Challenges and opportunities in generating microvasculature using bioprinting techniques Authors: Junjie Yu, Daiki Murata, Manabu Itoh, Koichi Nakayama Published: 2026 ABSTRACT AbstractSuccessful tissue transplantation requires the formation of microvasculature within engineered tissues. However, the formation of perfusable microvascular networks and their integration with host tissues after transplantation remain major challenges. We overview current bioprinting approaches for engineering microvascular structures and discuss their advantages and limitations. We further highlight applications of the Kenzan method developed in our laboratory for scaffold-free microvascular tissue fabrication and discuss future perspectives for vascularized tissue engineering.Subject terms: Biological techniques, Biotechnology, Engineering Introduction Tissue engineering involves the creation of artificial tissues by seeding patient-derived cells onto biocompatible and biodegradable scaffolds, followed by transplantation into the patient. The clinical success of these constructs, particularly for highly metabolic organs such as the liver, heart, and kidneys, depends on the establishment of hierarchical vascular networks. Therefore, achieving rapid and functional vascularization has emerged as a critical challenge in tissue regeneration. Early approaches to microvascular induction primarily involved seeding endothelial cells (ECs) onto growth factor-loaded scaffolds or coculturing ECs with other cell types, such as fibroblasts and myoblasts, within scaffold structures1,2. Despite these efforts, several challenges remain. For example, establishing stable perfusion culture systems, which are essential for long-term endothelial viability, remains difficult. In addition, reliance on sustained growth factor release may delay vascular formation and result in disorganized vascular structures. To overcome these challenges, researchers have developed various biofabrication strategies that enable precise control over microvascular architectures with diameters below 100 μm. Among these, microfabrication-based techniques for constructing microvascular channels can be categorized according to the method used for channel formation. These include template-based methods, in which glass capillaries are embedded within hydrogels and mechanically extracted after gelation, and sacrificial 3D printing strategies, in which complex tubular architectures are fabricated using fugitive inks that are subsequently removed to generate internal perfusable channels3,4. Additionally, laser ablation enables direct channel formation through localized laser irradiation, whereas soft lithography offers advantages in mass production and reproducible patterning of polydimethylsiloxane (PDMS) microfluidic chips5,6. These techniques can generate microchannels with dimensions of only a few micrometers ( ~ 10 μm), thereby enabling more physiologically relevant models of microvascular networks3,5. In this Perspective, we focus on bioprinting approaches for the fabrication of microvascularized tissues. 3D bioprinting enables the fabrication of biological constructs through the precise layer-by-layer deposition of biomaterials and/or cells. Over the past decade, this technology has developed into diverse bioprinting platforms. Extrusion-based bioprinting is highly versatile, enabling the deposition of highly viscous biomaterials and high-cell-density bioinks for large-scale fabrication, whereas light-based bioprinting techniques generally provide higher-resolution microvascular patterning7. Here, we first introduce the structural organization and key components of native vasculature to provide a biological basis for multi-scale vascular design. We then examine various scaffold-based approaches for fabricating microvascular networks that support the viability of artificial tissues. Next, we describe the mechanisms of the Kenzan method and its applications in fabricating vascularized tissues. Recognizing the current limitations of microvascular construction, we present our perspective on addressing these challenges through the design of spheroids and structural architectures. Finally, we discuss future directions and challenges, including tuning the mechanical strength of bioinks and spheroids, achieving clinically relevant cell densities, and integrating in situ bioprinting and artificial intelligence (AI) for advanced tissue engineering. Structure features of the vasculature Blood vessels are tubular structures that form intricate networks throughout the human body and play fundamental roles in maintaining homeostasis by delivering blood to various tissues (Fig. 1). Major arteries can reach diameters of up to 30 mm, whereas the largest veins extend to 25 mm8. Both vessel types possess a distinct trilayered architecture: the tunica intima, composed of a monolayer of ECs; the tunica media, consisting of densely packed smooth muscle cells (SMCs) and elastic fibers; and the tunica adventitia, the outermost layer comprising a loose connective tissue matrix in which collagen and fibronectin are deposited by vascular resident cells8,9. Structurally, arteries are characterized by thicker walls and circular lumina to withstand high hemodynamic pressure, whereas veins typically exhibit thinner walls and elliptical luminal cross-sections. Classification and size ranges of blood vessels. Created in BioRender. Yu, J. (2026) https://BioRender.com/32t44q0. Arterioles, the smallest arteries, measure 10–300 μm in diameter and are composed of three distinct layers10,11. Similar to arteries, the tunica intima is composed of a single layer of ECs supported by a basement membrane, providing a thromboresistant surface. The tunica media, composed of a dense layer of SMCs arranged concentrically with elastic fibers, regulates blood flow and blood pressure. Finally, the tunica externa, a perivascular connective tissue layer, contributes to mechanical stabilization. Capillaries, measuring 6–14 μm in diameter, consist of a single layer of ECs and a basement membrane10,12. Pericytes surrounding this layer regulate vascular stability, barrier function, and maturation through interactions with ECs13. This structure, characterized by low blood flow and a large surface area, is optimized for material exchange. Venules are divided into postcapillary and muscular venules and range in size from 10 to 200 μm10,12. While muscular venules possess a very thin tunica media composed of one or two layers of SMCs and are surrounded by an adventitia consisting of connective tissue with few elastic fibers, postcapillary venules lack both tunica media and adventitia12. Similar to veins, some larger venules also possess valves to prevent backflow14. From an engineering perspective, reproducing vascular networks involves more than simply creating tubular structures. It is essential to ensure robust barrier function through junctions between ECs and maintain biological function under shear stress generated by blood flow. These complex intercellular interactions and the precise control of the microenvironment remain key challenges in building microvascular networks within artificial tissues. However, current bioprinting approaches primarily focus on the geometric formation of microchannels and therefore still face limitations in fully recapitulating these biological functions. Nevertheless, precise structural control is considered a key requirement for the formation of functional vasculature. Accordingly, the following section examines various bioprinting approaches, including extrusion-based and light-based techniques, for fabricating microscale vascular structures. Scaffold-based techniques for engineering microvascularized tissues Extrusion-based bioprinting Extrusion-based bioprinting, which mechanically deposits bioinks onto a substrate, is an accessible technology that offers considerable material flexibility. To fabricate perfusable and endothelialized vascularized constructs, extrusion-based bioprinting utilizes multi-nozzle or coaxial printing heads in combination with sacrificial inks such as Pluronic F-127 or polyethylene glycol to create tubular structures15,16. Gu et al. used extrusion-based bioprinting equipped with a coaxial nozzle to fabricate a perfusable vascular construct with an approximate diameter of 900 µm17. ECs were encapsulated within methacrylate gelatin (GelMA) bioinks, whereas gelatin served as a sacrificial component that was subsequently removed to generate a hollow channel. To enhance the overall mechanical stability of the construct, polycaprolactone was used as a supporting structure (Fig. 2A). In general, conventional extrusion-based bioprinting often struggles to maintain structural stability because low-viscosity hydrogels tend to collapse under their own weight, thereby requiring additional support structures to fabricate complex 3D constructs with high structural fidelity. Extrusion-based bioprinting. A Fabrication of an EC-laden tubular construct using a coaxial nozzle and insertion of a polycaprolactone frame to improve the mechanical properties of the structure. Adapted under CC BY-NC 4.0 from17. B Fabrication of biomimetic blood vessels using a combination of SWIFT and a coaxial printhead. Scale bars: 10 mm (top) and 50 µm (bottom). Reproduced with permission from20, Copyright 2024, John Wiley and Sons. C Induction of angiogenic sprouting through the fabrication of endothelialized and fibroblast-laden channels. Scale bars: 200 µm (left) and 500 µm (right). Reproduced with permission from22, Copyright 2021, John Wiley and Sons. Embedded bioprinting based on conventional extrusion printing has been developed to address these challenges. This approach utilizes a support bath that rapidly stabilizes printed filaments, providing the buoyancy and mechanical support necessary to prevent the sagging or collapse of thin hydrogels owing to gravity. This environment effectively provides the “time and space” required for stable crosslinking. For instance, the freeform reversible embedding of suspended hydrogels (FRESH) technique, which utilizes an optimized slurry of gelatin microparticles, enables the extrusion of filaments with diameters as small as 20 μm18. Using the FRESH method, vascular structures measuring approximately 100 μm could be fabricated. Furthermore, micropores ( ~ 25 μm) created within the collagen scaffold following the removal of gelatin microparticles contributed to endothelial migration and angiogenesis. Similarly, the sacrificial writing into functional tissue (SWIFT) method developed by Skylar-Scott et al. utilizes sacrificial printing within high-density cell spheroid baths19. Upon removal of the sacrificial filaments, perfusable microchannels are created within the functional tissue matrix. In another study, the feasibility of fabricating hierarchical vascular networks using a coaxial sacrificial ink printing strategy combined with SWIFT was demonstrated (Fig. 2B)20. However, the size of the printed vessels remains limited to a few hundred micrometers. Fang et al. developed the sequential printing in a reversible ink template technique, which employs a microgel-based biphasic system that serves as both a support bath and a cell-laden bioink, thus enabling the fabrication of complex 3D architectures and engineered microvascular networks21. Notably, this technique allows the integration of bulk tissue constructs with hierarchical vascular networks within a single printing process. Although embedded bioprinting technologies can stably fabricate complex 3D geometries without structural collapse by utilizing support media, they still share the fundamental limitation that resolution is constrained by the internal diameter of the extrusion nozzle, similar to conventional extrusion-based approaches. Therefore, regardless of the presence or absence of support media, realizing complex microscale structures remains a significant technical challenge for all nozzle-based bioprinting methods. In particular, although extrusion-based bioprinting can rapidly generate structures on the scale of hundreds of micrometers, comparable to arterioles, it still faces substantial challenges in replicating capillary-scale architectures ranging from 6 to 14 μm. An alternative strategy involves leveraging the intrinsic ability of ECs to undergo sprouting morphogenesis to form microvascular networks. For example, Son et al. induced angiogenesis by positioning fibroblast-lined channels adjacent to endothelialized channels, thereby leveraging angiogenic gradients to promote endothelial sprouting (Fig. 2C)22. In a follow-up study, they transplanted microvascular patterns induced by adipose-derived stem cells (ADSCs) and ECs into an arteriovenous bundle model23. This resulted in a two-fold increase in angiogenesis, even in regions distant from host vessels, highlighting the important role of angiogenic gradients in promoting tissue vascularization. Light-based bioprinting Light-based bioprinting techniques, such as stereolithography (SLA) and two-photon polymerization (2PP), offer high-resolution, non-contact, and nozzle-free approaches that enable the creation of intricate channel networks with diverse diameters. SLA precisely crosslinks liquid hydrogels using a laser, solidifying the material point-by-point and line-by-line to form horizontal cross-sections. In a study by Melhem et al., cell-laden patch-type structures featuring 500-μm channels were bioprinted using a UV laser (Fig. 3A)24. As a projection-type SLA technique, digital light processing (DLP) bioprinting utilizes a digital micromirror device to project entire 2D patterns onto photosensitive materials simultaneously, rather than scanning line-by-line, thereby enabling faster printing than conventional SLA. This high-precision approach also allows the fabrication of highly complex structures. For instance, Grigoryan et al. printed entangled, non-intervening vascular networks within hydrogels, along with functional bicuspid-like valves that dynamically respond to fluid flow25. Moreover, a hydrogel carrier loaded with hepatocyte aggregates was bioprinted in combination with a vascular network, which enhanced hepatic function and was associated with improved integration and perfusion within host tissue after transplantation. Additionally, DLP enabled the fabrication of 200 μm-wide microchannels and the creation of intricate branched vascular networks26,27. Mazari-Arrighi et al. reported that encapsulated cells within printed photopolymerizable hydrogels spontaneously self-assembled into vascularized branched tubular structures with diameters of 10–100 μm (Fig. 3B)27. In contrast with the microscale precision of SLA and DLP, 2PP achieves nanoscale accuracy using a femtosecond pulsed laser. This technique focuses energy strictly on a single focal point, curing only the targeted 3D pixels to construct intricate, high-resolution architectures. Dobos et al. fabricated microchannels with diameters of 10–30 µm within hydrogels on a microfluidic chip, demonstrating the high-resolution capability of 2PP (Fig. 3C)28. Specifically, vascular-like network formation was achieved using EC spheroids as a source of endothelial sprouting within hydrogels. Additionally, predefined microchannel structures were endothelialized by seeding endothelial spheroids to form vessel-like linings, resulting in spatially separated endothelial and supporting-cell compartments (GFP-ADSCs), thereby establishing a microvascular barrier structure. These results demonstrate that 2PP enables the precise fabrication of complex microvascular architectures on microfluidic chips and highlight its potential for organ-on-a-chip applications. Light-based bioprinting. A Schematic of the SLA process and printing of cell-laden gel patches. Reproduced with permission from24. Copyright 2017, American Chemical Society. B Schematic of the DLP process and fabrication of vascular structures. Adapted under the Creative Commons CC BY license from27. C a Microchannels filled with FITC-dextran (top) and fabrication of vascular units incorporating RFP-HUVECs and supportive cells (bottom). b Endothelial sprouting after 5 days (left) and intertwined spiral channels composed of GFP-ADSCs and RFP-HUVECs (right). Adapted under the Creative Commons CC BY license from28. Challenges Both bioprinting methods have clear limitations, particularly in reliably achieving and maintaining physiologically relevant cell densities (108–5 × 108 cells/mL)29. Extrusion-based bioprinting allows for high cell densities and a wide range of material applications because of its physical dispensing mechanism. However, maintaining such cell concentrations may often lead to nozzle clogging, significantly reducing printability. Moreover, high extrusion-induced shear stress and chemical crosslinking agents can impose substantial stress on cells, resulting in impaired cell viability and functionality. Conversely, light-based bioprinting offers relatively higher resolution and is nozzle-free; however, it faces challenges related to potential DNA damage, reduced cell viability caused by UV exposure, and a limited selection of biomaterials. Furthermore, mismatches in refractive indices between cells and surrounding biomaterials can lead to light scattering, thereby compromising resolution. To address this issue, He et al. utilized a curcumin-Na photoinhibiting additive to suppress free radicals generated by light scattering, thereby preventing unintended curing and significantly improving printing resolution30. A further complication is the endothelialization of bioprinted constructs. Although in situ fabrication improves vascular lumen integrity compared with post-seeding approaches31, promoting the rapid establishment of functional, leakage-free microvascular structures remains challenging. Alternatively, technologies capable of achieving precise spatial arrangement and high cell density are required. Scaffold-free methods, such as magnetic levitation and biopaper, have been proposed as alternative approaches32,33. However, the difficulty of designing complex structures, the potential effects of magnetic fields on cells, and the incomplete removal of hydrogels remain important considerations. Here, we present the Kenzan method, a scaffold-free biofabrication strategy developed in our laboratory. This approach enables tissue reconstruction without the use of scaffolds. Scaffold-free method for engineering microvascularized tissues Kenzan method and bio-3D printer Kenzan method is a scaffold-free technology in which spheroids are sequentially placed onto a fine-needle array (Kenzan). The fine-needle array serves only as a temporary support, while spheroids, which act as tissue building blocks, secrete extracellular matrix (ECM) and undergo fusion and self-organization to form integrated cellular constructs that are structurally and functionally similar to native tissues, including cartilage, blood vessels, nerves, trachea, heart, and liver34–39. Because this technique does not rely on biomaterials, it may reduce the risk of material-related complications, such as biofilm formation on implanted materials and inflammatory responses induced by biomaterial degradation byproducts. Moreover, this method may reduce concerns regarding biomaterial-related immunogenicity while enabling direct cell-to-cell interactions without interference from exogenous materials. Furthermore, it may minimize problems associated with the premature or excessively slow degradation of scaffolds after implantation40,41. This method was invented by Koichi Nakayama, and the resulting automated device, “Regenova,” was commercialized by Cyfuse Biomedical. Unlike existing printers that extend 2D scaffold-based approaches into 3D structures, this system was designed from the outset to directly place cell aggregates for the construction of 3D tissues; hence, the term “bio-3D printer.” As shown in Fig. 4, the bio-3D printer mainly consists of units such as magazines, a well plate carrier, upper and under cameras, a nozzle, the Kenzan, and a control unit42. Notably, the path-generation software of the bio-3D printer supports a wide range of geometric designs, from simple tubes and patches43 to multi-layered trachea-like constructs44, articular surface constructs45, and loop-type constructs, thus providing a versatile platform for regenerative medicine (Fig. 5). Recently, clinical trials using the Kenzan method for nerve regeneration have been reported, demonstrating the feasibility and efficacy of the fabricated tissues36. In particular, this approach may avoid donor-site sensory loss and pain associated with conventional autologous nerve grafts and demonstrated safety and nerve recovery in treated patients. Moreover, preclinical studies showed that the bio-3D conduit promoted superior nerve regeneration, as evidenced by increased numbers of myelinated axons, enhanced Schwann cell expression, and consistent nerve bridging compared with conventional silicone conduits46. These findings suggest a promising therapeutic strategy for neurological diseases. Components of a bio-3D printer. a Bio-3D printer hardware. Adapted under the Creative Commons CC BY license from42. b Control unit. Fabrication of cellular constructs using a bio-3D printer. A a Schematic diagram of tubular tissue conversion into a patch. b Cellular patch. Reproduced with permission from43. Copyright 2018, Elsevier. B a Design data of a trachea-like tube model. b Multi-layered trachea-like construct. Reproduced with permission from44. Copyright 2019, John Wiley and Sons. C a Design data of an articular surface. b Articular surface construct. Adapted under the Creative Commons CC BY license from45. D a Design data of a loop-type structure. b Loop-type construct. Despite this potential, the technical complexity and high operational cost of bio-3D printers may limit their widespread adoption compared with other bioprinting technologies. In particular, research on the construction of functional microvascular structures within artificial tissues remains limited. Nevertheless, this Perspective aims to outline potential strategies and future directions for scaffold-free microvascular construction by optimizing key parameters, such as spheroid composition and structural design, based on existing studies of vascularized tissues. Design of spheroids and structures for construction of microvascularized tissues One strategy involves fabricating multicellular spheroids by combining ECs with organ-specific cells to engineer target constructs35,38,39. Itoh et al. utilized spheroids composed of ECs, SMCs, and fibroblasts to construct tubular structures with an inner diameter of 1.5 mm, in which self-organized vessel-like structures were observed via immunohistochemical staining (Fig. 6A, a–c)35. The cellular constructs were first cultured on microneedles to allow spheroid fusion, followed by further maturation after removal from the needle array. These constructs were then transplanted into nude rats, resulting in grafted vessels that remained patent and exhibited excellent short-term biocompatibility (Fig. 6A, d). However, the tensile strength was only approximately half that of native vessels, and elastic fibers were not observed within the reconstructed tissue. In a subsequent study, Itoh et al. transplanted human dermal fibroblast-based tubular constructs into immunodeficient minipigs, where the grafts underwent in vivo remodeling into vascular constructs through interactions with host cells47. Another study suggested that the inclusion of mesenchymal stem cells (MSCs) and human umbilical vein endothelial cells (HUVECs) with hepatocytes allowed the expansion of spheroid size to 500 μm, exceeding the typical size limitation of hepatocyte spheroids (100–150 μm) imposed by hypoxia (Fig. 6B)39. Notably, a cell-sorting phenomenon occurred, in which hepatocytes migrated to the surface, whereas MSCs and HUVECs localized to the core. The authors proposed that this core localization may mitigate hypoxic stress in hepatocytes and may support vascularization through the secretion of growth factors, such as vascular endothelial growth factor (VEGF) and hepatocyte growth factor. These findings are consistent with other studies demonstrating that organ-specific cells migrate to the periphery, whereas stromal cells and ECs form the core to support spheroid stability and maintenance of organ-specific cells (Fig. 6C)38,48. These results suggest that bioassembled microvascular fragments formed within multicellular spheroids may contribute to enhanced core viability, thereby demonstrating their potential for constructing large-scale vascularized tissues48,49. Design of spheroid composition and construct structure. A a–b Skewering multicellular spheroids onto the Kenzan and fabrication of a scaffold-free vascular graft. c Distribution of CD31-positive cells within the construct. d Anastomosis with the rat aorta and Doppler ultrasonography showing vessel patency on day 5. Adapted under the Creative Commons CC BY license from35. B Spheroids composed of three cell types, including hepatocytes (Heps), HUVECs, and MSCs. Adapted under the Creative Commons CC BY license from39. C a Spheroids composed of three cell types, including human induced pluripotent stem cell-derived cardiomyocytes (hiPSC-CMs), HUVECs, and normal human dermal fibroblasts (NHDFs). b Histological and immunohistochemical staining of cell-specific markers. Adapted under the Creative Commons CC BY license from38. D a Design of porous liver-like tissue using micro-gaps. b Changes in the construct during tissue culture. c Histological and immunostaining of the gaps between spheroids and the distribution of HUVECs, indicating the formation of endothelial networks within the cellular construct. Adapted under the Creative Commons CC BY license from39. A porous structural design may be applied to promote angiogenic induction. Yanagi et al. utilized “micro-gaps” within tubular structures to facilitate the outward migration of vascular cells, thereby forming interconnected microvascular structures within the printed constructs (Fig. 6D)39. These results are particularly interesting because, although the differential adhesion hypothesis suggests that highly adhesive cells localize to the core and less adhesive cells to the periphery, the opposite distribution was observed, with ECs positioned internally within the multicellular spheroids. However, in structures incorporating micro-gaps, culture medium flow through these gaps generates localized shear stress, which may induce cellular rearrangement and drive ECs toward the spheroid periphery. In a similar study, Oh et al. demonstrated the formation of microvascular networks using a core–shell bioprinting strategy, in which organ-specific cells and ECs were precisely compartmentalized50. Under dynamic perfusion culture conditions, the printed microtissues underwent fusion and evolved into a unified construct. While the core–shell approach achieved vascularization through artificial compartmentalization of cell types and controlled perfusion, the Kenzan method relied on the predesigned porosity of the constructs to facilitate medium flow, thus enabling spontaneous endothelial redistribution and network formation. Despite these different fabrication approaches, both strategies underscore the importance of porous structural design coupled with dynamic culture conditions in promoting functional vascularization. While optimizing the composition and structural design of spheroids is essential for promoting capillary formation, strategies to ensure the long-term stability of the resulting capillary networks remain equally important. Several studies have incorporated MSCs, which serve pericyte-like roles and promote vascularization through pro-angiogenic factors such as VEGF and ECM-remodeling enzymes including matrix metalloproteinases51,52. De Moor et al. showed that the addition of ADSCs to multicellular spheroids promoted endothelial migration, vascular formation, and ECM secretion53. Beyond their role in vascular support, co-cultured MSCs have also been reported to exert paracrine effects post-transplantation, thereby contributing to reduced scar formation and enhanced vascularization at the graft site54,55. Overall, strategies such as optimizing multicellular spheroid composition, improving structural design, and incorporating supporting cells, including stem cells, may further enhance the potential for constructing functional capillary networks using the Kenzan method. Challenges The Kenzan method faces challenges such as low throughput and the potential for physical damage caused by skewering spheroids onto microneedles. To address the inherent scalability challenges of spheroid-based bioprinting, Kim et al. recently developed a high-throughput system utilizing multiple nozzles29. In a previous study, aspiration-assisted bioprinting (AAB) was introduced, achieving a cell viability that exceeded 80% and a positioning accuracy of approximately 11% relative to spheroid size56. However, similar to the Kenzan method, AAB is limited by the inefficiency of the individual placement process. For example, the Kenzan method requires approximately 13.8 s per spheroid, whereas AAB requires approximately 30 s57,58. Recently, a parallelized approach was introduced to address these limitations, allowing the precise and simultaneous placement of multiple spheroids29. This system is approximately 8 times faster than previous techniques while maintaining a viability rate of over 90%. Specifically, this approach improves the speed, efficiency, and scalability of spheroid-based tissue engineering by assembling 600 cartilage spheroids in less than 40 min ( ~ 4 s per spheroid). Although the parallelized nozzle array demonstrated by Kim et al. could potentially be adapted to the bio-3D printer to enhance throughput, implementing such an approach in practice may involve significant risks related to pick-and-place reliability. From our perspective, spheroid miss-pickups or handling errors may lead to poor structural fidelity in the final assembled construct. To ensure practical applicability, bio-3D printing system equipped with multi-nozzle arrays must be supported by sophisticated control system upgrades and the integration of high-sensitivity sensors capable of real-time feedback. Another challenge is that relatively large spheroids, typically ranging from 500 to 700 μm in diameter, are generally employed in the Kenzan method42,43,45,47. Although multicellular spheroids containing ECs offer the potential for large-scale tissue fabrication, the outcomes of engineered organs composed solely of single-cell-type spheroids remain uncertain. This is because diffusion limitations for oxygen and nutrients occur when spheroid diameter exceeds approximately 400–500 μm, leading to central hypoxia and necrosis59,60. Given these factors, the Kenzan method may be inherently susceptible to the development of a hypoxic microenvironment due to these diffusion constraints. Although Zhang et al. reported 98% viability using mixed-cell spheroids with a diameter of approximately 660 μm, this result was observed only after a short-term culture period of 2 days43. Therefore, future research should focus on spheroid control technologies that ensure high viability while maintaining an appropriate spheroid size to support the long-term viability and engraftment efficiency of the constructs. In addition to the aforementioned issues, the Kenzan method has inherent limitations in achieving the high spatial resolution attainable with scaffold-based techniques. This limitation primarily arises from the physical pitch and diameter of the microneedle array used for assembly. Given that most previous studies utilized spheroids approximately 500 μm or larger in diameter, the geometries presented in Fig. 5 remain relatively simple compared with those produced by other high-resolution bioprinting techniques. Consequently, these limitations may challenge the translational application of this method in fields requiring intricate, high-resolution structural complexity. However, the scaffold-free nature of the Kenzan method may provide unique advantages for tissue reconstruction and maturation through self-organization and direct cell–cell interactions. In this Perspective, both scaffold-based approaches and the Kenzan method demonstrate significant potential in the field of tissue engineering. Although each approach possesses distinct advantages and limitations, neither can be considered universally superior. Specifically, scaffold-based methods can define microvascular structures through precise spatial control and hierarchical scaling, whereas the Kenzan method primarily relies on spheroid fusion and self-organization to generate microvasculature. Since both approaches can achieve vascularization within engineered constructs, the choice of method should depend on whether the intended application prioritizes predefined structural complexity or autonomous cellular behavior without physical scaffold interference. For a clearer comparison, the key characteristics of both methods are summarized in Table 118,24,26,28,37–39,42,45,61–66. Table 1Comparison of scaffold-based and Kenzan methodsScaffold useScaffold-basedScaffold-freeTechniqueExtrusionSLA2PPKenzan methodBuilding unitCells and biomaterialsSpheroidsCell density108–109 cells/mL610.4 × 106–2 × 107 cells/mL24,625 × 106–1 × 107 cells/ml63,641 × 104 cells–4 × 104 cells/spheroid45,65Geometry freedomHighModerateMicrovascular formationDirect microchannel fabrication and endothelial self-organizationSpheroid fusion and endothelial self-organizationFeature sizeFabricated microchannelCell spheroid~100 μm18~200 μm2610 μm28≥500 μm42Building 3D structureDrawingAssemblyFabrication riskShear stress-, crosslinking-, and UV-induced cell damagePotential mechanical stress from needle insertionMechanical strengthTunability via hybridization with synthetic polymers or polymerization via crosslinkingLow initial mechanical strength requiring maturationMinimum time to implantation after printingImmediate (in situ bioprinting)66≥3 days37–39Translation riskMaterial-related limitationsNo exogenous biomaterial requirementAssuming ideal spheroid geometry, the approximate cell density of spheroids is higher than 108 cells/mL. However, actual values may vary depending on cell type, culture duration, degree of compaction, and morphological sphericity. This approximation is provided solely for comparison with scaffold-based methods. Comparison of scaffold-based and Kenzan methods Assuming ideal spheroid geometry, the approximate cell density of spheroids is higher than 108 cells/mL. However, actual values may vary depending on cell type, culture duration, degree of compaction, and morphological sphericity. This approximation is provided solely for comparison with scaffold-based methods. Future perspectives and conclusion Vascularization remains one of the most formidable barriers to successful clinical applications. This Perspective explores the potential of bioprinting methods for constructing hierarchical microvascular structures. However, simple geometric fabrication alone cannot ensure long-term vascular function. Without appropriate biophysical stimulation, fabricated microvasculature may be prone to vascular regression and functional collapse67. Microfluidic platforms and perfusion systems are essential tools for modulating mechanotransduction, a key factor in vascular development and maturation. These physical stimuli promote not only the morphological maturation of vasculature but also ECM remodeling, thereby supporting a more physiological vascular phenotype. Specifically, fluid shear stress generated by luminal flow activates mechanotransduction pathways involving Piezo1 and PECAM-1/VE-cadherin/VEGFR2, contributing to endothelial alignment and vascular stabilization68,69. Shiwarski et al. combined FRESH 3D bioprinting with a microfluidic system to fabricate centimeter-scale, collagen-based scaffolds with embedded perfusable channels70. In EC-laden chips, endothelialization and the formation of capillary-like microvascular networks were observed under dynamic perfusion conditions, while functional insulin secretion was confirmed in chips containing pancreatic-like cells. In addition, Sexton et al. developed a vascular generation algorithm that achieved approximately a 230-fold improvement in computational speed compared with conventional full-tree-based vascular branching optimization methods, enabling the generation of up to one million terminal vessels. Furthermore, they fabricated perfusable structures using FRESH bioprinting, resulting in a 417-fold increase in viable cell counts compared with avascular controls71. While scaffold-based techniques can control endothelial lining and orientation through the fabrication of microchannels and computational vascular modeling, the Kenzan method primarily utilizes the self-organizing capacity of ECs to fabricate tissues containing microvasculature. However, controlling the stochastic growth of these vessels and ensuring functional perfusion remain significant challenges in vitro. Alternatively, in vivo vascularization can be achieved through the transplantation of cellular constructs, such as cell sheets or organoids, into host tissues, where anastomosis with the host vasculature enables subsequent perfusion and the formation of microvascular networks72,73. Although studies utilizing the Kenzan method have reported vascularization after transplantation, this process is often associated with structural remodeling of the implanted constructs, and host-derived cell infiltration35,47,74. However, direct evidence demonstrating functional microvascular formation through anastomosis between engineered vessels and host vasculature remains limited. Therefore, further analyses, such as evaluation of host-derived blood vessel infiltration and perfusion assays (e.g., dextran- or lectin-based assays), would be valuable for demonstrating functional vascular integration. To sustain endothelial sprouting and angiogenesis induced by perfusion culture, constructs must possess sufficient mechanical strength to withstand hemodynamic forces and maintain structural integrity until functional integration with host tissue is achieved. From this perspective, mimicking the physiological mechanical properties of the in vivo environment is essential for clinical success. Therefore, considerable attention has been directed toward optimizing the mechanical properties of bioinks in scaffold-based approaches and spheroids in the Kenzan method, both of which represent key structural components. The inherent properties of hydrogels utilized in scaffold-based methods, characterized by highly hydrated polymer networks, provide a biomimetic 3D environment that promotes cell proliferation, growth, and differentiation. Paradoxically, the same properties that ensure excellent biocompatibility, such as high water content and low solid content, are also the primary causes of the reduced mechanical strength of hydrogels. This structural vulnerability remains a significant challenge and severely limits their practical application, particularly in vascular engineering, where constructs must withstand pulsatile blood flow and clinically relevant blood pressures (60–160 mm Hg)75. To overcome these challenges, Hong et al. developed dual-crosslinked hydrogels composed of GelMA, hyaluronic acid modified with N-(2-aminoethyl)-4-(4-(hydroxymethyl)-2-methoxy-5-nitrosophenoxy) butanamide, and a lithium acylphosphinate photoinitiator, which showed burst pressure resistance up to 290 mm Hg75. Beyond this approach, the development of double-network hydrogels or the integration of nanomaterials into polymer matrices serves as an effective strategy for enhancing the structural and functional performance of bioinks76. Similarly, scaffold-free methods also face challenges related to structural integrity. One strategy involves incorporating stromal cells, such as fibroblasts, during spheroid design to promote collagen deposition and ECM formation, thereby enhancing construct integrity. Another approach focuses on maturing cellular constructs on catheters or highly oxygen-permeable materials (e.g., PDMS) following removal from the microneedles to further improve structural maturity and integrity44,77,78. As aforementioned, the microneedles serve as temporary supports, whereas the “self-healing” process and cellular rearrangement after microneedle removal have been shown to enhance construct integrity, making them suitable for transplantation36,44,47. However, conventional supports such as PDMS or catheters may impede internal perfusion of culture medium, potentially hindering the self-organization of ECs. Therefore, the development of novel approaches that can simultaneously enhance mechanical strength while maintaining functional perfusion remains a critical frontier for the successful clinical translation of scaffold-free vascular grafts. Another major challenge is the need for scalable cell production, as the number of cells required for clinical applications can vary widely, typically ranging from 5 × 10⁴ to 6 × 10⁹ cells per patient79. However, conventional 2D planar culture methods are labor-intensive and have limitations in large-scale production owing to restricted culture areas. Therefore, it is essential to establish stable processes capable of producing high cell concentrations while maintaining cell quality, including stemness and viability. Recently developed stirred-tank and vertical reciprocal motion-based bioreactor systems enable efficient cell proliferation by minimizing shear stress applied to cells and creating a uniform nutrient supply environment80,81. There have been reports of the successful harvesting of approximately 10 billion induced pluripotent stem cells using a 2-liter bioreactor, highlighting the potential of these scale-up technologies to support clinically relevant high-cell density cell production81. Although various bioprinting techniques have been explored, the ultimate goal of biofabricated constructs is to provide patients with timely and effective therapeutic solutions. Current in vitro maturation strategies often require prolonged culture periods, significantly increasing patient wait times. In contrast, in situ 3D bioprinting eliminates the need for external maturation and allows direct application to defect sites66,82. However, maintaining printing accuracy requires real-time process control, and AI technologies have recently been introduced as a potential solution. For example, Zanderigo et al. applied an AI-based Segment Anything Model to enable in situ real-time monitoring and geometric analysis during embedded bioprinting83. Accordingly, integrating advanced imaging with AI-driven closed-loop parameter tuning may further strengthen in situ 3D bioprinting for the fabrication of patient-specific constructs with enhanced reproducibility and clinical reliability. In conclusion, although vascular tissue engineering has made remarkable progress over the past decade, many challenges remain before its clinical application can be accomplished. As discussed above, various technological strategies offer potential pathways toward achieving functional and transplantable artificial tissues beyond simply mimicking anatomical morphology. Therefore, rather than relying on a single technology, an “optimized strategy” integrating multifaceted approaches, such as microfluidics, hybrid bioprinting processes, and the development of highly functional biomaterials, may represent the most reliable approach for achieving patient-tailored clinical goals.