The use of fat in plastic surgery is a vital part of the surgeon’s armamentarium. The harvest, preparation, and reimplantation of fat grafts into a second surgical site are straightforward, minimally invasive, and cost-effective. These qualities have made fat grafting a common procedure in plastic surgery.

Despite the obvious utility of fat grafting, strikingly little objective data have been accumulated with regard to the best method of performing this procedure. A practitioner reviewing the literature in an attempt to decide how to best perform the technique will find largely anecdotal experience and recommendations. Furthermore, much of the existing literature sets forth confusing and contradictory information. For example, Coleman advocates centrifuging the cells at 3000 rpm to consolidate them, whereas the work of Chajchir, involving histologic analysis, suggests that centrifugation at 1000 or 5000 rpm destroys almost all living fat. Saline-solution washing of harvested fat is a standard technique for many practitioners; Chajchir [10] insists, however, that exposure of the cells to saline solution “causes a loss of the fibrin which aids in the adhesion of the injected material to surrounding tissue. Ellenbogen [11] advocates bathing the cells in insulin for 45 to 60 minutes, as well as implanting 4- to 6-mm “pearls” of fat instead of performing needle or cannula harvesting. [2] Other variables that may bewilder the surgeon searching for the optimal technique include the diameter of the harvesting cannula, the diameter of the injection needle, and whether warming or chilling the cells is beneficial. Gatti [4] advises his patients to expect a minimum of 3 fat grafting sessions for an appreciable augmentation, whereas Coleman [8] has found, with few exceptions, that “fat placed stably maintains its shape and consistency over time and has every indication of permanence.” No matter the technical differences, however, most surgeons agree that injecting grafts with a high percentage live fat cells improves permanence in fat grafting.

Testing the viability of free fat grafts in vitro is a challenging task that has not been satisfactorily achieved. Current methods of testing adipocyte viability (eg, glucose uptake, trypan-blue staining, and G6PDH uptake assay [reduced glucose-6-phosphate dehydrogenase]) rely on cell isolation procedures. These procedures do not accurately reflect fat viability in the intact tissue because isolation of adipocytes eliminates the stromal component surrounding the fat cells. The stromal component provides the structural and biochemical support necessary for the maintenance of the adipocytes. Therefore any estimate viability in isolated fat cells does not reflect the viability of fat cells in the clinical environment. A model for testing fat grafts in the actual form in which they are used clinically would be useful in the evaluation of fat-handling techniques.

One well-known and simple cell viability test is the MTT [3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyl tetrasodium bromide] assay, a colorimetric assay in which active mitochondria cleave the pale-yellow substrate to yield a dark-purple formazan product (Chemicon International, Temecula, CA). Even newly dead cells will not cleave MTT, so the assay is very sensitive to the presence of living cells. However, no reports have been made of the MTT assay’s being used to assay adipocyte viability. Because the major criticisms of fat grafting are the tendency for grafts to be resorbed and the inability to predict long-term results, knowing whether a particular handling technique tends to increase or decrease the likelihood of fat survival could predict a graft’s long-term survival.

The purpose of this study was to develop a method for assaying the viability of fat in its clinically used form and then to test several common techniques used in fat grafting for their effects on fat’s viability.

Methods

Sample collection

We obtained fresh tissue, collected during abdominoplasty, after it had undergone routine pathologic analysis in the University of Virginia Department of Surgical Pathology. Fat was then suctioned by hand with a 3.7-mm–diameter Tulip cannula (Tulip Medical, San Diego, CA) on a 35-mL syringe.

Time-course viability determination

Samples (2 mL of whole-fat grafts) were placed in a 10-mL Falcon tube (Fischer Scientific, Pittsburgh, PA) and incubated at 37°C for 24 hours. Next, 200 ?L of MTT solution AB (MTT plus phosphate-buffered saline solution) was added to the “day 1” samples, and the tubes were gently inverted. A similar procedure was performed 24, 48, and 72 hours later for the “day 2,” “day 3,” and “day 4” samples. After being incubated for 1 hour, the fat was divided into aliquots of 500, 250, and 125 mg (8 samples of each weight and incubation day) in 1-mL Eppendorf tubes. These aliquots were sonicated to disrupt the cell membranes and diluted with 200 ?L of corn oil (Sigma Chemical Co, St. Louis, MO). The samples were then centrifuged at 16,000 g for 10 minutes in a microcentrifuge. At this point, each sample consisted of an upper lipid layer (with the purple formazan product) and a pellet containing cellular debris. The lipid layer was carefully pipetted from the top of the sample and read at 470 nm in a spectrophotometer. We graphed the results using a Microsoft Excel spreadsheet (Microsoft Corp, Redmond, WA).

Viability testing of handling techniques

Fat was subjected to a variety of treatments, described below. Fat collected with the 3.7-mm cannula and not treated in any way was considered the control. Control tissue was compared with tissue collected with the use of a 2.1-mm cannula or by means of sharp dissection alone in 4- to 6-mm pearls. Some control tissue was subjected to the following treatments: addition of insulin (2 ?L/mL fat), centrifugation at 1000 and 5000 rpm for 5 minutes, and extrusion through a Coleman needle, an 18-gauge needle, or a 23-gauge needle. Another sample was frozen (24 hours at –20°C), and yet another was treated with Triton-X 100 detergent (1:1 mL). The fat supplemented with Triton-X 100 served as a negative control because detergent causes lysis in fat cells. The samples, 2 mL of fat, were placed in Falcon tubes with 2 mL of Dulbecco’s modified Eagle medium and gently inverted. This allowed the fat to be suspended in medium and optimized contact between the fat particles and the MTT solution. After 24 hours of incubation (or, in 1 sample, –20° freezing), 100 ?L of MTT was added and allowed to react for 1 hour, with gentle mixing of the samples. The samples were then processed as described above for spectrophotometric analysis. We compared the data generated in this fashion using standard analysis of variance.