Authors: Chi-Pin James Wang, Chun Gwon Park, Se-Na Kim
Categories: Review Article, Hydrogel, Biomaterial, Injectable, Endoscopy, Endoscopic Submucosal Dissection
Source: Biomedical Engineering Letters
Authors: Chi-Pin James Wang, Chun Gwon Park, Se-Na Kim
Therapeutic endoscopy, including endoscopic submucosal dissection (ESD), has transformed the management of early gastrointestinal (GI) cancers by enabling the resection of large and complex lesions. However, its widespread adoption remains limited by technical difficulty and the high risk of complications such as bleeding, perforation, and stricture. Submucosal injection materials (SIMs) are critical to ESD procedures as they provide mucosal elevation, but the elevation achieved by solution-based SIMs is typically short-lived due to rapid absorption or diffusion, often necessitating multiple injections during surgery. Recent advances in injectable hydrogels present a promising strategy to overcome these limitations. Owing to their tunable physicochemical and biological properties, hydrogels can prolong submucosal elevation, reduce injection force, and actively contribute to the management of complications. Preclinical investigations of two-component, thermoresponsive, and shear-thinning hydrogels have demonstrated favorable biocompatibility and compatibility with endoscopic delivery systems. Moreover, functional hydrogels engineered with hemostatic, adhesive, antifibrotic, or regenerative properties have shown potential to mitigate intraprocedural bleeding, accelerate wound healing, and prevent stricture formation. This review aims to provide a comprehensive overview of preclinical hydrogel-based SIMs, focusing on how these novel materials may reshape the tools and strategies available in advanced endoscopic practices, particularly ESD.
Endoscopy is an integral tool in modern gastroenterology that allows rapid diagnosis and management of gastrointestinal (GI) diseases [1]. The therapeutic aspects of endoscopy have evolved rapidly by incorporating technological and clinical advancements to enable complex surgical procedures. While initially developed as a diagnostic tool [2], endoscopy now allows clinicians to safely identify, monitor, and treat various conditions of the GI tract (e.g., inflammation, hemorrhage, and neoplasm) [3–5]. The role of therapeutic endoscopy is particularly prominent in the field of oncology, successfully addressing GI cancers (e.g., esophageal, gastric, and colorectal cancers) and other neoplasms without the need for open surgery [6, 7]. Despite widespread colonoscopy and endoscopic screening, GI cancers still account for approximately 5.26 million cases and 3.70 million deaths worldwide [8], further emphasizing the importance of rapid detection and removal of early GI malignancies.
Among endoscopic procedures, endoscopic submucosal dissection (ESD) stands out as a transformative technique for the management of early GI cancers. By facilitating en bloc removal of the submucosal layer, ESD allows margin-negative (R0) resection of lesions regardless of size or location [9]. As a result, ESD has become an important alternative to conventional endoscopic mucosal resection (EMR). The trade-off between the clinical outcomes and procedural complexity of ESD remains controversial among experts [10], but meta-analyses have revealed that ESD yields higher curative rates and lower recurrence rates than EMR in esophageal [11] and colonic lesions [12]. Nevertheless, ESD carries greater risks of complications (e.g., procedural bleeding and perforation), requires longer operations, and presents a steeper learning curve than EMR [12, 13], which collectively limit its widespread use, especially in Western countries [14].
Submucosal injection material (SIM) is a critical component of ESD that enhances the efficiency of endoscopic procedures. SIMs are typically injected beneath the target lesion to elevate the mucosa and to create a protective cushion. This cushion allows for improved visualization and precise dissection of the lesion while preserving the underlying muscularis propria. Various SIMs have been used in clinical practices because of their accessibility and ease of administration. However, the elevation provided by solution-based SIMs is typically short-lived due to rapid absorption or diffusion, often necessitating multiple injections during surgery [15]. Such limitations contribute to the difficulty and prolonged duration of current ESD procedures; hence, there is a need to develop SIMs with enhanced duration and functionality.
Recently, injectable hydrogels have emerged as promising materials, demonstrating favorable mechanical and biological properties compared to conventional SIMs. Their shear-thinning behavior and sol–gel transitions triggered by physiological cues make injectable hydrogels optimal candidates for endoscopic injection. In contrast to traditional solutions, injected hydrogels can form three-dimensional polymer networks that maintain their mechanical integrity for longer periods. Most importantly, hydrogels can be engineered not only to support tissue elevation but also to impart biological functions for managing conditions such as hemorrhage, perforation, or stricture. Herein, this review provides a comprehensive overview of SIMs ranging from traditional solution-based formulations to recent injectable hydrogels, categorized based on their physicochemical properties. This review also evaluates the potential of hydrogel-based SIMs as therapeutic platforms, focusing on how these novel materials may reshape the tools and strategies available in advanced endoscopic practice, particularly ESD (Scheme 1).Scheme 1Overview of formulations used as submucosal injection materials in therapeutic endoscopy
SIMs are solutions injected beneath GI lesions prior to endoscopic resections to provide local elevation and protection. Injection-assisted resection was first reported for EMR in 1955 [16], and the materials used in this process have since become essential tools in therapeutic endoscopy. Even today, SIMs are actively used in EMR to aid the resection of anatomically complex lesions [17]. However, their significance in routine polypectomy has gradually diminished over time due to greater availability of improved instruments and increased clinician dexterity. Unlike the past, the use of SIMs in EMR is now optional and depends more on the clinician’s discretion upon the lesion’s location, size, and shape [18]. Likely owing to these reasons, clinical meta-analyses have also revealed that the success rates and outcomes of cold snare polypectomy are no longer dependent on the use of submucosal injections [19].
ESD, on the other hand, still requires the injection of SIMs regardless of size or location of the lesion, due to incisions directed at the submucosal layer. Scheme 2 outlines the basic processes of ESD, highlighting the roles of SIMs during the procedure. A major advantage of ESD is its potential to achieve R0 resection even for neoplasms that have invaded deeper margins. This positions ESD as a preferred technique when invasive components are present, as it enables precise histopathological evaluation of the resected specimen [20]. Unfortunately, despite the growing number of SIMs cleared by the U.S. Food and Drug Administration (FDA) (Scheme 3), the use of these marketed SIMs does not entirely eliminate the risks associated with ESD. Post-procedural bleeding and perforation remain two of the most common complications, accounting for up to 17% and 5% of ESD cases, respectively [21, 22]. Other complications observed during colorectal ESD include electrocoagulation syndrome and stricture [23]. Given these circumstances, SIMs remain irreplaceable components of ESD, yet still present a considerable margin for improvement to both patients and clinicians.Scheme 2Overview of the endoscopic submucosal dissection procedureScheme 3Timeline of submucosal injection agents cleared by the U.S. Food & Drug Administration
In this regard, the requirements for an ideal SIM in ESD are highly multifaceted. Optimal injection materials should provide effective mucosal elevation with sufficient mechanical stability and durability to withstand the entire procedure. They should also be readily injectable through the channels of endoscopic knives, as the use of separate injection needles may prolong operative time and increase technical complexity [24]. Biocompatibility is essential; SIMs must not elicit local toxicity, immunogenicity, or other adverse in vivo responses. From a practical perspective, SIMs should support clear visualization and delineation of lesion margins to facilitate precise dissection. With advancements in modern ESD techniques, biological functionality has recently emerged as another key requirement, reflecting the need for materials that can actively contribute to both procedural safety and postoperative outcomes [25]. In preclinical research, injectable materials incorporating advanced functionalities are receiving growing interest, and this dual capacity—combining mechanical performance with biological activity—represents a defining feature of next-generation SIMs.
Considering the requirements outlined in the previous section, it is notable that most SIMs used in clinical practice today still belong to some of the earliest materials introduced for submucosal elevation [13]. Developed primarily to achieve mucosal elevation and to protect the muscularis propria, these conventional agents comprise a range of solutions that differ in viscosity and rheological behaviors. The following section discusses these established SIMs, highlighting representative examples and their roles in facilitating ESD procedures. A summary of conventional and marketed injection materials used in therapeutic endoscopy is provided in Table 1.Table 1Examples of conventional and marketed submucosal injection materials for therapeutic endoscopyClassificationMaterial / ProductDurationCharacteristicsLimitationsNon-viscous SolutionsNormal Saline + Biocompatible, low price, easily availableShort duration, rapid absorption50% Dextrose + + Low price, easily availableModerate duration, risk of inflammation when used at high concentrationsGlycerol (10% Glycerin + 5% Fructose) + + Biocompatible, enhanced duration compared to salineModerate duration, higher price compared to salineViscous Solutions0.4% Sodium Hyaluronate + + + Long-lasting elevationHigh price, potential risk of stimulating resident tumor cellsSodium CarboxymethylcelluloseLong-lasting elevation, lower price than sodium hyaluronateHigher price compared to saline, high viscosityHydroxypropyl MethylcelluloseLong-lasting elevation, lower price than sodium hyaluronate, easily availableHigher price compared to saline, risk of antigen–antibody reactionsBiological FormulationsAutologous Blood + + Biocompatible, enhanced duration compared to saline, easily available, hemostatic effectsLimited use across patients, risk of premature coagulationFibrinogen Mixture + + + Long-lasting elevation, hemostatic effectsHigher price compared to saline, risk of viral contamination and transmissionMarketed ProductsMucoUp® (Seikagaku, Japan) + + + Based on sodium hyaluronate, long-lasting elevationHigh price, low availabilityEleview® (Cosmo Pharmaceuticals, Ireland)Based on Poloxamer 188 and polyoxyl-15-hydroxystearate, methylene blue added as contrast agent, long-lasting elevationHigh priceORISE™ Gel (Boston Scientific, USA)Based on Poloxamer 188 and polyoxyl-15-hydroxystearateRisk of granulomatous mass-like effects at resection site, retracted from market in 2022EndoClot® SIS (Olympus, Japan)Based on absorbable starch particles, long-lasting elevation, hemostatic effectsHigh price, requires additional contrast agentEverlift® (Laborie Medical Technologies, USA)Based on cellulose, methylene blue added as contrast agent, long-lasting elevationHigh priceBlueBoost™ (Micro-Tech Endoscopy, USA)Based on sodium hyaluronate, methylene blue added as contrast agent, long-lasting elevationHigh price, limitation of 50 mL per patientLiftUp® (Ovesco Endoscopy, Germany)Based on poloxamer, methylene blue added as contrast agent, long-lasting elevation, thermoresponsive propertiesHigh price, risk of premature gelationBlue Eye™ (The Standard, South Korea)Based on sodium hyaluronate, methylene blue added as contrast agent, long-lasting elevationHigh price
Normal saline (NS) has been traditionally used in ESD due to its safety, availability, and low cost. Its clinical efficacy, however, is greatly limited by the transient nature of the elevation it provides. Because saline rapidly diffuses into surrounding tissues, the cushion typically lasts only several minutes, necessitating multiple reinjections when addressing larger and more complicated lesions [26, 27]. To overcome this rapid absorption, various hypertonic solutions have been introduced as alternative materials capable of prolonging mucosal elevation while maintaining injectability.
Among these, dextrose solutions were one of the first hypertonic solutions to be validated in ESD through randomized controlled trials. In such studies, 50% dextrose solutions were more effective in maintaining mucosal elevation compared to NS [28, 29]. Unfortunately, subsequent studies reported potential risks of transmural inflammation associated with high-concentration dextrose solutions, limiting their practical use in clinical settings [28, 30, 31]. When injected into animal models, 50% dextrose solutions induced mucosal and muscle damage that developed into submucosal ulcerations a week after injection, whereas normal saline did not cause apparent mucosal changes (Fig. 1A) [30]. Another notable example is glycerol, commonly prepared as a mixture of 10% glycerin and 5% fructose. Originally employed for the treatment of intracranial hypertension in Japan, glycerol solutions were soon adapted in ESD as SIMs and demonstrated favorable elevation as well as high tissue compatibility (Fig. 1A) [30, 32]. Clinical studies have shown that ESD performed with glycerol solutions exhibits significantly higher en bloc resection rates and complete resection rates compared to those performed using NS [33, 34]. Nonetheless, despite their enhanced duration, hypertonic solutions still easily dissipate, making reinjections inevitable during prolonged ESD procedures.Fig. 1Characteristics of conventional submucosal injection materials. A Endoscopic views of injection sites of NS, 50% dextrose, and glycerol. Reproduced with permission from [30] *Copyright **© *2005 American Society for Gastrointestinal Endoscopy. B Comparison of submucosal elevation heights between saline, HA-Mc, HA-Ks, and HA-Ks-80%. Reproduced from [39] under the Creative Commons CC BY license. C Endoscopic ultrasound and histological evaluation of the submucosal layer after injection of saline or 2.5% SCMC. Reproduced with permission from [44] *Copyright **© *2006 American Society for Gastrointestinal Endoscopy. D Mucosa elevation immediately after injection of various materials and change in height over time. Reproduced from [48] under the Creative Commons CC BY-NC license. Abbreviations: NS: normal saline; HA-Mc: MucoUP®; HA-Ks: Ksmart™; SCMC: sodium carboxymethylcellulose; mucosal layer; submucosal layer; muscular layer; G: glycerol; HA: hyaluronic acid; HES: hydroxyethyl starch; S: serum; P: plasma; WB: whole blood
Polymeric solutions are recognized for their high water-retention capacity and viscosity, allowing them to generate stable and durable mucosal elevations. Sodium hyaluronate (SH) is a naturally occurring polymer that exhibits high biocompatibility and viscoelasticity in vivo [35]. Typically prepared as 0.4% solutions, SH was first adopted in Japan as a potential SIM and demonstrated successful ESD outcomes in various anatomical regions such as the esophagus [36], stomach [37], and colorectum [38]. Owing to its favorable performance as a SIM (Fig. 1B), SH was rapidly commercialized in Japan under the product names MucoUP® (Seikagaku, Tokyo, Japan) and Ksmart™ (Olympus, Tokyo, Japan) [39]. Although these products were initially limited to the Japanese market, comparable formulations have since become available in Western countries as well, with BlueBoost™ (Micro-Tech Endoscopy, Ann Arbor, MI, USA) receiving FDA clearance [31]. Despite these advantages, the widespread use of SH remains limited by its high cost (U.S. $495–1280 per 10 mL in the United States as of 2021) [40] and conflicting data regarding its potential to stimulate tumor growth [41, 42].
In Western countries, various solutions based on cellulose derivatives were developed as alternatives owing to their relatively low cost and availability. Sodium carboxymethylcellulose (SCMC) and hydroxypropyl methylcellulose (HPMC) are biocompatible and are commonly used as additives in food and cosmetics [43]. Due to their high viscosity, various concentrations (0.5–3.5%) of SCMC have been evaluated in vivo as potential SIMs. Submucosal injection of 2.5% SCMC effectively separated the mucosal layer from the underlying connective tissue without major complications such as bleeding or perforation (Fig. 1C) [44]. In randomized trials, gastric ESD using 1.0% SCMC achieved short procedural times and high therapeutic outcomes comparable to those obtained with 0.4% SH solutions [45]. Similarly, in a porcine model, 0.25% HPMC demonstrated significantly longer submucosal elevation than NS, although not significantly superior to 0.5% SCMC [46]. Following these studies, EverLift® (Laborie Medical Technologies, Portsmouth, NH, USA) was introduced to the U.S. market in 2020 as a cellulose-based SIM [47].
Beyond conventional non-viscous and viscous agents, several biological formulations have also been investigated as SIMs to improve interaction with the submucosal environment. One such example is autologous blood, which has been reported in a few studies as an effective injection material. In vitro comparisons of blood components (i.e., serum, plasma, and whole blood) with conventional SIMs demonstrated that whole blood produced the most durable cushions, retaining more than 70% of its original height at 60 min post-injection (Fig. 1D) [48]. When injected into the esophageal submucosa of porcine models, autologous blood produced elevations that persisted significantly longer than those produced with either NS or HPMC [49]. In a series of clinical studies, both whole blood and plasma solutions administered as SIMs exhibited superior duration and safety compared to NS. In patients, the use of whole blood was associated with reduced tissue damage, as reflected by lower histological scores for hydrops and tearing [50]. Although endogenous procoagulants in whole blood may offer a unique functionality by promoting hemostasis, they also pose a practical limitation as delayed injection may lead to premature clotting within the syringe [48]. Perhaps for these reasons, recent studies have increasingly reported the potential of autologous blood as a localization marker or tattooing agent prior to endoscopic procedures, rather than as a SIM [51, 52].
Fibrinogen is another biological formulation that provides a promising platform for therapeutic endoscopy. Fibrinogen mixtures are viscous solutions that not only provide durable submucosal elevation comparable to that of SH but also promote hemostasis to prevent microvascular bleeding [53]. Early randomized trials comparing fibrinogen mixtures and NS for EMR of early gastric neoplasms reported no significant differences in en bloc and complete resection rates; however, procedures using fibrinogen mixtures required fewer injections and resulted in shorter procedures [54]. Despite these advantages, fibrinogen is derived from human serum coagulation proteins and, unlike autologous blood, carries potential risks of viral contamination [55].
By overcoming the limitations of conventional SIMs, hydrogels offer a versatile platform that enables precise tuning of both the physical and biological characteristics required for EMR and ESD. In contrast to solution-based materials, these crosslinked polymers can be engineered to exhibit controlled gelation behavior, desirable viscoelastic properties, and therapeutic functionalities. This versatility allows hydrogels to serve not only as durable SIMs but also as therapeutic materials capable of directly addressing complications associated with endoscopic procedures. This section reviews recent studies on injectable hydrogels, with a particular focus on in situ gelling hydrogels (Table 2), shear-thinning hydrogels (Table 2), and therapeutic hydrogels (Table 3) designed for endoscopic applications.Table 2Preclinical hydrogels developed as submucosal injection materials for therapeutic endoscopyClassificationNameCompositionFunctionalityPreclinical modelRefTwo-componentO1A3Oxidized hyaluronic acid / Hydrazide hyaluronic acidMucosal elevation, tissue adhesionPorcine (in vivo)[56]A7C3Maleimide-based oxidized sodium alginate / Sulfhydryl carboxymethyl-chitosanMucosal elevation, tissue adhesion, hemostatic, wound healingCanine (in vivo)[57]ThermoresponsiveCS/β-GPChitosan / β-glycerophosphateMucosal elevationPorcine (in vivo)[62]CSLA/CS/GPLactobionic acid-modified chitosan / Chitosan / β-glycerophosphateMucosal elevation, tissue adhesion, acid resistancePorcine (ex vivo)[63] [64]CSLA/CS/GPLactobionic acid-modified chitosan / Chitosan / β-glycerophosphateMucosal elevation, tissue adhesion, acid resistance, wound healingPorcine (in vivo)[65]HBC-SASuccinylated hydroxybutyl chitosanMucosal elevation, hemostaticPorcine (ex vivo)[67]HpHCS-PVP-GPhigh pH chitosan / polyvinylpyrrolidone / β-glycerophosphateMucosal elevationPorcine (in vivo)[68]HydrogelPluronic® / Hyaluronic acid / Sodium alginate[71]FSF-127 / Sodium alginate[72]PEDOT:PSS/F127Poly(3,4-ethylenedioxythiophene):polystyrene sulfonate / Pluronic F127Mucosal elevation, electrical conductivity[74]Block blendPLGA-b-PEG-b-PLGA copolymersMucosal elevation. Sol–gel transitionPorcine (ex vivo)[75]Shear-thinningISAHGSH-modified sodium alginate / Oxidized sodium alginateMucosal elevation, self-healing, antioxidantLeporine (in vivo)[79]CMS/LapSodium carboxymethyl starch / LaponiteMucosal elevation, self-healingPorcine (in vivo)[80]CCS@AgCatechol modified chitosan / Silver nanoparticlesMucosal elevation, self-healing, antibacterial, wound healingMurine (in vivo)[81]HA–Fe^3^ ⁺Hyaluronic acid / Ferric ionsMucosal elevation, self-healingPorcine (ex vivo)[82]Cat-PBA-ApGltnCatechol group-modified Alaska pollock gelatin / Phenylboronic acid-modified Alaska pollock gelatinMucosal elevation, self-healing, tissue adhesion[83]G-OALGGelatin-oxidized alginateMucosal elevation, self-healingPorcine (in vivo)[84]GGHGellan gum[85]βCP-TET-ISOβ-cyclodextrin grafted with poly(oligo(ethylene glycol) methacrylate) / Tetracycline or IsoproterenolMucosal elevation, self-healing, drug-loadable[86]DGMSHDiglycerol monostearate / Paclitaxel or 6-ACA[87]Table 3Preclinical hydrogels developed as therapeutic materials for endoscopyTarget ConditionNameCompositionFunctionalityPreclinical modelRefHemorrhageAA/AA-NHSAcryloyl-6-aminocaproic acid / AA-g-N-hydroxysuccinimideHemostatic, wound healing, tissue adhesivePorcine (in vivo)[91]HACN-PEGHyaluronic acid (HA)-catechol-NCSN / 4-arm thiolated polyethylene glycolHemostatic, tissue adhesive[92]CMCS/Lap/PVACarboxymethyl chitosan / Laponite / Poly(vinyl alcohol)Hemostatic, wound healing, tissue adhesive, pH-sensitive, antibacterialMurine (in vivo)[95]DTGLDendritic mesoporous silica nanoparticles / Gelatin methacrylate / ThrombinHemostatic, photo-responsive, drug loadablePorcine (in vivo)[96]Perforation / Mucosal woundODA-CSOxidized dextran / Chitosan hydrochlorideTissue adhesive, antibacterial[102]GastroShieldOxidized dextran / polyethyleneimine(PEI)-modified PluronicTissue adhesive, sprayable, wound healing[103]hMPsHydrophobic-modified Alaska pollock gelatinTissue adhesiveMurine (in vivo)[104]THMAIonized N-acryloyl phenylalanine / N-[tris (hydroxymethyl) methyl] acrylamideTissue adhesive, pH-responsive, wound healing[105]PCACAcryl aspartate / Cysteine-grafted carboxymethyl chitosan / C16N-DCATissue adhesive, pH-responsive, wound healing, antibacterialPorcine (in vivo)[106]MCHDimethylacrylamide / Sodium alginate / vonoprazan fumarate or acidic fibroblast growth factorTissue adhesive, drug loadable, plug-type[107]Stricture / StenosishMPsGelatin-derived hydrophobized microparticlesTissue adhesive, sprayable, anti-fibrotic[115]Alg/GelaAlginate / Gelatin / Transglutaminase / Calcium chloride ionsTissue adhesive, anti-fibroticMurine (in vivo)[116]CS/β-GP/HPCChitosan / β-glycerophosphate / Hydroxypropyl cellulose / TriamcinoloneTissue adhesive, thermoresponsive, drug loadable, anti-fibrotic[117]EVs + GelPluronic® F-127 / Extracellular vesiclesThermoresponsive, EV loadable, anti-fibrotic, anti-inflammatoryPorcine (in vivo)[118]EISCHPyrogallol conjugated hyaluronate / ADSCsTissue adhesive, cell loadable, anti-fibrotic, anti-inflammatory[119]
One critical drawback of viscous fluids, despite their mechanical durability, is the high injection pressure required during endoscopic delivery. In situ gelling hydrogels have emerged as practical alternatives, as these materials can initially be delivered as low-viscosity solutions, regardless of the rheological properties they acquire after solidification. A straightforward strategy for generating such systems is the use of two-component hydrogels, in which two crosslinkable precursor solutions are sequentially injected to form hydrogels in situ.
Hyaluronic acid (HA), a natural material already used in numerous marketed products, has also been adapted into two-component hydrogel systems through functional modification. Two derivatives, oxidized hyaluronic acid (OHA) and hydrazide-modified hyaluronic acid (AHA), were formulated as reactive precursors capable of forming crosslinked hydrogels. When sequentially injected into the porcine submucosa, OHA and AHA solutions rapidly formed a stable hydrogel that maintained mucosal elevation for up to one hour in both the stomach and esophagus (Fig. 2A) [56]. A similar strategy was employed by Lei et al. using oxidized sodium alginate and carboxymethyl chitosan (CMC). Unlike the aforementioned example [56], this study used two precursor solutions sequentially injected as separate functional agents (Fig. 2B**)**. The more viscous sodium alginate solution was first injected into the submucosa as an independent SIM, followed by the application of low-viscosity CMC solutions as postoperative sealants. Notably, sodium alginate derivatives alone provided elevations that were more durable than those produced with either NS or glycerol, and the application of sulfhydryl-modified CMC rapidly formed a postoperative hydrogel that accelerated wound healing and re-epithelialization at the surgical site (Fig. 2C) [57].Fig. 2Characteristics of two-component hydrogels used for therapeutic endoscopy. A Endoscopic ultrasound images of the submucosal pad under the stomach submucosa and the esophagus submucosa after injection of the OHA/AGA hydrogel. Red lines indicate the lifting heights. Reproduced with permission from [56] *Copyright **© *2024 American Chemical Society. B Schematic diagram of the canine esophageal ESD procedure using AM and CS solutions. Reproduced from [56] under the Creative Commons CC BY-NC-ND license. C Upper Injection of AM solution into the submucosa of the canine esophagus; Upper Mucosal dissection; Upper Removal of blood stains and hydrogel from the wound; Lower Spraying of CS solution after dissection; Lower Gel morphology after CS solution in contact with AM solution at the wound at 0 min; Lower Morphology of the hydrogel after gelling and stabilizing in the wound at 10 min. Reproduced from [57] under the Creative Commons CC BY-NC-ND license. Abbreviations: OHA: oxidized hyaluronic acid; AHA: hydrazide hyaluronic acid; AM: Maleimide-based oxidized sodium alginate; CS: Sulfhydryl carboxymethyl chitosan
As another approach, in situ hydrogels can be designed to respond to various environmental cues, providing solid platforms for physiological applications [58]. Particularly for endoscopic use, chitosan-based systems have been widely studied for their ability to form thermoresponsive hydrogels when mixed with β-glycerophosphate (GP) [59, 60]. Jeon et al. evaluated the efficacy and safety of hydrogels composed of chitosan (CS) and GP in porcine models. Compared to conventional NS or 0.4% SH solutions, CS/GP hydrogels generated more durable cushions with lower injection volumes, although histological analysis revealed partial inflammatory responses and submucosal hypertrophy, indicating that further safety evaluation is needed [61]. In another study, the efficacy of CS/GP hydrogels was directly compared to commercial SIMs, including Eleview® (Cosmo Pharmaceuticals, Dublin, Ireland) and ORISE™ Gel (Boston Scientific, Marlborough, MA, USA). The injection forces of CS/GP hydrogels were comparable to those of Eleview® but significantly lower than those of ORISE™ Gel, demonstrating their favorable injectability (Fig. 3A) [62].Fig. 3Characteristics of thermoresponsive hydrogels used for therapeutic endoscopy. A Injectability of the chitosan thermosensitive solutions and commercial submucosal injection agents. Reproduced from [62] under the Creative Commons CC BY license. B The injection force of various CSLA/CS/GP hydrogel precursor solutions. Reproduced with permission from [63] *Copyright **© *2021 Elsevier Ltd. C Scanning electron microscope images of CS-GP. CSLA/CS-GP and CSLA-GP hydrogels. Reproduced with permission from [64] *Copyright **© *2021 Elsevier B.V. D Three-dimensional response surface plots showing synergistic effect of the concentrations of PVP and GP on the gelation time of thermosensitive hydrogel. Reproduced with permission from [68] *Copyright **© *2024 Published by Elsevier B.V. E A schematic diagram showing that the temperature sensitivity of the PEDOT: PSS/F127 hydrogel enables in situ gelation, while its electrical conductivity mitigates excessive local heating during high-frequency electrosurgical cutting. Reproduced with permission from [74] *Copyright **© *2025 Wiley‐VCH GmbH. F Conductivity of F127 hydrogel and PEDOT: PSS/F127 hydrogels. Reproduced with permission from [74] *Copyright **© *2025 Wiley‐VCH GmbH. Abbreviations: CS: chitosan; GP: β-glycerophosphate; CSLA: lactobionic acid-modified chitosan; PVP: polyvinylpyrrolidone; PEDOT: PSS: poly(3,4-ethylenedioxythiophene):polystyrene sulfonate
Other modified systems, such as hydrogels incorporating lactobionic acid-modified chitosan (CSLA) with CS and GP, also showed enhanced endoscopic injectability. The addition of CSLA improved the fluidity of the hydrogel solution and significantly reduced injection force compared to unmodified CS precursors (Fig. 3B). Both gelation time and gelation temperature decreased with increasing CSLA content but still maintained practical LCSTs and high stability under gastric conditions [63]. When CSLA/CS/GP hydrogels were further lyophilized into powders, the reconstituted solutions exhibited improved injectability compared to the original solutions while retaining their thermoresponsive properties and structural integrity (Fig. 3C) [64]. More recently, the practicality of both original and powder-based CSLA/CS/GP hydrogels was evaluated under conditions resembling clinical ESD. In porcine models, both forms demonstrated strong durability and acid resistance, successfully facilitating ESD in the stomach and esophagus, reducing operation time, and promoting wound healing [65].
Hydrogels derived from hydroxybutyl chitosan (HBC) also exhibit thermoresponsive behaviors; however, their low LCSTs limit endoscopic use [66]. Instead, hydrogels developed from succinylated HBC displays enhanced LCSTs owing to their improved hydrophilicity. The LCSTs of succinylated HBC hydrogels ranged from 23 °C to 29 °C, providing a practical thermoresponsive platform for SIM applications [67]. Despite these advances, thermoresponsive hydrogels require precise optimization of gelation time and temperature. Liu et al. addressed this by establishing a mathematical model predicting the gelation time of hydrogels with varying ratios of CS, polyvinylpyrrolidone, and GP (Fig. 3D). Experimental validation showed over 86% agreement with observed values, supporting the model’s predictive accuracy. In vivo, these CS hydrogels provided significantly greater mucosal elevation and prolonged cushion duration compared to NS [68]. In addition to natural polymers, synthetic polymers also offer distinct advantages due to their chemical tunability [69]. Poloxamers are amphiphilic triblock copolymers widely used in injectable platforms [70], including Eleview® and LiftUp® (Ovesco Endoscopy, Tübingen, Germany). In one preclinical study, poloxamer 407 (F127) was combined with sodium alginate and HA to develop a thermoresponsive SIM. Whereas sodium alginate and HA alone exhibited limited in situ gelation, the addition of F127 imparted rapid thermoresponsive behaviors and improved stability, maintaining performance for up to three weeks at room temperature or 105 days at 2–8 °C [71]. The efficacy of F127-based hydrogels in mucosal elevation and ESD were also similarly reported in other studies as well [72, 73].
Building on these advances, an injectable hydrogel composed of poly(3,4-ethylenedioxythiophene):polystyrene sulfonate (PEDOT:PSS) and F127 was introduced as a SIM with both conductive and thermosensitive properties (Fig. 3E). Compared to conventional F127 hydrogels, PEDOT:PSS/F127 hydrogels exhibited higher storage modulus while maintaining thermoresponsive and shear-thinning characteristics. PEDOT:PSS/F127 hydrogels also demonstrated high electrical conductivity (1.00 ± 0.03 S·m⁻^1^) due to the intrinsic properties of PEDOT:PSS (Fig. 3F), effectively dispersing electrosurgical currents and reducing local temperature by 8.7 °C compared to NS. In vivo validation in porcine models further showed reduced postoperative tissue burns and bleeding, collectively indicating improved efficacy and safety during endoscopic electrosurgery [74]. Additionally, Cui et al. introduced a “block-blend” approach to create a thermoresponsive hydrogel using non-thermogellable copolymer solutions. The resulting hydrogel, despite its non-thermoresponsive precursors, was injectable through 23-gauge needles and exhibited thermoresponsive properties sustaining mucosal elevations up to 90 min. While this strategy offers a novel route for thermoresponsive SIM development, further in vivo studies are needed to confirm clinical translation [75].
Taken together, these studies underscore the potential of thermoresponsive hydrogels as SIMs, showing that both natural and synthetic polymers can achieve favorable injectability and long-lasting mucosal elevations. Nonetheless, their reliance on external triggers can present drawbacks, particularly the risk of premature gelation within catheters. Despite their tunable gelation behavior, thermoresponsive hydrogels generally exhibit a narrow responsive window, necessitating further optimization for broader clinical applicability.
Shear-thinning hydrogels are non-Newtonian fluids that exhibit promising characteristics for therapeutic endoscopy. In contrast to two-component or thermoresponsive hydrogels that rely on secondary cues for gelation, shear-thinning hydrogels transition between solid-like and liquid-like states in response to shear force [76]. These unique behaviors arise from reversible crosslinks formed through dynamic interactions within the polymer network [77], thereby overcoming the traditional trade-off between mechanical strength and injectability.
When evaluated under experimental conditions simulating clinical settings, shear-thinning hydrogels based on laponite (LAP), Carbopol, or xanthan gum all demonstrated favorable characteristics for endoscopic applications. Among the three, LAP hydrogels required the lowest injection pressure and showed minimal post-delivery dripping, maintaining intact morphology at a 90° slope. Although injection pressures were dependent on both catheter length and diameter, the most critical determinant was the hydrogel 6% LAP, 4% xanthan gum, and 3% Carbopol hydrogels each required pressures exceeding 667 kPa, a threshold considered as the practical upper limit for manual injection [78].
Natural polysaccharides are commonly employed to develop shear-thinning SIMs due to their high biocompatibility and ability to form reversible polymer networks. For example, sodium alginate hydrogels prepared from oxidized sodium alginate and glutathione-modified alginate exhibited injectability and self-healing properties, generating mucosal elevations superior to those of 0.4% SH. In porcine colorectal models, these hydrogels not only maintained 92% of their initial elevation height for 120 min but also demonstrated notable antioxidant activity owing to glutathione [79]. Similarly, Wang et al. developed a polysaccharide-based system exploiting electrostatic and hydrogen-bonding interactions between LAP and carboxymethyl starch (Fig. 4A). Compared to the commercial product MucoUp®, LAP hydrogel prepared at 15 mg/mL preserved 90% of its initial height after two hours, whereas MucoUp® lost 72% of its original height in the same timeframe [80].Fig. 4Characteristics of shear-thinning hydrogels used for therapeutic endoscopy. A Diagram outlining the electrostatic mechanism of CMS/Lap hydrogel formation and its shear-thinning properties. Reproduced with permission from [80] *Copyright **© *2024 Elsevier Ltd. B (Left) Scanning electron microscope image of HA foam and HAF hydrogels. (Right) Photographs of the injection process of HAF hydrogel in PBS after methylene blue staining. Reproduced with permission from [82] *Copyright **© *2025 Elsevier B.V. C Injection setup and cross-sectional views of the porcine stomachs following injection of different materials for 30 min (scale bar = 5 mm). Reproduced with permission from [83] *Copyright **© *2024 Elsevier B.V. D (Upper) Raw images and (Lower) histological analysis of injection site of rat skin at each time point. Red outline in histology images indicates residual hydrogel. Reproduced from [84] under the Creative Commons CC BY-NC-ND license. E Comparison of GGH with other injection materials in terms of injection pressure. Reproduced from [85] under the Creative Commons CC BY-NC-ND license. Abbreviations: CMS: carboxymethyl starch; NS: normal saline; GF: glycerol fructose sodium chloride; DS: dextrose solution; SA: sodium alginate; SH: sodium hyaluronate; SCC: sodium carboxymethylcellulose; GGH: gellan gum hydrogel
As natural polymers, CS and its derivatives also provide versatile platforms for shear-thinning SIMs. Recently, catechol-modified chitosan (CCS) hydrogels, crosslinked via Michael addition and Schiff base reactions, were loaded with silver nanoparticles to serve as antibacterial SIMs. In addition to enhanced mucosal elevation and accelerated wound closure demonstrated in vivo, CCS hydrogels loaded with silver nanoparticles effectively inhibited the growth of both gram-negative E. coli and gram-positive S. aureus [81]. Using a similar approach, HA can also be formulated as a shear-thinning hydrogel with self-healing properties. Although HA chains inherently form hydrogen bonds to create weak crosslinks, these intermolecular interactions can be further strengthened by adding trivalent ferric ions (Fe^3^⁺) as crosslinking agents. Due to the additional ionic coordination between HA and Fe^3^⁺, HA foams soaked in aqueous Fe^3^⁺ solutions formed hydrogels that were injectable through conventional syringes (Fig. 4B). The resulting hydrogels produced greater mucosal elevation than buffered saline, HA alone, or F127 when injected into the porcine stomach [82].
Gelatin is another example that can be chemically modified as catechol-functionalized gelatin or boronic acid-functionalized gelatin to serve as precursors for shear-thinning hydrogels. These hydrogels demonstrated superior mucosal elevation compared to both saline and MucoUP® (Fig. 4C). Following submucosal injections in mice, these gelatin-based hydrogels completely degraded within 28 days and induced minimal accumulation of inflammatory cells [83]. In line with this approach, Fan et al. also introduced a gelatin-based hydrogel for ESD. However, instead of using modified gelatin, this study utilized functionalized alginates, crosslinking gelatin with reactive alginate polymers to develop shear-thinning hydrogels. These hydrogels exhibited strong durability for up to one hour in porcine models. When subcutaneously injected into rat skin, these hydrogels demonstrated gradual degradation across seven days, with the diameter of the injected hydrogel decreasing from 1.7 cm to 0.4 cm in three days (Fig. 4D). When examined one week after injection in porcine models, no hydrogel residue was observed at the injection site, confirming excellent biodegradability [84]. Unlike other polymers that often require chemical functionalization to impart crosslinking capability, gellan gum hydrogels (GGHs) exploit the natural ability of gellan polymers to form double-helix aggregates via intra- and inter-chain interactions. Despite their strong mechanical properties, GGHs required significantly lower injection forces (127.8 kPa) compared to commonly used viscous SIMs such as 50% dextrose and 0.4% SH (Fig. 4E). This hydrogel showed greater elevation retention than either NS or 0.4% SH when injected into the porcine colon and was compatible with epinephrine loading, enabling effective hemostasis during endoscopic resection [85].
In a different study, the synthetic polymer poly(oligo(ethylene glycol) methacrylate) (PEGMA) was grafted onto β-cyclodextrin cores to develop a multi-arm star polymer hydrogel. The reversible entanglement of polymer brushes within this hydrogel imparted shear-thinning properties, enabling optimal endoscopic injection. Importantly, its amphiphilic structure—composed of hydrophobic β-cyclodextrin cores and hydrophilic PEGMA side chains—allowed encapsulation of both hydrophilic drugs (e.g., isoproterenol) and hydrophobic drugs (e.g., tetracycline) [86]. The FDA-approved amphiphilic molecule diglycerol monostearate (DGMS) was also utilized to develop an injectable hydrogel network, leveraging its amphiphilic structure for dual drug loading. Because DGMS solutions readily form gel-like structures through reversible interactions such as π–π stacking and hydrophobic forces, these hydrogels were easily scalable in vitro and exhibited a long shelf life up to six months. This hydrogel effectively encapsulated both hydrophilic and hydrophobic drugs, underscoring the versatility of amphiphilic materials not only as SIMs but also as functional drug delivery platforms [87].
Beyond two-component and thermoresponsive systems, shear-thinning hydrogels represent a class of SIMs that combine injectability with mechanical robustness, both of which are critical for endoscopic applications. Preclinical studies employing a wide range of materials—from natural polysaccharides to synthetic amphiphilic polymers—have consistently demonstrated their practicality as SIMs, underscoring their ability to enhance procedural efficiency while functioning as multifunctional drug delivery platforms. Although further studies are needed to evaluate their safety and performance with repeated use, the procedural and therapeutic advantages of shear-thinning hydrogels over conventional SIMs emphasizes their growing importance in the advancement of endoscopic injection materials.
In addition to the physicochemical and rheological properties of injectable hydrogels, another important reason for their significance in endoscopy is their potential as therapeutic platforms. Recent studies have sought to engineer injectable hydrogels as bioactive materials, not only enabling them to serve as structural supports, but also as active safeguards against complications associated with EMR or ESD. While hydrogels discussed herein also employ gelation strategies described previously, this section shifts focus to discuss and outline hydrogel-based SIMs from a therapeutic perspective. These hydrogels, developed for the treatment or prevention of complications such as hemorrhage, perforation, or stricture, provide novel insights beyond their conventional role as structural supports (Table 3).
Intraprocedural bleeding is frequently associated with ESD, arising from damage to blood vessels during deep resections. While minor bleeding is generally considered an inherent event rather than a complication, significant bleeding may still disrupt en bloc resection, induce hemorrhagic shock, or necessitate surgical intervention, placing patients at considerable risks [13]. Electrocoagulation is widely used as both a prophylactic and intraprocedural measure because it can be directly performed with endoscopic knives. However, when performed under impaired visibility, it carries the risk of unnecessary thermal injury that can lead to perforation, making the use of additional mechanical tamponade [88] or hemostatic agents [89] indispensable.
Injectable hydrogels based on chemically modified polymers are well suited for these purposes because they provide a structurally stable yet chemically flexible basis for local hemostasis. A representative example is the acryloyl-6-aminocaproic acid (AA) hydrogel, which has been reported to exhibit self-healing behavior under acidic conditions [90]. Leveraging this pH responsiveness, AA hydrogels have been evaluated as gastric hemostatic agents, where they demonstrated superior bleeding control and therapeutic efficacy compared to proton pump inhibitors. When applied to mouse liver incision models, a composite hydrogel containing both AA and AA-NHS resulted in significantly less blood loss than hydrogels composed solely of AA monomers or blank controls (Fig. 5A) [91]. Similarly, hydrogels developed from catechol-modified HA and polyethylene glycol (PEG) exhibited strong hemostatic properties comparable to those of fibrin gels (Fig. 5B). In contrast to AA hydrogels, which required several minutes to naturally restore their mechanical properties, hemostatic HA hydrogels were able to instantly form a solid network in situ by sequentially applying sodium periodate as an additional oxidizing agent [92].Fig. 5Characteristics of hydrogels used for management of hemorrhage and mucosal wounds. A Schematic representation of the mouse liver incision model and blood loss after application of AA/AA-NHS hydrogels. Reproduced from [91] under the Creative Commons CC BY license. B Illustration of the effects of fibrin gels and hybrid hydrogels on heparinized rat blood. Reproduced with permission from [92] *Copyright **© *2021 Wiley‐VCH GmbH. C Loss of stomach blood after treatment with CMCS/Lap hydrogels and photographs of the stomach hemostasis process. Reproduced with permission from [95] *Copyright **© *2023 American Chemical Society. D GastroShield application via a catheter onto a perforated colon ex vivo. Reproduced from [103] under the Creative Commons CC BY-NC-ND license. E Comparison of the burst pressure between Org and Hm-ApGltn gels on the duodenum, large intestine, and stomach. Reproduced with permission from [104] *Copyright **© *2021 Elsevier B.V. F Images of in situ gelation and adhesion performances on gastric wound ex vivo. Scale 1 cm. Reproduced with permission from [105] *Copyright **© *2024 Elsevier Ltd. G X-ray photograph of the hydrogel plug and gastroscopic follow-up images of the perforations in groups treated with VF-loaded or AFGF-loaded hydrogels. Reproduced with permission from [107] *Copyright **© *2023 American Chemical Society. Abbreviations: AA: acryloyl-6-aminocaproic acid; CMCS: carboxymethyl chitosan; Lap: laponite; Hm-ApGltn: hydrophobic-modified Alaska pollock gelatin; VF: vonoprazan fumarate; AFGF: acidic fibroblast growth factor
Building on the preclinical success of LAP hydrogels in endoscopy [78, 80, 93, 94], CS–LAP composites have also been formulated as injectable adhesive patches for local hemostasis. Reinforced with poly(vinyl alcohol), these hydrogels remained stable in acidic environments and achieved complete hemostasis within 90 s in murine liver bleeding models, outperforming conventional hemostatic powders. Such hemostatic property may be attributed to the abundant surface amino groups that can interact with erythrocyte membranes, accelerating the coagulation process. When further applied to a gastric perforation model in rat, the CS-LAP hydrogels demonstrated strong tissue adhesion, significantly reducing the blood loss from 172.6 ± 39.8 mg to 29.2 ± 5.8 mg (Fig. 5C) [95]. In another study, injectable hydrogels were utilized as depots for the local delivery of hemostatic agents. Highly porous nanoparticles were loaded with thrombin and coated with gelatin methacrylate to produce injectable nanogels, which were successfully crosslinked into a bulk scaffold in situ when irradiated with blue-laser endoscopy. These photo-crosslinked hydrogels achieved effective hemostasis not only through thrombin release but also by forming a dense physical clot at the bleeding site [96].
Perforation remains a major complication of ESD with potentially life-threatening outcomes if not properly managed. ESD-induced perforations can occur throughout the GI tract, but colorectal perforations are considered the most critical because they may lead to secondary contamination from colonic flora and fecal contents [97]. The likelihood of such events is lower with gastric perforations, although they can still cause peritonitis and require urgent intervention [98]. Even in the absence of full-thickness perforations, extensive submucosal injuries can leave large mucosal defects that may progress into iatrogenic ulcers. While iatrogenic ulcers are known to heal more rapidly than peptic ulcers [99], up to 20% of ESD patients still experience persistent ulcers beyond eight weeks, which increases the risk of delayed bleeding and perforation [100]. Endoscopic management of perforations typically relies on mechanical clips [101], but few intraoperative approaches are currently available for the prevention of such events. Under these circumstances, endoscopically deliverable hydrogels with adhesive or pro-regenerative properties offer a rational strategy for the management of ESD-related wounds.
Recently, an adhesive hydrogel formed from oxidized dextran (ODA) and chitosan hydrochloride was evaluated in porcine models as a wound-shielding material. When co-delivered to esophageal wound lesions, the precursor solutions rapidly formed hydrogels in situ, generating bioadhesive sheets that maintained their structural integrity and adhesion for up to two weeks. However, despite their effectiveness as a protective layer, the hydrogels did not directly promote tissue regeneration [102]. Taboda et al. reported a similar study, utilizing a sprayable two-component hydrogel for sustained protection of GI wounds. In this system, ODA and polyethyleneimine-modified micelles were delivered through a tri-lumen catheter, producing a thin, adherent coating upon contact with the tissue (Fig. 5D). These sprayable hydrogels demonstrated sustained adhesion and barrier function in acidic and enzymatic environments, making the platform suitable for both colonic and gastric wounds. Although the ODA hydrogel in this study also lacked intrinsic therapeutic activity, it effectively protected the lesion against acids and enzymes that may delay gastric wound healing, resulting in faster regeneration compared to untreated controls [103].
Hydrophobically modified gelatin microparticles can also be utilized as sprayable precursors that rapidly form adhesive hydrogels when applied to gastric wounds. In this approach, gelatin was functionalized with alkyl chains, enabling the formation of self-assembled microparticles with an adhesive interface that strongly adhered to wet tissue via hydrophobic interactions. Under physiological conditions, hydration and fusion of the hydrophobic colloids formed a robust hydrogel layer with burst pressures significantly greater than those of control hydrogels in the duodenum, large intestine, and stomach (Fig. 5E). Although further preclinical evaluation in large animals is required, this endoscopically sprayable hydrogel naturally degraded within 14 days, providing sustained wound protection without long-term residue [104].
Several systems have also been developed specifically for the treatment of gastric wounds by utilizing pH-responsive materials. One such approach involved the development of an acid-responsive platform by crosslinking phenylalanine and acrylamide derivatives into injectable precursors. When exposed to acidic environments, the phenyl groups of phenylalanine underwent hydrophilic–hydrophobic transitions, repelling interfacial water and promoting hydrogen bond–mediated gelation in situ (Fig. 5F). The resulting phenylalanine-based hydrogel demonstrated strong tissue adhesion and stability for up to 14 days under acidic conditions, effectively protecting the lesion surface from acid and pepsin exposure. When tested in a rat model of acid-induced gastric ulcers, these hydrogels not only accelerated ulcer healing but also suppressed local inflammation and promoted capillary formation at the damaged lesion [105].
A different acid-triggered hydrogel was later introduced as a multifunctional therapeutic platform with antibacterial properties. Here, the injectable hydrogel backbone was developed using acryl aspartate and cysteine-grafted CMC, into which the molecule C16N-DCA was loaded as an antibacterial agent against H. pylori. Interestingly, β-carboxy amide was employed as an acid-responsive headgroup for C16N-DCA, successfully inducing charge reversal under acidic pH. This pH-mediated shift from anionic to cationic charge enhanced mucus permeation and cellular uptake of the antibacterial agent, actively disrupting bacterial cell walls and membranes. This injectable hydrogel exhibited high injectability and served as an effective depot for C16N-DCA, accelerating wound closure and promoting tissue recovery in porcine wound models [106].
Beyond sheet-like hydrogels, Liu et al. introduced a hyperboloid-shaped adhesive hydrogel inspired by the shape of mushroom caps as a therapeutic plug for gastric perforations. Although not injectable, these acrylamide–sodium alginate hydrogels exhibited strong mechanical strength and compressibility, enabling effective closure of perforations for up to 28 days. Surface modification with nanosilica further endowed the hydrogel with a robust bioadhesive interface suitable for gastric applications. In addition to its mechanical robustness, the hydrogel was further loaded with therapeutic agents such as vonoprazan fumarate and acidic fibroblast growth factor (aFGF) to actively modify the wound environment (Fig. 5G). Real-time pH monitoring after application of the hydrogel plugs revealed that delivery of vonoprazan fumarate successfully maintained elevated gastric pH, which further promoted hemostasis compared to untreated acidic environments. Subsequent immunohistochemical analyses confirmed that this drug-laden hydrogel plug minimized edema and necrosis while simultaneously reducing local inflammation and promoting angiogenesis to accelerate perforation repair [107].
In addition to perforation, stricture is another complication that frequently arises from postoperative wounds, especially in the esophagus. Unlike the stomach or colorectum, postoperative perforations are relatively rare in the esophagus [108]. Instead, the incidence of esophageal stricture has been reported to range from 66–75% following ESD of lesions exceeding three-quarters of the esophageal circumference [109]. The precise mechanism underlying stricture development remains unclear, although it is generally accepted that excessive fibroblast activation and dysregulated regeneration during mucosal repair are key mediators [110]. Because of its delayed and progressive nature, preventive measures are typically directed toward inhibiting initial inflammation, regulating epithelial regeneration, and suppressing fibrosis [111]. Clinically, these have been attempted through the administration of corticosteroids such as triamcinolone acetonide or prednisolone, which reduce the rates of strictures and dilation sessions, respectively [112]. However, inadequate use of steroids may induce mural necrosis [113], while other mechanical approaches, including endoscopic balloon dilation or esophageal stents, carry risks of swallowing difficulties or hyperplasia [110]. Endoscopically administrable hydrogels have emerged as promising candidates in this context, providing not only wound protection but also a platform for the delivery of bioactive molecules that can modulate the inflammatory and fibrotic microenvironment.
A clinical study recently evaluated the efficacy of a hemostatic self-assembling peptide (SAP) gel for stricture prevention in patients undergoing esophageal ESD. In this multicenter study, the SAP gel, designed to mimic the extracellular matrix, was endoscopically applied after ESD and monitored for a median of two months across 43 patients. Notably, stricture formation occurred in 20.9% of all patients and in only 19% of high-risk patients with mucosal defects involving more than three-quarters of the esophageal circumference. Considering that administration of steroids typically yields stricture rates ranging from 19 to 67%, the outcomes achieved with SAP gels were highly comparable to those of conventional steroid therapy. Nonetheless, while these findings support the practicality of SAP gels as effective prophylactic agents, it should be noted that all patients in this study also received variable doses of proton pump inhibitors (PPIs) for at least 30 days post-ESD, which may have influenced the results [114].
In preclinical settings, gelatin-based hydrogels have been investigated as versatile platforms for stricture management. Yano et al. reported the anti-inflammatory and anti-fibrotic effects of endoscopically sprayable hydrogels prepared from hydrophobized gelatin microparticles. Similar to their earlier application to perforation [104], these sprayable microparticles adhered to esophageal wound surfaces and provided a protective barrier to support enhanced regeneration. In porcine esophageal ESD models, hydrogel treatment significantly reduced stricture rates compared to untreated controls, as verified by decreased inflammatory infiltration, fibrosis, and muscular layer thickening. Although gelatin hydrogels did not completely prevent stricture formation, the stenosis rate — defined as the ratio of the most stenotic diameter to the normal diameter — decreased to 76% after treatment compared to 90% in untreated controls (Fig. 6A) [115]. In another study, a mixture of alginate and gelatin powders co-delivered with calcium and transglutaminase crosslinkers was evaluated for stricture management. In this model, submucosal exfoliation in rats was introduced as a novel platform to mimic ESD-induced stenosis in patients. Under these conditions, treatment with the hydrogel effectively attenuated collagen accumulation and myofibroblast differentiation, highlighting the preventive potential of gelatin-based hydrogels in stricture management [116].Fig. 6Characteristics of hydrogels used for management of stricture. A Esophageal stricture rates compared to the basis of macroscopic findings after treatment of gel-forming hydrophobized microparticles. Reproduced from [115] under the Creative Commons CC BY license. B Therapy follow-up by radiologic evaluation assessing the stricture rate after treatment. Reproduced from [118] under the Creative Commons CC BY-NC license. C Schematic illustration of the paracrine effects of the ADSCs encapsulated in the EISCH hydrogel. Release profiles of VEGF and IL-10 from the ADSCs encapsulated in the EISCH hydrogels crosslinked with different concentrations of HRP. Reproduced from [119] under the Creative Commons CC BY license. Abbreviations: EVs: extracellular vesicles; ADSCs: adipose cell derived stem cells; VEGF: vascular endothelial growth factor
Regarding stricture management, other studies have also attempted to incorporate bioactive molecules by utilizing injectable hydrogels as sustained delivery platforms. Wang et al. developed a thermosensitive hydrogel composed of CS, β-glycerophosphate (GP), and hydroxypropyl cellulose for the delivery of triamcinolone, a corticosteroid used to prevent stricture formation. Owing to its thermoresponsive characteristics, this hydrogel exhibited excellent initial injectability and strong mechanical strength after gelation while sustainably releasing the loaded drug for more than 10 days. Although not directly evaluated in esophageal stricture models, the hydrogel effectively suppressed fibroblast proliferation and fibrogenesis in vitro and attenuated keloid formation in rat dermal defect models [117]. A complementary approach also employed thermosensitive hydrogels as injectable depots for delivering extracellular vesicles (EVs) derived from adipose stromal cells as therapeutic agents. In a porcine model, a follow-up analysis of esophageal strictures 21 days after ESD revealed that endoscopic delivery of EVs encapsulated within F127 hydrogels significantly reduced stricture rates compared to untreated controls. While passage of closed biopsy forceps (2 mm) were successful in all groups, the passage of open forceps (7 mm) were only possible in 66%, 33%, and 0% of EV-hydrogel-treated, hydrogel-treated, and untreated groups, respectively. While blank F127 hydrogels alone moderately reduced local inflammation, the therapeutic effect was markedly enhanced with EVs. Gastroscopes were able to move through the esophageal stricture only in porcine models treated with EV-loaded hydrogels, demonstrating its successful therapeutic efficacy against esophageal strictures (Fig. 6B) [118].
Beyond bioactive molecules, direct stem cell therapy has also been explored using injectable hydrogels for stricture treatment. Chung et al. developed a self-crosslinkable hyaluronate hydrogel functionalized with pyrogallol groups as an injectable scaffold for adipose-derived stem cells (ADSCs). These self-crosslinking hydrogels served as biocompatible carriers, maintaining strong adhesion despite swallowing and peristalsis in the esophagus. In porcine ESD models, administration of ADSC hydrogels reduced stricture rates to 38%, compared to 62% and 79% observed in free ADSC and untreated control groups, respectively. In vitro, ADSCs encapsulated within the hyaluronate hydrogel secreted vascular endothelial growth factor (VEGF) and interleukin-10 (IL-10) for up to 14 days, suggesting their pro-regenerative potential (Fig. 6C). Further evaluation in porcine models demonstrated reduced cell proliferation and myofibroblast activation in the damaged esophagus, confirming the therapeutic efficacy of injectable hydrogels against post-ESD strictures [119].
Collectively, functional hydrogels have emerged as important therapeutic adjuncts in endoscopy, providing capabilities that go beyond structural mucosal elevation. By incorporating hemostatic, adhesive, anti-inflammatory, antifibrotic, and regenerative functions, these hydrogels offer promising strategies for preventing and managing intraprocedural and postprocedural complications that continue to challenge the safety and efficiency of therapeutic endoscopy. Although much of the current evidence remains preclinical, clinical studies — such as those employing SAP gels — demonstrate the translational feasibility of these early developments. Further optimization of material design and delivery strategies, along with rigorous evaluation in clinical trials, will be critical to further potentiate hydrogel-based SIMs as next-generation tools in therapeutic endoscopy.
Submucosal injections in endoscopy have traditionally relied on solutions that provide mucosal elevation without therapeutic function. Recent advances in hydrogel-based systems have introduced new opportunities to redefine this role by enabling controlled gelation and imparting bioactive functionalities. Across two-component, thermoresponsive, and shear-thinning hydrogels developed from both natural and synthetic polymers, preclinical studies have consistently demonstrated reduced injection forces, prolonged cushion persistence, and improved compatibility with endoscopic workflows. Importantly, functional and cargo-laden hydrogels that can be endoscopically administered to GI wounds extend beyond simple protection and effectively manage complications associated with endoscopic procedures.
Despite encouraging preclinical outcomes, several translational challenges remain. Since the FDA clearance of Eleview® as a class II medical device in 2015, various SIMs have been introduced to the market using Eleview® as a predicate device (Scheme 3). However, ORISE™ Gel was withdrawn from the market in 2022 due to its association with granulomatous mass-like reactions at the injection site. This underscores the importance of biosafety and highlights the limitations of safety assessments based solely on large-animal studies, indicating the need for long-term, well-controlled clinical trials. Furthermore, cost remains a major barrier to clinical translation. Even today, commercially available products prepared in ready-to-use syringes are not standard materials in routine procedures because of their high price, with many clinicians still preferring economical in-house preparations. For novel hydrogels to achieve widespread adoption as SIMs, scalability and economic competitiveness must be addressed alongside safety.
From a procedural standpoint, two-component and thermosensitive hydrogels also present aspects that require further optimization before routine endoscopic use. Although two-component hydrogels provide tunable gelation through external initiators or crosslinkers, they often require sequential injections or multi-channel catheters to avoid premature gelation, inevitably increasing procedural complexity. Thermosensitive hydrogels, while effective, remain vulnerable to premature gelation within narrow endoscopic catheters if not applied rapidly. Although extended durability and therapeutic benefits may offset these drawbacks, further optimization is needed to ensure reliable clinical administration. Similar considerations apply to therapeutic hydrogels. Unlike many hydrogel-based SIMs that generally require minimal injections, therapeutic hydrogels described in this review often required multiple applications, particularly for stricture management. This repeated administration may affect patient compliance and efficacy, echoing the challenges previously seen with conventional SIMs for ESD.
Looking forward, future research should not only focus on improving mechanical and biological properties but also attempt to address long-term biosafety and biodegradation of hydrogel components. Personalized design strategies tailored to lesion type, patient-specific healing responses, and anatomical site could further maximize safety and efficacy. Integrating imaging capability, controlled drug delivery, and immunomodulation into hydrogel systems may pave the way toward true theranostic applications in endoscopy. Additionally, the establishment of standardized preclinical models will be essential to ensure reproducibility and accelerate market adoption. Emerging computational tools, including artificial intelligence, may hold promise for predictive optimization of hydrogel formulation and simulation of clinical outcomes, ultimately bridging the gap between bench and bedside. However, preclinical evaluation using animal models are yet irreplaceable and many studies discussed in this review have utilized either porcine or murine models (Table 2, 3) for these purposes. While no single animal can completely mimic the human GI tract, porcine and canine models are few examples that resembles the morphology found in human with limitations [120]. Due to these advantages, both models have also been thoroughly evaluated by clinicians as practical models for endoscopic practices [121, 122]. Unfortunately, the translational significance of murine models still remain questionable. As murine models lack key physiological features found in larger animals, further evaluation regarding their significance as preclinical models of hemorrhage, perforation, or structure is required.
Interestingly, the selection of hydrogel systems appears closely aligned with clinical context, underscoring their versatility as therapeutic platforms. For gastric applications, pH-responsive hydrogels have been designed to address perforations, whereas thermosensitive systems have been applied to esophageal strictures as resistant dressings. This selective and tailorable use highlights the distinct advantages of polymeric hydrogels over conventional injection agents and points to new directions for functional hydrogel development. Advances in therapeutic endoscopy are now shaping hydrogel designs that balance efficacy and procedural practicality by optimizing physicochemical, biological, and injectable properties. Although challenges related to safety, cost, and practicality remain, the preclinical evidence outlined in this review provides a strong foundation and promising direction toward the integration of injectable hydrogels into clinical endoscopy.