A bioadhesive pacing lead for atraumatic cardiac monitoring and stimulation in rodent and porcine models

Abstract

Existing clinically-adopted bioelectronic implants mostly rely on surgical suturing or insertion of electrodes to the target tissue for diagnostic and therapeutic applications. However, these approaches can cause tissue trauma during the application and/or retrieval of the bioelectronic implants, potentially causing detrimental complications such as bleeding, tissue damage, and/or device failure. Here, we report a bioadhesive pacing lead for atraumatic epicardial monitoring and stimulation of the heart in vivo to overcome the limitations of existing bioelectronic implants. We use multi-material 3D printing to fabricate the bioadhesive pacing lead, which offers rapid atraumatic application and retrieval as well as robust mechanical and electrical interfacing with the heart. We systematically validate the mechanical and electrical properties, biocompatibility, continuous epicardial monitoring and pacing capability, and rapid on-demand atraumatic application and removal of the bioadhesive pacing lead based on in vivo rat and porcine models. These findings may offer a promising platform for atraumatic bioelectronic diagnosis and treatment and inspire the future development of bioadhesive electronics.

One Sentence Summary:

In this paper, we report a novel bioadhesive pacing lead for atraumatic implantation and on-demand retrieval, and continuous cardiac monitoring and stimulation to address limitations of the clinically-adopted cardiac pacing leads while readily compatible with the standard clinical setup.

INTRODUCTION

Electrophysiological sensing and stimulation require an intimate bioelectronic interface between the implant and the target organ (1–5). Existing bioelectronic implants are usually surgically fixed to the target tissue by suturing or inserting the electrodes (6–11). However, the traumatic and invasive nature of existing approaches can occasionally lead to undesirable outcomes such as bleeding and tissue damage during implantation and removal of the devices in clinical practice (12, 13). This can potentially compromise the reliability and accuracy of electrophysiological functionality and lead to scarring, cardiac chamber perforation, and associated life-threatening complications (Fig. 1A). In particular, conventional methods can impose substantial risk in crucial organs such as the heart with various adverse outcomes (14, 15) including device failure (16, 17), heart chamber perforation (particularly on the atrium) (13), hemorrhage (18), pericardial effusion, and cardiac tamponade (12). Hence, bioelectronic implants capable of atraumatic interfacing with organs are highly desirable yet remain an unmet need in the field.

Fig. 1. Concept of a bioadhesive pacing lead.

(A) The placement of bioelectronic devices on tissues based on surgical suture as a supplemental step and the subsequent traumatic retrieval process. (B) Fabrication of the bioadhesive pacing lead with bioadhesive pattern and all-in-one configuration via multi-materials 3D printing (left); One-touch application of the bioadhesive pacing lead on the tissue with a magnified cross-sectional view at the interface (middle); The on-demand atraumatic retrieval process of the bioadhesive pacing lead with a magnified cross-sectional view at the interface (left). (C) The synergetic advantages of the bioadhesive pacing lead.

Tissue adhesives have recently been adopted as an alternative fixation method to facilitate the sutureless integration of implantable bioelectronic devices with tissues (19–23). However, existing tissue adhesives are pre-manufactured separately from bioelectronic devices and applied adjunctively to the targeted devices/tissues during the implantation process (22–24) (fig. S1). The characteristics of this approach can substantially increase the complexity of the implantation process, especially when dynamic and/or non-uniform tissues like the heart are targeted. This can lead to inconsistent integration and unreliable electrical interfacing. Furthermore, the ad hoc application of separately manufactured tissue adhesives may result in low compatibility with various designs of bioelectronic devices and may impose the potential risk of tissue damage or residual materials upon removal of the device (fig. S1).

Here, we introduce a bioadhesive pacing lead to address the limitations of existing devices and methods. The bioadhesive pacing lead takes the form of an all-in-one device capable of atraumatic implantation, continuous cardiac monitoring and pacing, and on-demand atraumatic retrieval in in vivo rat and porcine models. We further develop a multi-material 3D printing method to print all components of the bioadhesive pacing lead including the electrode, insulation, and bioadhesive in an integrated manner (Fig. 1B). The unique design of the bioadhesive in the device allows atraumatic and non-invasive implantation as well as on-demand retrieval of the device from the epicardial surface. Importantly, the bioadhesive pacing lead is compatible with existing clinical equipment and protocols for cardiac pacing (18, 25). Furthermore, the high charge injection capacity (CIC, over 420 µC cm^-2) of the bioadhesive pacing lead provides stable electrophysiological performance with higher R wave amplitude for cardiac monitoring and lower capture threshold for cardiac pacing compared to commercially-available epicardial pacing leads throughout the clinically-relevant period for temporary epicardial pacing (2 weeks). The bioadhesive pacing lead may offer a promising concept and platform technology not only for bioelectronic diagnosis and treatment of cardiac patients, but also for the future development of next-generation implantable electronics and their applications (Fig. 1C).

RESULTS

Design of bioadhesive pacing lead

Implantable temporary epicardial pacing leads have been clinically adopted to protect patients from the risk of postoperative arrhythmias including complete heart blocks during the recovery period after cardiac surgery. These pacing leads usually remain in place for 1–2 weeks (15, 18, 25, 26). After 1–2 weeks, the temporary pacing leads are required to be removed — typically by pulling on the external portion of the leads (18, 25, 27) (Fig. 2A). This forceful removal puts the patients at risk of aforementioned bleeding complications. To avoid such complications, the proposed bioadhesive pacing lead consists of a bioadhesive interface for mechanical and electrical integration with epicardium and a built-in reservoir for on-demand detachment of the pacing lead. An electrode wire and a fluidic catheter are connected to the bioadhesive interface and the reservoir, respectively. Notably, the electrode wire can be connected to the standard cardiac pacing system, allowing the proposed design to be readily incorporated into the existing clinical setups. On top of the bioadhesive pacing lead, minimally-invasive implantation tools (a balloon catheter, an adapter, and a sheath catheter) can be additionally combined to allow minimally-invasive implantation. (Fig. 2B).

Fig. 2. Design and fabrication of the bioadhesive pacing lead.

(A) The typical clinical application scenario (implantation, on-demand pacing, and removal) of a temporary epicardial pacing lead. (B) Schematic illustrations for the bioadhesive pacing lead and minimally-invasive implantation tools. (C) Schematic illustration for the implantation, epicardial pacing, and on-demand retrieval process of the bioadhesive pacing lead. (D) Representative images of nonconductive bioadhesive and conductive bioadhesive inks, 3D printing nonconductive bioadhesive and conductive bioadhesive, respectively. (E) Apparent viscosity as a function of shear rate for the nonconductive bioadhesive and conductive bioadhesive inks, respectively. (F) Shear storage modulus (G’) as a function of shear stress for the nonconductive bioadhesive and conductive bioadhesive inks, respectively. (G) Sequential snapshots for 3D printing process of the bioadhesive pacing lead. Different color dyes (green for sacrificial ink; red for insulation ink) are mixed with the transparent inks to improve visualization in the snapshots. (H) Representative images of the folded and inflated bioadhesive pacing lead with the minimally invasive implantation tools.

For implantation, the folded bioadhesive pacing lead can be inserted through the sheath catheter and delivered to the epicardium. Then inflation of the balloon catheter unfolds the pacing lead, and subsequently the adapter provides a gentle pressure to facilitate the adhesion. The bioadhesive interface allows rapid, robust, and atraumatic integration of the device to the epicardial surface by temporally absorbing the water on the tissue surface and forming both physical and chemical crosslinks (19, 21) (fig. S2). After implantation, the bioadhesive interface allows targeted bi-directional electrical communication for electrocardiographic monitoring and continuous cardiac pacing. For on-demand removal of the bioadhesive pacing lead, a biocompatible detachment solution can be injected into the built-in reservoir to activate fast cleavage of physical and chemical crosslinks between the bioadhesive interface and the underlying epicardial tissue for atraumatic detachment (Fig. 2C).

Fabrication of bioadhesive pacing lead

To allow flexible, customized, and streamlined fabrication of the proposed bioadhesive pacing lead, we use a multi-material 3D printing method to print all components of the device including the electrode, insulation, and bioadhesive (fig. S3 and movie S1). To develop the 3D printable bioadhesive, polyacrylic acid is first grafted on a hydrophilic polyurethane. These are two widely-used polymers in biomaterials and medical devices without allergic reactions (28–33). The bioadhesive polymer is then dissolved in 70% ethanol solution and mixed with 1-ethyl-3-(-3-dimethylaminopropyl) carbodiimide (EDC) and N-hydroxysulfosuccinimide sodium salt (sulfo-NHS) to obtain the 3D printable nonconductive bioadhesive ink (Fig. 2D and fig. S4). In the grafting step, a functional monomer with disulfide bond can be added to provide the on-demand detachment capability to the bioadhesive polymer (figs. S4 and S5). Notably, the conductive bioadhesive ink can be prepared by mixing the nonconductive bioadhesive ink with an aqueous solution of a conducting polymer (5 % w/w poly(3,4-ethylenedioxythiophene):poly(styrenesulfonate) (PEDOT:PSS)); see Methods for detailed procedure). Rheological characterization of the nonconductive and conductive bioadhesive inks shows shear-thinning (Fig. 2E) and shear-yielding (Fig. 2F) properties, which allow the fabrication of bioadhesive electronics via direct-ink-writing 3D printing(34, 35). Thereafter, the printed bioadhesive inks can be converted into the bioadhesive interface by evaporating solvents (ethanol and water) in the inks (Fig. 2G). SEM images show homogeneous distributions of PEDOT:PSS and 3D printable bioadhesive polymer. NMR spectroscopy shows representative peaks of PEDOT:PSS and printed bioadhesive polymer (fig. S6). After connecting a lead wire to the bioadhesive interface, a built-in reservoir made of electrically insulating polyurethane is printed on the top of the bioadhesive interface and connected with a fluidic tube (Fig. 2H). Due to the polymer chain entanglements in the printing and evaporating process, each component can be robustly bonded with each other in the printed bioadhesive pacing lead.

Mechanical properties and adhesive performance

Upon contacting wet epicardial tissue surfaces, the bioadhesive interface absorbs the interfacial water and subsequently forms physical (hydrogen bonds, electrostatic interactions) and covalent (amide bonds) crosslinks with the tissue surface to establish rapid and robust adhesion within 10 s (fig. S2). The fully swollen bioadhesive interface shows tissue-like softness (Young’s modulus of 156 kPa for the nonconductive bioadhesive and 437 kPa for the conductive bioadhesive, respectively), stretchability (>1.8 times), and high toughness (fracture toughness of 156 J m^-2 for the nonconductive bioadhesive and 437 J m^-2 for the conductive bioadhesive, respectively) (Fig. 3A, figs. S7, and S8). While slightly different in mechanical properties, the polymer chain entanglements between the nonconductive and conductive bioadhesive can ensure the integrity of the bioadhesive interface. As a result, the bioadhesive interface can provide highly conformal and mechanically imperceptible integration to the underlying tissue (Fig. 3B) while providing robust adhesion sufficient to lift a whole porcine heart (Fig. 3C).

Fig. 3. Mechanical properties and adhesive performance.

(A) Nominal stress vs. stretch curve of the nonconductive bioadhesive and conductive bioadhesive fully swollen in PBS, respectively. The slope in the curve is used to calculate the Young’s modulus of fully swollen nonconductive bioadhesive and conductive bioadhesive. F indicates tensile force applied to the sample. (B) Nominal stress vs. stretch curve of the cardiac tissue with and without the adhered bioadhesive pacing lead, respectively (C) Photograph of an ex vivo porcine heart lifted by the adhered bioadhesive pacing lead. (D) The bioadhesive pacing lead robustly integrates to (top) and atraumatically detaches from (bottom) an ex vivo porcine heart, respectively. (E) Interfacial toughness (left) and shear strength (right) between heart tissues and polyurethane substrates adhered by the nonconductive bioadhesive or conductive bioadhesive, 5 min after applying PBS or the detachment solution. (F) Fluorescent microscope images for validation of on-demand detachment based on fluorescent primary amine-coupled microbeads. The bioadhesive interface in the initial state (top) and 10 min after incubation in the detachment solution. Values in (E) represent the mean and the standard deviation (n = 3 independent samples).

The bioadhesive interface can further offer on-demand and atraumatic removal of the bioadhesive pacing lead to avoid potential complications from traumatic removal processes. A biocompatible aqueous detachment solution (50 mM sodium bicarbonate and 50 mM L-glutathione reduced in phosphate buffered saline (PBS)) can be delivered through the fluidic channels to the built-in reservoir to trigger on-demand detachment (36) (Fig. 3D). The detachment solution can diffuse through the bioadhesive interface within 5 min (fig. S9) and subsequently cleave both physical and covalent crosslinks with the underlying tissue (fig. S5). This on-demand detachment process provides substantially decreased interfacial toughness (11 J m^-2 for the nonconductive bioadhesive; 6 J m^-2 for the conductive bioadhesive) and shear strength (1.6 kPa for the nonconductive bioadhesive; 1.2 kPa for the conductive bioadhesive) of the adhered bioadhesive. In contrast, high interfacial toughness (309 J m^-2 for the nonconductive bioadhesive; 220 J m^-2 for the conductive bioadhesive) and shear strength (49 kPa for the nonconductive bioadhesive; 36 kPa for the conductive bioadhesive) of the adhered bioadhesive are maintained when PBS is used instead of the detachment solution (Fig. 3, E and F, and fig. S10).

Electrical properties and pacing performance

To achieve effective sensing and stimulation in vivo, the bioadhesive pacing lead should provide high performance and stable electrical properties in wet physiological environments. CIC and electrical impedance are two key parameters that are associated with cardiac pacing and monitoring performance in clinic applications (26). The CIC of the bioadhesive pacing lead can reach ~420 μC cm^-2 which is substantially higher than that of a commercially-available temporary cardiac pacing lead (Medtronic 6500, 150 μC cm^-2) (Fig. 4, A and B). A cyclic CIC measurement validates the electrochemical stability of the bioadhesive pacing lead, showing a stable CIC value after 2 million charging and discharging cycles (Fig. 4c). We further evaluate the stability of CIC (Fig. 4D), impedance (Fig. 4E), and conductivity (fig. S11) for 2 weeks incubation in PBS at 37°C, which show no significant statistical difference in CIC (P = 0.89), impedance (P = 0.94) and conductivity (P = 0.82) during the incubation period.

Fig. 4. Electrical properties.

(A) Schematic diagram of a programmed stimulation waveform for charge injection capacity measurement. (B) Voltage transients acquired by the bioadhesive pacing lead and Medtronic 6500 in PBS, respectively. Geometric surface area, GSA; current injection limit, CIL; charge injection capacity, CIC. (C) Charge injection capacity of the bioadhesive pacing lead and Medtronic 6500, respectively. (D) Voltage transients acquired by the bioadhesive pacing lead after 2 million charge-discharge cycles in PBS. (E and F) Charge injection capacity (E) and impedance at 1 kHz (F) of the bioadhesive pacing lead in PBS over 14 days at 37°C, respectively. Charge injection capacity, CIC. Values in (C), (E) and (F) represent the mean and the standard deviation (n = 4 independent samples). Statistical significance and P values are determined by the two-sided unpaired t-test between two groups and one-way ANOVA and Tukey’s multiple comparison test; ns, not significant; **** P ≤ 0.0001.

To visualize the pacing capability of the bioadhesive pacing lead, we apply the bioadhesive pacing lead on the surface of an ex vivo porcine heart that is periodically pressurized to simulate heartbeats (fig. S12 and movie S2). A light-emitting diode (LED) is placed near the bioadhesive pacing lead as an indicator of charge injection. Once the electrical pulse is delivered through the bioadhesive pacing lead to the porcine heart, the LED is periodically lit up following the injected electrical pulse, indicating the successful charge injection to the porcine heart (fig. S12A). We compare the threshold of the current pulse to activate the LED between the bioadhesive pacing lead and a commercially-available epicardial pacing lead under the same experimental conditions. The current pulse is delivered to the epicardium for both of the pacing leads. The bioadhesive pacing lead can activate the LED at a lower electrical pulse (1.5 mA, 100 ms) than that delivered by commercial epicardial pacing lead (3.0 mA, 100 ms), further validating the high CIC of the bioadhesive pacing lead compared to the commercial counterpart (fig. S12, B and C). In addition, the electrical pulse with different frequencies can be effectively delivered without delay by the bioadhesive pacing lead, which further indicates that the bioadhesive pacing lead can meet the clinical needs of different pacing conditions (movie S2).

Atraumatic implantation and retrieval

To evaluate the atraumatic implantation and retrieval capability of the bioadhesive pacing lead, we demonstrate minimally-invasive implantation and retrieval of the bioadhesive pacing lead in an ex vivo porcine model (fig. S13 and movie S1). First, the bioadhesive pacing lead combined with an adapter and a balloon catheter can be inserted through a sheath catheter to approach the epicardium (fig. S13, A and B). The balloon catheter is then inflated to unfold the bioadhesive pacing lead, and the adapter is used to apply gentle pressure to form adhesion on the epicardium (fig. S13, C and D). Subsequently, saline can be injected through the balloon catheter to dissolve the sacrificial layer (polyvinyl alcohol) between the balloon catheter and the bioadhesive pacing lead to remove the balloon catheter. The adhered bioadhesive pacing lead provides robust integration on the epicardium for bi-directional electrical communication (fig. S13, E and F). To achieve atraumatic retrieval, the detachment solution can be injected into the built-in reservoir, and the bioadhesive pacing lead can be removed without tissue damage or residues on the heart. (fig. S13, G and H).

In vivo atraumatic cardiac monitoring and pacing in rat model

The atraumatic implantation, continuous cardiac pacing, and retrieval capabilities of the bioadhesive pacing lead are further evaluated based on an in vivo rat model (Fig. 5). The bioadhesive pacing lead adheres to the ventricle of the rat heart with an electrode wire and a fluidic tube connected to a dorsal subcutaneous port (Fig. 5, A and B). Upon contact with the beating rat heart, the bioadhesive pacing lead can adhere to the epicardium within 10 s under the application of gentle pressure (Fig. 5C). The bioadhesive forms adhesion with extracellular matrix components of epicardium such as collagen with abundant primary amine and carboxylic acid group (fig. S2). After the initial atraumatic implantation, the bioadhesive interface provides stable mechanical integration and bi-directional electrical communication between the bioadhesive pacing lead and the cardiac tissue (movie S3). In contrast, a commercial temporary pacing lead (Medtronic 6500) requires insertion into the myocardium for mechanical fixation and causes tissue trauma and bleeding (Fig. 5D). As control groups, non-adhesive pacing leads are prepared by fully swelling the devices in PBS before implantation and implanted based on sutures or commercial-available polymerizable tissue glue (BioGlue) (fig. S14).

Fig. 5. In vivo atraumatic cardiac monitoring and pacing in rat model.

(A) Schematic illustrations for in vivo continuous epicardial pacing with surface ECG recording in rat model. (B) Representative image of a rat with the implanted bioadhesive pacing lead and subcutaneous port. (C) Representative images of in vivo implantation of the bioadhesive pacing lead to the rat heart and continuous epicardial pacing on day 0 post-implantation. (D) Representative images of in vivo implantation of a commercial-available temporary pacing lead (Medtronic 6500) to the rat heart and continuous cardiac pacing on day 0 post-implantation. (E to G), Representative ECG waveforms before and after ventricular pacing by the bioadhesive pacing lead on day 0 (E), 7 (F), and 14 (G) post-implantations. (H and I) Representative ECG waveforms before and after ventricular pacing by the commercial temporary pacing lead on day 0 (H) and 14 (I) post-implantation. (J) Ventricular capture threshold of the bioadhesive pacing lead and Medtronic 6500 on day 0, 7, and 14 post-implantations, respectively. (K) R wave amplitude recorded by the bioadhesive pacing lead and Medtronic 6500 on day 0, 7, and 14 post-implantations, respectively. (L) Representative images for atraumatic retrieval of the bioadhesive pacing lead from subcutaneous port on day 14 post-implantation. (M) Representative images for retrieval of the commercial temporary pacing lead from the rat heart on day 0 post-implantation. (N and O) Representative histology images stained with hematoxylin and eosin for the bioadhesive pacing lead (N) and the commercial temporary pacing lead (O) implanted to rat hearts on day 14 post-implantation. (P and Q) Representative optical image (P) and histology image (Q) stained with hematoxylin and eosin of a rat heart after the retrieval of implanted bioadhesive pacing lead on day 14 post-implantation. Values in (J) and (K) represent the mean and the standard deviation (n = 4 independent samples). Statistical significance and P values are determined by two-sided unpaired t-test between two groups and one-way ANOVA and Tukey’s multiple comparison test for more than three groups; ns, not significant; * P < 0.05.

Notably, in contrast to the bioadhesive pacing lead, tissue damage from suturing (fig. S14A) and undesirable spreading of unpolymerized glue (fig. S14B) substantially decrease the reproducibility and effectiveness of the implantation process. We further perform in vitro characterizations to evaluate cytotoxicity and biocompatibility. A cell culture media (Dulbecco’s modified eagle medium, DMEM) conditioned with the bioadhesive pacing lead shows comparable in vitro cytotoxicity to Medtronic 6500 and a control (pristine DMEM) after 24 h culture (fig. S15).

To evaluate continuous monitoring and pacing capability in vivo, the cathode and anode electrodes of an external pulse generator are connected to the bioadhesive pacing lead through the dorsal subcutaneous port and left foreleg, respectively (Fig. 5A). Surface electrocardiogram (ECG) signals are recorded simultaneously by subdermal needle electrodes to evaluate the pacing performance. We find that the bioadhesive pacing lead can successfully provide ventricular pacing of the rat heart confirmed by the ventricular pacing spikes preceding each rhythm (i.e., QRS complex in ECG) and the increased heart rate from natural sinus rhythm (~260 beats per minute (bpm)) to overdrive pacing rate (420 bpm) (Fig. 5E and movie S4). Owing to the robust and stable integration of the bioadhesive pacing lead to the epicardial tissue, we further confirm the successful ventricular pacing on day 7 (Fig. 5F) and day 14 (Fig. 5G) post-implantation with the consistent ventricular-paced ECG rhythm. As a control, a commercially-available temporary pacing lead (Medtronic 6500) can also perform ventricular pacing at a higher capture pulse than the bioadhesive pacing lead (Fig. 5H), while it fails to capture on day 7 post-implantation (Fig. 5, I and J). The average capture threshold of the bioadhesive pacing lead (1.3 mA, 0.5 ms) is significantly lower (P = 0.01) than that of the Medtronic 6500 (1.7 mA, 0.5 ms) on day 0 (Fig. 5J), consistent with the higher CIC of the bioadhesive pacing lead than the commercial pacing lead (Fig. 4C). The capture threshold of the bioadhesive pacing lead is stable during 14 days of implantation (P = 0.19) (Fig. 5J).

In contrast, 25% of sutured and 50% of BioGlue-fixed pacing leads fail to capture on day 14 post-implantation (fig. S16, A to F) due to trauma-induced immune response around the lead-tissue interface (fig. S14C) and/or unstable fixation over time (fig. S14D). Moreover, the survived sutured or glued pacing leads show a significantly higher capture threshold than that of the bioadhesive pacing lead (fig. S16G). Furthermore, we evaluate the cardiac monitoring capability by average R wave amplitudes measured by the pacing leads. The average R wave amplitudes monitored by the bioadhesive pacing lead show no significant difference (P = 0.23) over 2 weeks (Fig. 5K and fig. S13). In contrast, the average R wave amplitudes are significantly decreased for Medtronic 6500 (P < 0.0001) (Fig. 5K and fig. S17), BioGlue-fixed pacing lead (P = 0.0004) (fig. S16, H and I), and sutured pacing lead (P = 0.0132) (fig. S16, H and J).

To simulate the clinically-relevant postoperative cardiac condition, we design a myocardial ischemia-reperfusion model by left anterior descending coronary artery ligation (fig. S18, A to D), and evaluate the pacing capability of bioadhesive pacing leads (fig. S18, E to I). ECG waveform featuring ST-segment elevation substantiates the effective induction of myocardial ischemia. Bioadhesive pacing leads show the capability to ventricular pace the rat heart after ischemia-reperfusion, exhibiting a capture threshold comparable to that observed in a healthy rat heart.

For on-demand detachment of the bioadhesive pacing lead, the detachment solution is delivered through the fluidic channel from the dorsal port to the built-in reservoir of the device after 14 days post-implantation. The on-demand and localized delivery of the detachment solution allows the non-invasive and atraumatic removal of the pacing lead (Fig. 5L). In contrast, the standard retrieval process of the pacing lead is based on the mechanical pulling of the pacing lead (15, 18), which can cause tissue trauma and bleeding (Fig. 5M).

The atraumatic characteristics of the bioadhesive pacing lead are further evaluated by histological analysis. Histological assessments by a blinded pathologist indicate that the bioadhesive pacing lead shows minimal to mild inflammatory response on day 14 post-implantation to the rat heart (Fig. 5N). In contrast, the control groups of commercially-available pacing leads (Medtronic 6500) (Fig. 5O), and sutured or BioGlue-fixed pacing leads (fig. S19) exhibit a higher degree of inflammation and scarring of the myocardial tissue, potentially due to the tissue damage during implantation. The immunofluorescence analysis shows the expressions of inflammation-related biomarkers (CD3, CD68, Collagen-I, and α-smooth muscle actin (α-SMA)) for the bioadhesive pacing lead are significantly lower than those for Medtronic 6500 on day 14 post-implantation (fig. S20), which is in agreement with the histological evaluation. Moreover, the macroscopic (Fig. 5P) and histologic (Fig. 5Q) observations of the rat heart do not show signs of residual implants or tissue trauma after on-demand retrieval of the bioadhesive pacing lead.

The blood analyses are performed to evaluate long-term biocompatibility after detachment. Complete blood counts (CBCs) and comprehensive blood chemistry panels show comparable results to those of untreated animals without notable systemic toxicity after 2 weeks of implantation and 1 week after removal of the bioadhesive pacing leads (fig. S21). We further evaluate the systemic immune response via Luminex quantitative analysis of serum samples on day 14 post-implantation for bioadhesive pacing lead and Medtronic 6500 (fig. S22). The pro-inflammatory cytokines (IL-6 and TNF-α) show lower expression for the bioadhesive pacing leads compared to the Medtronic 6500. The serum samples of bioadhesive pacing lead implantation show higher amounts of anti-inflammatory cytokines (IL-13) compared to Medtronic 6500. All the other cytokines are comparable values for both bioadhesive pacing lead and Medtronic 6500 implantation.

In vivo atraumatic cardiac monitoring and pacing in porcine model

To further evaluate the proposed bioadhesive pacing lead in a large animal model, we demonstrate atraumatic implantation, cardiac monitoring, single- and dual-chamber pacing, and retrieval of the bioadhesive pacing lead in a proof-of-concept in vivo porcine model (Fig. 6A). The bioadhesive pacing lead can be robustly adhered to the atrium and ventricle of the porcine heart in vivo by application of gentle pressure for 10 s without causing tissue trauma (Fig. 6B). In contrast, the insertion of Medtronic 6500 or suturing of non-adhesive pacing lead result in myocardial damage and bleeding (Fig. 6C and fig. S23, A to E). Furthermore, the device implantation based on BioGlue exhibits an undesirable spread of the uncured glue to the nearby tissues and the tissue-device interface due to movement and the non-flat surface of the heart (fig. S23, F to J). Notably, the design of the bioadhesive pacing lead allows ready compatibility with a clinical-grade pacing system (GE Mac-Lab hemodynamic recording system and Medtronic 5330 pulse generator) for bi-directional electrical communication, including R-wave amplitude monitoring in real-time and continuous pacing of the porcine heart (Fig. 6, D to G, and movie S6).

Fig. 6. In vivo atraumatic cardiac monitoring and pacing in porcine model.

(A) Schematic illustrations for in vivo cardiac pacing of porcine heart by the bioadhesive pacing lead and a commercially-available temporary pacing lead (Medtronic 6500). (B and C) Representative images for the placement, continuous cardiac pacing, and retrieval of the bioadhesive pacing lead (B) and the commercial temporary pacing lead (C), respectively. (D) Representative natural sinus rhythm with a heart rate of 60 beats per minute (bpm). (E) Schematic illustration and representative ventricular paced ECG waveforms with an overdrive pacing rate of 120 bpm. The capture pulse of the bioadhesive pacing lead is 4 mA on the porcine ventricle. (F) Schematic illustration and representative atrial paced ECG waveform with an overdrive pacing rate of 110 bpm. The capture pulse of the bioadhesive pacing lead is 4 mA on the porcine atrium. (G) Schematic illustration and representative atrioventricular sequential paced ECG waveforms with an overdrive pacing rate of 120 bpm. The capture pulse is 10 mA on the porcine atrium and 4 mA on the porcine ventricle, respectively. (H) R wave amplitude sensed by the bioadhesive pacing lead and the commercial temporary pacing lead from the porcine heart, respectively. (I) Ventricular capture threshold of the bioadhesive pacing lead and the commercial temporary pacing lead, respectively. Values in (H) and (I) represent the mean and the standard deviation (bioadhesive pacing lead: n = 3 independent samples; Medtronic 6500: n = 3 independent measurement). Statistical significance and P values are determined by two-sided unpaired t-test; ns, not significant.

Owing to the robust adhesion (Fig. 3C) and high CIC (Fig. 4C), the electrical pulses generated by the external pulse generator can be effectively delivered to the porcine heart via the bioadhesive pacing lead for atrial, ventricular, and atrioventricular pacing with high R wave amplitude and low capture threshold (Fig. 6, D to H). All paced rhythms are regular and stable, and we observe increased heart rate from natural sinus rhythm (~60 bpm) (Fig. 6D) to an overdrive pacing rate (110-120 bpm) (Fig. 6, E to G). To perform ventricular pacing, we apply the bioadhesive pacing lead on the left ventricular free wall with successful ventricular pacing at a current pulse of 4 mA. A typical ventricular-paced rhythm with a wide QRS complex can be observed and the heart rate increase to 120 bpm, indicating successful ventricular capture. As a control, a commercial temporary pacing lead (Medtronic 6500) is inserted into the myocardial tissue at the same location on the left ventricular free wall to perform ventricular pacing (fig. S16 and movie S7). We compare the R wave amplitude and ventricular capture threshold, which are two key parameters to evaluate the sensing and pacing capability of cardiac pacing leads in clinical applications(26, 37). The average R wave amplitude (6 mA) of the bioadhesive pacing lead is higher than that of the commercial pacing lead (4.7 mA) (P = 0.46). In addition, the average capture threshold (4 mA) of the bioadhesive pacing lead is lower than that of the commercial pacing lead (7.3 mA). For atrial pacing, the ECG waveform shows a typical narrow QRS complex and pacing spike before the non-sinus P wave as well as the increased heart rate (110 bpm), indicating successful atrial capture at 4 mA (Fig. 6G and movie S6). Furthermore, atrioventricular pacing can also be successfully performed by two bioadhesive pacing leads adhering on the atrium and ventricle, validated by the characteristic sequentially paced rhythm with two spikes and wide QRS complex (Fig. 6G).

Taking advantage of the on-demand removal capability of the bioadhesive pacing lead, the adhered bioadhesive pacing lead can be atraumatically retrieved from the porcine heart by injecting the detachment solution to the built-in reservoir of the device (Fig. 6B). This on-demand removal strategy can also be used if the lead needs to be repositioned during the initial application process. After on-demand removal of the bioadhesive pacing lead, the underlying epicardial tissue does not show observable tissue damage, bleeding, or residual implants (movie S6). In contrast, the commercial temporary pacing lead shows obvious myocardial damage and bleeding during both implantation to and retrieval from the porcine heart (Fig. 6C and movie S7).

To continuously evaluate the performance of the bioadhesive pacing lead, the bioadhesive pacing leads are connected to the implantable pacemaker and adhered to the ventricles in a survival porcine model (Fig. 7, A and B). Implantable pacemaker allows for continuous monitoring of the capability of the bioadhesive pacing leads, and can be telemetrically detected and adjusted via a portable programmer (Fig. 7C). The average R wave amplitudes of epicardial ECG, as measured by bioadhesive pacing lead on the ventricle, show no significant difference on day 0 and day 10 post-implantation (Fig. 7D). The stable sensing capability is attributed to the robust adhesion on the ventricular tissues and no obvious inflammatory response on the tissue-device interface over 10 days (Fig. 7, E and F). The portable programmer demonstrates that continuous ECG telemetry remains stable without abnormal events detected during the implantation period, and the epicardial ECG waveforms coincide with the surface ECG waveforms (Fig. 7G). In addition, the bioadhesive pacing leads enable to perform ventricular pacing with a low capture threshold of 4.2 V on day 10 post-implantation. Clear pacing spikes and representative ECG waveforms during ventricular pacing can be real-time recorded via implantable pacemaker and portable programmer (Fig. 7H).

Fig. 7. Continuous and telemetric cardiac monitoring and pacing in a porcine model.

(A) Schematic illustration for continuous and telemetric cardiac monitoring and pacing by the bioadhesive pacing leads connected to an implantable pacemaker and adjusted via a portable programmer. (B) Representative image for the bioadhesive pacing leads connected with implantable pacemaker via IS-1 connector. (C) Representative image for the telemetric monitoring and adjustment via a portable programmer. (D) R wave amplitudes sensed by the bioadhesive pacing lead on day 0 and day 10 post-implantation. The R wave amplitudes were extracted from the epicardial ECG on ventricles and monitored telemetrically via a portable programmer. (E and F) Representative histology images stained with hematoxylin and eosin (E) and Masson’s trichrome (F) for the bioadhesive pacing lead implanted to porcine heart on day 10 post-implantation. (G) Representative surface ECG waveform (top) and epicardial ECG waveform (bottom) were recorded via subdermal electrodes and bioadhesive pacing lead on porcine limbs and ventricle on day 10 post-implantation, respectively. (H) Representative ventricular-paced ECG waveforms on day 10 post-implantation. The pulses (4.2 V, 0.4 ms) were delivered by the implantable pacemaker through the bioadhesive pacing leads and adjusted using the portable programmer. Values in (D) represent the mean and the standard deviation. n = 3 independent samples. Statistical significance and P values are determined by two-sided unpaired t-test; ns, not significant.

DISCUSSION

Implantation of temporary epicardial pacing leads is the standard-of-care for patients undergoing cardiac surgery (15, 18, 38). Temporary epicardial pacing helps to manage perioperative electrophysiological abnormalities such as surgery-related high-degree heart blocks that can result in severe bradycardia and hemodynamic collapse (25). Furthermore, optimal epicardial pacing can improve overall hemodynamics for both adult and pediatric patients in postcardiotomy (39, 40). While various pacing lead technologies have been in development (41), existing devices still rely on the traumatic screw-in or insertion fixation. Existing pacing devices and traumatic methods can cause various undesirable and detrimental complications (42). At implantation, the need for surgical fixation of conventional devices puts patients at risk of hemorrhage, cardiac chamber perforation, and arrhythmia due to myocardial irritation associated with the device fixation. In post-implantation, inflammatory response and fibrosis can cause pacing-threshold elevation and potential loss of sensing or capture capability. During removal of the pacing lead, the traumatic pulling on the external portion of the leads is associated with risks of hemorrhage, pericardial effusion, and tamponade (43–45). Furthermore, these risks are substantially increased for neonatal and pediatric patients considering the small size and thin walls of the heart and blood vessels (17, 46). Hence, an atraumatic epicardial pacing system that does not cause cardiac trauma yet provides stable and effective pacing capability has great potential to improve the safety and efficiency for patients.

In this study, we report a bioadhesive pacing lead for atraumatic implantation, continuous cardiac monitoring and pacing, and on-demand retrieval in vivo. The current study aims to address the limitations of traumatic implantable bioelectronic devices and methods. A synergistic set of design, materials, and fabrication advances in bioadhesive electronics can overcome tissue-traumatic implantation and retrieval of devices, unreliable communication, and associated problems. Specifically, the 3D printed bioadhesive interface provides robust, facile, and straightforward mechanical integration and bi-directional electrical communications with the heart, while offering flexibility in the fabrication of the device that is readily compatible with advanced manufacturing techniques such as multi-material additive manufacturing. The bioadhesive interface provides stable ECG monitoring and a low capture threshold for cardiac pacing superior to commercially-available cardiac pacing leads. The unique all-in-one electro-fluidic design of the bioadhesive pacing lead provides continuous cardiac monitoring, pacing, and atraumatic on-demand retrieval in a minimally-invasive manner in vivo.

In addition, the bioadhesive pacing lead demonstrates various clinical advantages and possible benefits which warrant further investigation to achieve clinical translation. The multi-material 3D printing-based fabrication capability is greatly beneficial as it allows the customization and personalization of cardiac devices including for adult and pediatric applications. The atraumatic characteristics of the bioadhesive pacing lead may offer a promising option for atrial pacing and synchronous atrioventricular pacing which are technically challenging with existing devices due to the relatively thin wall of the atrium (13, 43) and associated high risk of tissue trauma and massive hemorrhage (15). Furthermore, the bioadhesive pacing lead can readily be incorporated into the existing clinical settings and protocols for temporary cardiac pacing as demonstrated in the proof-of-concept in vivo porcine study, potentially benefiting future translation and adoption into clinics.

While we systematically validate the feasibility and efficacy of the proposed bioadhesive pacing lead based on proof-of-concept in vivo rat and porcine models with promising outcomes, the current study also has several limitations and areas for future work. First, we demonstrate the efficacy of the bioadhesive pacing lead up to two weeks post-implantation reflecting the typical clinical treatment regimens for temporary cardiac pacing (18). The long-term performance of the bioadhesive pacing lead, especially regarding the stability of adhesion and electrical performance for cardiac monitoring and pacing, may require further studies to investigate the feasibility and efficacy of the proposed platform for potential prolonged cardiac pacing applications. Second, long-term in vivo biocompatibility and tissue-implant interactions such as fibrosis beyond the timeframe of temporary epicardial cardiac pacing (1-2 weeks) would require follow-up studies. Notably, further improvements in materials and design of the bioadhesive pacing lead can also be explored to reduce the risk of fibrosis and pericardiac adhesion around the implanted lead for long-term applications (47). Third, the adoption of manufacturing technologies in the industry for scalable fabrication would further increase the mechanical robustness, electrical reliability, and reproducibility of bioadhesive pacing lead prior to clinical translation of the technology.

MATERIALS AND METHODS

Study design.

The aim of this study was to develop a bioadhesive pacing lead for atraumatic implantation and on-demand retrieval, and continuous cardiac monitoring and stimulation. We hypothesized that the bioadhesive pacing leads can address the limitations of the clinically-adopted cardiac pacing leads while being readily compatible with the standard clinical setup. Systematic mechanical characterizations were performed using ex vivo porcine epicardium to evaluate the capabilities of robust adhesion and atraumatic detachment to the epicardial tissues. Structural and electrical characterizations were performed in the physiological conditions to validate the structural and electrical stability. In vivo efficacies of continuous monitoring and stimulation were investigated based on rat and porcine heart implantation models in comparison with a commercial epicardial pacing lead, assessing the capture threshold and sensing amplitudes. The atraumatic implantation and on-demand retrieval were assessed based on animal monitoring, histopathological evaluation, immunofluorescence analysis, and blood analysis. The appropriate sample size for each study was used based on the literature on similar evaluations. All tests were performed with randomly allocated experimental groups, and no data were excluded from the analyses.

Materials.

All chemicals were obtained from Sigma-Aldrich unless otherwise mentioned and used without further purification. All porcine tissues and organs for ex vivo experiments were purchased from a research-grade porcine tissue vendor (Sierra Medical, Inc.).

Preparation of 3D printable bioadhesive ink.

Hydrophilic polyurethane (PU, HydroMed D3, AdvanSource Biomaterials), 4,4′-Methylenebis (phenyl isocyanate) (MDI) were dried overnight in the vacuum condition before use. N,N-dimethylformamide (DMF) was distilled before use. PU (10g) was dissolved in DMF (30 mL) for 2 h at 50 °C with a mechanical stirrer in a nitrogen environment to obtain a homogeneous mixture. Then, 2-hydroxyethyl methacrylate (HEMA, 0.6 mL) dissolved in 10 mL of DMF was added to the reaction mixture and stirred for 1 h. To synthesize non-detachable bioadhesive, azobis (isobutyronitrile) (AIBN, 0.3 mL) and acrylic acid (30 mL) were slowly added to the reaction mixture to prevent a sudden increase in viscosity. The reaction was continued for 3 h at 70 °C. To synthesize detachable bioadhesive and introduce disulfide bonds, 6-(2-(methacryloyloxy) ethoxy) hept-6-enoic acid (3g) and AIBN (0.3 mL) were added to the reaction mixture and stirred for 1 h before adding acrylic acid monomer. The product was precipitated in distilled water to terminate the polymerization, and the product was cut into small pieces and thoroughly washed in distilled water to remove the remaining reactants. The final product was filtered and dried in the fume hood for 3 days to obtain non-detachable and detachable bioadhesive resin. The non-detachable and detachable bioadhesive resin was stirred and dissolved in 70% ethanol (15 w/w %) to obtain non-detachable and detachable bioadhesive ink, respectively. To introduce NHS ester groups into the polyacrylic acid network, the 1-ethyl-3-(-3-dimethylaminopropyl) carbodiimide (0.5 w/w %, ThermoFisher Scientific) and N-hydroxysulfosuccinimide sodium salt (0.25 w/w %, ThermoFisher Scientific) were mixed with the detachable bioadhesive ink, and then mixed with non-detachable bioadhesive ink in a ratio of 1:1 v/v to obtain a 3D printable bioadhesive ink.

To prepare 3D printable electrically conductive bioadhesive ink, PEDOT:PSS pellets (Orgacon DRY5, AGFA) were dispersed in a deionized water-DMSO mixture (water: DMSO = 85:15 v/v) at the concentration of 5 w/w %, and then was mixed with the 3D printable bioadhesive ink in a ratio of 1:2 v/v to obtain a 3D printable conductive bioadhesive ink. The 3D printable bioadhesive and conductive bioadhesive inks were filtered with 52-µm, 31-µm, and 18-µm nylon membrane filters (TISCH Scientific) in order before printing.

To prepare 3D printable ink for the fabrication of the reservoir, hydrophilic polyurethane (25 % w/w, HydroThane, AdvanSource Biomaterials) was dissolved in a DMF-tetrahydrofuran (THF) mixture (DMF:THF = 50:50 v/v) at 70 °C for 2 h, and then cooled down to room temperature to obtain an insulation ink. Polyvinyl alcohol (PVA, 30 w/w %, M_w 13,000–23,000) was dissolved in distilled water at 95 °C for 2 h, and then cooled down to room temperature to obtain a PVA sacrificial ink.

Fabrication of bioadhesive pacing lead.

3D printing of the bioadhesive pacing lead were conducted based on a custom-designed 3D printer based on a Cartesian gantry system (AGS1000, Aerotech) with various sizes of nozzles (200- and 100-µm nozzles from Nordson EFD). Printing paths were generated by drawings (Adobe Illustrator) and converted into G-code by a commercial software package (CADFusion, Aerotech) to command the X-Y-Z motion of the printer head. The detailed printing paths are provided in Fig. S3.

The bioadhesive interface was printed on the glass substrate based on the bioadhesive and electrically conductive bioadhesive inks, and an electrode lead wire (AS633, Cooner Wire) was connected to the printed bioadhesive interface via the silver paste. The PVA sacrificial layer was then printed on the dried bioadhesive interface. The insulation layer was printed on the top and connected with a PU-based fluidic channel (MRE25, Braintree Scientific, Inc.) to obtain the bioadhesive pacing lead with a built-in reservoir. Nylon membrane filters (3 µm, TISCH Scientific) were added to replace the PVA sacrificial layer to physically separate the bottom adhesive and top insulation layers and facilitate the detachment solution injection in the reservoir for terminal porcine studies.

To prepare a balloon catheter, a PU film was prepared by drop-casting PU solution (10 % w/w, HydroThane, AdvanSource Biomaterials) on the glass and evaporating the solvent. The PU film was thermally formed into a hemispherical balloon shape by a vacuum forming molding machine (JINTAI). The mold for the balloon was prepared by a stereolithography 3D printer (Form2, Formlabs). The balloon-shaped PU film was connected with a PU tube (MRE37, Braintree Scientific, Inc.) and combined with the bioadhesive pacing lead by a printed ring-shaped sacrificial layer to serve as a balloon catheter. An adapter was prepared by thermal forming a PU tube (McMaster Carr). The balloon catheter, adapter, and sheath catheter were assembled as an implantation tool for minimally-invasive implantation of the bioadhesive pacing lead.

Mechanical characterization.

To characterize the mechanical properties of the bioadhesive interface in the wet physiological condition, the nonconductive bioadhesive and conductive bioadhesive were equilibrated in 37 °C PBS. The tensile properties and fracture toughness were measured using tensile tests of a rectangular sample (10 mm in length, 20 mm in width, and 0.2 mm in thickness) with a mechanical testing machine (20 N load-cell, Zwick/Roell Z2.5). All tests were conducted with a constant tensile speed of 50 mm min^-1. The fracture toughness was calculated based on tensile tests of unnotched and notched samples(48).

For adhesion characterization, porcine epicardial tissues and polyurethane substrates (HydroMed D3, AdvanSource Biomaterials) were adhered by the nonconductive bioadhesive or the conductive bioadhesive by pressing for 5 s (applying a pressure of ~ 1 kPa by either a mechanical test machine or an equivalent weight). All adhesion tests were performed 24 h after applying the pressure to provide sufficient time for equilibrium swelling of the bioadhesive in wet environments. Poly (methyl methacrylate) films (50 µm thick, Goodfellow) were applied by using cyanoacrylate glue (Krazy Glue) as a stiff backing for the tissues and polyurethane substrates for the measurement of interfacial toughness and shear strength.

To measure the interfacial toughness, adhered samples (25 mm in width) were tested by the standard 180-degrees peel test (ASTM F2256) with the mechanical testing machine (2.5 kN load-cell, Zwick/Roell Z2.5). Interfacial toughness was calculated by dividing two times the plateau force by the width of the sample following the corresponding ASTM standard. To measure the shear strength, the adhered samples (25 mm in width and 10 mm in length) were tested by the standard lap-shear test (ASTM F2255) with the mechanical testing machine (2.5 kN load-cell, Zwick/Roell Z2.5). Shear strength was determined by dividing the maximum force by the adhesion area of the sample following the corresponding ASTM standard. All tests were conducted with a constant tensile speed of 50 mm min^-1.

Rheological characterization.

The bioadhesive and conductive bioadhesive inks were characterized by using a rotational rheometer (AR-G2, TA Instruments) with a 40 mm diameter and parallel-plate geometry. Apparent viscosity was measured as a function shear rate by steady-state flow tests with a logarithmic sweep of shear rate (0.01–100 s^-1). Shear storage modulus (G′) was measured as a function of shear stress via oscillation test with a logarithmic sweep of shear stress (1–1,000 Pa) at a frequency of 1 Hz. All rheological characterizations were conducted at 25 °C with a preliminary equilibration time of 30 s.

Electrical characterization.

To measure the charge injection capacity, the bioadhesive pacing lead and a platinum counter electrode were connected to a potentiostat (Autolab PGSTAT204, Metrohm). The bioadhesive pacing lead and an Ag/AgCl reference electrode were connected to the oscilloscope (Siglent Technologies). The potentiostat provided biphasic current inputs where a cathode phase of charge given by a certain amplitude and constant width (i_c, t_c) was followed by an inter-phase delay before symmetrical charge-balancing anodic phase with the same amplitude and width (i_a, t_a) (Fig. 4a). The current amplitude (i_c, i_a) was increased until the interface polarization (E_p) on the oscilloscope reached the electrochemical safety limit (cathodic water electrolysis, -0.8 V). The charge injection capacity was calculated by equation 1, and the charge injection capacity of commercial temporary myocardial pacing lead (Medtronic 6500) was measured based on the same method(49).

To measure the electrical impedance, the bioadhesive pacing lead was adhered on a gold-coated polyester film (McMaster Carr) and was equilibrated in PBS before tests. The sensing and working electrodes of the potentiostat were connected to the lead wire of the bioadhesive pacing lead, and the counter and reference electrodes were connected to the gold-coated polyester film, respectively. The impedance and phase angle were measured in a range from 1 Hz to 100,000 Hz in frequency. The conductivity of conductive bioadhesive was measured based on the previously reported method(21). The length, width, and thickness of the sample were measured with a vernier caliper or a microscope (LV10, Nikon). To measure the electrical stability, samples were stored in PBS with 0.01 % w/v sodium azide to prevent the growth of microorganisms during the test. The impedance of bioadhesive pacing lead for pig study with 200 µm thick adhesive layer is 680 ohms, which is measured by Biotronik pacemaker and programmer.

Characterization of on-demand detachment.

The detachment solution was prepared by dissolving 0.05 M sodium bicarbonate and 0.05 mM L-glutathione reduced in PBS. To validate the cleavage of chemical crosslinks by the detachment solution, the bioadhesive interface was incubated in PBS with primary amine-coupled fluorescent microbeads (FluoSpheres; Thermo Fisher Scientific) for 30 min at room temperature. The samples were further incubated in the detachment solution for 5 min. The surface of the sample was characterized by a fluorescence microscope (LV; Nikon) after thorough washing with PBS to remove the non-crosslinked microbeads.

To characterize the diffusion of detachment solution through the bioadhesive interface to the underlying tissues, the bioadhesive pacing lead adhered on a gelatin hydrogel, and the detachment solution with a fluorescent dye (Rhodamine B) was injected into the bioadhesive pacing lead for 1, 3 and 5 min, respectively. The detachment solution was fully removed before further characterization. The cross-section of the sample was imaged by a laser confocal microscope (SP 8, Leica). To prepare the gelatin hydrogel, 10 % w/w gelatin powder was dissolved in deionized water at 37 °C, followed by cooling down to room temperature for gelation.

Minimally invasive implantation and retrieval in ex vivo porcine model.

All ex vivo experiments were reviewed and approved by the Committee on Animal Care at the Massachusetts Institute of Technology. To simulate a minimally invasive surgical setting, the experiment was conducted inside a dark chamber with a porcine abdomen tissue on the top. Periodic pressured air inputs were introduced to the porcine heart to simulate heartbeats by a programmable pressure dispenser (Ultimus V, Nordson EFD). Two full-thickness holes in the porcine abdominal wall were created by a biopsy punch (10 mm, Integra™), and two trocars (15 mm, Medtronic) were used to place an endoscope camera (DESPTECH) for visualization and the bioadhesive pacing lead through each trocar. The sheath catheter with a folded device was inserted through the trocar. A pressure-controlled syringe (Mercury Medical, AnapnoGuard Cuffill) was used to inflate the balloon catheter and open the bioadhesive pacing lead. An adapter was used to provide gentle pressure and form adhesion on the epicardium. Saline was injected through the balloon catheter to dissolve the PVA sacrificial layer, followed by robust adhesion of the bioadhesive pacing lead to the epicardium and retrieval of the balloon catheter. To achieve on-demand retrieval of the adhered bioadhesive pacing lead, the detachment solution was injected into the reservoir. After 5 min, the bioadhesive pacing lead was atraumatic detached from the epicardium and retrieved through the trocar.

Simulated cardiac pacing in ex vivo porcine model.

One LED was used as an indicator to quantitatively evaluate the efficacy of charge injection through pacing leads by comparing the current threshold to light up the LED. One electrode of the LED was inserted into the epicardium of an ex vivo porcine heart and another electrode was connected to a potentiostat (Autolab PGSTAT204, Metrohm). The bioadhesive pacing lead or a commercially-available pacing lead (Medtronic 4968) were placed on the epicardium and connected to the potentiostat. The distance between the LED and the pacing leads remained consistent (2 cm) for both the bioadhesive pacing lead and the commercial pacing lead to ensure the same conductive pathway. The current pulse at different amplitudes and frequencies were provided by the potentiostat to evaluate the current threshold to light up the LED. Periodic pressured air inputs were introduced to the porcine heart to simulate heartbeats by the programmable pressure regulator (Ultimus V, Nordson EFD).

In vitro biocompatibility.

We conducted in vitro biocompatibility tests using device-conditioned medium for cell culture. To prepare device-conditioned media, the bioadhesive pacing lead or Medtronic 6500 were incubated in 5 ml DMEM at 37 °C for 24 h, respectively. Pristine DMEM was used as a control. Wild-type mouse embryonic fibroblasts were plated in a confocal dish (n = 3 per group). The cells were then treated with the bioadhesive pacing lead-conditioned medium or Medtronic 6500-conditioned medium, and incubated at 37 °C for 24 h in 5% CO₂, respectively. The cell viability was determined by a LIVE/DEAD viability/cytotoxicity kit for mammalian cells (Thermo Fisher Scientific). A laser confocal microscope (SP 8, Leica) was used to image live cells with excitation/emission at 495 nm/515 nm, and dead cells at 495 nm/635 nm, respectively. The cell viability was calculated by counting the number of live (green fluorescence) and dead (red fluorescence) cells by using ImageJ 1.8.0.

In vivo rat model.

All studies in rats were approved by the MIT Committee on Animal Care and all surgical procedures and post-operative care were supervised by the MIT Division of Comparative Medicine (DCM) veterinary staff. Female Sprague Dawley rats (250-300g, Charles River Laboratories) were used for all in vivo studies.

Prior to implantation, the bioadhesive pacing lead was prepared using aseptic techniques and was further sterilized for 1 h under UV light. For continuous ventricular pacing, the rats were anesthetized using isoflurane (1-3% isoflurane in oxygen) in an anesthetizing chamber. Back hair and chest hair were removed. Endotracheal intubation was performed, and the rats were connected to a mechanical ventilator (Model 683, Harvard Apparatus) and placed over a heating pad for the duration of the surgery. For measurement of surface ECG, anodic and cathodic subdermal needle electrodes were inserted into the skin of the left hindleg and right foreleg of a rat, respectively. A subdermal needle electrode was inserted into the skin of the right hind leg as a shared ground for surface ECG recording and epicardial pacing. The ECG signals were collected with data acquisition hardware (PowerLab and BioAmp, AD Instrument) and software (LabChart Pro 7, AD Instrument) for the duration of the surgery. The heart was exposed via a thoracotomy and the pericardium was removed using fine forceps. A subdermal needle electrode was inserted into the skin of the left foreleg as an anodic pacing electrode. A bioadhesive pacing lead adhered on the left ventricle (n = 4) as a cathodic pacing electrode. The lead wire and fluidic tube of the bioadhesive pacing lead were then tunneled subcutaneously from a ventral exit site close to the left fourth intercostal space to the dorsal side. The dorsal end of the lead wire and fluidic tube was inserted through a subcutaneous port. The subcutaneous port was placed by interrupted sutures (4-0 Vicryl, Ethicon) between the shoulder blades of the rat and covered by a protective aluminum cap (VABRC, Instech Laboratories)(50). An external stimulator (FE180, AD Instrument) was connected with the anodic and cathodic pacing electrodes and controlled by the data acquisition hardware and software (AD Instrument). Unipolar rectangular current pulses (0.5 ms, 0-3 mA, 5-7 Hz) were applied by the stimulator with continuous ECG recording to evaluate the capture threshold. The cardiac signals were sensed by the implanted pacing leads and recorded by the data acquisition hardware and software to evaluate the R wave amplitude. The incision was closed using interrupted sutures (4-0 Vicryl, Ethicon) and 3–6 ml of saline was administered subcutaneously. The animal was ventilated with 100% oxygen until autonomous breathing was regained, and the intubation catheter was removed.

On day 7 post-implantation, each animal was anesthetized and connected to the data acquisition hardware (AD Instrument) for surface ECG recordings according to the procedure described above. Anesthesia was maintained using a nose cone and the rats were placed over a heating pad for the duration of tests. The external stimulator was connected through the dorsal subcutaneous port. The continuous ventricle pacing and the capture threshold were examined by the surface ECG recordings. The animal was ventilated with 100% oxygen until autonomous breathing was regained.

On day 14 post-implantation, the continuous ventricle pacing and capture threshold evaluation was conducted according to the procedure described above. For atraumatic retrieval of the implanted bioadhesive pacing lead, the dorsal subcutaneous port was removed firstly, and the detachment solution was injected from the dorsal end of the fluidic tube to the built-in reservoir. Five minutes after applying the detachment solution, the bioadhesive pacing leads were gently pulled out from the incision at the dorsal side. The animals were euthanized by CO₂ inhalation, and the hearts were excised and fixed in 10% formalin for 24 h for histological analyses.

For evaluation of in vivo biocompatibility, the bioadhesive pacing leads (n = 4) and commercially-available temporary pacing leads (Medtronic 6500, n = 4) were used. The bioadhesive pacing lead was applied to the heart according to the procedure described above. For the commercial pacing lead, the tip of lead was inserted into the myocardium by a tapered point curved needle and fixed by a monofilament coil on the ventricle. The incision was closed using interrupted sutures (4-0 Vicryl, Ethicon) and 3–6 ml of saline was administered subcutaneously. The animal was ventilated with 100% oxygen until autonomous breathing was regained, and the intubation catheter was removed. At day 14 post-implantation, the animals were euthanized by CO₂ inhalation and the hearts with the implants were excised and fixed in 10% formalin for 24 h for histological analyses. All rats in the study were monitored daily by the MIT DCM veterinary staff and maintained normal health conditions.

All fixed tissue samples were placed in 70% ethanol and submitted for histological processing and hematoxylin and eosin staining at the Hope Babette Tang (1983) Histology Facility in the Koch Institute for Integrative Cancer Research at Massachusetts Institute of Technology.

In vivo porcine model.

All studies in pigs were approved by the Mayo Clinic Institutional Animal Care and Use Committee at Rochester. Female domestic pigs (100-110 kg, Manthei Hog Farm) were used for all in vivo porcine studies. All animals were acclimatized in the holding facilities for seven days before the study. Anesthesia was induced with tiletamine/zolazepam HCl (Telazol, 5 mg/kg, Zoetis) xylazine (2 mg/kg, Akorn animal house), and atropine (0.04 mg/Kg intramuscularly, West-ward), and maintained with isoflurane (1-3%, Baxter) in oxygen. Continuous ECG tracings, arterial blood pressure, and peripheral capillary oxygen saturation (SpO2) percentage were monitored during the surgery (GE Mac-Lab hemodynamic recording system). Animals were intubated and placed on mechanical ventilation using volume-cycled ventilation. A left lateral thoracotomy was used to access the chest and the pericardium was incised to expose the epicardium for device implantation. A commercially-available pacing lead (Medtronic 6500) was inserted into the myocardium by a tapered point curved needle and fixed on the left ventricular free wall. The bioadhesive pacing lead was applied and adhered to the epicardium in the same region. Ventricular pacing was performed by the commercial pacing lead (n = 3) and the bioadhesive pacing lead (n = 3), respectively. Then the bioadhesive pacing lead was applied and adhered to the right atrium. Atrial pacing and atrioventricular pacing were performed by the bioadhesive pacing leads. A dual chamber pulse generator (Medtronic 5330) was connected with pacing leads to perform all types of epicardial pacing to alter the heart rate (70-120 bpm). The pulse current was gradually increased to measure the capture threshold. The R wave amplitude of the commercial pacing lead and the bioadhesive pacing lead were recorded as the highest sensitivity setting for which intrinsic R waves were appropriately recognized on the pulse generator. Following pacing, leads were removed either by manual traction (commercial leads) or instillation of the lead detachment solution through the fluidic port (bioadhesive pacing leads). At the end of the trial (maximum length was four hours), the animal was euthanized by an intravenous injection of sodium pentobarbitol (Fatal- Plus, 150mg/kg, Baxter).

For survival porcine study, bioadhesive pacing leads were applied and adhered to the ventricular free wall. The bioadhesive pacing leads modified with Oscor C/IS adaptors were tunneled and connected to an implantable pacemaker (Dextronix, PetPacer-DR), which was buried subcutaneously. The chests were surgically closed, and a RENAMIC portable implant programmer and interrogator were used to measure the sensing amplitude and pacing capture threshold telemetrically. On day 10 post-implantation, the animal was euthanized, and the lead-cardiac tissue interface was collected for detailed histological evaluation.

Immunofluorescence analysis.

The expression of targeted proteins (CD68, CD3, Collagen-I, α-SMA) were analysed after the immunofluorescence staining of the collected tissues. Before the immunofluorescence analysis, the paraffin-imbedded fixed tissues were sliced and prepared into slides. The slides were deparaffinized and rehydrated to deionized water. Antigen retrieval was performed using steam method during which the slides were steamed in IHC-Tek Epitope Retrieval Solution (IW-1100) for 35 min and then cooled for 20 min. Then the slides were washed in three changes of PBS for 5 min per each cycle. After washing, the slides were incubated in primary antibodies (1:200 mouse anti-αSMA for fibroblast (ab7817, Abcam); 1:200 mouse anti-CD68 for macrophages (ab201340, Abcam); 1:100 rabbit anti-CD3 for T-cells (ab5690, Abcam); 1:200 rabbit anti-collagen-I for collagen (ab21286, Abcam)) diluted with IHC-Tek Antibody Diluent for 1 h at room temperature. The slides were then washed three times in PBS and incubated with Alexa Fluor 488 labelled anti-rabbit or anti-mouse secondary antibody (1:200, Jackson Immunoresearch) for 30 min. The slides were washed in PBS and then counterstained with propidium iodide solution for 20 min. A laser confocal microscope (SP 8, Leica) was used for image acquisition. ImageJ (version 2.1.0) was used to quantify the fluorescence intensity of expressed antibodies. All the images were transformed to the 8-bit binary images, and the fluorescence intensity was calculated with normalized analysis. All analyses were blinded with respect to the experimental conditions.

Luminex quantitation analysis.

On 14 days post-implantation, the whole blood samples were collected for bioadhesive pacing lead and Medtronic 6500. The serum samples were separated from the whole samples following the manufacturer’s instructions. A Luminex multiplex assay was used to measure the concentrations of immune response-related cytokines and chemokines (RECYTMAG-65K-07, Milliplex). Values were normalized to untreated animals and presented on a log10 scale (n = 3 animals per group).

Statistical analysis.

Prism GraphPad (version 8) was used to assess the statistical significance of all comparison studies in this work. Data distribution was assumed to be normal for all parametric tests, but this was not formally tested. In the statistical analysis for comparison between multiple samples, one-way analysis of variance (ANOVA) followed by Tukey’s multiple comparison test was conducted with thresholds of *P < 0.05, **P ≤ 0.01, ***P ≤ 0.001, and ****P ≤ 0.0001. For the statistical analysis between two data groups, an unpaired two-sided t-test was used, and the significance thresholds were *P < 0.05, **P ≤ 0.01, ***P ≤ 0.001, and ****P ≤ 0.0001.

Supplementary Material

Fig. S1. Tissue-device integration based on existing tissue adhesives.

Fig. S2. Schematic illustrations for the adhesion mechanism of bioadhesive pacing lead.

Fig. S3. Design and printing path of the bioadhesive pacing lead.

Fig. S4. Chemical schemes for the synthesis of printable bioadhesive.

Fig. S5. On-demand detachment mechanism.

Fig. S6. Structural characterization of the printable conductive bioadhesive.

Fig. S7. Swelling of the nonconductive bioadhesive and conductive bioadhesive.

Fig. S8. Fracture toughness of the nonconductive bioadhesive and conductive bioadhesive.

Fig. S9. Built-in reservoir filled with the detachment solution.

Fig. S10. Adhesion performance characterization.

Fig. S11. Electrical stability of the conductive bioadhesive.

Fig. S12. Ex vivo demonstration for cardiac pacing.

Fig. S13. Minimally-invasive implantation and retrieval of the bioadhesive pacing lead in ex vivo pig model.

Fig. S14. Application of non-adhesive epicardial pacing leads based on sutures and BioGlue to rat heart in vivo.

Fig. S15. In vitro cytotoxicity.

Fig. S16. In vivo cardiac monitoring and pacing performances of sutured and BioGlue-fixed non-bioadhesive pacing leads.

Fig. S17. Cardiac monitoring by the bioadhesive pacing lead and Medtronic 6500.

Fig. S18. In vivo epicardial pacing in ischemia-reperfusion rat model.

Fig. S19. Representative histology images of pacing leads.

Fig. S20. Immunofluorescence analysis of bioadhesive pacing lead and Medtronic 6500.

Fig. S21. Long-term biocompatibility after detachment of bioadhesive pacing leads.

Fig. S22. Luminex quantitative analysis of systemic immune response.

Fig. S23. Application of non-adhesive epicardial pacing leads based on sutures and BioGlue to porcine heart in vivo.

Movie S1. 3D printing of the bioadhesive pacing lead.

Movie S2. Stimulation of epicardial pacing by the bioadhesive pacing lead and Medtronic 4968 on an ex vivo porcine heart.

Movie S3. Minimally invasive implantation and retrieval of the bioadhesive pacing lead on an ex vivo porcine heart.

Movie S4. Atraumatic implantation, continuous ventricular pacing, and on-demand retrieval of the bioadhesive pacing lead on an in vivo rat heart.

Movie S5. Implantation, continuous ventricular pacing, and retrieval of Medtronic 6500 on an in vivo rat heart.

Movie S6. Atraumatic implantation, continuous atrial pacing, and on-demand retrieval of the bioadhesive pacing lead on an in vivo porcine heart.

Movie S7. Implantation, continuous ventricular pacing, and retrieval of Medtronic 6500 on an in vivo porcine heart.

Data and materials availability

All data are available in the main text or the supplementary materials.