |
| 1000-2000 | 487 | 53.2% | 12 |
| 2001-4800 | 292 | 31.8% | 1 |
| Total volume of fat graft (mL) | No. of patients | Percentage of population | Complications |
|---|---|---|---|
| 1-999 | 137 | 15% | 0 |
| 1000-2000 | 487 | 53.2% | 12 |
| 2001-4800 | 292 | 31.8% | 1 |
Patient and Complication Breakdown by Volume of Fat Grafted
| Total volume of fat graft (mL) | No. of patients | Percentage of population | Complications |
|---|---|---|---|
| 1-999 | 137 | 15% | 0 |
| 1000-2000 | 487 | 53.2% | 12 |
| 2001-4800 | 292 | 31.8% | 1 |
| Total volume of fat graft (mL) | No. of patients | Percentage of population | Complications |
|---|---|---|---|
| 1-999 | 137 | 15% | 0 |
| 1000-2000 | 487 | 53.2% | 12 |
| 2001-4800 | 292 | 31.8% | 1 |
Fourteen patients (1.5%) had subsequent fat grafting procedures of the buttocks secondary to patient preference for greater volume, half of which had their initial procedure during our early experience and had less than 1000 mL grafted per buttock. Four patients (0.44%) had secondary grafting procedures due to postoperative asymmetry, either persistent or graft-take related. One patient (0.1%) was identified to have developed a lateral contour defect secondary to patient postoperative positioning from nighttime side-sleeping and required additional fat grafting to correct the hip contour. The patient who had the large oil cyst aspirated will require additional fat grafting to improve the symmetry of the buttocks (Figure 4).
A contour deformity of the right buttock secondary to aspiration of a large oil cyst. The arrow indicates the location where there is flattening of the anterior-posterior projection secondary to loss of volume.
DISCUSSION
Our practice has seen a dramatic increase in the number of patients wishing for autologous buttock augmentation as well as an increase in the volume required to meet their goals. The roller pump injection technique has helped us to be able to meet this demand by facilitating large volume fat grafting in a way that is both rapid and safe. The setup for this method is completed in under a minute and can be performed at any time in the sterile field. The minimal manipulation required to prepare the graft also takes less than a minute and therefore does not occupy a surgical assistant during time under anesthesia for any longer than that.
Our dataset begins with the start of our experience with the roller pump injection technique. Since utilization of this method for all patients undergoing buttock augmentation with fat in 2009, we have gone from 58 cases in that first year to 178 this year with two months left remaining. Figure 1 shows the overall trend of volume injected has steadily increased, starting with an average of 1290 mL total graft in 2009 up to 2004 mL total graft today. This mirrors the trends from the series published by Cardenas-Camarena et al6 in which they demonstrated the progression in their experience in different phases. There were no patients in their series that had more than 500 mL per buttock grafted in the initial phase of the study and then by the time the accrual of their last phase had concluded, there was a considerable constituent with 501 to 1160 mL grafted per side.
As demonstrated by the increase in the number of cases and average amount of graft volumes over the course of our experience, we affirm that there is an evolution in buttock aesthetics. Demands for higher volumes are influenced by the increasing awareness of the “Brazilian Butt Lift” from popular culture and through social media. At the time that this manuscript is being prepared, there are 157,652 hashtags for “BBL” and 39,998 for “Brazilian Butt Lift” on Instagram alone and 3,510,000 results on Google for the search “celebrity Brazilian Butt Lift.”
Patient satisfaction with this technique has been validated by survey12,13 and is also evident in our hands by the low revision rate. Though we did not otherwise measure patient-reported outcomes. Since implementing the roller pump injection technique, only 18 patients (1.96%) have opted to repeat the procedure and one additional patient is planning to in the future. Three patients had asymmetry after the procedure and one patient had a preoperative contour deformity that went undercorrected and remained persistent through the first operation. One patient who was found to have a lateral contour flattening early in the postoperative period was identified as a side-sleeper. We would recommend that patients who are known to sleep on one side should sleep in the supine position with a pillow beneath their knees for the first two weeks to prevent turning during sleep. Otherwise, we do not implement a postoperative regimen of positioning, we simply advise patients to ambulate frequently during the day and make a conscious effort to shift their weight often when at rest. Of the 14 patients who had revisions secondary to patient preference for greater volume, about half of these patients underwent their initial operation early in our experience with this technique and their total graft volumes were less than 1000 mL per buttock. This, combined with lower volume goals set by patients years ago, would explain why this subset of patients returned for a second procedure.
Even though we have found an increasing amount of total volume of fat grafting over time, we do not show a trend of increasing volumes of liposuction across the population to match this, nor is there a relationship between the amount of liposuction and the amount of fat grafting performed on each individual patient. In some patients, the maximum amount of fat grafting performed is limited by the amount of fat provided by liposuction. In others, the volume of fat grafted is capped by the buttocks capacitance to receive fat, or by patient desire for subtle augmentation and there is often volume of graft left to be discarded. Cardenas et al do not recommend overcorrection of the buttock aesthetic, they recommend optimal contour on table because they feel that overcorrection leads to fat necrosis.6 Many of our patients request that we achieve the largest volume of grafting possible and as our comfort has grown with this concept, we often reach the point of maximal capacitance in each buttock while maintaining the patients’ desired buttock shape (Figures 5 and 6 and Supplemental Figure 1, available online at www.aestheticsurgeryjournal.com). Despite this, we have only discovered three (0.33%) clinically detectible cysts and four (0.44%) cases of postoperative asymmetry requiring intervention.
(A, C, E) Preoperative and (B, D, F) 7-year postoperative photographs of this 34-year-old woman who underwent liposuction of the abdomen, flanks, back, inner thighs, and knees with augmentation of the buttocks with fat grafting in 2010. Total of 3100 mL of fat was removed and fat grafting of 1100 mL per buttock was performed. She underwent subsequent abdominoplasty in 2011. Postoperative pictures were taken in 2017 at 7 years.
(A, C, E) Preoperative and (B, D, F)15-month postoperative photographs of this 24-year-old woman who underwent abdominoplasty with liposuction of the back, flanks, hips, and axillae with augmentation of the buttock and trochanteric region with fat grafting. Total of 2400 mL of fat was removed and 1200 mL was grafted per side.
The rate of complications that we have shown in our series, 1.4%, is lower than that previously established in the literature by any single center or meta-analysis. Recent meta-analyses reporting complication rates of 9.9%22 and 7%23 include low volume studies and centers reporting earlier experiences. Despite having treated over 900 patients, we were unable to identify specific variables associated with elevated complication risk due to the power of this study and the low percentage of complications overall and within groups.
We attribute our low complication rate to strict adherence to the safety standards that govern our practice. These standards are created by a set of guidelines that apply to preoperative, intraoperative, and postoperative decision making. Patients are warmed with forced air gowns and hydrated with intravenous fluids ahead of the operation in the preoperative holding area. Patients identified preoperatively as high risk for DVT by the assessment model created by Caprini31 and validated for plastic surgery patients by Pannucci et al,32 undergoing multiple or high risk procedures, or have high volume liposuction are started on either oral or injectable prophylactic anticoagulation on postoperative day 1 as per our routine.30 These patients are also kept overnight in our facility for observation. Sterile sequential compression devices are applied to the calves on all patients during the operation and patients are encouraged to ambulate every two hours postoperatively. There were no suspected or confirmed DVTs or PEs in the study population.
Fat embolism is the most devastating complication in buttock augmentation with fat grafting and we have been fortunate not to have seen this pathology first hand.33 Last year, Cardenas-Camarena et al compiled a thorough examination of a group of patient mortalities in Mexico and Colombia sustained over the past 15 years.25 They describe the pathology as either a macroscopic or microscopic process, which have different mechanisms of infiltration and injury, presenting times and symptomatology, and may result in different outcomes. In this review, it was found that 19 of 27 deaths occurred intraoperatively and the remaining 8 occurred within the first 24 hours. Therefore, early recognition is essential to provide supportive care as soon as possible to attempt to alter the outcome of this process. The average amount of fat grafted in the mortality group was just 214 mL per buttock. Some of the autopsies demonstrated macroscopic fat in the vena cava, heart, and lungs. Fat globules found in the gluteus muscle was a common finding. In early techniques, this was intentional. This comes as no surprise as Guerrerosantos has shown in both animal and clinical models show better take of fat grafting if infiltrated into a highly vascular bed such as muscle tissue.34-37 It was not initially considered that this would be suspected as the route for infiltration for microscopic fat emboli or the proximity allowing injury to the gluteal vessels permitting macroscopic fat emboli. In the cases of macroscopic fat embolism and on-table death, traumatic injury to the gluteal vein was identified on autopsy.25 The gluteal veins enters the pelvis through the greater sciatic foramen and converge into the internal iliac vein. For a patient in the lateral or prone position, this foramen is open towards the inferior direction. The gluteal veins are thus vulnerable to injury as they enter the foramen by a cannula passed from an inferior to superior direction. When the cannula is passed into the buttock from a superior to inferior direction, access to the foramen is shielded by the ileum and by a network of connective ligaments (Figure 7). In order to reduce risk of causing microscopic fat embolic syndrome, our intent is to avoid injection of fat into the gluteus muscles. To reduce risk of occlusive macroscopic fat emboli, injection is only performed in the tissue superficial to the muscle, using a 4-mm or 5-mm cannula from incisions at the superior margin of the buttocks in a superior to inferior direction to avoid direct injury to the large bore veins entering the pelvis. We also no longer use an open-ended cannula for grafting, replacing it with a rounded-tip basket cannula only. These safety measures have been our standard of practice and are inclusive but not limited to the recommendations recently published by the Aesthetic Society Education and Research Foundation (ASERF) Task Force on Mortality from Gluteal Fat Grafting.38
(A, B) Access to the sciatic foramina is shielded from injection approaches from the sacral and lateral access points. (C) A dangerous trajectory is shown as cannula passed from infragluteal crease may enter into the open foramina.
Prevention of seroma is directed mainly at fat harvest locations by employing compression garment strategies and by leaving incisions for liposuction and fat grafting open to drain. With just a few moments of gentle pressure, the fat grafting sites do not have loss of the fat graft. The liposuction incisions may drain a copious amount over the first 24 hours and may persist to drain lesser amounts over the next 48 to 72 hours. This is by intention as coupled with compression garments selected to cover appropriate regions, the body is tasked with having less fluid to have to resorb and thus lowering the likelihood of seroma. Klein found that this strategy also resulted in fewer incidences of panniculitis by not only encouraging the rapid drainage of residual tumescent fluid but also the elimination of proinflammatory postsurgical subcutaneous exudate.39
To further control the incidence of postoperative infections, we have employed a number of strategies. Each patient is provided with chlorhexidine soap and instructed to use for whole body the night before and morning of the procedure. When the patients arrive to the surgery center, they are then to use chlorhexidine wipes across the whole body in the preoperative dressing area. We then perform a standing, circumferential prep with chlorhexidine scrub and the incision sites are treated with chlorhexidine alcohol-based prep after the time-out is performed. Appropriate intravenous antibiotics are administered prior to initiation of the case. Prior to fat injection, 1 gram of cephalexin is added to each canister of lipoaspirate after evacuation of settled liquid. Additionally, the minimal manipulation and closed nature of the roller pump injection system limits exposure to any equipment or surfaces outside of the circuit. Regarding the use of steroids in the tumescent solution, Klein no longer includes the triamcinolone that was part of the original tumescent formulation and caveats the possible increased risk of infection secondary to steroid-induced impairment of immune function.39 We believe the observed benefit of reduced soreness and bruising outweighs the potential risk infection and we continue to include it in our tumescent formula.
The purpose of this study was to introduce our methods for achieving fat grafting to the buttocks and demonstrate that when performed with certain safety measures in mind, the minor complication rate is low and major complications can be avoided. This study did not assess graft retention rate, or compare outcomes against other methods of performing this procedure. MRI or 3-dimensional topographic studies would be feasible, although expensive methods that could be utilized to compare these outcomes. We did not directly measure patient satisfaction by means of survey, although others have in the past and demonstrated high patient satisfaction rates.12,13 We assume patient satisfaction based on the increasing number of procedures performed each year by our group by referral, our low overall revision rate, and our patients return for additional nonrelated procedures and continue with spa services. Our patients are routinely seen during the first six months after the procedure, but many do not return beyond that. Some patients are lost to follow up earlier than several months and others have been seen and photographed as far as seven-years postoperative. We have a large constituency of patients who travel to our practice for the procedure and leave the city after an initial postoperative check. Patients that are either from out of town and leave after initial postoperative check or those whom are lost to follow up earlier than anticipated may cause an underreporting of some complications such as asymmetry but major complications such as occlusive fat embolism, fat emboli syndrome, pulmonary embolism, deep vein thrombosis, sepsis, dehydration, and infection would all have occurred during the period in which all patients are still under our care. Long-term (multiple year) patient satisfaction has not been thoroughly studied in buttock fat grafting performed by any method and considering the current popularity of the procedure, we as a community should aim to do so to continue to refine our techniques for future patients.
CONCLUSION
For the last seven years, we have performed our gluteal augmentation with fat grafting using the roller pump injection technique. This method of lipoinjection is safe and both time and cost saving, requiring little in terms of specialized equipment and spending less time in the operating room reducing the amount of time spent under anesthesia. Our complication rate is low because of strict adherence to safety measures inclusive of, but not limited to, those recommended by the ASERF Task Force.38 Despite being the largest series found in the literature to date, our complication rate is low, and additional randomized, prospective and multicenter studies would be required to identify risk factors associated with elevated complication risk.
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 financial support for the research, authorship, and publication of this article.
Acknowledgements
We would like to acknowledge Nicole Lemberg, Gunner West, Sarah Hays and James Mentz for their contributions towards the management of our dataset and preparation of this manuscript and accompanying figures.
REFERENCES
Author notes
Dr Everett is a fellow at a private plastic surgery practice in Houston, TX.
Drs Morales, Newall, Fortes, Hustak, Patronella, and Mentz are plastic surgeons in private practice in Houston, TX
