Dual Extrusion Patterning Drives Tissue Development Aesthetics and Shape Retention in 3D Printed Nipple-Areola Constructs

Abstract

Breast cancer and its most radical treatment, the mastectomy, significantly impose both physical transformations and emotional pain in thousands of women across the globe. Restoring the natural appearance of a nipple-areola complex directly on the reconstructed breast represents an important psychological healing experience for these women and remains an unresolved clinical challenge, as current restorative techniques render a flattened disfigured skin tab within a single year. To provide a long-term solution for nipple reconstruction, this work presents 3D printed hybrid scaffolds composed of complementary biodegradable gelatin methacrylate and synthetic non-degradable poly(ethylene) glycol hydrogels to foster the regeneration of a viable nipple-areola complex. In vitro results showcased the robust structural capacity and long-term shape retention of the nipple projection amidst internal fibroblastic contraction, while in vivo subcutaneous implantation of the 3D printed nipple-areola demonstrated minimal fibrotic encapsulation, neovascularization, and the formation of healthy granulation tissue. Envisioned as subdermal implants, these nipple-areola bioprinted regenerative grafts have the potential to transform the appearance of the newly reconstructed breast, reduce subsequent surgical intervention, and revolutionize breast reconstruction practices.

Graphical Abstract

The nipple-areola complex constitutes an important landmark on the breast and its loss due to breast cancer treatment can have devastating psychological effects on the patient. A 3D-printed tissue engineered dermal scaffold, composed of complementary biodegradable and synthetic hydrogels, is developed to promote the formation of a lasting nipple-areola complex for mastectomy patients.

With the latest estimate of 276,480^[1] new cases of invasive breast cancer occurring each year in the US, many female patients will face a malignant form of this disease even with an early diagnosis by annual mammographic screening. Breast cancer most commonly begins in the cell lining of the lactiferous ducts^[2] and can quickly spread into the surrounding breast tissue, infiltrate into axillary lymph nodes, and further migrate to distant organs. Most patients must undergo surgical treatment (modified radical mastectomy - MRM, breast conserving surgery - BCS, or nipple-sparing mastectomy - NSM) on top of adjuvant therapies to resect the affected tissue and eradicate any metastasized tumor cells. A significant number of patients (58%)^[3] choose mastectomy to avoid the negative side effects of adjuvant therapy and decrease the likeliness of cancer recurrence. In most cases, the cancer has spread through the ductal tract of the mammary glands and negatively impacts the health of the nipple tissue, resulting in the inability to undergo NSM^[4]. The most common mastectomy performed is the MRM, which entails the removal of the entire breast tissue.

Mastectomies can be life-saving surgeries, yet breasts are so deeply tied to the patient’s identity and self-esteem that for many women, losing one or both of their breasts is undeniably difficult to mentally process. Reports indicate that >70% of patients experience tremendous pain and depression posttreatment of the disease^[5]. Trauma experienced during the initial treatments of chemotherapy and its side effects, including radiation fatigue, nausea, hair loss, changes in taste, headache, and anxiety, can be recalled with a mere glance at the patients’ scars^[6]. Many patients have found breast reconstruction increases their sense in quality of life, and a majority (63%)^[7] of patients enduring a mastectomy in the US undergo these procedures. Plastic surgeons have found that the additional restoration of the critical visual landmark -the nipple-areola complex- is influential in the emotional recovery of their patients^[8]. Thus, nipple reconstruction using Skin Flap Suturing (SFS) techniques has become a regular post-operative procedure, where the skin located at the highest point of the reconstructed breast mound is incised to small skin flaps and sutured together to build an elevated skin tab. Unfortunately, the outcome of the sutured knot is an inevitable loss of projection with contracture due to wound healing effects. Studies have demonstrated a disappointing 70% loss of projection within 12 months due to complications in scar contracture, retraction forces from surrounding skin, and inadequate vasculature^[9–11]. Regardless of the skin flap incision pattern used, the result remains a flattened skin tab scar that these women repeatedly fix throughout life^[12–17]. The act of repetitive reconstructive surgery is yet another detrimental psychological challenge these women face, adding even more difficulty to overcoming their previous disease. Consequently, nipple reconstruction remains an unresolved clinical challenge and can benefit from a reimagined, tissue regenerative approach.

Several attempts have been made to address this problem, yet these methods fail to accurately capture the relevant size, shape, and tactile properties of the nipple. Research groups have investigated a variety of materials to create a nipple projection, including autologous materials (cartilage^[18], fat^[19], calcium hydroxyapatite^[20]), alloplastic materials (hyaluronic acid^[21], artificial bone substance^[22]), decellularized matrices (rolled dermal grafts^[23], cadaveric nipple tissue^[24]), and synthetic materials^[25]. Although each approach might have some advantages, undesirable aspects persist across the spectrum that limit their long-term clinical applications such as rapid remodeling, in vivo collapse, and donor site morbidity of autologous materials; interference of breast tissue oncological surveillance and high maintenance costs of alloplastic injections; poor shape retention and low rates of patient adoption of cadaveric tissue for decellularized dermal matrices; and a myriad of complications for fully silicone implants such as implant migration and rupture, fibrotic encapsulation, hematoma, local flap necrosis, and implant extrusion. Overall, these materials result in low patient satisfaction due to the absence of long-term shape retention, natural tactile properties, and a reliable aesthetic restoration.

The goal of this study is to tissue engineer a dermal scaffold that promotes the formation of a lasting nipple-areola complex for mastectomy patients, thus avoiding multiple restorative surgeries. As additive manufacturing via 3D printing has become a popular and advantageous practice to produce scaffolds with complex, patient-specific structures, this technology holds great promise for the fabrication of custom shaped nipple-areola grafts per any breast size. Groundbreaking work recently published by Samadi et al^[26] describes the beauty of custom-designed 3D printed products with their polylactic acid (PLA) cylindrical scaffolds embedded with auricular cartilage for nipple reconstruction; while their findings are promising, the stiff PLA cage was shown to degrade minimally and its rigidity may negatively impact patient satisfaction with the product. Our previous work on a dual hybrid printing technique of synthetic and cell-laden bioinks developed an elegant approach to fabricate long-term shape-retaining soft tissue scaffolds with more desirable mechanical properties^[27]. Here we present a bioengineered system specifically designed for the nipple-areola complex that incorporates entangled methylcellulose (MC) and poly(ethylene) glycol (PEG) networks integrated with a gelatin methacrylate (GelMA) bioink to provide a combination of strength and resistance to nipple projection degradation, while aiding dermal regrowth. We first engineered the MC-PEG bioink to obtain a hybrid scaffold that mirrors the mechanical properties of nipple tissue. The printed scaffolds were evaluated for fibroblastic contraction over an extended period of in vitro culture to investigate the impact of printing parameters on nipple projection. We further tested the biocompatibility of these hybrid matrices in a rat subcutaneous model over a 4-week period. Histological analysis revealed a mild inflammatory, the development of healthy granular tissue, and signs of neovascularization. Overall, our results suggest that the dual extrusion print patterning of GelMA and MC-PEG produced an innovative strategy for aesthetic nipple restoration through a viable nipple-areola composite. The proposed system holds significant promise for clinical applications as it can be seamlessly applied to current breast reconstruction practices post patient healing of silicone implantation and thus revolutionize current standards of breast reconstruction.

Results

Synthetic Ink Development and Hybrid Nipple-Areola Constructs

Because of its unique thermal gelation, methylcellulose is a prime candidate for pressure-driven extrusion printing. We created a bioink containing 15% w v⁻¹ methylcellulose homogenously mixed with a 20% w v⁻¹ PEG solution. The MC-PEG ink was strategically combined with 7% w v⁻¹ GelMA, a popular biomaterial that is known for its natural cell binding motifs, ease in printability, and UV photopolymerization. By themselves, GelMA hydrogels have a tendency to fracture when subjected to compressive loads due to their brittle and swollen nature, which can result in a structural collapse and ultimately scaffold failure. Hence, we established a hybrid printing technique that capitalizes on the strengths of both natural (GelMA) and synthetic (MC-PEG) hydrogels to create printed structures containing tunable mechanical, chemical, physical, and biological properties (Figure 1). For dermal mimicry and further in vitro evaluation, primary human dermal fibroblasts were encapsulated within the GelMA matrix.

Figure 1: Components of the Hybrid Nipple-Areola Implant.

Two bioinks (GelMA and MC-PEG) are co-printed in various patterns to create a hybrid regenerative scaffold. The GelMA bioink is composed of gelatin methacrylate polymers that can be crosslinked via UV light exposure when photoinitiator LAP is present. Fibroblasts can be encapsulated within the GelMA bioink and serve as a biodegradable region for host dermal integration. The MC-PEG synthetic ink is a double polymer network composed of methylcellulose autonomously interacting with its hydrophobic groups and poly(ethylene) glycol covalently crosslinked upon UV exposure.

In Vitro Nipple Implant Functionality

Physiologically sized nipple-areola scaffolds were fabricated by alternating MC-PEG and GelMA bioinks in various print patterns (Figure 2). The hydrogel compositions of these scaffolds were dictated by the print patterns used during fabrication, specifically where MC-PEG:GelMA varied from 1:0 (no tissue regenerative capacity, MC-PEG control) to 0:1 (completely degradable GelMA control). These hybrid constructs were analyzed in terms of internal cell viability and mechanical properties, as it was equally imperative that the graft 1) provided adequate diffusion to support cellular proliferation and 2) exhibited a stiffness that mirrors the natural value of nipple tissue to account for patient preference.

Figure 2: In Vitro Implant Functionality.

Hybrid nipple-areola scaffolds were 3D printed using the Envisiontec Bioplotter. A. CAD designs of each hybrid print, displaying the deposition of GelMA (pink) and PEG (purple) bioinks. B. Visual representation of the various print patterns used to fabricate the hybrid scaffolds. C. Representative images of each hybrid print, both before and after collagenase digestion, scale bar 10 mm. D. Cell viability visualized via Live/Dead (Calcein AM green, live cells; Ethidium Homodimer red, dead cells). E. Uniaxial compression testing was performed on hybrid scaffolds both before and after experiencing collagenase digestion with respective compressive modulus recorded (n=3, p<0.001). F. Fluorescent images (n’=5, over n=3 biological samples) were taken using a fluorescent microscope (Nikon) and processed using a configured MATLAB code. Cell viability was determined from the ratio of the number of live cells to the total number of cells. Data was analyzed using single factor analysis of variance (ANOVA) followed by Tukey’s Multiple Comparison Test assuming normal data distribution with a confidence of 99.9% (p<0.001).

Across print patterns 1:2 and 1:3, the primary adult human dermal fibroblasts (2 million cells mL⁻¹ concentration) displayed a high range of viability throughout the entire scaffold, specifically presenting 78.12 ± 13.06% and 85.25 ± 6.58% for day 1 and 78.35 ± 9.59% and 74.98 ± 9.52% for day 14 respectively (Figure 2D, F). These results indicate that 1:2 and 1:3 hybrid scaffold compositions provide a matrix conducive for cellular proliferation, as no significant difference in cell viability is noted throughout the 2-week in vitro culture. Pattern 1:1 was the only hybrid that showed a significant decrease in cell viability, with an initial value of 81.42 ± 12.64% on day 1 that decreased to 57.37 ± 6.67% by day 14. This drop in cell viability could be caused by either 1) the decreased GelMA volume within this 1:1 print pattern, which lowers the potential adhesion sites for the encapsulated human dermal fibroblasts throughout the scaffold; or 2) the dense MC-PEG polymer network may have affected the internal diffusion of cytokines, media, and cellular waste throughout the construct. Previous investigators have similarly found that a densely-crosslinked hydrogel network can impede vital diffusion throughout cell-laden constructs, which can negatively impact cellular proliferation^[28–30]. Thus, these results showcase the necessary balance between the synthetic and biodegradable components in the scaffold’s design to support high cell viability throughout the graft.

We further evaluated the compressive moduli of the hybrid nipple-areola scaffolds. The achievable range was bound by the modulus of the two single component hydrogels: MC-PEG at 152.2 ± 2.4 kPa and GelMA at 60.7 ± 4.3 kPa (Figure 2E). The hybrid moduli were 136.50 ± 15.6 kPa for the 1:1 pattern, 111.4 ± 4.7 kPa for the 1:2 pattern, and 101.4 ± 14.1 kPa for the 1:3 pattern. Acellular scaffolds were chosen for this analysis for a multitude of reasons: 1) to exhibit the ‘weakest’ properties of the construct (which would define the construct’s properties directly upon implantation); 2) to reduce experimental cost; and 3) to expedite pattern design and in vitro testing. This evaluation describes the graft’s compressive capabilities prior to any tissue integration with the host. All values aligned with the stiffness of porcine teat tissue, which served as our representative in vivo tissue sample, within one order of magnitude. As the GelMA components are replaced with the patient’s infiltrating cells, the graft is predicted to increase in strength and better match those properties displayed in the porcine control tissue. To determine the influence of the internal MC-PEG structure on scaffold stiffness, the GelMA strands were digested via collagenase incubation for 24 hours at 37°C and the remaining MC-PEG scaffold was further reassessed for stiffness. Although the 1:1 pattern (50.4 ± 3.3 kPa) was significantly stiffer than both the 1:2 sample (13.3 ± 2.8 kPa) and the 1:3 sample (11.6 ± 3.0 kPa), these digested values fell below the GelMA control modulus. These results indicate that the MC-PEG strands can provide structural integrity to the nipple projection during any potential remodeling of the GelMA phase and that the hybrid scaffold can be readily tailored to exhibit stiffness values similar to those observed in vivo while simultaneously providing a biocompatible matrix to enhance the implant’s integration.

Nipple Projection Shape Retention

Nipple flattening is the most notable pitfall of SFS reconstruction and a primary cause for repeated surgical interventions (Figure 3A). In order to improve upon these clinical outcomes, our hybrid implant must be able to maintain nipple projection while being subjected to a high degree of cellular contractile forces. To test this phenomenon in vitro, a high density of primary dermal human fibroblasts (10 million cells mL⁻¹) were encapsulated in the GelMA bioink and printed with the MC-PEG ink into nipple projection scaffolds with various print patterns (MC-PEG:GelMA – 1:0, 1:1, 1:2, 1:3, 0:1). Since the areola is relatively flat and not a part of the nipple projection, it was excluded from the scaffold design to help minimize experimental costs and print time. After 3 weeks of culture, the scaffolds were 3D scanned with a Hexagon ROMER Absolute Arm, and their surface topographies were compared to freshly printed nipple projection scaffolds. Finite element mesh analysis was performed with CloudCompare software and the results are illustrated through color maps, in which regions that display a white color signified coincident points, while a red spectrum (positive deviation) or blue spectrum (negative deviation) indicated a lack of coincidence (Figure 3B). The color maps aligned with our compressive testing results, where the print patterns containing higher MC-PEG content had a greater retention of nipple shape. Patterns 1:0 and 1:1 exhibited minimal scaffold swelling and shrinkage, while pattern 1:2 displayed a modest shrinkage of up to 0.5mm, and patterns 1:3 and 0:1 displayed the highest shrinkage of up to 1.2 mm. The consistent blue hue displayed in maps for 1:3 and 0:1 indicated a uniform contraction across the projection diameter.

Figure 3: Nipple Projection Shape Retention.

A. Projection flattening is a common occurrence in clinical nipple reconstruction while skin flap suturing techniques are used. B. Hybrid scaffolds with nipple geometry were printed with a high concentration of primary human dermal fibroblasts (10 million cells mL⁻¹) and cultured for 3 weeks in vitro under submerged conditions. The scaffolds were then 3D scanned, and their respective point clouds were compared. Color maps (center) display deviations of the contracted scaffold to its original form. Histograms (left) graphically display the percentage of points that deviated from their initial positions. Scale bar displays 5mm. Scaffold dimensions were recorded and compared to its original fabricated form, specifically C projection height, D projection diameter, and E projection curvature. Percent differences display a gradual decrease in scaffold shape, as the scaffolds experienced contraction over the 3-week in vitro culture. n=3 and mean ± propagated error depicted. Data was analyzed using single factor analysis of variance (ANOVA) followed by Tukey’s Multiple Comparison Test assuming normal data distribution with a confidence of 95% (p<0.05).

Point clouds were also compared quantitatively amongst all pattern types (Figure 3C, D, E) to obtain distances along the nipple projection height and diameter, and radius of the nipple projection curvature. As expected, GelMA control samples exhibited the greatest deviation in shape (−16.8 ± 5.7% change in projection diameter, −9.5± 5.3% change in projection height, −16.9 ± 6.3% change in radius of curvature), and the non-degradable PEG control samples exhibited the least (5.0 ± 3.4% change in projection diameter, −0.4 ± 3.0% change in projection height, and −0.8 ± 3.2% change in radius of curvature). Across all patterns, a high degree of shape retention was observed with the highest deviation of −11.31 ± 3.4% change in projection diameter, −10.3 ± 3.7% change in projection height, and −9.33 ± 1.7% change in radius of curvature for the 1:3 hybrid scaffold.

In Vivo Application of Hybrid Nipple Areola Constructs

Following sterile fabrication, four of each hybrid nipple-areola scaffold type (MC-PEG:GelMA 1:0, 1:1, 1:2, 1:3, and 0:1) were randomly implanted in the dorsal subcutaneous regions of Lewis rats (3 scaffolds per animal) to assess the local biological response to these hybrid materials (Figure 4A). In this study, the implanted scaffolds were acellular to limit the recorded immune response to an innate one. In addition, the Lewis strain (an inbred type) was particularly used to minimize the differences in the initial subcutaneous tissue inflammation grade between each animal. All animals exhibited an exceptional pain recovery, active behavior, and healthy weight throughout the 4-week study. Using a modified scoring system following the Standard ISO 10993–6: Biological Evaluation of Medical Devices - Tests for Local Effects After Implantation, the following criteria were considered for a semi-quantitative analysis of the tissue surrounding the implant: inflammatory infiltrate, phagocytosis, degradation of the implant, fatty infiltrate, neovascularization, and capsule characterization. Slight modifications were made to the ISO10994–6 to tailor the analysis to our nipple-areola scaffold and these changes are highlighted in Supplementary Table 1.

Figure 4: Animal Surgery and Method for Capturing Images.

A. Three hybrid scaffolds were subcutaneously implanted in the dorsal lumbar region of each rat. Two incisions were made, the first was the initial 20 mm dorsal midline incision over the thoracolumbar area, and the second was through the superficial cervical fascia (a thin layer of subcutaneous connective tissue). This allowed the implant to be stabilized, as tacking sutures resisted significant movement of the implant. B. Six images were taken for each central section in every resected implant. We defined the location of each image by the length of the scaffold’s areola to normalize this process and ensure that each implant interaction was holistically represented. C. Representative images taken before implantation and D. after resection, where all samples appear to have maintained a nipple projection with minimal fibrotic capsule development. Scale bars represent 3 mm.

For each scaffold, tissue response was evaluated at the implant-subcutaneous space interface and at the implant-fascia interface (Figure 4B), providing us a holistic representation of the foreign body and vascular responses induced by the host. Figures 5, 6, and 7 graphically depict the score averages in every evaluation per scaffold type, where each color (red, yellow, green, blue, and purple) represents the results from a single blind scorer and each violin curve (gray or pink) represents the peak population incidence of each interface. In general, there appeared to be a lower immune response from the underlying fascial tissues compared to the upper subcutaneous interface. This may be due to the imbalanced, yet natural, cellular presence in the matrix of each neighboring tissue type; the fascia is predominantly composed of extracellular matrix (tightly wound collagen and elastin fibrils) with little vasculature and few cellular constituents of mainly fibroblasts^[31], while the subcutaneous pocket experiences a high influx of immune cells as it directly contacts the rat’s skin. To serve its function as a protective barrier, the epidermis is highly composed of immune cells, such as Langerhans cells, neutrophils, mast cells, and dendritic cells, that defend against microbial pathogens, physical, or chemical threats to the host^[32]. Therefore, the subcutaneous interface is more likely to experience a heightened foreign body response and thus receive higher scores due to the increased immune cell migration rate from the epidermal layer.

Figure 5: Cellular Characteristics.

A. Representative model, H&E 2X, and H&E 20x images pertaining to each scaffold type that portrayed the cellular characteristics in and around the scaffolds. B. Semiquantitative scoring evaluations for inflammatory infiltrate (pertaining to lymphocytes, plasma cells, and polymorphonuclear granulocytes) were evaluated by five blind scorers. On the left, colored dots (red, yellow, green, blue, and purple) represent the average score results in every scaffold type evaluation. On the right, a violin curve (type: kernel density estimation) portrays the probability density function of the averaged scores and represents the peak population incidence of each interface (SC refers to the implant-subcutaneous interface, while FA refers to the fascia-implant interface). Results show a lower inflammation of immune cells in FA than in SC interfaces, with minimal-moderate infiltrate present. C. Phagocytosis refers to the presence of macrophages, and the results display a similar trend with lower phagocytosis present in FA than in SC interfaces.

Figure 6: Degradation of Implant and Neovascularization of Surrounding Tissue.

A. Degradation of the biodegradable GelMA is expected in patterns containing the highest amount of this biomaterial (i.e. 1:3 and GelMA 0:1 control), and the average score results reflect this prediction. A striking contrast exists between FA and SC interfaces, which may be due to the increased shear stress induced by the contraction of underlying fascial muscles. B. Neovascularization is present in all scaffold types tested, signifying that granulation tissue is beginning to form. C. Blood vessels were counted in each 20X (n’=6 with n=4 biological samples) image and recorded to validate our previous blind score results. These quantitative image results display mean ± standard deviation and were analyzed using a two-group Paired Comparison plot followed by Tukey’s Multiple Comparison Test assuming normal data distribution with a confidence of 95% (p<0.05). Representative H&E 20X images that display evidence of neovascularization in every scaffold type, where a red arrow points to the location of blood vessels and/or capillaries present at the tissue interface. Scale bars represent 100 μm.

Figure 7: Presence of Fatty Infiltrate and Fibrous Capsule Formation.

A The subcutaneous interface experiences a proliferation of adipose tissue, which can be visualized by the high peaks of the gray violin curves. With all foreign objects introduced to the body, a fibrotic capsule forms at the boundary of the implant and is composed of giant nucleated macrophage cells and surrounding fibroblasts. B Scoring results show that the capsule exists in a mild-moderate thickness. This was further quantified in C, where each image was evaluated at 2 separate points. These quantitative image results display mean ± standard deviation and were analyzed using a two-group Paired Comparison plot followed by Tukey’s Multiple Comparison Test assuming normal data distribution with a confidence of 95% (p<0.05). Representative Masson Trichrome images for capsule quantitative assessment. Scale bars represent 100 μm.

Cellular Characteristics of Surrounding Host Tissue

The score results recorded a range of mild to moderate cellular infiltration that was dependent on the implanted hybrid scaffold type. Inflammatory infiltrate referred to the cells that appeared to be either lymphocyte, plasma, or polymorphonuclear granulocyte-like in appearance, which were generally cells exhibiting a single and rounded and/or lobed nucleus with little cytoplasm. Phagocytosis referred to the macrophage-like cells, such as those with multiple nuclei (multinucleated giant cells) and a large cytoplasm loaded with debris. The infiltration of both cell groups was primarily observed in the implant-subcutaneous interface, as visualized by the higher peaks of the gray violin curves (Figure 5). The presence of lymphocytes and polymorphonuclear neutrophilic granulocytes is frequently associated with the early acute inflammatory response to the surgical procedure^[33]. Macrophages, on the other hand, phagocytose the degradable parts of the implant, which is reflected in the high macrophage presence in hybrid scaffolds 1:1, 1:2, 1:3, and GelMA controls. For non-degrading implants, such as the MC-PEG control, phagocytosis is not possible and the implant becomes encapsulated, which is also reflected in both the pink and gray violin curves and H&E images (lack of macrophage cell presence) shown in Figure 5. Overall, across all hybrid scaffold types, both inflammatory infiltrate and phagocytosis were recognized along the Implant-Subcutaneous interface.

Characteristics of Surrounding Host Tissue

The tissue characteristics of the surrounding host tissue is in congruence with the recorded cellular response. The implants containing the most GelMA experienced the highest degradation; score trends depict the GelMA control exhibited the most degradation, followed by the 1:3 pattern, 1:1 and 1:2 patterns with fairly equal amounts of degradation, and finally the MC-PEG control exhibiting the least degradation (Figure 6A). There appears to be a striking contrast in score trends between the subcutaneous and fascia interface within the GelMA implant test group, where the fascia-implant interface displayed a higher amount of degradation. One possible explanation for this finding would be that the fascia interface experiences a higher shear load along its boundary as the animal moves, and this contraction of underlying muscle may cause mechanical friction that enhances the degradation rate of the implant. Although all scaffolds have experienced relatively the same amount of friction along the fascial interface, the GelMA components of the graft are far more brittle than the MC-PEG components, and thus have a greater potential of experiencing fracture and further enhanced degradation. Figure 2E displays GelMA’s weaker mechanical properties when compared to those of MC-PEG. GelMA contains a loose, covalently-bonded and highly hydrated network, while MC-PEG contains highly crosslinked, enzymatically-resistant double network that displays resilient mechanical properties due to its entangled covalent bonds and hydrophobic interactions. Therefore, patterns containing a higher MC-PEG volume ratio are less prone to degradation. Degradation is favorable in regard to our implant, as its hybrid material makeup is designed to promote integration with the surrounding host tissue and thus lead to an interlocked placement for a stable nipple projection.

Additionally, neovascularization (Figure 6B) signified the formation of granulation tissue. Granulation tissue, a young connective tissue that grows from the base of a wound and contains proliferative fibroblasts and delicate capillaries, indicates that inflammatory responses (both acute and chronic) have subsided, and new, healthy tissue is beginning to form. Neovascularization is present in all the implant types tested, with GelMA controls, 1:1, and 1:2 scaffolds exhibiting the highest number of capillaries present. These results were later validated with quantitative measurements taken from the H&E images, by which the number of blood vessels within each image was recorded (Figure 6C).

Adipocyte proliferation is often stimulated as granulation tissue forms^[34], which aligns with the variable amount of fatty infiltrate that was recorded among the hybrid implant types (Figure 7A). Large adipose vacuoles were visualized along the subcutaneous interface in hybrid patterns 1:1, 1:2, and 1:3. This fatty infiltrate may have stemmed from the migration and proliferation of adipocytes derived from the hypodermis of the rat skin, and its presence could ultimately be beneficial in the implant’s ability to match the natural tactile properties of nipple tissue.

The last and most promising recorded tissue characterization was the minimal capsule thickness surrounding each implant. Capsular contracture is a very common, painful, and detrimental complication that accompanies many soft tissue implants^[35], especially those used in breast reconstruction. In fact, the overall incidence of capsular contracture of silicone breast implants is reported as high as 37% and increases up to 100% in studies with patients receiving post-mastectomy radiation therapy (PMRT)^[36], requiring a further surgical procedure (capsulotomy or capsulectomy). Our hybrid scaffolds exhibited an average fibrotic capsule score in a mild – moderate range in severity. As before, the implant-subcutaneous interface experienced a higher inflammation and thus foreign capsule response around the implant than the fascia-implant interface due to the migration of epidermal immune cells. Scaffolds with the highest GelMA component (1:3 hybrid, GelMA control) and PEG controls displayed trends in a mild fibrotic capsule formation (Figure 7B). Hybrid scaffolds 1:1 and 1:2 appeared to have a more moderate capsule, where several layers of fibrous tissue were present. To validate the semiquantitative scoring results, sample sections were stained with Masson’s Trichrome (MT) and subsequent quantification of the fibrotic capsule was measured. The mean thickness of the capsules was larger at the subcutaneous interface than the fascial interface, mirroring the score results. Both blind score and quantitative studies confirms the effectiveness of these hybrid materials in subsiding significant acute and chronic inflammations and displays characteristics of a mild foreign body reaction.

Discussion

In this study, we combined tissue engineering with a versatile 3D printing platform to reimagine effective strategies for nipple reconstruction. The technical objective of nipple reconstruction is to create the appearance of a nipple-areola complex that maintains projection and is symmetrical with the contralateral breast in terms of pigmentation and size. Symbolically, it represents the final chapter of the breast reconstruction process and lends a meaningful sense of closure for the patient. Previous reports have documented that nipple reconstruction enhances the self-esteem in patients and decreases the feeling of distress due to the mastectomy procedure itself^[37–39]. Some have even reported that women are more likely to undergo a mastectomy if the nipple can be reconstructed^[40]. Overall, this process undeniably provides psychological benefit to the patient, and therefore is an essential component in the breast reconstruction process.

An ideal reconstruction of the nipple-areola complex requires symmetry in position, size, shape, texture, pigmentation, and permanent projection. The nipple itself may project as much as 1cm, with a diameter of approximately 4–10 mm, and the surrounding areola averages 3.0–4.5 cm in diameter^[41,42]. To ease the physical, psychological, and social challenges breast cancer survivors face, nipple reconstruction using SFS has become a regular post-operative procedure following patient healing of the initial silicone implantation^[42]. The mammary glands and the lactiferous ducts, responsible for the production and delivery of milk to the surface of the skin, are removed during the MRM procedure. Thus, nipple reconstruction does not require the constructed nipple to act as a vessel to transport milk and properly function. SFS techniques to restore the nipple are inadequate as these procedures are associated with high rates of infection, multidirectional scarring, and severe nipple flattening^[14,43]. By replacing standard SFS techniques with a tissue engineered nipple-areola graft, the resulting nipple is designed to exhibit less flattening, provide an opportunity to regenerate part of the breast, and result in an appearance that more closely mimics the previous natural tissue.

The first objective of this work was to develop a hybrid printing platform that incorporated the strengths of natural and synthetic materials for the fabrication of dermal regenerative nipple-areola scaffolds. We developed a synthetic ink comprising of methylcellulose (MC) to act as a second self-interacting network with poly(ethylene) glycol. MC, a water-soluble derivative of the polysaccharide cellulose, has been FDA-approved for use in medicinal capsules and tablet coatings as clinical studies have proven the material to be non-allergenic, non-toxic, and biocompatible^[44]. MC has previously been shown to exhibit stable hydrophobic chain interactions throughout a large range of pH (3–11)^[45] and resist enzyme degradation^[46], two critical properties that are necessary for our synthetic material to maintain the scaffold’s projection as the biodegradable components of the graft are reconstituted by infiltrating host cells. Since the material and biological properties of MC greatly align with those envisioned for our scaffold, we then optimized MC concentrations with PEG to produce a bioink with reproducible swelling and extrusion characteristics.

By coupling the synthetic MC-PEG ink with a biodegradable GelMA bioink in various print patterns, we were able to produce hybrid constructs that mirrored the mechanical properties of natural nipple tissue. Once the design was established, we wanted to determine the effect of the hybrid print pattern on internal cell viability and projection shape maintenance when human dermal fibroblasts were encapsulated in the GelMA bioink. These aspects are crucial in dictating the successful integration of a scaffold with a host, as commercial products have been shown to fail in vivo because of inadequate diffusion between the wound bed and autograft^[9,23,47]. Additionally, dermal fibroblasts have been recorded to displace significant contractile forces on their substrate, and this behavior has resulted in the warping of many documented skin grafts^[48]. We found that the majority of the tested print patterns produced a composite matrix that provided adequate diffusion to support cellular proliferation and exhibited minimal projection deformation when subjected to high levels of cellular contractile forces. Notably, print patterns suggest that the internal MC-PEG lattice structure aids in the shape retention of the scaffold’s nipple projection by preventing lateral and vertical contraction of the nipple, thereby retaining the curvature of its intended geometry. Without this synthetic component, the projection decreases in radius, height, and curvature, as seen in the 0:1 GelMA control. Overall, the hybrid scaffolds demonstrate promising results in biocompatibility, mechanical properties, and shape retention and has strong potential to serve as an effective strategy for the reconstruction of nipple-areola constructs.

While in vitro development of tissues is the starting point of a tissue-engineered product, it is critical to analyze the outcomes of animal in vivo studies to properly predict its success upon human implantation. The inevitable host immune response is a factor that greatly influences an implant’s successful integration^[49]. If the reaction is significant, it can lead to total rejection of the implant. Biologically active acellular scaffolds remain an attractive alternative to current strategies as it involves in situ tissue regeneration; by providing an instructive biomaterial matrix, the scaffold helps guide cell migration and informs the regeneration process. To enhance nipple reconstruction, an acellular 3D printed scaffold whose matrix recruits host dermal cells and relies on the body’s innate regenerative ability could be used to better reproduce a well-integrated and stable nipple projection.

Our investigation focused on the integration of acellular nipple hybrid prints with host tissue in a rat subcutaneous model. The hybrid scaffolds were fabricated with alternating patterns of MC-PEG and GelMA bioinks investigated in vitro, implanted into the dorsal subcutaneous regions in Lewis rats, and further retrieved after 4 weeks for histological analysis. A mild local foreign body reaction was observed, with minimal encapsulation and a moderate infiltration of inflammatory cells. The apparent formation of granulation tissue signified that both inflammatory responses (acute and chronic) had subsided, and a new connective matrix with proliferative fibroblasts and delicate capillaries was beginning to form. Given that the foreign body response is a multifactorial event cascade, and that we have only examined the process after 4 weeks, further studies are needed to evaluate the long-term success of the implant.

The success of our in vitro and in vivo trials suggest that these hybrid nipple-areola complex scaffolds can provide a promising solution to nipple reconstruction. The acellular 3D printed scaffold contains a complementary matrix that can both support the infiltration of host dermal cells and withstand contractile forces, which can be used to better reproduce a well-integrated and stable nipple projection. Overall, the customization and the regenerative potential of this strategy prioritizes the emotional health of the mastectomy patient and may finally offer a meaningful sense of closure for these women.

Methods

MC-PEG Synthetic Ink

Our previous double network ink was modified with the substitution of alginate with methylcellulose. It was synthesized by first dissolving 15% (w v⁻¹) methylcellulose dehydrated powder with 10% (w v⁻¹) four-arm PEG thiol and 10% (w v⁻¹) four-arm PEG norbornene in deionized water. Lithium phenyl-2,4,6- trimethylbenzoylphophinate (LAP, Sigma-Aldrich) with a 0.2% (w v⁻¹) ratio was added and thoroughly mixed under 80°C.

GelMA Bioink

For in vitro experiments, lyophilized GelMA was dissolved at 7% (w v⁻¹) in fibroblast media at 50°C for 10 mins. LAP initiator was added to the GelMA solution at a concentration of 0.1% (w v⁻¹) at 50°C for 15 min, and later sterifiltered with a 0.4μm filter. Primary adult, normal human dermal fibroblasts (Passage 3, Lonza) were then added at a concentration of 2×10⁶ cells mL⁻¹ (for cell viability assay) or 10×10⁶ cells mL⁻¹ (for the shape retention assay) and were homogenously mixed throughout the solution. The prepolymer solution was then loaded into a sterile syringe barrel and allowed to equilibrate for 30 min at 23°C.

Hybrid Scaffold Fabrication

Physiologically relevant human-shaped nipple areola constructs were printed with various print patterns (MC-PEG:GelMA- 1:0, 1:1, 1:2, 1:3, 0:1). All model scaffolds were sliced into layers with a slicing thickness equal to 80% (0.32 mm) of the needle size (0.4 mm) before printing). Each printed layer was exposed to UV light (5 mW cm⁻²) for 5 s, and the final print was placed in a UV box (5 mW cm⁻²) and exposed for 3 more min. For the both the cell viability and shape retention assays, the hybrid nipple-areola prints were immediately transferred into 6-well tissue culture plates and were fully submerged in media, with media changes every 2 d for 2/3 weeks. For mechanical testing, half of the produced scaffolds were exposed to collagenase IV (500 Units mL⁻¹, Worthington Biochemical Corporation) to evaluate the synthetic material impact on the composite mechanical properties.

Mechanical Testing

Compression testing was performed on a Dynamic Mechanical Analyzer (DMA Q800, TA Instruments) with a strain sweep (0–15%) and a load of 0.01N at 1Hz frequency. Elastic modulus for each sample was calculated by determining the slope of the linear region of the stress-strain curve (strain region 3–5%). These compressive moduli were compared against fresh pig nipple teat (Tissue Source, Lafayette) for a close in vivo representation of nipple mechanical properties.

Cell Viability Assay

Cell viability was assessed using a Live/Dead assay (Invitrogen) following the manufacturer’s protocol. Briefly, 3D printed hybrid nipple-areola constructs (35 mm diameter, 8 mm projection height) were incubated with 4μM calcein AM and 2μM ethidium homodimer for 1 h at multiple timepoints (Day 0, Day 14). Fluorescent images (n=8 per sample) were taken using a fluorescence microscope (Nikon) and processed using Nikon’s NIS Elements AR Software. Cell viability was determined from the ratio of the number of live cells to the total number of cells.

Point Cloud Generation

The ROMER Absolute Arm (Hexagon) was used to generate point clouds containing high geometric detail of the nipple-areola surfaces. Scaffolds were sprayed with water soluble paint prior to scanning for greater ease in detecting surfaces. The nipple-areola scaffolds were sequentially scanned at multiple angles and the surfaces were selected using the software’s editing brush tool. After the total region of interest was created, a volume was generated and exported as a stereolithography (STL) file.

Surface Alignment

To maintain sterile in vitro culture conditions for the contracted nipple scaffolds, we could not spray/scan the same scaffold twice. Therefore, we could not compare the same scaffold from before/after contraction to visualize the exact scaffold deformation. STL files of all unique samples were imported into the open-source software MeshLab (Visual Computing Lab- ISTI CNR) for an averaged comparison among projection contraction amongst hybrid scaffold type. Propagation of error was necessary in the quantitative cloud comparisons, which is further discussed in the supplemental methods section. With MeshLab software, the alignment tools (‘point-based alignment’ and ‘process tool alignment’) were used by selecting four possible points along the base areola for each hybrid model. This provides fine superimposition of both point clouds, as the ‘process’ tool feature is an ICP (iterated closed point) algorithm capable of enhancing alignment by finding corresponding points on both models. Once both models are superimposed, visual comparison of the coincidence in anatomical structures is possible. This final alignment is frozen and exported as a new file, which will keep the achieved alignment intact once opened in the second open-source software CloudCompare.

Cloud Comparison

The two repositioned STL meshes for each pattern were uploaded into the CloudCompare software and point cloud comparison was performed by selecting the precontracted nipple projection as the reference and the contracted nipple projection as the compared object. A color map was generated, which calculated the distance of points in the compared cloud to the nearest point in the reference precontracted cloud. The color map colors were further modified in this chapter to better visualize the contraction of the nipple scaffold. White color signified coincident points, while a red spectrum (positive deviation) or blue spectrum (negative deviation) indicated a lack of coincidence. The accompanying histogram to the right of the colormaps provide the concentration of points analyzed by its distance. CloudCompare’s Point Picking function allowed distances to be obtained along the diameter of the nipple projection, projection height, and the projection’s radius of curvature. The Fit Sphere function rendered dependable approximations in curvature of the nipple projection.

In Vivo Experimental Design

This animal study was approved by the Institutional Animal Care and Use Committee (IACUC) at the University of Maryland, College Park. The main goal of our study was to evaluate the cellular and tissue responses to novel hybrid biomaterial implants. All of the recorded metrics could indicate the functionality of the material for a nipple-areola specific tissue engineered scaffold. Hybrid nipple prints with a projection diameter of 9.0mm, projection height 8.0mm, and areola diameter 20.0mm were implanted under the superficial cervical fascia in 10 Lewis strain female rats. The in vivo analysis is critical due to the situational degradation expected from these scaffolds. As the biodegradable materials of the scaffolds degrade, we are interested in investigating the cell and surrounding matrix and the remaining synthetic material presence.

Scaffold Fabrication for Animal Implantation

3D models of nipple-areola constructs were first designed in SolidWorks (Waltham, MA). The two bioinks used during fabrication are 1) 7% (w v⁻¹) gelatin methacrylate solution dissolved in phosphate buffered solution and 2) an aqueous solution containing 15% (w v⁻¹) autoclaved methylcellulose, 0.2% (w v⁻¹) LAP photoinitiator, 10% (w v⁻¹) four-arm PEG norbornene, and 10% (w v⁻¹) four-arm PEG thiol. All bioinks are sterilized prior to printing via 0.4 μm sterifilter and were fabricated in a sterile tissue culture hood and treated with ethylene oxide for surface sterilization prior to implantation. Scaffolds are programmed to be printed with inner patterns (similar to those described in Chapter 3 previously: PEG Control 1:0, hybrid 1:1, hybrid 1:2, hybrid 1:3, GelMA control 0:1) under controlled conditions, including temperature, pressure, print speed, fiber spacing, and needle size.

Animal Surgery

10 Adult female Lewis rats (200g, Covance Inc, NJ) were sedated for surgery in an induction chamber under 5% isoflurane and 2 L min⁻¹ oxygen. Anesthetic was maintained with 2% isoflurane and 2 L min⁻¹ oxygen through a nose cone for the duration of the procedure. Breathing rate was monitored, and the rats were placed on a water recirculating heating pad and distant heat lamp. For the dorsal implantation, the fur was clipped with electric shears. The exposed skin and surgical site were prepared with alternating iodophor antiseptic solution and 70% alcohol washes for 3 cycles, and further covered with sterile drapes. Using strict aseptic techniques, a 20 mm midline incision was created over the spine at the scapular level. The incision was deepened with blunt dissection, which allowed for the creation of a subcutaneous pocket. A second, shallow incision was made on the superficial cervical fascia- the primary subcutaneous connective tissue that lies between the dermis of the rat’s skin and the deep dorsal fascia. The scaffolds were then placed within this pocket, and tacking sutures using 4–0 C-14 reverse cutting resorbable sutures closed the fascial pocket. This helped stabilize the nipple implant in position to ultimately enhance host cell migration into the scaffold. The first incision was further closed by subcuticular suturing along the midline incision lines, using the same resorbable suture mentioned above. Wound sites were further cleaned with saline, and the animals were transferred to a recover cage (2 animals per cage) warmed with a heating lamp. General health and wound checkup were performed daily for 2 weeks, followed by weekly checkups until the end of the study (4 weeks). All animals exhibited exceptional pain recovery, active behavior, and healthy weight, and no implant exposure occurred.

Histological Analysis

The rats were euthanized 30 days after the implantation with 10% CO₂ chamber for 10 mins. The animals were shaved and a dorsal midline incision was made. Both the scaffold implant and the surrounding subdermal pocket tissue were explanted and further fixed with a 4% paraformaldehyde solution for 12 hours at room temperature. After this fixation, the tissues were soaked in solutions containing a gradual increased concentration of sucrose (10% - 30% w v⁻¹) and embedded in OCT gel for cryosectioning.

Sample sections underwent Hematoxylin-Eosin (H&E) and Masson’s trichrome staining per manufacturer’s protocol. Histopathological features were scored in a blind fashion according to a semi-quantitative scoring system suggested in Annex E of International Standards of Organization (ISO 10993–6-2016, which takes into consideration the extent of fibrosis and capsule formation, the extent and nature of the inflammatory reaction, and the presence of vascularization. Five blind scorers evaluated a total of 150 images with a modified ISO10994–6 protocol to tailor to the available microscope 20X capability.

The extent of fibrosis and neovascularization was further analyzed using a histomorphometric approach. The fibrotic capsule thickness was determined using the ‘Freehand’ tool in Image J to manually select the fibrotic tissue layer. The Measurements tool allowed 2 measurements to be obtained across the thickness of the fibrotic capsule. The measurement protocol foresaw taking 12 individual thickness measurements per implant analyzed. The independent measurements were then averaged to obtain the mean fibrotic capsule thickness. Neovascularization was also quantified per 20X images by counting the number of blood vessels present in each image. All the images were captured using 10X and 20X magnification by light microscopy (Nikon).

Statistical Analysis

Data were analyzed using single factor analysis of variance (ANOVA) followed by Tukey’s Multiple Comparison Test assuming normal data distribution with a confidence of 95% (p< 0.05).