Microvascular Angiogenesis and Apoptosis in the Survival of Free Fat Grafts


  • Supported by grant 10770878 from the Japanese Ministry of Education.


Objectives/Hypothesis Autologous fat is an ideal material for augmentation in plastic surgery because of its minimal tissue reaction and easy availability, but its long-term graft survival is somewhat unpredictable. This study was conducted to determine how fat grafts get their vascular supply from the recipient bed and why they keep reducing in volume and weight.

Study Design Experimental study using animal models.

Methods The expression of vascular endothelial growth factor (VEGF) in grafted fat tissue was examined by using immunohistochemical staining, and apoptotic cell death in the grafted fat was studied by using terminal deoxynucleotidyl transferase (TdT)–mediated deoxy-uridine triphosphate (dUTP)-biotin nick end-labeling method. Twenty-five Wistar rats were used as models of free fat grafts. Fat tissue taken from inguinal fat pads was grafted to the back skin with an 18-gauge needle injection.

Results The weight of the injected fat was significantly reduced on the 180th day compared with the original weight (32% ± 10%). VEGF+ cells were observed in fibrous connective tissue of the grafts on days 7 and 30 but not after day 90. Apoptotic cells were also observed on days 7 and 30.

Conclusions Angiogenic factors including VEGF started to revascularize the graft around day 7, and the extent of the vasculature was not reduced after the revascularization. In addition to necrosis in the graft's early stages, apoptosis induced by many factors in the graft's environment is also, at least in part, a cause of long-term volume reduction of the fat graft. Thus clinical application of angiogenic factors such as VEGF to fat grafts and control of apoptosis may contribute to improvements in fat-grafting techniques.


Free autologous fat graft has been mentioned for augmentation in the treatment of unilateral vocal cord palsy 1,2 and for cosmetic improvement of facial contours. 3–5 Although synthetic graft materials such as collagen and Teflon are available, injected collagen is absorbed quickly and Teflon causes foreign-body granulomatous reaction. 2 Autologous fat is an ideal material because of its minimal tissue reaction and easy availability. However, its long-term graft survival is somewhat unpredictable; in fact, 20% to 80% of the volume of grafted fat was found to have been lost in long-term follow-up studies. 6–10 Moreover, little is known about the mechanism involved in the uptake and absorption of grafted fat. 11–13

For skin or free bone grafts, revascularization of grafted tissues is essential for the grafts to take, and to maintain their viability. Several clinical and experimental studies have reported continuing volume loss of the grafted fat even after the grafts appeared to have been revascularized. 3,6 Apoptotic cell death may be responsible for this graft loss during long-term follow-up after transplantation.

In the present study, animal models of free fat grafts were used to study the revascularization of grafted fat from the viewpoint of microvascular angiogenesis, and to investigate apoptotic cell death among grafted fat cells. This study attempts to explain how grafted fat cells survive and long-term volume loss occurs.


Animal Model

Twenty-five Wistar rats weighing between 180 and 220 g were used in accordance with the protocol approved by the Animal Care and Use Committee of Kanazawa University (Kanazawa, Japan). The animals were anesthetized with intraperitoneal pentobarbital sodium (50 mg/kg). Fat tissue was excised surgically from the clipped and shaved inguinal region on the right side. The entire surgical procedure was carried out under sterile conditions.

For the augmentation model, 300 mg of fat was grafted between the back skin and underlying muscle because there is no adipose tissue between the dermis and underlying muscle. After the fat was minced with a scalpel and washed twice in physiological saline, graft placement was accomplished by means of percutaneous injection (18-gauge needle with 1-mL syringe).

Graft Specimen Sampling

After 2, 7, 30, 90, and 180 days, the rats (five rats for each observation period) were sacrificed with an overdose of intraperitoneal pentobarbital sodium (200 mg/kg), and their body weights recorded. The grafts were removed and weighed for comparison of their weight with that at the time of placement. The ratio of the percentage of weight change of the graft over the percentage of weight gain for each animal (corrected weight ratio [CWR]) was used to determine the actual weight change of the grafts.

Because the surrounding fibrous tissue was often difficult to remove from the fat graft, the graft, together with the tissue that could not be removed, was considered to be a specimen. After the sample was weighed, it was fixed in acetone with the acetone methyl-benzoate xylene (AmeX) fixation method 14; then the specimens were embedded in paraffin. The AMeX fixation method was used because it has been reported to preserve antigenicity of specimens very well.

Histological Evaluation

Sections with a thickness of 4 μm were cut from the paraffin blocks and processed with H&E staining for histopathological examination. The slides were reviewed by a pathologist without knowledge of the method of graft transplantation or the duration of the postgraft period.


Immunohistochemical studies used the avidin-biotin-peroxidase complex (ABC) method. 15 The primary antibodies used were rabbit polyclonal antibody specific for 121, 165, and 189 amino acid variants of vascular endothelial growth factor (VEGF) (Santa Cruz Biotechnology, Santa Cruz, CA) at a 1:200 dilution and the antibody specific for von Willebrand factor (vWF) (DAKO, Glostrup, Denmark), also at a 1:200 dilution. For positive controls, human placenta was stained for VEGF, and normal human palatine tonsil for vWF. A negative control was included for each specimen by replacing the primary antibody with normal goat serum.

Microvessel Count

All microvessels were highlighted by staining the endothelial cells with vWF. The area of intense vascular staining was located under light microscopy in a field of original magnification × 100, and the microvessels were counted in a field of original magnification × 400. Endothelial cells or endothelial clusters staining positive for vWF and clearly separated from adjacent microvessels were considered to represent single, countable vessels. Morphologically bifurcating microvessels were also counted as single vessels. These assessments were performed by two observers using a double-headed microscope. Results represent the mean value for three fields of original magnification × 400.

Evaluation of Immunohistochemical Staining of Vascular Endothelial Growth Factor

Staining was assessed by two independent observers, and the cells with VEGF immunoreactivity were counted. At least three representative fields in each section were examined, and mean value for three fields was recorded.

Detection of Vascular Endothelial Growth Factor Messenger RNA Using Reverse Transcriptase–Polymerase Chain Reaction

Poly-adenylated RNA (polyA RNA) was isolated from 50 mg of the injected graft fats at the back skin with the aid of a QuickPrep Micro messenger RNA (mRNA) Purification Kit (Amersham Pharmacia Biotech, Piscataway, NJ). Five microliters of the eluted solution was used to make complementary DNA (cDNA) by using a First-Strand cDNA Synthesis Kit (Amersham Pharmacia Biotech) with random hexadeoxynucleotides according to the manufacturer's protocol. Reverse transcriptase–polymerase chain reaction (RT-PCR) was performed as described in the protocol. The primers used were sense: 5′CCATGAACTTTCTGCTCTCTTG and antisense: 5′GGTGAGAGGTCTAGTTCCCGA. 16 The thermal cycler condition consisted of 2 minutes of denaturation at 94°C followed by 35 cycles of 45 seconds at 94°C, 1 minute at 58°C, and 90 seconds at 72°C. Samples were then treated at 72°C for 7 minutes. RT-PCR products were analyzed by means of electrophoresis on 3% agarose gels. The sizes of the expected products were 698, 626, and 493 base pairs, depending on different splicing forms of the mRNA of VEGF. 16

In Situ Labeling of Apoptotic Cells

Apoptosis was detected by labeling the 3′ OH ends of DNA by means of digoxigenin incorporation using terminal deoxynucleotidyl transferase (TdT)–mediated deoxy-uridine triphosphate (dUTP)-biotin nick end-labeling method). 17 Antidigoxigenin antibodies and immunoperoxidase staining were used to demonstrate digoxigenin-nucleotide incorporation with a commercially available in situ apoptosis detection kit (Apop Tag, Oncor, Gaithersburg, MD). The tissue sections were visualized by counterstaining them with 1% methyl green.

Statistical Analysis

Mean values and standard deviations were calculated for fat graft weight, the number of VEGF+ cells, and the number of microvessels. The unpaired Student t test was used to compare the values obtained. Statistical significance was defined as P < .05.


Graft Weight Change After Transplantation

Graft weights including surrounding fibrous tissue that could not be removed and corrected for body weight gain (CWR) increased for the first week because of fibrous tissue growth (Fig. 1). After 1 month, however, CWR has started to decrease and was reduced to 0.32 ± 0.10 of the original weight 180 days after transplantation. The weight of the graft on day 180 was significantly less than the original weight.

Figure Fig. 1..

Changes in the weight of the back skin fat graft. Points represent mean value ± SD. The weight increased during the first 7 days, then decreased gradually. The weight loss of the injected graft became significant on day 180 (0.32 ± 0.10 of original weight).

Histological Findings

Edematous swelling of fat cells, infiltration of inflammatory cells, and growth of fibroblasts were obvious on day 7, and lipid drops consisting of lipid escaped from dead fat cells were observed on day 30 (Fig. 2). Fibrosis between fat cells and a reduction in the number of fat cells were the predominant histological findings on days 90 and 180.

Figure Fig. 2..

Injected fat grafts in the back skin (top left). Excised fat tissue used as control (top right). At day 7, leukocyte and macrophage infiltration, as well as proliferation of fibroblasts, can be seen (bottom left). On day 30, necrosis of fat cells and smaller adipocyte-like cells were observed (bottom right). On day 180, connective tissue had increased over time (H&E stain, original magnification × 140).

Expression of Vascular Endothelial Growth Factor and Microvessel Counts

Vascular Endothelial Growth Factor–positive cells were observed in fibrous connective tissue peripheral to the graft on days 7 and 30 but not after day 90 (Fig. 3A). Moreover, the number of VEGF+ cells on day 7 (Fig. 4) was greater than on day 30 (P < .01, unpaired Student t test).

Figure Fig. 3..

A. Immunohistochemical staining for VEGF after 7 days. VEGF+ cells were most often found in connective tissue around the fat graft. B. Immunohistochemical staining for von Willebrand factor 7 days after the transplantation. Microvessels are represented by clusters, which stand out sharply from other tissues. The vessels with the smaller diameter were considered to be newly generated microvessels (black arrows).

Figure Fig. 4..

Number of vascular endothelial growth factor (VEGF)-positive cells in the fat grafts observed under the microscope in high-power fields. Points represent mean ± SD. VEGF expression was seen on posttransplantation day 7, then decreased. The VEGF+ count on day 7 was significantly higher than on day 30 (P < .01, unpaired Student t test).

On day 7, RT-PCR analysis showed the expected products from the cDNA of the fat graft injected under the back skin (Fig. 5). This suggests that VEGF was synthesized de novo at the graft.

Figure Fig. 5..

Detection by reverse transcriptase–polymerase chain reaction (RT-PCR) of VEGF messenger RNA (mRNA) in minced fat graft on day 7 (3.0% agarose gel stained with ethidium bromide). Products of three different sizes, depending on the splicing forms, were observed in the minced graft.

The microvessel counts increased until day 7 (Fig. 3B), after which they decreased slightly, but not to a statistically significant extent (Fig. 6).

Figure Fig. 6..

Number of microvessels observed under the microscope in high-power fields. Points represent mean ± SD. The number of microvessels had markedly increased on posttransplantation day 7 and had not decreased after 30 days. The number of microvessels on day 7 was significantly higher than on day 2 (P < .05).

Detection of Apoptotic Cells

Apoptotic cells were observed on days 7 (Fig. 7) and 30. The number of apoptotic cells on day 30 (Fig. 8) was greater than on day 7, but the difference was not statistically significant (P = .67).

Figure Fig. 7..

Detection of apoptosis in the fat graft. Apoptotic cells were stained with the terminal deoxynucleotidyl transferase–mediated dUTP-biotin nick end-labeling method. Positive signals were detected in the nuclei of fat cells. The dark brown stained nucleus was considered to be apoptotic.

Figure Fig. 8..

Number of apoptotic cells in fat grafts. More apoptotic cells were observed on day 7 than on day 30, but the difference was not statistically significant (P = .67).


The present study was conducted to determine how fat grafts obtain their vascular supply and why they keep reducing in volume and weight during long-term follow-up. Many studies have reported on the histological aspects of these phenomena, 1–13 but this study adds a new analytical viewpoint in terms of angiogenesis and apoptosis.

It is difficult to compare the findings for experimental and clinical fat grafts because different studies have used various types of animal models, different methods of transplantation, and different methods to estimate the volume and weight reduction of the grafts. Using a rabbit model, Kononas et al. 11 directly weighed the retrieved grafts and reported graft weight reductions after 9 months to 32.7% ± 11.4% in the case of suctioned fat and to 42.2% ± 19.3% in the case of surgically excised fat. Niechajev and Sevcuk 12 reported that in clinical cases grafted fat had retained 40% to 50% of its volume after 2 years and based their observation on esthetic clinical impressions. Mikus et al. 7 reported an injected-graft take of only 20% in a canine vocal fold model. Despite specific differences in their findings, these studies reported that only a small part of the fat grafts survived and that fat graft volume was constantly reduced over time.

Our study also showed a significant reduction in fat graft weight. However, mincing of the graft may have damaged cell membranes of adipose cells and interstitial connective tissue, including the vascular structures. Nguyen et al. 6 reported that only 10% of adipocytes remained intact in suctioned adipose tissue. Such damage constitutes a major factor in the higher failure rate of injection grafts. In the clinical situation, however, fat graft injection is more convenient than surgical placement. Therefore it is recommended that a weak negative pressure (0.5 atm) is used to collect fat grafts to avoid cell damage.

Generally speaking, revascularization at the recipient bed is essential for any kind of free graft to survive. In our model, VEGF was expressed in the interstitial mononuclear cells, and most significantly on day 7. Fat grafts are believed to obtain nourishment from interstitial fluid for the first 4 days after transplantation. During this period, the supply of oxygen and nutrition may not be enough for the graft to survive, and the resulting hypoxia could induce angiogenic factors, including VEGF production of cells in the interstitial connective tissue. VEGF 16,18,19 is a potent and specific mitogen vascular endothelium excreted from tumor cells, macrophages, keratinocytes, and smooth muscle cells. We consider it to be one of the major factors contributing to revascularization of the grafted fat in our study. Although many other angiogenic cytokines must be involved in the revascularization process, our speculation is supported by the number of newly generated microvessels, which increased until day 7 in parallel with VEGF expression. The RT-PCR assay results also suggested de novo synthesis of VEGF mRNA in the graft.

Apoptosis has been called programmed cell death caused by genetically transduced signals. 17 In our study, the number of grafted fat cells decreased after revascularization had been established. Although acute necrosis can explain the early volume loss of the fat graft, apoptosis should be, at least in part, responsible for the volume loss in the later period. Apoptotic cells were most often observed on day 30, and this could explain the prolonged adipocyte reduction. After 30 days, dead fat cells and lipid drops were removed by macrophages, which would explain the continued weight loss of the grafted fat. Apoptosis of fat cells 20 is induced by cytokines, insulin, corticosteroids, heat, and serum deprivation. In clinical cases, emaciation of cancer patients in advanced stages of disease could be caused by the apoptosis of fat cells induced by tumor necrosis factor-α excreted by tumor cells. 21 Grafted fat cells exist under hypoxic and hyponutritional conditions during the early graft period, and such conditions might cause genetic signals to initiate apoptosis. The histological findings of the present study indicate that leukocytes which infiltrated the connective tissue in the grafts may have caused inflammation leading to the excretion of cytokines and thus the induction of apoptosis.


The results of this study suggest that grafted fat may be considered to exist under poor microvascular circulation during the early post-graft period and this condition may induce angiogenic cytokines, including VEGF, which facilitates revascularization of the graft. Such angiogenic factors revascularized the graft after roughly 7 days, and the extent of the vasculature was not reduced in the long term. In the present study, in addition to necrosis during early stages of the graft, apoptosis induced by many factors in the graft environment was also, at least in part, a cause of long-term volume reduction of the fat graft. Thus clinical applications of angiogenic factors such as VEGF to fat grafts and control of apoptosis may contribute to improvements in fat-grafting techniques.