Fat has the potential to be the ideal soft-tissue COSMETIC. Fine Tuning Lipoaspirate Viability for Fat Grafting.

Transcription

1 COSMETIC Fine Tuning Lipoaspirate Viability for Fat Grafting J. Lauren Crawford, M.D. Bradley A. Hubbard, M.D. Stephen H. Colbert, M.D. Charles L. Puckett, M.D. Columbia, Mo. Background: The efficient harvest of abundant viable adipocytes for grafting is of considerable interest. Hand aspiration, low-g-force, short-duration centrifugation, and harvest of the lower sublayer of fat centrifugate maximize viable adiopocytes, but this process is cumbersome with conventional equipment. The Lipose Corporation (Greenwich, Conn.) has produced special syringes, filters, and a low-g-force centrifuge (Viafill system) to facilitate this process. The adipocyte viability using this system is presented. Methods: Six women underwent fat graft harvest using the Viafill system from the lower hips (n 6) and/or upper hips (n 3). After centrifugation for 2 minutes at 50 g, the lower, middle, and top sublayers of the adipose layer were analyzed for viable adipocyte counts using trypan blue vital staining. Additional samples from standard power-assisted liposuction were obtained and analyzed similarly. Results: The mean difference in square-root transformation of cell counts between the bottom sublayer of centrifuged fat and the middle sublayer was 0.95 (95 percent CI, 0.61 to 1.3), and the difference between the middle and top sublayers was 0.67 (CI, 0.50 to 0.84). Thus, the bottom sublayer had approximately 2.5 to 3 times more cells than the top sublayer. The difference between the hand aspirate samples and the power-assisted liposuction samples was significant (1.62; CI, 1.35 to 1.90). Conclusions: This study reconfirms the authors early findings that atraumatic harvest of lipoaspirate yields high cell counts and that adipocyte density is greatest at the lowest sublayer of centrifuged fat. The Viafill system provides a more efficient and user-friendly system for fat grafting while maintaining cell counts similar to the authors technique using conventional equipment. (Plast. Reconstr. Surg. 126: 1342, 2010.) From the Division of Plastic Surgery, University of Missouri. Received for publication August 17, 2009; accepted April 2, Copyright 2010 by the American Society of Plastic Surgeons DOI: /PRS.0b013e3181ea44a9 Fat has the potential to be the ideal soft-tissue filler because it is abundant, easily accessible, inexpensive, and host compatible, and because it can be harvested repeatedly. Successful and long-term results from free fat grafting used in soft-tissue augmentation have long been a goal for plastic surgeons. A large array of applications has been reported previously: cosmetic enhancement and rejuvenation, body contour improvement, and reconstruction of scarred and radiation-damaged sites, among others. 1 7 The realization that facial aging is partially a result of soft-tissue atrophy has spurred renewed interest. An exciting development has been the demonstration that lipoaspirate contains stem cell concentrations exceeding those of bone marrow. 8,9 The regenerative qualities of lipoaspirate stem cells have already found clinical applications by reversing the damage of radiation injury 6,7 and as a source of stem cells in tissue engineering. 10 The applications of adipose progenitor cells will undoubtedly continue to grow in number. The historical drawback to fat grafting has been the unpredictable nature of its outcomes. The most probable argument for this phenome- Disclosures: None of the authors holds a financial position with the Lipose Corporation. The Lipose Corporation supplied funding for the facility fees and equipment used in the liposuction procedures and the use and modification of the equipment discussed. No professional fees were obtained from the patients or the Lipose Corporation

2 Volume 126, Number 4 Fine Tuning Lipoaspirate Viability non is Peer s cell survival theory. 11 Simply stated, the number of viable cells grafted correlates with the ultimate volume of graft survival. Antiquated techniques have used traumatic approaches and, in retrospect, the poor or inconsistent results are not surprising. Using the Coleman technique or similar atraumatic approaches, acceptable and reliable resorption rates have been obtained However, there is a large variability in the methods used in terms of harvest, processing, grafting method, and recipient location. In addition, there may be differences in the intrinsic quality of the adipose tissue from person to person or regionally within the same person. Thus, the published resorption rates vary wildly from 20 to 90 percent Because of such variability in techniques and inconsistencies in resorption rates, the ability to efficiently harvest large quantities of viable fat cells continues to be of great interest. In 2001, we reported a method for obtaining the maximum number of viable fat cells to be used for grafting. 19 Since that study, our approach has been that of hand aspiration; low-g-force, short-duration centrifugation; and use of the lower centrifuged sublayer of fat cells for our fat grafting cases. However, the process has been cumbersome and time consuming using commercially available equipment. Since that time, the Lipose Corporation (Greenwich, Conn.) has expressed an interest in producing a line of equipment to facilitate this process. The purpose of the present study is to investigate the technology of the Viafill (Lipose) fat graft harvest system and to assess lipocyte viability as compared with our previous results. PATIENTS AND METHODS Six female subjects were enrolled in the study. University of Missouri Health Sciences Institutional Review Board approval was obtained before patient enrollment. The average age of the subjects was 36 years (range, 24 to 47 years). None of the patients had significant medical histories and none had undergone previous liposuction or body contouring procedures. Half of the subjects had liposuction on bilateral upper and lower hips, and half had liposuction on bilateral lower hips alone, for a total of 18 body areas. Hand lipoaspirate and power-assisted lipoaspirate specimens were obtained from each body area. For each patient, our standard tumescent fluid (1 ml of 1:1000 epinephrine and 50 ml of 1% lidocaine per liter of lactated Ringer s solution) was injected into the harvest site. Fat was aspirated by hand using the Viafill system with a 2.10-mm 12-cm hydrophilic coated disposable cannula (Tulip Biomed, San Diego, Calif.) and the 20-ml Viafill syringe. Samples (15 to 20 ml) were harvested from each area, and these syringes were then placed in a 37 C warm water bath. After the hand-suctioned specimens were taken, standard power-assisted liposuction (Accelerator-III; Byron Medical, Inc., Tucson, Ariz.; MicroAire Surgical Instruments, Charlottesville, Va.) using 4-mm cannulas was performed at the same site and fat was collected in 1.5-liter lipoaspirate canisters using atmosphere negative pressure. Lipoaspirate volumes ranged between 600 and 1000 ml. The lipoaspirate canisters were maintained within the water bath with the other samples for less than 60 minutes. The power-assisted liposuction lipoaspirate canisters were gently stirred into a homogeneous mixture, and one 15- to 20-ml sample was drawn into a Viafill syringe to serve as a control. Each Viafill syringe was then placed into a filtered Viafill centrifuge tube (Figs. 1 through 3) and placed within the specially fitted 50-g centrifuge (Lipose). All tubes were spun at 50 g for 2 minutes. The filtered Viafill centrifuge tube allows the aqueous layer and erythrocytes to flow out of the syringe tip during the centrifugation process, leaving the adipose layer at the bottom, near the tip, and the oil layer at the top (Fig. 4). One-milliliter samples of adipose were taken from the bottom, middle, and top sublayers of the centrifuged fat layer. Each of these samples was then processed with 1 mg/ml of collagenase (collagenase type Fig. 1. The syringe and filter of the Viafill (Lipose) system prototype are shown. The same syringe used for harvest is placed within the filtered centrifuge tube and also can be used for distribution of the fat to smaller syringes for injection. After hand suctioning, the syringe is placed into the filter for centrifugation. 1343

3 Plastic and Reconstructive Surgery October 2010 Fig. 2. A syringe from the Viafill system with a plunger handle in place and a syringe with the plunger unlocked and removed. Fig. 3. The centrifuge prototype of the Viafill (Lipose) system prototype is shown with two filters and syringes in place. 1, filtered; Worthington, Lakewood, N.J.) in a warm water bath. After 1 hour of collagenase digestion, the specimens were diluted 1:1 with trypan blue vital stain 0.4% solution (product T8154; Sigma, St. Louis, Mo.). A 100- l sample was taken from each digested and stained preparation and the numbers of viable fat cells were counted with a hemocytometer under 400 magnification. Five different squares from the hemocytometer grid were recorded three separate times for each 100- l sample. After the trypan blue dilution was factored in, the number of viable fat cells from each sample was determined. This fat cell quantification method has been used previously by our laboratory and has been described by Moore et al.19, Fig. 4. Viafill syringe and centrifuge tube after centrifugation. Note that the heavier aqueous layer and erythrocytes (arrow) have passed out of the syringe through the filter and the adipose and oil layers remain. Statistical Analysis Statistical analysis was completed with assistance from the Division of Biostatistics, University of Missouri. All statistical analysis was performed using SAS v9 (SAS Institute, Inc., Cary, N.C.). Subgroup analysis was performed based on body area (upper or lower hips), laterality (right or left), and the sublayer from within the centrifuged adipose layer from which the specimen was taken (bottom, middle, or top). The cell counts were skewed to the high end (i.e., few values were close to 0). We therefore found it useful to analyze the data using the square-root transformation of the cell counts. Differences between square-root transformations were then averaged. Because measurements were taken on the same patient, there are dependencies in the data attributable to multiple observations on the same subject. Consequently, the patient was included in the analysis as a random effect and a mixed model analysis of variance was carried out. RESULTS Table 1 lists the mean and median cell counts for each centrifugate adipose sublayer from each body area and by harvest method. One body area in one patient had abnormally low cell counts; otherwise, the counts were distributed normally. The outlier was not found to adjust conclusions when left out or when kept in and was considered insignificant. Figure 5 shows a box plot of the square root transformation of the cell counts from

4 Volume 126, Number 4 Fine Tuning Lipoaspirate Viability Table 1. Lipocyte Cell Counts* Harvest Method No. Mean Median SD Minimum Maximum Top Upper PAL Hand aspirate Lower PAL Hand aspirate Middle Upper PAL Hand aspirate Lower PAL Hand aspirate Bottom Upper PAL Hand aspirate Lower PAL Hand aspirate PAL, power-assisted liposuction. *Lipocyte cell counts (in millions of cells per milliliter) separated by centrifuge sublayer (bottom, middle, and top), location (upper or lower hip), and method of harvest (PAL or hand aspirate). each centrifugate sublayer of the hand and powerassisted liposuction aspirates. In all body areas, the hand aspirate samples had significantly higher mean cell counts than power-assisted liposuction samples (p ). As expected, the laterality, right compared with left, did not have a significant effect (p 0.90). Consistent with our previous work, the cell counts at the bottom of the centrifugate fat sublayer were greater than those of the middle sublayer, which were greater than those of the top sublayer (p ). Somewhat surprisingly, the upper hip area had significantly greater cell counts compared with the lower hip (p ). Considering all sublayers and all body areas by square-root analysis, the hand aspirate samples had greater numbers of viable cells than powerassisted liposuction samples (mean difference in square roots, 1.62; 95 percent CI, 1.35 to 1.90). Similarly, the mean of the square roots from the bottom sublayer was greater than that of the middle sublayer (0.95; 95 percent CI, 0.61 to 1.3), and the mean of the middle sublayer was greater than that of the top sublayer (0.67; 95 percent CI, 0.50 to 0.84). By syllogism, the bottom sublayer squareroot transformation was also greater than that of the top sublayer (mean difference, 1.62; 95 percent CI, 1.45 to 1.79). The mean difference in square-root transformations from the upper hip area compared with the lower hip area was statistically significant at 1.63 (95 percent CI, 1.47 to 1.80). DISCUSSION Debate continues in the literature as to the best method of fat graft harvest. Despite the absence of an undisputed best method for long-term volume restoration, a consensus seems to have developed. A recent survey has shown that a total of 70 percent of plastic surgeons performing fat grafts use either the Coleman technique or another atraumatic syringe aspiration technique, 54 percent and 16 percent, respectively. 21 In addition, nearly half of these surgeons (49 percent) use centrifugation as opposed to other techniques of fat preparation such as washing or gravity. 21 As mentioned, there are many conflicting publications comparing the different techniques for fat graft harvest. It is beyond the scope of this article to revisit all of these studies. However, our results do reinforce the current consensus view, using centrifugation and atraumatic harvest. Disagreement continues to surround the centrifugation of the lipoaspirate before grafting. In 2008, Kurita et al. found a decrease in potentially detrimental erythrocytes, but found no difference in viable adipose cells after centrifugation at 1200 g. 22 However, the authors did not make a distinction between the sublayers of centrifuged fat. The filter system in the Lipose centrifugation tube allows erythrocytes to be removed with the aqueous layer, thus addressing the potential erythrocyte problem. This makes available the direct transfer of the highest quantity of 1345

5 Plastic and Reconstructive Surgery October 2010 Fig. 5. Box plot of the square-root transformation of adipose cell counts for hand aspirate and power-assisted liposuction (PAL). The area of harvest was ignored; therefore, there are 18 samples (12 lower and six upper hip) represented by each box plot. Both groups were separated into the bottom, middle, and top centrifuged samples. Box demarcates the first and third quartiles;, mean;, median; asterisks, outliers. viable fat cells from the capture syringe to an injection syringe for grafting. Our current study reconfirms our previous work (Boschert et al.) and that of Butterwick that low-force, short-duration centrifugation does increase viable adipose cell counts. 19,23 In addition, we have reconfirmed our finding that viable cell density is greater at lower sublayers of centrifuged fat. The purpose of the study was not to perform a head-to-head comparison of harvest techniques; thus, no true control group was created. We are convinced from our previous work and that of others that atraumatic hand aspiration is superior to other methods (e.g., power-assisted liposuction or conventional liposuction). 19,24 Nevertheless, in this study, atraumatic hand aspiration provided a greater number of viable fat cells relative to conventional liposuction. This finding is limited by our harvesting of the hand-aspirated samples first in all body areas and the larger volume of powerassisted liposuction aspirate. Although this has not been studied previously, the order of harvest could potentially have an effect on cell numbers. The first pass of liposuction might obtain higher fat cell counts because of less trauma or could obtain lower numbers because of the dilutional effect of the tumescent fluid. This would be an interesting concept for future study. In addition, a head-tohead comparison would be flawed because the power-assisted liposuction samples represent the average cell counts from a larger volume of aspirate. Regardless, it was not our intent to compare power-assisted liposuction to Lipose syringes per se, but merely to show whether the Lipose system is efficient and user-friendly while maintaining cell counts similar to our previous work. Despite the introduction of potential confounding variables, order of harvest and volume of harvest, the present study found that atraumatic hand aspiration provides 150 percent greater viable fat cells relative to conventional liposuction. It was not the intention of our protocol to investigate the superiority of donor sites for fat grafting. Recently, such a study has been performed. Rohrich et al. compared thigh, knee, flank, and abdominal donor sites. 25 They found no difference in viable adipocytes obtained from these areas. Surprisingly, our results seem to contradict this finding, and other studies would suggest a potential difference among donor sites. Padoin et al. examined the effect of donor site on quantity of mesenchymal stem cells in fat aspirates and found significant differences between some regions. 26 In our patients, the upper hip area adipocyte cell counts were significantly higher than those of the lower hip. It is important to note our 1346

6 Volume 126, Number 4 Fine Tuning Lipoaspirate Viability small patient sample size and that both areas were harvested in only half of the patients. Although statistical significance was reached, with this small sample we feel that the potential differences in donor sites should be the focus of additional investigation. As mentioned previously, we have been dissatisfied with the commercially available equipment for fat graft harvest, centrifugation, and grafting. The Viafill system relieves many of these complaints. A single 20-ml syringe is connected by means of a Luer lock to a harvesting cannula of choice. After the sample is acquired, the syringe is then secured into a filtered centrifuge holder and the plunger handle is removed. During centrifugation, the aqueous layer and erythrocytes are filtered from the syringe into the centrifuge holder. With the plunger handle reattached, the same syringe can be connected to smaller syringes (connectors provided) that can then be used for grafting. When comparing the results of this work to our previous studies, the Viafill system did yield greater cell counts, 20 compared with 8 million cells per milliliter. Obviously, such a comparison is scientifically flawed and therefore statistical analysis was not performed. However, the difference in cell count by sublayer (bottom, middle, and top) was similar. Both studies found the bottom fat sublayer to contain more viable cells than the middle, and the middle to contain more than the top fat sublayer by approximately 150 percent each. Both studies also found that the bottom fat sublayer has 2.5 to 3 times more viable cells relative to the top fat sublayer. Thus, a gradient exists, with the density of viable cells increasing from top to bottom of the centrifugate fat layer. This study is limited by the lack of confirmation of viable cell counts obtained with trypan blue staining with other methods. For example, 3-(4,5- dimethylthiazol-2-yl)-2,5-diphenyl tetra sodium bromide or glycerol-3-phophatase dehydrogenase assays would have to be performed for the absolute viable cell counts to be comparable to other works. A viable cell count is not the only factor predicting the success of fat grafting techniques. Peer s cell survival theory has scientific merit but certainly does not account for all the variables involved. 11 The popularity of cell counts as the endpoint for fat graft research is certainly related to the ease by which they can be obtained. Adipose cell volume, area of harvest, method of grafting, recipient location, and person-to-person variability all will affect the success of fat grafting procedures. Possibly, with the renewed popularity of fat grafting procedures, these clinical answers are on the horizon. CONCLUSIONS This study reconfirms our early findings that atraumatic harvest of fat grafts continues to yield high viable cell counts and that the adipose cell density is greatest at the lowest level of centrifuged fat. The Viafill system provides an efficient and user-friendly system for fat graft harvest while maintaining cell counts similar to those of our technique using conventional equipment. Charles L. Puckett, M.D. Division of Plastic Surgery University of Missouri One Hospital Drive Columbia, Mo bolandd@health.missouri.edu REFERENCES 1. Coleman SR. Facial recontouring with lipostructure. Clin Plast Surg. 1997;24: Coleman SR. Hand rejuvenation with structural fat grafting. Plast Reconstr Surg. 2002;110: Gatti JE. Permanent lip augmentation with serial fat grafting. Ann Plast Surg. 1999;42: Guerrerosantos J. Autologous fat grafting for body contouring. Clin Plast Surg. 1996;23: Roberts TL III, Weinfeld AB, Bruner TW, Nguyen K. Universal and ethnic ideals of beautiful buttocks are best obtained by autologous micro fat grafting and liposuction. Clin Plast Surg. 2006;33: Rigotti G, Marchi A, Galie M, et al. Clinical treatment of radiotherapy tissue damage by lipoaspirate transplant: A healing process mediated by adipose-derived adult stem cells. Plast Reconstr Surg. 2007;119: Phulpin B, Gangloff P, Tran N, Bravetti P, Merlin JL, Dolivet G. Rehabilitation of irradiated head and neck tissues by autologous fat transplantation. Plast Reconstr Surg. 2009;123: Zuk PA, Zhu M, Ashjian P, et al. Human adipose tissue is a source of multipotent stem cells. Mol Biol Cell. 2002;13: Strem BM, Hicok KC, Zhu M, et al. Multipotential differentiation of adipose tissue-derived stem cells. Keio J Med. 2005; 54: Mischen BT, Follmar KE, Moyer KE, et al. Metabolic and functional characterization of human adipose-derived stem cells in tissue engineering. Plast Reconstr Surg. 2008;122: Peer LA. Cell survival theory versus replacement theory. Plast Reconstr Surg. 1955;16: Billings E Jr, May JW. Historical review and present status of free fat graft autotransplantation in plastic and reconstructive surgery. Plast Reconstr Surg. 1989;83: Sommer B, Sattler G. Current concepts of fat graft survival: Histology of aspirated adipose tissue and review of the literature. Dermatol Surg. 2000;26: Coleman SR. Structural fat grafts: The ideal filler? Clin Plast Surg. 2001;28:

7 Plastic and Reconstructive Surgery October Coleman SR. Structural fat grafting: More than a permanent filler. Plast Reconstr Surg. 2006;118: Nguyen A, Pasyk KA, Bouvier TN, Hassett CA, Argenta LC. Comparative study of survival of autologous adipose tissue taken and transplanted by different techniques. Plast Reconstr Surg. 1990;85: Boyce RG, Nuss DW, Kluka EA. The use of autogenous fat, fascia, and nonvascularized muscle grafts in the head and neck. Otolaryngol Clin North Am. 1994;27: Mattsudo PK, Toledo LS. Experience of injected fat grafting. Aesthetic Plast Surg. 1988;12: Boschert MT, Beckert BW, Puckett CL, Concannon MJ. Analysis of lipocyte viability after liposuction. Plast Reconstr Surg. 2002;109: Moore JH Jr, Kolaczynski JW, Morales LM, et al. Viability of fat obtained by syringe suction lipectomy: Effects of local anesthesia with lidocaine. Aesthetic Plast Surg. 1995;19: Kaufman MR, Bradley JP, Dickinson B, et al. Autologous fat transfer national consensus survey: Trends in techniques for harvest, preparation, and application, and perception of short- and long-term results. Plast Reconstr Surg. 2007;119: Kurita M, Matsumoto D, Shigeura T, et al. Influences of centrifugation on cells and tissues in liposuction aspirates: Optimized centrifugation for lipotransfer and cell isolation. Plast Reconstr Surg. 2008;121: Butterwick KJ. Lipoaugmentation for aging hands: A comparison of the longevity and aesthetic results of centrifuged versus noncentrifuged fat. Dermatol Surg. 2002;28: Pu LL, Coleman SR, Cui X, Ferguson RE Jr, Vasconez HC. Autologous fat grafts harvested and refined by the Coleman technique: A comparative study. Plast Reconstr Surg. 2008; 122: Rohrich RJ, Sorokin ES, Brown SA. In search of improved fat transfer viability: A quantitative analysis of the role of centrifugation and harvest site. Plast Reconstr Surg. 2004;113: ; discussion Padoin AV, Braga-Silva J, Martins P, et al. Sources of processed lipoaspirate cells: Influence of donor site on cell concentration. Plast Reconstr Surg. 2008;122: Customer Service Contact Information All correspondence concerning business matters, including subscription information, orders, or changes of address, should be directed to: Lippincott Williams & Wilkins Hunters Green Parkway Hagerstown, MD Tel: Fax: customerservice@wolterskluwer.com 1348