Tissue engineering and regenerative medicine strategies for the female breast

Abstract The complexity of mammary tissue and the variety of cells involved make tissue regeneration an ambitious goal. This review, supported by both detailed macro and micro anatomy, illustrates the potential of regenerative medicine in terms of mammary gland reconstruction to restore breast physiology and morphology, damaged by mastectomy. Despite the widespread use of conventional therapies, many critical issues have been solved using the potential of stem cells resident in adipose tissue, leading to commercial products. in vitro research has reported that adipose stem cells are the principal cellular source for reconstructing adipose tissue, ductal epithelium, and nipple structures. In addition to simple cell injection, construct made by cells seeded on a suitable biodegradable scaffold is a viable alternative from a long‐term perspective. Preclinical studies on mice and clinical studies, most of which have reached Phase II, are essential in the commercialization of cellular therapy products. Recent studies have revealed that the enrichment of fat grafting with stromal vascular fraction cells is a viable alternative to breast reconstruction. Although in the future, organ‐on‐a‐chip can be envisioned, for the moment researchers are still focusing on therapies that are a long way from regenerating the whole organ, but which nevertheless prevent complications, such as relapse and loss in terms of morphology.


| Macroscopic anatomy
Together with the adipose tissue, the mammary gland contributes the connective tissue and other minor components, to form the breast.
The mammary gland distinguishes mammals from other animals and appears as tissue in the shape of a conical disk. It is present in both females and males, but only reaches complete development in females (Hassiotou & Geddes, 2013). The breast lies on the deep pectoral fascia and varies in shape and size. In an adult female, its base extends vertically from the second or third rib to the sixth rib and, into the transverse plane, from the sternal edge medially almost to the midaxillary line laterally. The mammary gland consists of between 15 and 20 lobes immersed in adipose tissue and buried by connective laminae, organized into a three-dimensional network (Figure 1). Each lobe, which contains alveoli arranged in bunches, ends in a lactiferous duct that generally has an independent outlet on the nipple (Figure 1).
Each duct has a dilated portion, the lactiferous sinuses, in which milk accumulates during lactation. The areola (the disk of skin that circles the base of the nipple) contains sebaceous glands, which increase in size during pregnancy. The nipples are almost conical prominences containing smooth muscles without adipose tissue positioned in the center of the areolas. The overall position of the nipples is related to the stage of pregnancy. A vascularized compact fibroadipose tissue surrounds the glands and lactiferous ducts (Gilroy, Ross, MacPherson, & Gaudio, 2014).

| Physiology
The development of the mammary gland is based on several stages.
During the lifetime of a female, the adult mammary gland is subjected to a cyclic expansion due to hormonal changes within the menstrual cycle, lactation after pregnancy and involution (Hassiotou & Geddes, 2013). The specific function of the gland is to synthetize, secrete and deliver milk to the infant for his/her nourishment, protection, and development (Hassiotou & Geddes, 2013). The oxytocin produced in response to the suckling infant causes the surrounding myoepithelial cells to contract in order to move the milk through the ductal tree to reach the nipple (Ercan, van Diest, & Vooijs, 2011).

| Microscopic anatomy
Multiple cell types make up the mammary tissue: epithelial cells, adipocytes, fibroblasts, vascular cells, immune cells, nerves on the gland surface, and bipotent mammary stem cells (MaSCs).
The mammary epithelium shows a regenerative potential supporting the cycle of growth and involution. The epithelial compartment contains the mammary stem cells (Hassiotou & Geddes, 2013). During puberty, cap cells (multipotent stem cells; Figure 2) cover the external layer of the bud developing into myoepithelial cells or luminal epithelials, whereas body cells fill the interior part (Woodward, 2005).
Adipocytes represent the reservoir of fat needed for the metabolic process of milk production, and they influence epithelial growth and mammary epithelium function. In addition, they produce vascular endothelial growth factors and regulate mammary gland angiogenesis, thanks to the stromal vascular fraction cells (SVF). The adipose tissue surrounds the stroma, which is characterized by dense fibrous connective tissue and contains mesenchymal cells.
Fibroblasts communicate with the epithelium and produce extracellular matrix components as collagens, proteoglycans, and fibronectin in addition to enzymes. Hassiotou et al. proved the possibility of detecting, MaSCs, progenitor cells, mature myoepithelial, and milk-secretory cells in human breast milk stem cells (Hassiotou & Geddes, 2013).
The mammary stem cells physiologically experience asymmetric cell division (Ballard et al., 2015), each subdividing into two daughters; one of which is identical to the mother and the other starts committing. This process differs from symmetric cell division, during which the mother cell two daughters identical to itself.

| Homeostasis
In the mammary gland, homeostasis involves the renewal of somatic stem cells (basal and luminal). Ballard et al. proved that the binding between the extracellular signal SLIT2 and the cellular receptor ROBO1 controls the balance between self-renewal based on asymmetric cell division and self-renewal based on symmetric cell division (Ballard et al., 2015). If SLIT2/ROBO1 signaling fails, these cells tend to undergo symmetric cell division alone.
In the homeostatic process, hormones also play an important role.
For example, the growth hormone, which stimulates the production of insulin-like growth factor-1 and estrogen (Macias & Hinck, 2012), controls homeostasis during puberty. Ovulation and the menstrual cycle then activate the MaSCs thus driving breast development. The mammary gland is completely remodeled during the pregnancyactation cycle, due to the activation of MaSCs by a circulating hormonal complex, which leads to changes in the ductal system (Hassiotou & Geddes, 2013). In the final phase, progesterone and prolactin, two hormones responsible for the differentiation of alveoli and for the increase in secondary and tertiary ductal branching, manage gland maturation and alveologenesis (Macias & Hinck, 2012).

| Spontaneous repair
In the case of damage or malfunction of the mammary's ductal tree, inflammation and scar formation occur. No studies have yet demonstrated the restoration of functionality, especially in the case of severe or extended injuries.

| Pathologies requiring regenerative therapies
Most breast cancers originate in the terminal duct-lobular unit of the mammary glands, driven by the activation of oncogenes (HER2) and the inhibition of tumor-suppressor genes (such as BRCA1, BRCA2, and p53; Antoniou et al., 2003).
Biochemical (Gudjonsson et al., 2002) and mechanical factors can also cause tumorigenesis. The atypical presence of stomelysin-3, around the tumor (Andarawewa et al., 2005), and the loss of the ability to produce functional laminin 332 (Kwon,

| Conventional therapies
Several conventional therapies are currently used both for breast reconstruction, in the case of tumors, and for breast augmentation.
The most common treatment for both purposes is mastectomytotal and conservative. The latter consist of skin-sparing mastectomy and nipple-sparing mastectomy: Both of which are used for tumors and cosmetic purposes. A total mastectomy removes the entire breast glandular tissue. The conservative mastectomy spares the healthy skin envelope with or without the nipple-area complex. A research group based in Ireland has shown that the remaining skin works as a pocket to facilitate the reconstruction through prosthesis (O'Halloran, Courtney, Kerin, & Lowery, 2017). Benson et al. suggested applying conservative mastectomy to women with a ductal carcinoma in situ, located at least 2 cm aside the nipple-areola complex (NAC). On the other hand, where tumors are bigger than 5 cm, patients need to undergo a total mastectomy (Benson, Dumitru, & Malata, 2016).
A conservative mastectomy is used only in the case of cancer development and is known as breast-conserving therapy. This treatment is a valid alternative to mastectomy, for patients with breast cancer Stages I or II, whose dimension needs to be smaller than 5 cm.
The first step in this procedure, consists in breast conserving surgery, also known as a lumpectomy, aimed at eradicating the breast cancer.
The second step is based on a controlled dose of radiotherapy to remove possible microscopic residuals of the tumor. Unlike mastectomy, breast conserving therapy provides both better esthetic results and reduced postoperative complications (Rahman, 2011).
The final procedure is autologous fat transfer, also called fat grafting. This is currently available for both breast reconstruction and cosmetic purposes. The core of the treatment is to harvest autologous adipose tissue, usually from the patient's abdomen or thigh and its consequent centrifugation to eliminate substances, such as lipase, which prevent the tissue from being viable for use. The processed tissue is then grafted onto the site of interest. Fat grafting may be a better alternative to the other therapies mentioned above, because adipose tissue is biocompatible, easy to obtain, and the procedure is repeatable and less expensive (Bellini, Grieco, & Raposio, 2017).

| Critical issues regarding conventional therapies and an introduction to regenerative therapies
Synthetic prosthesis has some shortcomings, such as infection of the surrounding tissues, wound dehiscence, capsular contracture, variations in shape, dislocation from the implant site, calcification, or even a further rupture of the prosthesis itself (Galimberti et al., 2017).
In the worst case, the patient requires a replacement of the implant through additional surgery. A research group based in the United States showed that almost 60% of patients who underwent breast conserving therapy, due to breast cancer, needed a re-excision to remove remaining diseased tissue (Waljee, Hu, Newman, & Alderman, 2008). Moreover, Fajdic and colleagues, reported cases of the local recurrence of tumor after treatment (Fajdic, Djurovic, Gotovac, & Hrgovic, 2013).
Finally, regarding fat transfer, Gentile et al. reported that the percentage of surviving volume of transplanted tissue with respect to the injected tissue reached 39%. This means that the adipose tissue alone cannot naturally maintain the breast volume (Gentile et al., 2012).
As an alternative to these conventional therapies, O'Halloran and collaborators showed the increasing interest in using adipose-derived stem cells (ADSCs) for breast reconstruction and augmentation (O'Halloran et al., 2017). These cells show a high regenerative potential and commitment mainly in mature adipocytes, which are able to maintain the viability of the transplanted adipose tissue and provide a more natural volume of the mammary gland.

| IN VITRO RESEARCH
The need for greater psychological and esthetic benefits has led to the development of innovative strategies for breast reconstruction after invasive interventions, such as the mastectomy. Each newly engineered medical strategy should be modeled in order to predict the efficacy of the culture approach (Galbusera, Cioffi, & Raimondi, 2008) and then tested in vitro under different controlled conditions. Below, we summarize the various strategies that have been used in the development of the first experimental step of in vitro modeling.

| Cell models
Embryonic stem cells are theoretically beneficial for regenerative therapy as a result of their pluripotential capacity; however because of ethical and regulatory issues, they have not been used to date.
Research has thus focused on the use of primary adult ADSCs, especially autologous ones. These are multipotent cells that have been used since 2001 as a major source of cells with a regenerative potential, becoming the gold standard for regenerative techniques and breast tissue engineering (O'Halloran et al., 2017;Sterodimas, de Faria, Nicaretta, & Pitanguy, 2010).
ADSCs secrete cytokines and growth factors that stimulate tissue recovery in a paracrine manner. ADSCs modulate the host local reaction by enhancing differentiation in the endogenous stem cells along the required lineage pathway. At the same time, they provide antioxidants, free radical scavengers, and chaperone/heat shock proteins. As a result, toxic substances released into the injured environment can be removed thus promoting the recovery of the surviving cells (Gimble, Katz, & Bunnell, 2007).
ADSCs are collected by the centrifugation of the SVF, the product of liposuction, previously digested by collagenase. The SVF contains various cell types such as endothelial precursor cells, endothelial cells, pluripotent vascular progenitors, macrophages, and preadipocytes (O'Halloran et al., 2017).
The advantages of SVF composition and the easy harvest method have led to the development of strategies involving the use of ADSCs enriched with SVF (Bielli et al., 2014;Bora & Majumdar, 2017;Sterodimas et al., 2010).
In addition, the adipose tissue is heavily vascularized, providing vital networks for the normal renewal of the tissue, which thus needs to be taken into account during tissue regeneration.
Several studies have demonstrated that the interaction between vascular endothelial growth factor, secreted by ADSCs, and the endothelial precursor cells, contained in SVF, leads to a more widespread angiogenesis compared with the use of ADSCs alone, differentiated in endothelial cells (Liu, Zhuge, & Velazquez, 2009;Suga et al., 2009).
The main disadvantages of using ADSCs enriched with SVF is related to SVF's heterogeneous composition, which could lead to immunological rejection (O'Halloran et al., 2017). Numerous interactions between adipocytes and epithelial cells are still being studied, encouraging researchers to use adipocyte-epithelium cocultures (Wang, Reagan, & Kaplan, 2010).
Several studies in tissue engineering technology have shown that three-dimensional adipocyte-epithelium cocultures enhance epithelial cell differentiation, with respect to two-dimensional cocultures. Cell injection in the implant site is a viable alternative in clinical use; however, it has some limitations such as the formation of cysts. The positioning of a construct (cells + scaffold) has thus been studied, which could improve the benefits of breast regeneration, while limiting the drawbacks of cell administration alone. Both human adipose stem cells (hASCs) and epithelial cells involved in these studies are from human sources and thus can be labeled as primary cells.
Although the current in vitro research in adipose tissue engineering has advanced, ADSCs or adipose stem cells (ASCs) mechanisms of interaction with other physiological cell sources are still not clearly understood. ADSCs help to repair damaged or failing organs. Such mechanisms have been studied as in case of cardiac diseases, endothelial formation mechanisms (with an effective lineage commitment; Gimble et al., 2007), bone tissue, and cartilage failures (Strem et al., 2005); however, no studies to date have described how ADSCs help in the recovery of the mammary gland tissue.
Researchers have also focused on using induced pluripotent stem cells (iPSCs). This kind of cell source is obtained by converting adult somatic cells, of autologous origin, to a pluripotent phenotype by genetic engineering, such as the retroviral transduction of four specific transcription factors (Oct3/4, Sox2, Klf4, and c-Myc; Takahashi et al., 2007).
iPSCs are usually obtained from adult fibroblasts or epithelial cells thus dramatically reducing problems related to the availability of the donor cellular source. This ability to reprogram somatic cells toward a pluripotent lineage has dramatically changed in vitro approaches in the regenerative field. Qu et al. (2017) were the first to develop a method to direct "human" iPSCs toward mammary differentiation.
They cultured the cells with the MammoCult medium (Hassiotou et al., 2012), usually employed to enrich breast stem cells with nonneural ectoderm ones susceptible for the formation of mammary-like cells, then directing the human iPSC differentiation toward the nonneural ectodermal type as a precursor of the desired tissue formation. Mechanical, molecular, structural cues, and scaffold properties are key (Nava et al., 2015) to maintaining the architecture and the functionality of this tissue (Nava et al., 2016;Wang et al., 2010) thus leading to ADSC proliferation and differentiation (Sterodimas et al., 2010).

| Effect of scaffold/matrix on 3-d models
For example, Tremolada and colleagues (Tremolada, Palmieri, and Ricordi, 2010) verified the expansion of hASCs in vitro: 1 g of adipose tissue provides approximately 5,000 stem cells, which is 500-fold greater than the yield from the same volume of bone marrow. This shows that the adipogenesis capability of ASCs is greater than the other differentiation lineages when supported by a suitable 3-D scaffold.
The choice of scaffold is thus fundamental for the success of the related regenerative therapies and for the analysis of the cell types.
Examination of the types of scaffold is essential in order to get the best results from in vitro analyses.
Good scaffolds, engineered for constructing synthetic adipose tissue, require biocompatibility, degradability, noncytotoxic byproduct formation, and soft tissue mechanical properties, which provide the correct vascularization. Many research teams have thus studied natural, synthetic, and biological scaffolds (Table 1).
Collagen gels have been used as a 3-D matrix on which preadipocytes were placed in a coculture with human autologous mammary epithelial cells (Wang et al., 2010). For example, researchers seeded a coculture of nonmalignant human epithelial cell lines (MCF10A) and predifferentiated hASCs, which were embedded in Matrigel and a 3-D porous collagen scaffold (O'Halloran et al., 2017;Wang et al., 2009;Wang et al., 2010). They verified that the presence of hASCs increased both alveolar and ductal morphogenesis, also inhibiting epithelial cell proliferation. This was done in order to propose a regeneration therapy for restoring all the various types of tissue that make up the breast. In contrast to cocultures, Wang et al. (2009) also showed that monocultures can only regenerate the alveolar structures, which led to the analysis of cocultures. Yao, Zhang, Lin, and Luan (2013) verified that collagen/alginate microspheres were suitable for ADSC encapsulation and delivery.
After 4 weeks of using an adipogenic differentiation medium, from a macroscopic point of view, the microspheres were similar to the adipose tissue lobules. 3-D silk scaffolds with hASCs and human umbilical vein endothelial cell cocultures depict human vascularized adipose tissue (Kang, Gimble, & Kaplan, 2009). Silk has a controllable degradation time, thanks to which it is possible to enhance neovascularization and prevent premature loss of nutrient transport through the matrices, which may lead to necrotic areas. This latter issue must be taken into account in choosing the best scaffold to degrade with a suitable kinetics mechanism.
A 2010 study used gelatin sponge scaffolds onto which human mammary stem cells were seeded. Thanks to the adipogenic differentiation medium, the authors observed the accumulation and expansion of lipid droplets (Wang et al., 2010).
F I G U R E 3 Bioengineered graft: isolation and differentiation of adipocytes from adipose tissue, cells seeding on a scaffold and suitable stimulations of the construct. The dotted line traces a possible orthotopic implantation, which does not strictly concern in vitro models, but regards clinical trials. bFGF, basic fibroblast growth factor; hASCs, human adipose stem cells; SVF, stromal vascular fraction [Colour figure can be viewed at wileyonlinelibrary.com] Polylactic acid, polyglycolic acid, polyethylene terephthalate, polylactic-co-glycolic acid, polyurethane, and polycaprolactone are suitable synthetic materials for mammary gland tissue regeneration (Zhu & Nelson, 2013). These kinds of scaffolds, with adipogenic stimulants, have shown significant adipogenic transcript levels and substantial lipid accumulation, which unfortunately decreases in vivo (Zhu & Nelson, 2013).
Adipose decellularized extracellular matrix is the key ingredient for biological scaffolds. This type of scaffold has suitable biological properties and an excellent maintenance of its 3-D architecture and biochemical composition, after decellularization.
Mature adipocytes were analyzed for 8 weeks after seeding within a decellularized human placenta scaffold. The results were positive; however, this procedure is expensive and time-consuming (Flynn, Prestwich, Semple, & Woodhouse, 2009).
The production of the three categories of scaffolds described above is based on different techniques: 3-D printing, sponge,

| Whole organ models
Due to the multiple cell types of the mammary gland, the major challenge is coculturing the epithelial cells, fibroblasts, endothelial cells, and adipocytes into a suitable matrix scaffold. Microfabrication approaches could lead then to the regeneration of a more complete mammary gland structure (Zhu & Nelson, 2013).
Organ-on-a-chip platforms have been described in the literature,  Campbell, Hume, et al., 2014).
For example, a combination of adipose tissue engineering and microfabrication techniques has been proposed to build up a 3-D engineered epithelium aimed at mimicking the physiological duct structure of the breast. The researchers started seeding preadipocyte in a collagen gel, in which they generated cavities using a stamp. They then induced the preadipocytes differentiation. Finally, epithelial cells covering the cavities formed tubules. They used this model to study gene expression, the mechanical stress profile during branching morphogenesis, and the consequences of a host microenvironment (Zhu & Nelson, 2013).
Future perspectives could focus on the potential of organ-on-achip or a staminal niche (Raimondi et al., 2013) (to expand stem cells in vitro); however, researchers are still far from regenerating the whole organ architecture in vitro.

| PRECLINICAL TRIALS
Animal models can represent human tissue and physiology; however, the efficacy and safety of newly developed strategies need to be demonstrated before clinical trials can occur. Preclinical inspections are thus mandatory (required by European law) on all human-pledged drugs or cellular therapy products.
This section describes the potential of ADSCs to provide new adipogenesis and vasculogenesis, the oncological risks that entail examining the interaction between breast cancer cells, and fat grafting enriched with ADSCs.

| ADSC effects on fat grafts
Autologous fat grafting is a well-known technique for breast reconstruction; however, recent studies have reported reabsorption rates from 40% to 60% of the injected volume. Much of the volume loss is probably due to inadequate vascularization and to intolerance of the fat tissue to hypoxia (Tsuji et al., 2018). In addition, ADSCs are suspected of supporting the growth and the progression of tumors, which could increase the risk of breast cancer recurrence (Schweizer et al., 2015).
To investigate the effects of ADSCs on fat grafts, Jiang, Li, Duan, Dong, and Wang (2015) carried out animal experiments in which different techniques were used on 18 mice in the following groups: control, ADSCs, and ADSCs + basic fibroblast growth factor (bFGF) method. Mice in the control group were injected with 2 ml of aspirated fat and 0.1-ml medium directly. Mice in the second group were coinjected with a mixture containing 1 × 10 6 ADSCs with 2 ml of aspirated fat treated in 0.1-ml medium. Mice in the third group were injected with a blend of 10 6 ADSCs with 2 ml of isolated aspirated fat and 100-U bFGF. The aspirated fat and ADSCs were isolated by liposuction from six healthy female human donors. The fat was injected subcutaneously into the mouse's back and, after 12 weeks, each fat graft was weighed, and the volume was measured in order to calculate the survival ratio of transplanted fat using the following formula: survived volume/previous volume (2 ml).
The results demonstrated that the weight and survival ratio in ADSCs and ADSCs + bFGF-treated fat grafts were significantly higher than the control group (Figures 5 and 6).
Jiang's study demonstrated that ADSC treatment improves fat graft survival and fat integration. In particular, ADSCs + bFGF treatment had the best beneficial effect on the fat graft weight and survival ratio due to the ability of bFGF to modulate stem cell self-renewal, differentiation, and survival.

| ADSC effects on breast cancer cells
To study the risk of tumor formation and cancer metastasis, ) transplanted breast cancer cell lines (MDA-MB-231) with and without ADSCs into two different mouse groups. The researchers F I G U R E 5 Weight in different fat grafts. *p < .01 versus the control group; **p < .05 versus ADSCs group, (Jiang et al., 2015). ADSCs, adipose-derived stem cells; bFGF, basic fibroblast growth factor F I G U R E 6 Survival ratio in different fat grafts. *p < .01 versus the control group; **p < .05 versus ADSCs group, (Jiang et al., 2015). ADSCs, adipose-derived stem cells; bFGF, basic fibroblast growth factor  Table 2 summarizes the experimental studies conducted on ADSCs when these cells are coinjected in fat grafts together with the interactions between breast cancer cells and ADSCs. These studies revealed that ADSCs have the capacity to improve the long-term F I G U R E 7 3 × 10 6 human MDA-MB-231 breast cancer cell lines were injected subcutaneously into the mammary fat pads with or without 3 × 10 6 human ASC cells from a donor with a body mass index of 25. *p < .05, (Rowan et al., 2014). ASC, adipose stem cells; BMI, body mass index F I G U R E 8 3 × 10 6 human MDA-MB-231 breast cancer cell lines were injected subcutaneously into the mammary fat pads with or without 3 × 10 6 human ASC cells from a donor with a body mass of 18.3. *p < .05, (Rowan et al., 2014). ASC, adipose stem cells; BMI, body mass index ADSCs seemed to have no effect when administered intravenously after the transplantation of cancer cells . Other factors could be the weight of the patients; thus, future studies using multiple donors with different BMIs will be able to determine whether donor BMI has an impact on the effects of ASCs on tumor growth (Rowan et al., 2014).
Although clinical trials have been carried out on the use of fat grafting combined with ADSCs, the data on the influence of ADSCs on breast cancer cells are still contrasting (Schweizer et al., 2015).

| CLINICAL TRIALS
This section focuses on the results of six clinical trials (Table 3)  In the cell-assisted lipotransfer method (Figure 9), which is performed in all the trials mentioned below, fat is enriched with ADSCs, contained in the SVF or with platelet-rich plasma, obtained after enzymatic digestion of fat or after cell culture, to improve the fat survival rate. In all the alternative treatments, fat grafting works only as a decellularized scaffold, on which cells are seeded, as described in the section related to in vitro and animal trials.
All these clinical trials suggest that ADSCs can survive without undergoing postsurgery hypoxia, which is thought to result in the necrosis of a conventional fat graft (Khouri et al., 2014). This T A B L E 2 In Vivo experiments studying the potential of ADSCs when coinjected in fat grafts and the effects of ADSCs on breast cancer

3-D imaging
Note. The letter "I" in the first column indicates the case in which the proposed approaches are clinically eligible.
provides the scientific rationale for the use of these procedures in breast surgery and consequently makes these techniques key in the repertoire of techniques available to surgeons (Yoshimura et al., 2008). All the reported trials have a common primary objective, which consists in breast reconstruction or augmentation and, as secondary objective, esthetic improvement. In each trial, patients signed an informed consent regarding "the aims, methods, sources of funding, any possible conflicts of interest, institutional affiliations of the researchers, the anticipated benefits, and potential risks" In the following section, three significant clinical trials are reported; all of which should be considered as an attempt to find a solution to the disadvantages of each of the conventional therapies mentioned above.

| Fat graft enrichment with SVF cells
In the context of fat graft enrichment, Yoshimura et al. analyzed 15 patients in a clinical single-arm trial that reached Phase II. After breast implant removal, a cell-assisted lipotransfer was carried out, based on the injection, into one single breast, of 264 ml of autologous adipose tissue extracted from the abdomen or thighs enhanced with 9.7 ± 1.7 × 10 7 stem cells (SVF) per liter of tissue, isolated, expanded, and seeded in the fat graft (Yoshimura et al., 2006).
One quantitative parameter used to verify the efficacy of this therapy is the surviving fat volume, after a 1-year follow-up (ε[mL]), calculated as Equation (1): Yoshimura et al. used three-dimensional measurement tools to quantify each volume of the expression above. The outcomes ( Figure 10) show that the average surviving volume of the transplanted adipose tissue in the right breast varies during the follow-up. In the first 5 weeks, there is a gradual absorption of fat, whereas in the remaining period, the volume decreases slightly, reaching 155 ± 50 ml, more than half of the injected adipose tissue.
No calcifications and microcalcifications were observed. In conclusion, these results suggest that the enrichment of fat grafting with SVF cells, in a cell-assisted lipotransfer, may not be a safe and effective treatment, but might also be a suitable alternative to the breast reconstruction or augmentation (Yoshimura et al., 2010).
Considering grafts enhanced with SVF, another research group based in Italy (University of Rome "Tor Vergata") carried out a F I G U R E 9 Cell assisted lipotransfer procedure: harvested fat is separated from the donor (the most common sites are the lower abdomen and inner thigh) and processed in one part (lipoaspirate) that is used as lipofilling and in another that is digested by enzymes in order to isolate stem cells (stromal vascular fraction, SVF) and expand them. Fat is then enriched with SVF or platelet-rich plasma (PRP) F I G U R E 1 0 Average trend and standard deviation of the remaining fat volume for the right breast in a follow-up period of 12 months. This parameter takes into account the volume of the removed prosthesis and the variations in breast volume (Yoshimura et al., 2006) [ This clinical trial also shows only a minimal loss of adipose tissue and the absence of calcifications; only in one case was a small cyst and microcalcification found (Gentile et al., 2012).

| Fat graft enrichment with autologous ADSCs
In All the clinical trials reported in Table 3 reached Phase II, except for the study conducted by a research group based in Italy. Through a randomized placebo-controlled trial (Phase III), Domenis and collaborators, using lipotransfer techniques for breast reconstruction, concluded that autologous fat grafts enriched with cells could significantly contribute to the long-term preservation of the transplanted adipose tissue. The cell-assisted techniques used by Domenis, to apply the regenerative treatment, are also currently available on the market (Phase IV; Domenis et al., 2015).
In terms of safety and efficacy (Phase I), all these preliminary procedures have had good results; however, additional trials with longer follow-up periods are essential for the evaluation of the possible risks introduced by these methods in several clinical circumstances.

| COMMERCIAL PRODUCTS
With regards to regenerative therapies for breast reconstruction and/or augmentation, three devices are currently available on the market: Celution ® 800/CRS, LipoKit II, and Fastem Coris (Table 4).
The common procedure is based on the harvesting (liposuction) and centrifugation of autologous fat tissue, from which ADSCs are separated and extracted. The harvested cells are then expanded in specific culture conditions. Lastly, cells are seeded on the harvested adipose tissue, which is grafted on the patient's adipose tissue. Once in situ, cells start committing.
F I G U R E 1 1 Trend of the remaining fat volume in a 1-year followup. The fat grafting with platelet-rich plasma shows the best performance with respect to conventional fat grafting and enriched SVF fat grafting (Gentile et al., 2012). PRP, platelet-rich plasma; SVF, stromal vascular fraction F I G U R E 1 2 Number of colony-forming units: comparison between adipose tissue alone and adipose tissue enriched with stem cells through three commercial devices (Domenis et al., 2015) [Colour figure can be viewed at wileyonlinelibrary.com] [Colour figure can be viewed at wileyonlinelibrary.com] There are two main differences between the three devices: first, the method whereby stem cells get isolated from the tissue allowing them to aggregate in colony-forming units; and second, the cell lineages toward which primitive cells commit. As regards the first method, Celution ® 800/CRS and LipoKit II use an enzymatic separation of the SVF cells, whereas the Fastem Coris technique involves a mechanical procedure to carry out the same goal.
Concerning the second method, Celution ® 800/CRS allows cells to commit into adipocytes, epithelial cells, and muscle cells, LipoKit II enables differentiation into adipocytes, and the Fastem Coris technique does not allow significant commitment of cells, which tend to maintain a stable stemness phenotype.
Due to all these nonminimal manipulations (such as cells extraction and manipulation), these techniques are a feasible cellular therapy product alternative to conventional therapies, such as lipofilling ( Figure 12).
A commonly employed parameter to quantify the efficacy of each system is the number of colony-forming unit cells: an index of cell viability and an ability to differentiate (Domenis et al., 2015).
As reported in Figure 12, Celution ® 800/CRS and LipoKit-IIenriched lipoaspirates are characterized by a higher number of colony-forming unit cells, which is not so evident for the Fastem procedure.

| CONCLUSIONS AND FUTURE PERSPECTIVES
The aim of regenerative medicine applied to breast tissue engineering is to overcome the main criticalities experienced with conventional therapies (mastectomy, breast conserving therapy, and lipofilling), regenerating fat tissue in the breast and improving the esthetic outcome (Visscher et al., 2017). To overcome the side effects of the standard treatments, a combination of lipofilling with tissue engineering has become a promising new technique.
Owing to the presence of multiple cell types in the breast tissue, mammary gland regeneration is quite complex. Although devices that exploit regenerative strategies are available on the market, the existing techniques are constantly being improved.
The key factor is the use of scaffolds, either biodegradable or integrated with the host tissue, to support cell growth until the formation of a stable and mature tissue.
In vitro models verify the importance of different cellular sources to regenerate both adipose and epithelial tissues. ADSCs have several advantages over other sources of stemness. They provide an accessible and easily available source of multipotent stem cells, and their harvest procedure is repeatable and safe, thanks to the negligible morbidity. Their capacity to differentiate both into adipocytes and epithelial cells makes them important for future applications in this field. Further studies are needed to verify the safety and efficacy of cells derived from ADSCs, before reaching preclinical and then clinical trials (Zhu & Nelson, 2013).
Finally, human iPSCs could be a promising cellular source. They are obtained by converting adult somatic cells of autologous origin, into pluripotent stem cells (Takahashi et al., 2007). This thus avoids ethical and technological controversies that hinder the application of human embryonic stem cells. A recent promising study demonstrated the possibility of obtaining mammary-like cells and organoids from human iPSCs (Qu et al., 2017). Future research by this group will further examine the effects of growth factors on iPSC differentiation into mammary cells to refine the generation of a two-layer mammary ductal structure. However, the research work is still in the in vitro phase; thus, many other studies are required before obtaining preclinical models. Moreover, criticalities related to the teratoma formation after reinfusion of differentiated cells from iPSCs need to be examined.
Many scaffolds have shown suitable properties as found in in vitro research. The main advantage of this strategy is that biomaterials already approved by the Food and Drug Administration, or other regulatory organisms (e.g., European Medicines Agency), can be adapted to create brand new cellularized constructs, which would speed up the development of all trials up to the commercialization phase (Jewell, Daunch, Bengtson, & Mortarino, 2015;Visscher et al., 2017).
It is worth noting that the principal purpose is not to replicate the physical structure, but rather promoting tissue regeneration by designing a customized scaffold architecture and using the patient's body as a bioreactor.
Adipose tissue has a high metabolic function; thus, the energy demand needs to be met, preventing any necrosis and volume loss  (Tsuji et al., 2018). Contradictory results on the other hand, have shown that human ADSCs harvested from patients with a body mass index of 18.3, and injected subcutaneously into the mammary fat pad of mice, promoted the growth and proliferation of breast cancer cells; thus, the weight of the patient can be a factor for cancer-promoting effects. Further studies, using different BMIs, are needed to study the effect of ADSCs on tumor growth (Rowan et al., 2014).
Despite the occasional unsatisfactory results obtained in preclinical trials, some regenerative strategies reach the stage of human trials, most using the stromal vascular fraction to enrich autologous fat tissue (e-SVF). The common target of these procedures is the long-term assessment of the actual effects of SVF cells on fat graft volume retention and its maintenance.
Optimal results have been obtained, in particular for the enrichment of adipose tissue with platelet-rich plasma cells, which increase the number of ASCs, maintaining the expression of CD34, an indicator of stemness (Gentile et al., 2012;Yoshimura et al., 2006).
Almost all patients involved in these trials were pleased with the esthetic outcome of the treatment, in terms of breast volume and contour preservation. The improved survival of breast tissue may be due to the ability of SVF cells to increase vascularization and secrete pro-surviving growth factors. The crucial result is the absence of tumor recurrence in those patients who have undergone regenerative therapy due to breast cancer.
The overall results also highlight the superiority of regenerative therapies over standard procedures in obtaining both reliable and esthetically effective outcomes.
Clear evidence of the progress made by regenerative medicine in terms of breast tissue treatment is the availability on the market of three devices exploiting the cell-assisted lipotransfer technique: Celution ® 800/CRS, LipoKit II, and Fastem Coris.
In conclusion, breast tissue engineering can be considered as a viable alternative to the criticalities related to conventional therapies, potentially leading to a personalized medicine paradigm (Visscher et al., 2017).