Modular stimuli-responsive hydrogel sealants for early gastrointestinal leak detection and containment

Gastrointestinal anastomotic leaks occur in 4–21% of abdominal surgeries and are associated with high morbidity and mortality (up to ~50% in septic peritonitis), prolonged hospital stays, and high costs. Early detection remains challenging because current approaches monitor nonspecific, late-arising clinical signs or surrogate biomarkers. Existing drains or stomas mitigate risk but do not provide early, site-specific monitoring. Many surgical sealants and adhesives have limited efficacy in the gastrointestinal tract due to digestion and lack integrated monitoring capabilities. The study proposes a modular, intelligent hydrogel suture-support sealant patch that both seals and enables early, unambiguous detection of impending gastrointestinal leaks at the anastomotic site via ultrasound-readable stimuli-responsive sensing elements. The central hypothesis is that a robust, tissue-anchored, layered hydrogel with integrated pH- or enzyme-responsive echogenic elements can provide early, point-of-need detection and containment of leaks across multiple GI tissues.

Prior suture supports (e.g., fibrin-based Tachosil) often underperform in GI applications due to susceptibility to digestion. Recent adhesive technologies (e.g., NHS-activated, catechol/dopamine-coordinated, layered/folded sealants) emphasize adhesion and biocompatibility but are rarely tested against active digestive effluents and lack monitoring capabilities. Early attempts at suture monitoring without sealing rely on implanted electronics and RF setups, hindering translation. Gas vesicles have been explored as stable ultrasonic reporters, and various antimicrobial additives (e.g., ZnO nanoparticles) are known for antibacterial action. These gaps motivate an integrated sealing-and-sensing approach operable under harsh GI conditions and readable by portable ultrasound.

Study design: Develop a layered hydrogel sealant patch integrating (i) a non-adhesive PNHEA backing to minimize postsurgical adhesions, (ii) a polyanionic PAMPS adhesive/support layer to contact tissue and host sensing/therapeutic elements, and (iii) a PNAGA-based mutually interpenetrating network (mIPN) formed in situ to robustly anchor the patch to GI tissues under digestive conditions. Two sensing modalities were engineered: enzyme-responsive TurnOFF gas vesicles (GVs) embedded in soft polyacrylamide (PAAm) and pH-responsive TurnON sodium bicarbonate embedded in agar to generate CO2 bubbles under gastric pH. Optional therapeutic elements (e.g., ZnO nanoparticles or gentamycin) were included in PAAm within the PAMPS layer for antimicrobial function. Materials and hydrogel fabrication: Monomers (NHEA, AMPS/PAMPS, AAm), crosslinker (mBAA), photoinitiator (Irgacure 2959), and PNAGA precursors were prepared. Sensing elements were patterned as 5 µL PAAm drops containing 20 vol% GVs (Halo GVs or Ana) for TurnOFF; antibacterial elements (e.g., ZnO at 2.5 mg/mL) were similarly patterned (12.5 µL). Layers were polymerized sequentially in Teflon molds (UV-visible lamp with H2 filter). For TurnON elements, 2 wt% agar with 2 wt% NaHCO3 was gelled, cored (8 mm), sliced to 0.2 cm disks, and embedded into a prepolymerized PAMPS layer; fused by casting and curing a PNHEA backing mix. mIPN anchoring: Prior to application, premade layered gels were soaked 10 min in a PNAGA precursor solution (33 wt% NAGA with LAP initiator). The tissue serosa received a measured aliquot of the precursor; the gel was applied and irradiated with visible/near-UV (365 nm) for 5 min to form a PNAGA mIPN traversing patch and superficial tissue, anchoring the construct without additional crosslinkers or UV exposure harmful to tissue. Biological fluids: Simulated intestinal fluid (SIF; USP protocol, pancreatin, pH 6.8) and simulated gastric fluid (SGF; 35 mM NaCl, pH 2.0) were prepared; bile and PBS were also used. Characterization:

• Adhesion and mechanics: T-peel and lap-shear (ASTM F2256-05, F2255-05) on porcine stomach, small intestine, and colon quantifying adhesion energies; tensile tests (Zwick/Roell) on individual layers and assembled patches pre- and post-mIPN; rheology (oscillatory sweeps).
• Structural/chemical analyses: SEM/EDXS (layer interfaces; ZnO localization), FTIR (PNAGA signatures within PAMPS), and confocal Raman spectromicroscopy with k-means clustering on histological sections to map PNAGA penetration into tissue and confirm layered architecture.
• Swelling and stability: Swelling ratios of individual layers and assembled patches in SIF, SGF, bile, PBS over time; effect of PNAGA mIPN on PAMPS swelling and overall stiffness.
• Antimicrobial/therapeutic release: Zn2+ release kinetics via ICP-OES after acid/peroxide digestion of aliquots from incubations in biological fluids; antibacterial activity against E. coli via growth inhibition assays; optional gentamycin incorporation.
• Cytocompatibility: Perfused-medium assays using NHDF fibroblasts with LDH and ATP-based viability measurements after exposure to hydrogel-conditioned media (multiple perfusion iterations).
• Leak sealing model: Cup-based anastomotic leak assay on porcine intestine sheets with 4 mm defects; patches applied via mIPN; leakage of SIF/SGF/PBS quantified gravimetrically over time at 37 °C and humidified conditions with shaking.
• Adhesion under digestive challenge: Custom 3D-printed molds to immerse patch-sealed tissues in digestive fluids (SIF, SGF, bile, PBS) with time-resolved T-peel after incubation to quantify adhesion retention.
• Ultrasound sensing: Ex vivo tests on patch-sealed tubular intestinal segments with or without 4 mm perforations filled with SIF (TurnOFF) or SGF (TurnON), imaged at intervals using a handheld Clarius L7 HD probe (tablet/smartphone controlled); phantom studies to benchmark echogenicity.
• In vivo feasibility: Piglet model with a 4 mm small-intestine defect sealed by a TurnOFF circular patch and a contralateral reference patch on intact intestine; abdomen closed and patches imaged immediately and after 2 h using handheld ultrasound; post-mortem visual inspection and histology to assess attachment and tissue response.

• Robust, tissue-agnostic adhesion via PNAGA mIPN: T-peel adhesion energies comparable to first-generation mIPN adhesives (observed 124 ± 21 vs. 92 ± 24 J/m²) with higher adhesion on colon and stomach (134 J/m² and 229 J/m², respectively). Minimal adhesion from the PNHEA backing (non-adhesive side), supporting anti-adhesion design.
• Mechanical performance and stabilization by mIPN: PAMPS support layer was highly extensible and soft (strain 967 ± 56%, stress 169 ± 12 Pa). Assembled pre-application double-layer patch: strain 551 ± 78%, stress 554 ± 71 Pa. After mIPN anchoring, tensile strength increased markedly (patch with mIPN: 4515 ± 1033 Pa) approaching PNAGA mIPN alone (5245 ± 210 Pa). High tensile strength largely retained after 24 h incubation in digestive fluids.
• Confirmed interpenetration and anchoring depth: FTIR showed PNAGA signatures within PAMPS (decrease at 1039 and 1183 cm⁻¹; increased amide peaks). Raman mapping demonstrated PNAGA penetration into serosa with depths: stomach 64 ± 4 µm > colon 49 ± 10 µm > small intestine 35 ± 4 µm; PAMPS/PNHEA were not detected in tissue, confirming selective anchoring network.
• Swelling control and porosity: PNAGA mIPN reduced PAMPS superabsorbent swelling to non-unitary levels, stabilizing the structure while allowing PAMPS to absorb fluids and permit interaction of effluents with embedded sensing/therapeutic elements.
• Antimicrobial function and stimuli-responsive release: ZnO nanoparticles embedded in PAAm within PAMPS enabled Zn²⁺ release dependent on contacting fluid; extracts inhibited E. coli growth relative to controls (qualitative data reported). Gentamycin integration was demonstrated as an alternative antimicrobial.
• Adhesion under digestive challenge and burst resistance: After immersion, adhesion retained at least ~50% of initial value after 8 h in enzymatically active SIF; under SGF (pH 2.0), adhesion remained close to initial levels. Burst pressures in SGF and SIF were similar and exceeded native maximal intestinal burst pressure.
• Ultrasound leak detection performance: • Enzyme-responsive TurnOFF (GVs in PAAm, intestinal leaks): Loss of echogenic pattern detectable ex vivo by 3 h post-exposure to SIF; progressive disappearance by 6 h; patches remained attached and contained the leak. • pH-responsive TurnON (NaHCO₃ in agar, gastric conditions): Clear increase in ultrasound scattering within 15 min of SGF contact ex vivo.
• In vivo feasibility (piglet): TurnOFF patch on a 4 mm small-intestine defect showed initial echogenic pattern that disappeared at 2 h, while the contralateral reference on intact tissue remained detectable. Upon reopening, patches were firmly attached and contained the leak; histology showed firm serosal attachment without detectable tissue damage.
• Biocompatibility: Perfused-medium cytotoxicity assays with NHDF fibroblasts indicated high cytocompatibility across patch variants (qualitative, no adverse effects reported).

The study addresses the critical need for early, unambiguous detection of gastrointestinal anastomotic leaks by combining robust sealing with embedded, ultrasound-readable sensing. The PNAGA mIPN provides strong, tissue-compatible anchoring across stomach, small intestine, and colon, preserving adhesion and mechanical integrity under harsh digestive conditions where many adhesives fail. The layered architecture enables the PAMPS support layer to become porous upon fluid contact, facilitating interaction of effluents with sensing and therapeutic elements while the PNHEA backing limits unwanted adhesions. The enzyme-digestible gas vesicle TurnOFF modality and the pH-triggered bicarbonate/agar TurnON modality yield contrasting echogenic signatures that can disambiguate leak types and locations: intestinal-fluid exposure results in loss of signal within hours, whereas gastric acid contact generates rapid echogenicity within minutes. These contrast changes occur prior to overt leak symptoms, enabling proactive monitoring using portable ultrasound without complex electronics. In vivo piglet data confirm feasibility, leak containment, and tissue compatibility, supporting translational potential for perioperative monitoring and timely intervention.

This work introduces a modular, layered hydrogel suture-support patch that integrates a robust PNAGA mIPN anchoring mechanism with stimuli-responsive ultrasound sensing to enable early detection and containment of gastrointestinal leaks. The system achieves strong adhesion on multiple GI tissues, mechanical stability in digestive environments, and rapid, distinguishable echogenic responses (3 h TurnOFF in intestinal conditions; 15 min TurnON in gastric conditions). Optional antimicrobial incorporation further expands functionality. The technology suggests a path toward radiation-free, point-of-need, drain-free postoperative monitoring using handheld ultrasound. Future work should include extended in vivo studies, optimization of long-term biocompatibility and biodegradation/fate, evaluation of chronic performance under leak conditions, and exploration of synergistic therapeutic release triggered by leak cues or external stimuli.

The study primarily demonstrates ex vivo performance with a proof-of-concept in vivo piglet experiment of limited duration. Long-term biocompatibility, biodegradation, and fate of materials and sensing elements were not fully characterized. Detailed quantitative clinical-scale performance metrics (e.g., prolonged implantation, chronic motion, and host responses) and multi-day in vivo leak detection sensitivity/specificity remain to be established. Performance under diverse physiological and pathological conditions (e.g., variable enzyme activities, pH fluctuations, bile exposure) warrants further investigation.