Injectable hydrogel electrodes as conduction highways to restore native pacing

Ventricular arrhythmias are the leading cause of sudden cardiac death. In scarred or diseased myocardium, slowed electrical conduction and spatiotemporal heterogeneities promote wave break, re-entry, and fibrillation. Current management—antiarrhythmic drugs, ablation, and implantable defibrillators—does not correct the underpinning mechanism of re-entry (delayed conduction), carries toxicity or high failure/recurrence rates, and can reduce quality of life due to painful shocks. The study addresses this unmet need by proposing an injectable, conductive hydrogel electrode that can be delivered via coronary veins to access the mid-myocardium, enabling simultaneous multi-site pacing along a linear path to normalize activation and potentially prevent re-entry. The hypothesis is that transforming coronary veins and tributaries into flexible electrodes will generate planar wave propagation across wide myocardial areas, reducing activation delays in heterogeneous tissue and restoring near-native conduction.

The authors contextualize the work by noting limitations of current pacing and ablation strategies with reported recurrence rates of 18–40% after ablation. Antiarrhythmics can be pro-arrhythmic by further slowing conduction. There is no existing modality that directly addresses delayed conduction underlying re-entry. Prior hydrogels and conductive polymers have been explored primarily as scaffolds for myocardial regeneration rather than as injectable electrodes; many conductive polymers face challenges of insolubility, toxicity, or processing barriers for endovascular delivery. Myocardial conductivity typically ranges from approximately 0.1–6.0 mS/cm, setting a target for electrode materials. Clinical experience suggests venous occlusion in the coronary sinus branches (as with CRT leads) is well tolerated, supporting the safety of venous-based approaches. Multi-electrode and planar-wave strategies have shown that distributing low-energy pulses across larger areas can terminate fibrillation below pain thresholds, including demonstrations with epicardial line electrodes producing planar activation. These data motivate the development of an endovascular, mid-myocardial, multi-site pacing approach for prevention of re-entry and potential painless defibrillation.

Materials and hydrogel synthesis: A biostable, conductive hydrogel system was engineered using a polyethylene glycol-based macromer, polyether urethane diacrylamide (PEUDAm), synthesized via a three-step process (PEG-CDI activation, conversion to PEG-diamine, and end-functionalization with acrylamide). Functionalization was confirmed by 1H NMR. N-acryloyl glycinamide (NAGA) served as a small-molecule crosslinker providing bidentate hydrogen bonding. Rapid in situ curing without external stimuli was achieved via redox initiation using ammonium persulfate (APS) and iron gluconate (IG). A double-barrel syringe with a mixing head (analogous to a dual-lumen catheter) delivered the two-component precursors, achieving target gelation under 1 minute and full network formation under 2 minutes at APS ≥0.75 mM. Ionic conductivity was imparted by including salts (1.35% NaCl) in the precursor solutions.

Hydrogel formulation and testing: Typical hydrogel contained PEUDAm (20 wt%), NAGA (1 wt%), APS (0.75 mM) and IG (1.5 mM). Conductivity was quantified from impedance spectroscopy (1 to 10^? Hz sweep; plateau region used to estimate resistance), using σ = L/(R×A). The ionic hydrogel achieved 13.5 ± 0.8 mS/cm, exceeding myocardial values; a water gel control was <0.1 mS/cm.

In vivo feasibility and deployment: Yorkshire pigs (~40–50 kg) were anesthetized and instrumented. Initial deployment used open-chest access. Hydrogel was injected into the anterior interventricular vein (AIV) and, in a subset, the middle cardiac vein (MCV), where it cured in situ. Postmortem, hydrogel was excised and segmented along its length to assess uniformity via gel fraction (NMR-based sol fraction analysis after D2O extraction) and equilibrium swelling ratio (mass swelling Q = W2/WD). Homogeneity along the AIV was assessed (n=3 animals; 5 segments/animal). Safety was monitored via echocardiography (ejection fraction), serial 12-lead ECGs, high-sensitivity troponin, necropsy, and histology (H&E, Masson’s Trichrome/Pentachrome) at 2 weeks (MCV, n=1) and 4 weeks (AIV, n=3).

Pacing protocols and thresholds: Epicardial pacing was performed using (i) a bare metal point electrode, (ii) a hydrogel point source (approximately 8 mm gel “disk” cured on a wire), and (iii) a hydrogel line source (cylindrical gel, 1.6 mm diameter by 8 cm length). After epicardial testing, the hydrogel was injected into the AIV, allowed to cure ~2 minutes, a pacing lead was attached to the proximal gel, and thresholds were re-measured. Capture thresholds (minimum current for reliable capture) were recorded at pulse widths of 0.5, 1, 5, and 10 ms, generating strength-duration curves. Unipolar pacing at 5 ms resulted in thresholds of 2.2 ± 1.2 mA (metal), 2.3 ± 0.6 mA (hydrogel point), 1.5 ± 1.28 mA (hydrogel line), and 2.7 ± 1.8 mA (AIV hydrogel). Two-way ANOVA indicated significant differences only for pulse widths ≤1 ms. Surface 12-lead ECGs were analyzed to compare paced QRS morphologies to native sinus rhythm.

Electroanatomical mapping in an ablation model: To model heterogeneous conduction, epicardial radiofrequency ablation was performed near the AIV to create lesions. Electroanatomical mapping (EnSite Precision) was performed at baseline, during point pacing post-ablation, and during hydrogel pacing post-ablation on epicardial and endocardial surfaces. Activation maps (0–150 ms) were used to compare capture area, activation timing, and wavefront uniformity between point pacing and hydrogel-in-vein pacing.

Ethics and animal care: All procedures were approved by the Texas Heart Institute IACUC (Protocols 2021-08 and 2022-01). Anesthetic and analgesic regimens, monitoring, and perioperative care are detailed. Sample sizes: AIV safety (n=3, 4 weeks), MCV safety (n=1, 2 weeks); acute pacing and mapping studies included n=3 animals, with supplemental replicates for mapping (Supplementary Figs. 30–32).

• Material performance: The ionic PEUDAm/NAGA hydrogel cured in situ in under 2 minutes via APS/IG redox initiation and achieved conductivity of 13.5 ± 0.8 mS/cm, exceeding myocardial conductivity (0.1–6.0 mS/cm).
• In vivo deployment and uniformity: Hydrogel precursors injected into the AIV (and MCV) cured and were retained within veins and tributaries. Postmortem analysis showed uniform gel fraction and swelling ratio across hydrogel segments with no significant segmental differences (two-way ANOVA, p>0.05).
• Pacing thresholds: At 5 ms pulse width, unipolar capture thresholds were comparable across modalities: metal electrode 2.2 ± 1.2 mA; hydrogel point 2.3 ± 0.6 mA; hydrogel line 1.5 ± 1.28 mA; hydrogel in AIV 2.7 ± 1.8 mA. Significant differences were observed only at pulse widths ≤1 ms (two-way ANOVA).
• Conduction phenotype: Epicardial point and line pacing inverted QRS morphology, indicating non-physiologic activation. In contrast, pacing through hydrogel in the AIV yielded QRS morphologies closely matching sinus rhythm (upright in leads I, II, aVL; biphasic in III, aVF; inverted in aVR) with a short isoelectric interval prior to activation, consistent with capture of deep septal bundle branches and Purkinje fibers.
• Mapping in ablation model: Post-ablation, point pacing produced delayed, heterogeneous focal activation. Hydrogel-in-AIV pacing generated broader early activation that reached mid-myocardium and endocardium earlier than point pacing, producing long linear regions of capture and a planar wavefront that normalized conduction across heterogeneous tissue.
• Safety: Echocardiography showed preserved ejection fraction without regional wall motion abnormalities at 4 weeks (AIV) and 2 weeks (MCV). High-sensitivity troponin levels remained within similar ranges pre- and post-implantation (e.g., pre 17–121 ng/L; post 45–54 ng/L at 4 weeks AIV; 28–97 ng/L at 2 weeks MCV), indicating no myocardial necrosis or ischemia. Lungs were free of emboli on necropsy. Histology showed mild perivascular/interstitial fibrosis, chronic epicarditis near injection/incision sites, and foreign body giant cell reaction within gel-containing vessels with focal replacement fibrosis immediately surrounding those vessels, but no myocardial necrosis.

The findings support the central hypothesis that transforming coronary veins into linear, conductive hydrogel electrodes enables planar, multi-site activation resembling native conduction, particularly by engaging mid-myocardial pathways (bundle branches/Purkinje network). The hydrogel’s higher conductivity relative to tissue creates a preferential conduction pathway—a "conduction highway"—facilitating rapid, broad activation across heterogeneous myocardium. In an ablation-induced heterogeneity model, hydrogel pacing minimized delayed conduction observed with point pacing, expanding early activation zones to reduce the substrate for re-entry. Comparable pacing thresholds to conventional electrodes suggest feasibility without excessive energy demands. Preservation of cardiac function and stable cardiac enzyme levels, combined with histological findings limited to local foreign body responses and mild fibrosis, indicate subacute safety of venous occlusion with the hydrogel. The approach integrates with existing clinical workflows (transvenous delivery, pacemaker interfaces) and could enable personalized, mid-myocardial pacing strategies to prevent re-entry and potentially deliver painless defibrillation through distributed, low-energy planar wavefronts.

This study demonstrates the first successful design and in vivo deployment of an injectable, biostable, conductive hydrogel electrode that converts coronary veins into flexible electrodes capable of capturing mid-myocardial tissue and restoring near-native activation patterns. The hydrogel achieves rapid in situ curing, high ionic conductivity, and uniform formation, enabling multi-site, planar pacing that normalizes conduction across heterogeneous regions and may prevent re-entrant arrhythmias. Electroanatomical mapping in an ablation model confirmed earlier and broader activation with hydrogel pacing versus point pacing. Subacute safety data show preserved cardiac function, stable biomarkers, and no embolic events. Future work will focus on: (i) developing and validating a dual-lumen transvenous catheter for clinical delivery; (ii) chronic large-animal pacing studies with implantable pacemakers; (iii) comprehensive assessment in myocardial infarct models with heightened arrhythmia risk; (iv) long-term biocompatibility, vessel remodeling, and optimization of hydrogel fill ratios; and (v) exploration of multi-stage, low-energy pacing paradigms for painless ventricular defibrillation.

• Disease model: The primary conduction heterogeneity model used epicardial ablation; effects in true infarct scar were not yet evaluated (myocardial infarct model under development).
• Delivery method: Initial studies used open-chest, epicardial vessel access rather than the intended transvenous catheter approach; surgical trauma likely contributed to localized inflammation.
• Sample size and duration: Small cohorts (AIV n=3 at 4 weeks; MCV n=1 at 2 weeks) limit generalizability; chronic pacing performance and long-term host response remain to be established.
• Histopathology: Chronic inflammation with foreign body giant cell reaction and focal replacement fibrosis around gel-containing vessels were observed; origins and long-term stability require further study.
• Device availability: No commercial catheter currently exists for in situ-curing hydrogel delivery; a dual-lumen catheter is in development.
• Electrophysiology nuances: QT intervals were not always identical to baseline during AIV hydrogel pacing; detailed T-wave analyses were not the study focus.
• Biological variables: Pilot data were not powered to assess sex as a biological variable.