The Role of Free Fat Graft in Breast Reconstruction After Radiotherapy

Radiotherapy is a fundamental therapeutic resource in a large majority of neoplastic diseases. But adverse reactions and complications can severely damage the irradiated tissues. Adverse effects can be distinguished into early (within few weeks to 90 days from treatment) and late ones (months to years from exposure). Late adverse effects are primarily the result of radiation-dependent reduction of stem cells or progenitors (Brush, 2007; Rodemann & Blaese, 2007). Fibrosis, teleangiectasias and atrophy are the most common late effects for skin and subcutaneous tissues and those most frequently observed in irradiated breast cancer patients. The mean concentrations of collagen are two times higher in irradiated skin than in non-irradiated skin thus leading to fibrosis (Autio et al, 1998; Riekki et al, 2002). There seems to be a genetic predisposition for fibrosis and teleangiectasias as a response to radiotherapy. Fibrosis risk is also associated with an inflammatory response, whereas telangiectasia is linked to endothelial cell damage. Atrophy is related to an acute response, but no genetic predisposing factors have yet been identified. (Quarmby et al, 2003; Andreassen et al, 2005; Giotopoulos et al, 2007). The reconstructive properties of adipose tissue are also significantly altered (Poglio et al, 2009) as the stromal microenvironment is unable to self-repair the injury suffered. Depletion of the stromal compartment leads to a loss of the precursor reservoir, thus compromising its ability to maintain tissue homeostasis. In response to ionizing radiation, fibroblasts and macrophages remain in an activated state, continuously generating growth factors and free radicals (Barcellos-Hoff et al, 2005) which are the main reasons for fibrosis. Radiations also severely harm the homeostatic network connecting parenchymal, mesenchymal, and vascular cells within tissues and normal interactions between cells are therefore altered (Barcellos-Hoff et al, 2005; Bentzen, 2006). Late adverse effects of radiotherapy in breast reconstructions can consequently cause flap failure, implant exposure, and capsular contracture in prosthetic breast reconstructions. The final aesthetic results can also be compromised due to liponecrosis.


Introduction
Free fat grafts in plastic surgery and in regenerative medicine are extremely promising. Their use in breast reconstruction and in particular after radiotherapy is radically changing the approach to the problem.

Radiotherapy and tissue damage
Radiotherapy is a fundamental therapeutic resource in a large majority of neoplastic diseases. But adverse reactions and complications can severely damage the irradiated tissues. Adverse effects can be distinguished into early (within few weeks to 90 days from treatment) and late ones (months to years from exposure). Late adverse effects are primarily the result of radiation-dependent reduction of stem cells or progenitors (Brush, 2007;Rodemann & Blaese, 2007). Fibrosis, teleangiectasias and atrophy are the most common late effects for skin and subcutaneous tissues and those most frequently observed in irradiated breast cancer patients. The mean concentrations of collagen are two times higher in irradiated skin than in non-irradiated skin thus leading to fibrosis (Autio et al, 1998;Riekki et al, 2002). There seems to be a genetic predisposition for fibrosis and teleangiectasias as a response to radiotherapy. Fibrosis risk is also associated with an inflammatory response, whereas telangiectasia is linked to endothelial cell damage. Atrophy is related to an acute response, but no genetic predisposing factors have yet been identified. (Quarmby et al, 2003;Andreassen et al, 2005;Giotopoulos et al, 2007). The reconstructive properties of adipose tissue are also significantly altered ) as the stromal microenvironment is unable to self-repair the injury suffered. Depletion of the stromal compartment leads to a loss of the precursor reservoir, thus compromising its ability to maintain tissue homeostasis. In response to ionizing radiation, fibroblasts and macrophages remain in an activated state, continuously generating growth factors and free radicals (Barcellos-Hoff et al, 2005) which are the main reasons for fibrosis. Radiations also severely harm the homeostatic network connecting parenchymal, mesenchymal, and vascular cells within tissues and normal interactions between cells are therefore altered (Barcellos-Hoff et al, 2005;Bentzen, 2006). Late adverse effects of radiotherapy in breast reconstructions can consequently cause flap failure, implant exposure, and capsular contracture in prosthetic breast reconstructions. The final aesthetic results can also be compromised due to liponecrosis. particular, the loss of self-repair properties, chronic inflammation and destruction of microcirculation. The fat grafts can replace atrophic functional niches (the complex made of cell, extracellular and biochemical elements whom the adipose cell interacts with) with physiologic ones, thus playing their normalizing role on the receiving tissue. The normalizing role of free fat grafts in tissue regeneration was pointed out by several clinical studies (Moseley et al, 2006;Rigotti et al, 2007;Locke & de Chalain, 2008;Panettiere et al, 2009;Sarfati et al, 2011). In a recent work (Panettiere, 2011), USG and MR of a free fat grafts reconstructed breast suggested that the proliferation and differentiation of the ADSCs allowed the formation of a perfectly normal structural tissue. Moreover, a quite high density of perivascular stem elements was found, demonstrating a persistent regenerative potential.
An important reduction of fibrosis was also observed (Rigotti et al, 2007;Panettiere et al, 2009Panettiere et al, , 2011. The immunoregulatory activity and the capacity to modulate inflammation displayed by ADSCs can at least partially explain such behaviour. Recent studies therefore started challenging the dogma of the relative contraindication of prosthetic reconstruction after radiotherapy (Percec, 2008;Panettiere et al, 2009;Salgarello et al, 2010).

Neoplastic degeneration of the ADSCs or activation of dormant neoplastic cells
The immunosuppressive effect associated with angiogenic properties of ADSCs, as well as their paracrine activity, raised questions about their interactions with cancer cells. A positive correlation between obesity and cancer is well-known (Roberts et al, 2010). Some authors observed that ADSCs can promote tumour growth (Zhang et al, 2009;Lin et al, 2010;Prantl et al, 2010;Zhao et al, 2010;Nomoto-Kojima et al, 2011;Zimmerlin et al, 2011). Tumour growth beyond the size of 1-2 mm is angiogenesis-dependent. Therefore, it was hypothesized that the angiogenic spike induced by MSCs or ADSCs could awake dormant cancer cells (Naumov et al, 2006;Vessella et al, 2007;Indraccolo et al, 2006;Favaro et al, 2008); but more recent studies (Donnenberg et al, 2010;Zimmerlin et al, 2011) concluded that ADSCs could trigger tumour growth from active cancer cells, but not from dormant ones. Moreover, ADSCs promoted tumour growth only when transplanted at the beginning of the neoplastic process. An unusual cell line was observed to emerge from a culture of h u m a n A D S C s . T h o u g h b e i n g p r o v e d t o b e u n a b l e t o f o r m t u m o u r s , t h e y p r e s e n t e d alarming similarities with human angiosarcoma (Ning et al, 2009). At present, there is no firm evidence that ADSCs can directly promote tumorigenesis and some works showed an even inhibitory effect of ADSCs when implanted in pre-existing tumours . It was thus suggested that reactive cross-talk can take place between ADSCs and other cell types that maintain a proper tissue development and a correct balance between proliferation and differentiation (Casteilla et al, 2011). Some works addressed the angiogenic capability and suspected that ADSCs respond to chemotactic factors and migrate to the sites of injuries, inflammation, and/or tumour (Kubis et al, 2007;J.M. Kim et al, 2007;Constantin et al, 2009;Lamfers et al, 2009;Lee et al, 2009;Lin, 2010;Gehmert et al, 2010;U. Kim et al, 2011). This property was used in a recent experimental work where engineered ADSCs (able to convert 5-fluorocytosine into the 5fluorouracil) engrafted into tumours and micro-metastases, activating prodrugs directly within the neoplastic mass (Cavarretta et al, 2010). The secretion of anti-apoptotic factors, and the T cell-mediated immune response suppression have been blamed for a tumoursupporting role (Jones & McTaggart, 2008;Fox et al, 2007;Wels et al, 2008). ADSCs can be www.intechopen.com found everywhere in the body and their ability to migrate is not a peculiarity of transplanted cells. So, any ADSC in the body could migrate and promote cancer growth and not only transplanted ones. In irradiated patients very few ADSCs can be found in the site of possible residual tumour cells due to radiotherapy itself. So fat grafts implanted directly in the irradiated site could restore a tumour-supporting environment. But this could happen also with flap reconstructions where the ADSCs present in the subcutaneous fat of the flap are transferred to the irradiated areas. Much concern could be raised by in vitro stimulated ADSCs. Anyhow, a potential different behaviour between native and cultured cells remains an open question. A recent limited clinical study with a good follow up found no increase in the recurrence rates in patients treated with fat grafts (Petit et al, 2011). A single case of a late osteosarcoma recurrence 18 months after free fat graft was reported (Perrot et al, 2010). In any case, the risk for possible cancer promotion in the context of cancer related diseases cannot still be ruled out at present.

The free fat grafts
Free fat grafts were first described by Neuber (Neuber, 1893). In 1914, Bruning (Bruning, 1914) was the first to inject autologous fat into the subcutaneous tissue. Liposuction provided large volumes of autologous fat that could be re-injected. Since 1985 the first works about fat re-implantation from liposuction were published (Illouz, 1985(Illouz, , 1986a(Illouz, , 1986bChajchir & Benzaquen, 1986).

Techniques for fat harvesting
The keystone of fat graft is to transplant as many vital cells as possible in the best survival conditions, while injecting more (nonviable) graft material is useless. The keys for graft survival are well known: adequate donor site, atraumatic harvesting, short time between collection and re-implantation, suitable host bed and adequate stabilization.
As far as regards the best donor site, there is no clear evidence in the literature. In a recent survey, the most preferred site for fat harvest was the abdomen (89%), followed by thighs (34%), flanks (25%), gluteal regions (12%), and knees (9%) (Kaufman et al, 2010). Rohrich found no significant difference in the viability of fat cells collected from the abdomen, flank, thigh, and medial knee (Rohrich et al, 2004). Similar conclusions were obtained by other authors using preadipocytes from the abdomen, breast, and buttock (Von Heimburg, 2004) or from thigh, abdomen, and breast . The donor site choice seems to play a negligible role in graft take. When they are used as fillers, site choice should be based on ease and safety of access, fat distribution and abundance, and patient's preference. When treating critical host beds (as irradiated tissue) or when ADSCs are paramount (as in regenerative surgery), donor site choice could be more significant. Some animal model works showed that the potential of ADSCs differs according to the location of adipose tissue from which they are purified (Prunet-Marcassus et al, 2006). In a recent experimental study, we reported function related structural specializations of adipose tissue, depending on the harvesting site. Large cells and few blood vessels with rare stem niches are present in deposit adipose tissue (large fatty depots like the abdominal area), while good vascularity and adequate staminality can be found in structural adipose tissue (limbs, hips, knees or the trochanteric areas) .
www.intechopen.com Another consideration stems from the possible effects of local or general cancer therapies on the number or the viability of ADSCs. A recent work proved that adipose tissue can be severely damaged by radiotherapy, so that the number of ADSCs is deeply reduced and its potential for use in regenerative therapies is dramatically limited ). Therefore, the donor site should be chosen distant from irradiated areas. No work studied the effects of chemotherapy, monoclonal antibodies or hormone suppression on the number, viability and functionality of ADSCs. Therapies addressing angiogenic activity of tumours or directly interacting with adipose tissue may be relevant and there may be differences in the regenerative effectiveness of fat grafts in patients treated with such therapies. Specific studies may be advisable.
The possible damage induced by harvesting technique has widely been studied. When maximum vacuum levels were comparable, mechanical suction and handheld syringe aspiration showed no difference in injury to fat cells. But histologic studies of human fat grafts demonstrated that relative vacuum levels greater than -500 hPa caused cell membrane expansion and deformation, while membrane rupture and fat cells vaporization occur with higher levels (Niechajev & Sevcuk, 1994). When using large syringes (20-60 mL), extraction normally produces less than -600 hPa vacuum levels and rapidly decreases as the fluid and fat are pulled. Fournier (Fournier, 1988a(Fournier, , 1988b(Fournier, , 1990a(Fournier, , 1990b(Fournier, , 1996(Fournier, , 2000 and Coleman (Coleman, 1995(Coleman, , 1997(Coleman, , 2001) clinically showed that relatively high level suction decreased viable cells concentrations compared to manual aspiration; therefore, they underlined the need for atraumatic harvesting. Tholen observed that very few viable fat cells were present in fat from standard liposuction (large cannula, high vacuum) due to a significant damage (mainly cell membranes rupture) in a large number of lipocytes. On the contrary, a significantly higher rate of intact and viable adipocytes was found when fat was harvested with atraumatic, low-vacuum syringe technique. (Tholen et al, 2010).
Surface adipocytes are nourished from the surrounding recipient bed before vascular ingrowth from the bed occurs, while the core cells rely on diffusion only. So the distance between the core and the surface is critical and therefore fat particles dimensions are vital for graft take as larger particles are more prone to central necrosis (Carpaneda & Ribeiro, 1993;Niechajev & Sevcuk, 1994). This is significant only when excisional harvesting is compared to liposuction. On the contrary, smaller liposuction cannulas provide less viable fat grafts (Erdim et al, 2009) probably due to higher cell damage or to damage to the microvascular structure and to the niche. Some liposuction techniques including ultrasonic emulsification, power-or laser-assisted lipoplasty, or high volume fluid administration ("tumescent" or "super-wet" technique) are blamed to additionally damage the aspirate thus reducing the viable cell rate. Our preference goes to syringe aspiration (20 cc) with a wet technique via open tip or 1-hole bullet tip, 3 mm large cannulas.

Fat processing
The rationale for fat processing before re-injection stems first of all from the need to remove blood, fluids, debris and free lipid fractions in order to improve the actual volume of viable fat, and to reduce the inflammatory reaction which may jeopardize long term uptake. Secondarily, processing can improve graft take by adding promoting agents or by in vitro activation. Finally, some studies are trying to preserve the adipocytes for future uses.

www.intechopen.com
The most common processing techniques are centrifugation (Asken, 1988;Toledo, 1991;Coleman, 1995Coleman, , 1997Coleman, , 2001Fulton et al, 1998;R. Amar, 1999;Locke & de Chalain, 2008), sedimentation and washing/rinsing. In a recent consensus survey, various spin rates centrifugation was preferred by 47% of the respondents, 29% favoured fat washings, 12% opted for "other" unspecified treatment techniques, whereas 12% used no preparation (Kaufman et al, 2010). The concentration of viable fat cells in centrifuged vs. sedimented samples seems to be similar. A reliably high concentration of fat cells after centrifugation was seen (Boschert et al, 2002), but maybe the cell damage was underestimated due to the low centrifugation spin rate. An improved long term graft persistence with decanted samples compared to centrifuged ones (1,500 rpm for 5 min) was suggested . A recent work showed interesting results with the "squeezing centrifugation lipotransfer system" to concentrate healthy fat cells, the ADSCs, the patient's own peptides (growth factors), and scaffolds (extracellular matrix) through a combination of squeezing, centrifugation and filtration (Yang & Lee, 2011).
Different processing methods significantly correlated with neither viable adipocytes rates nor grafts longevity (Sommer & Sattler, 2000), while other works (Rose et al, 2006) found that sedimentation was associated to almost double the mean concentration of intact cells than centrifugation (3,000 rpm for 3 min) with or without washing. The limit of almost all these in vitro studies is the assumption that morphologically intact cells are viable and that viable adipocytes rate correlates to graft survival.
Lidocaine was proved to reversibly inhibit glucose transport, causing lipolysis as the cells fully regain their function after washing (Moore et al, 1995) or centrifugation , regardless of exposure duration. Serial rinsing with normal saline can also significantly reduce the intracellular lidocaine concentration (Alexander, 2010).
Another reason for fat processing is to add substances that may increase graft take. Even if somewhat controversial, insulin can stabilize cell membrane and enhance the survival rate by increasing intracellular glycogen and lipid formation (Asaadi et al, 1993;Yuksel et al, 2000). Bovine fibroblastic growth factor in an animal model was shown to improve weight retention of the grafts (Eppley et al, 1992); IL-8 reduced fat necrosis due to its angiogenic action, cellular proliferation stimulation, and cytokine and growth factor synthesis . A higher survival rate was noted suspending the aspirate in enriched cell culture medium (Har-Shai et al, 1999) or in vascular endothelial growth factor (Nishimura et al, 2000). Platelet-rich plasma was added to the grafts by some authors with quite contradictory results: some works observed an effective reduction of the inflammatory response and a fat survival improvement (Pires Fraga et al, 2010;Nakamura et al, 2010), while others found no significant positive effect (Por et al, 2009;Salgarello et al, 2011). Freezing aspirate for later use causes the adipocytes rupture, further decreasing viable cell counts: after 8 weeks the cell survival rate is only 5% (Son et al, 2010).
Our preference goes to serial washing and decantation. Normal saline is aspired in the harvesting syringe (half filled with fat) that is gently tilted to improve washing and then it is decanted for about five minutes. The heaviest fraction is discarded and the procedure is repeated two or three times until the fat turns to yellow. The upper fraction (free lipids) is also discarded. This technique gave us good results (Panettiere et al, 2009(Panettiere et al, , 2011.

Techniques for implantation
Interstitial fluids nourish fat grafts for the first 4 days after transplantation, but during this period oxygen supply and nutrition may be insufficient for graft survival. Graft take can be improved by placing small grafts surrounded by as much as possible healthy recipient tissue, thus maximizing the interface between the graft and the recipient bed. This can be achieved by implanting the fat in multiple tunnels or in single spots (less than 0.1 mL), injecting it linearly while withdrawing the cannula or the needle. Large fat clogs can obstruct the needle/cannula during injection and morre pressure applied to the plunger can cause a large bolus to be inadvertently delivered. Injection guns were therefore proposed to improve control and precision of delivery (Agris, 1987;Newman & Levin, 1987;Niechajev, 1992;Asaadi & Haramis, 1993;Niechajev & Sevcuk, 1994;Berdeguer, 1995;Fulton et al, 1998;Niamtu, 2002Niamtu, , 2010.  (Tholen et al, 2010) or when tissue fibrosis is significant (Panettiere et al, 2011), but no sound evidence is available about the minimum interval between them. There is some evidence that newformed blood vessels are similar to normal ones 21 days after free fat graft (Missana et al, 2007). In our experience (Panettiere et al, 2009) a 20 days interval between sessions is generally safe and efficient.
Local anaesthetics should be injected only at the injection point in order to preserve the positive effects of sample washing (Shiffman, 2010). The injection site should be chosen far enough from the recipient area to prevent fat extrusion. In our experience, there is usually no need for suturing the injection point, while sterile taping is advisable. The diameter of the delivery cannula or needle should be at least as large as the one used for harvesting in order to limit graft damage. Although infection of the transplanted fat is not a common event (Valdatta et al, 2001;Dessy et al, 2006;Delay et al, 2009;Talbot et al, 2010), absolute sterility is imperative and antibiotic prophylaxis may be advised.
Our preference goes to large diameter (18-14G) needles for delivery. We gently mould the graft with the fingers to improve its uniform distribution and we place Steri-Strips ® to encircle the grafted area in order to limit graft's dislocation. Compression dressings can worsen graft take (blood flow reduction) and cause graft's displacement. On the other hand, mostly when using needles, local bleeding in the recipient site can occur thus hindering www.intechopen.com graft's survival. In our experience adhesive soft pads (Reston ® by 3M ® ) applied for five days can balance the risks from both recipient site bleeding and graft ischemia.

The survival rates of fat grafts
The differences in the method of harvest, processing, storage and so on created a great confusion and an extreme variability in the reported results and resorption rates which, mostly in the early lipofilling era, were extremely high, up to 70% (Fournier, , 2000Carpaneda & Ribeiro, 1993;Coleman, 1995Coleman, , 1997. But many studies relied on subjective evaluations that cannot provide comparable results. In an MR study, the long-term volume persistence of autologous fat grafts in facial defects was 51% 3 months after the implant and 45% at 6 months. Nine -twelve months after implantation the volume was stabile. Therefore, a one stage overcorrection protocol was suggested to balance the resorption rate (Horl, 1991). CT scan studies found a 47.5% survival rate nine months after injection (O. Amar et al, 2008). In a recent MR study by our group, eight months after the last session of serial fat grafting the fat survival rate was about 62% (Panettiere et al, 2011).
Several techniques were also proposed to improve grafts survival rates (Hiragun et al, 1980;Bircoll & Novack, 1987;Eppley et al, 1992;Niechajev & Sevcuk, 1994). An experimental study proved that fat resorption was lower with excised fat compared to lipoaspirate (Fagrell et al, 1996). It was suggested that the preservation of fat microvascular structure could improve adipocyte viability and subsequently graft take. In a very recent work, the survival rate of ADSCs was significantly improved when they were implanted along with their collagen scaffold (Mojallal et al, 2011).
Some considerations are crucial in our opinion about the concept itself of fat survival rate. The first ambiguous point is how many are the elements whose survival is about to be calculated. All the studies generally assume that the total injected volume is the same as the graft volume. But, while non-viable adipocytes contribute to the total injected volume, their persistence is obviously null, so they cannot contribute to the long term implanted volume at all. Another undetermined variable is the amount of fluids injected (and accounted) with fat (Alexander, 2010), so it is arduous to discriminate between carrier fluids extraction and fat resorption. In other terms, if the first term of a proportion is uncertain, what will it be the meaning of the proportion itself? Consequently, real fat survival rates (i.e. how many implanted vital cells survive) in in vivo studies are unpredictable in our opinion, and the comparison of such data between different studies is aleatory.

Applications of lipofilling to irradiated breast: personal experience
The personal experience with three different applications of free fat grafts to irradiated breasts is presented. Data are expressed as mean±95% confidence interval. Statistical evaluations were made using the Kruskal-Wallis test for rank variables, the Student's t test for continuous data and the Fisher's exact test for proportions.

The salvage of pre-exposed prostheses and expanders
Exposition is the worst adverse effect of radiotherapy in implant breast reconstruction. In the present preliminary study we hypothesize a possible role for free fat grafts in the rescue of pre-exposed expansion flaps and prosthetic reconstructions. www.intechopen.com

Patients and methods
14 patients presenting with pre-exposed expanders and 19 patients with pre-exposed prostheses were offered free fat grafts to prevent reconstruction failure. 10 patients with expander pre-exposure (expanders active branch, EAB, age: 53.2±5.5 years) and 8 ones with implant pre-exposure (prostheses active branch, PAB, age: 52.5±6.2 years) adhered. The remaining patients (4 pre-exposed expanders, age: 49.6±28.8 years, p=0,53 and 11 preexposed prostheses, age: 49.9±6.8 years, p=0,55) who refused were treated with local flaps (control branches). The expanders in the EAB were partially deflated before graft. Fat was harvested by syringe from the abdomen, the hips or the trochanteric areas through a 3 mm open tip cannula, washed with saline and implanted around the pre-exposure area and successively in the thinned area too, through an 18G needle.

Discussion
Expansion is undoubtedly critical in irradiated patients because even modest volumes can cause high tension, severe ischemia and a high exposition risk, due to flap stiffness. The present study demonstrates that free fat grafts can play an interesting role in expansion flaps salvage, while a significant positive effect in prosthetic reconstruction rescue failed to be proved. In our opinion, the key is that all expanders were at least partially deflated, while the same could not obviously be done with prostheses. This is a limited preliminary study, so the results should not be generalized. The main advantage of free fat grafts is that they are a closed procedure minimizing the risk of expander/prosthesis infection or further flap loss. The main risk is inadvertent expander/prosthesis rupture.

Patients and methods
137 irradiated breast implant reconstructed patients were offered free fat grafts. 48 of them (active branch, AB, mean age 50.6±2.8 years) adhered, while the remaining 89, who refused (control branch, CB, mean age 51.5±2.4 years, p=0.32), received a conservative treatment. The autologous fat was harvested by syringe from the abdomen, the hips or the trochanteric areas using a 3 mm cannula (1-hole, bullet tip) and processed by gentle washing with saline. Then it was implanted using a 14G needle in depressed areas (10-15% overcorrection), under the scars, and in dystrophic sites. Functional results were compared using the LENT-SOMA score, while a five-points scale (5: very good, 1: very poor) was used to compare the aesthetic results.

Results
The patients in the AB received 28.2±2.0 cc of fat per session (range: 8-70 cc) in 3.3±0.6 sessions (range: 1-9) with a 86.0±14.6 days interval (range: 20-392 days) between sessions. The initial LENT-SOMA scores in the two branches were comparable, but they significantly improved in the AB after lipograft ( fig. 1). 3 months after the last fat graft all the parameters except oedema were better in the AB than in the CB (table 1). A significant improvement was recorded also comparing homogeneous subgroups with similar initial LENT-SOMA ranks (excluding only breast oedema in the lower rank subgroup) as reported in table 2. Two patients in the AB presenting with Bk3 capsular contracture downgraded to Bk1 after respectively 2 and 4 fat graft sessions (fig. 2). The initial aesthetic outcome was similar in the two branches (2.5±0.2 vs. 2.5±0.2, p=0.624), while it significantly improved in the AB 3 months after the last fat graft session (2.5±0.2 vs. 2.9±0.2, p=0.00137). In particular, a great improvement in superficial irregularities and scars, reduction of fibrosis related deformities, and a general improvement of skin appearance and trophism were observed. The aesthetic result in the CB was significantly worse (2.5±0.2) than in the AB 3 months after the last session (2.9±0.2, p=0.00274).

Initial evaluation
Active

Discussion
The present study substantially confirms the results of our previous, more limited published series (Panettiere et al, 2009), but as the present series is more than two folds larger than the previous one, the results appear even more convincing. In particular, a reduced effect of lipofilling on the oedema was confirmed. A possible explanation could be that oedema depends more significantly on the axillary nodes status (sentinel node vs. total dissection, axillary radiotherapy) than on the local tissues. So, a reduced beneficial effect of free fat grafts can be expected. Anyhow, a significant improvement was observed in the subgroup where oedema was initially more severe. Improvement in tissue vascularization and reduction of local inflammatory factors could explain it. In 2 cases a significant improvement of capsular contracture was observed after fat graft. This may be an interesting option for breast augmentation too and, maybe, also to help understand the physiopathology of this challenging adverse event. A possible action mechanism is that ADSCs interact with the inflammatory response as leukotriene antagonists do. Such a hypothesis should obviously be addressed by specific studies. In one case, an accidental prosthetic rupture occurred, due to an inadvertent patient's movement during lipofilling. This was the only procedure-related complication observed, but it should be prevented mostly when dealing with very thin coverage tissues. The data about local and distant recurrence in the present series show no significant difference between the patients who received free fat grafts and those who did not, as far as regards neither incidence, nor distance from cancer surgery. The current series presents some possible biases besides the relatively small number of patients. First of all, 15 patients in the CB (16.9%) were lost at follow up before 3 months. So a higher recurrence rate in the CB could be expected. Secondarily, this study was not specifically designed to address oncologic data. In particular, the patients were not actively and uniformly studied in search for recurrences, and the data here reported relied only on their individual oncologic follow up programs. So, these results should be considered with caution and specific prospective studies should address the relationship between free fat grafts and cancer recurrence. In our opinion, autologous free fat grafts should be routinely offered to all the irradiated implant reconstructed breasts because they can greatly improve both functional and aesthetic results and reduce the incidence of complications.

The total breast reconstruction with fat grafts only: case reports
Prostheses and autologous flaps are the most common options in breast reconstruction after mastectomy. Autologous flaps are the gold standard in reconstruction failures, but when they are contraindicated, no validated option is available.

Case 1
A 36-year-old patient underwent prophylactic bilateral nipple sparing mastectomy and immediate prosthetic reconstruction 6 years after a left breast quadrantectomy and radiotherapy. Eight months later, the reconstruction failed due to fatal exposure even if several salvage attempts using local flaps were tried. Three years later, the patient asked for reconstruction, but obesity (BMI: 36), and severe asthma contraindicated general anaesthesia. Free fat graft was thus considered the only viable reconstructive option. This case was particularly tricky because of the large breast and the stiffness due to both radiotherapy and the multiple salvage attempts. 700 cc of fat were implanted in 9 sessions in 13.5 months (40 days minimum interval between sessions). The fat was harvested by syringe through a 3 mm open tip cannula from the abdomen, the hips, the thighs, the buttocks or the trochanteric areas and it was washed with saline. The grafts were then implanted using the same syringes and a 14G needle. In the first 4 sessions, small fat volumes were implanted accurately avoiding any significant skin tension (average volume 42.5 cc) under a quite wide surface, in order to release the scar and regenerate the tissues. A great reduction of fibrosis (LENT-SOMA score 3 before session #1; score 1 before session #4) was observed. In a second phase (the last 5 sessions), larger volumes (average: 106 cc) were implanted to improve volume and shape. The overall fat survival rate was 62% measured at MR eight months after the last grafting session. The aesthetic result was pleasant and stable 14 months after the last session ( fig. 3). A great improvement in irradiated skin quality and scars was also achieved (LENT-SOMA score before the first session: 8; 8 months after the last session: 3) with improvements in fibrosis (-2), pain (-1), atrophy (-1) and oedema (-1).

Case 2
A 58-year-old woman underwent modified radical mastectomy and prosthetic reconstruction. During expansion she received radiotherapy, but reconstruction was completed with a good result (total LENT-SOMA score: 0, 6 months after the prosthesis was implanted). 8 months later, an infection occurred (Pseudomonas sp.) imposing implant removal. 2 months later (negative blood and tissue cultures) a new reconstruction procedure was performed, but 15 months later the breast suddenly inflated. Pseudomonas was found, so the implant was removed once again. The patient was asking for breast reconstruction, but she then rejected general anaesthesia and local flaps. Therefore, she was proposed a free fat breast reconstruction. A total 367 cc of fat (52.9% larger than the explanted prosthesis) was implanted in 6 sessions (total time: 11.0 months, 61.2±8.5 cc per session) with a 54.8±10.1 days average interval between the sessions. The aesthetic result was pleasant 10.4 months after the last session ( fig. 4), with no sign of infection. Fig. 3. a, b: case 1, initial status; c, d: 14 months after the last fat graft session www.intechopen.com

Discussion
The cases here reported offered rather different reconstructive challenges. In the first one, quite a large volume needed to be restored, but a severe fibrosis was present and any attempt to place large volumes of fat under such a stiff skin could cause dangerous tension. In breast reconstruction using free fat only, the grafts usually act as both expanders and vital fillers. Del Khouri & Del Vecchio, 2009) proposed pre-expansion using the BRAVA ® system. Even if suction induced by external pre-expansion system was stated to improve blood supply and graft take, in our knowledge, the stiffness observed in the present case was never addressed before with BRAVA ® system in published works and its safety on irradiated tissues was not assessed. In case 1, we were concerned about possible risks of pre-expansion as the tissues were extremely thin (mostly in the lower pole). Moreover, we were worried that suction could worsen pain and oedema (both LENT-SOMA score 1). So we opted for a two-step procedure: in the first four sessions we aimed at tissue regeneration, hoping that ADSCs could reduce fibrosis and improve vascularity, so that in the last five sessions larger volumes of fat could be safely implanted. The stable final results demonstrate that this approach is a valuable option. In case 2, breast volume was not very large and the tissues showed no significant adverse effect of radiotherapy, but some fibrosis and stiffness developed later due to infections and www.intechopen.com multiple procedures. So the two-step approach was successfully applied in this case too. But the main concern was possible recurrent infection. In our knowledge there is no specific work addressing fat grafts in potentially infected areas. The patient was administered an antibiogram driven antibiotic treatment to eradicate the infection. A 5-days antibiotic prophylaxis was also empirically administered after the first fat graft session. Standard oneshot prophylaxis was then applied in the following sessions. No sign of re-infection was observed after a 10.4 months follow up. In the present case, free fat grafts proved to be safe also in high infection risk patients.
The long overall reconstruction time is undoubtedly the greatest limit of breast reconstruction using fat grafts. In the present cases respectively 9 sessions in 13.5 months and 6 sessions in 11.0 months were necessary; a mean of 3 sessions were needed in other studies (Delaporte et al, 2009). Moreover, no immediate reconstruction with fat grafts can be planned, even if grafting immediately after mastectomy could probably be an interesting option. Poor donor tissues are other possible limits. The main advantage of breast reconstruction using free fat grafts is the absence of significant contraindications, as it can be performed in almost all patients under local anaesthesia. The technique is undoubtedly easier and extremely less expensive than any other reconstructive option. It can also offer an autologous reconstruction option without major surgeries when other procedures failed.

Conclusion
Free fat grafts proved to be a remarkable alternative in irradiated breast reconstruction. Some doubts still remain about the risk of reactivation of dormant cancer cells and further studies should assess chemotherapy, hormone suppression and monoclonal antibodies effects on the regenerated tissues. Advances in adipocyte cryoconservation techniques could further improve the effectiveness by reducing the number of harvesting procedures.