Author + information
- Peter Kohl, MD, PhD∗ ()
- National Heart and Lung Institute, Imperial College London, London, United Kingdom; and the Department of Computing Science, University of Oxford, Oxford, United Kingdom
- ↵∗Reprint requests and correspondence:
Dr. Peter Kohl, National Heart and Lung Institute, Harefield Heart Science Centre, Hill End Road, London UB9 6JH, United Kingdom.
Conventional wisdom suggests: this should not be possible. You should not be able to successfully restore sequential activation of cardiac chambers in an adult animal with complete atrioventricular conduction block simply by gluing one end of a dish-grown tissue strand containing neonatal rodent myocytes to the outer surface of the atrium and the other end to the ventricle. However, that is exactly what Cingolani et al. (1) report in this issue of the Journal.
The authors (1) reconfirm earlier work (2) showing that myocardial tissue grafts can serve as conduits to “patch up” electrical conduction. Their electrical conduits are formed using paramagnetic beads with surface-conjugated antibodies that bind either neonatal cardiac myocytes or human stromal cells, mixed in varying ratios and attracted by a linear magnetic field to form an elongated, structurally and functionally integrated tissue strand. In vitro, these conduits synchronize spontaneous electrical activity of otherwise disconnected regions of cardiac neonatal cell cultures, as shown earlier in similar cell culture models for bridges containing cardiac myocytes (3) or connexin-expressing nonmyocytes (4). In vivo, the investigators attached the ends of preformed myocyte/nonmyocyte strands to right atrial and right ventricular epicardium, using fibrin glue but no other treatment of cardiac or conduit contact sites, which is perhaps the most elegant and most surprising aspect of their report. After 3 days, this procedure offered an atrioventricular conduction pathway, which supported sequential chamber activation when the intracardiac conduction system was taken out of action by methacholine perfusion. They report abundant connexin 43 expression at contact sites of native and engineered tissue, compatible with electrical connection via functional heterotypic cell coupling in the heart (5).
The present report (1) builds on and reinforces previous research, also conducted in rats, with myogenic progenitor cell grafts surgically implanted into the atrioventricular groove (2). These grafts supported atrioventricular conduction in one-third of animals, with the added benefit that they became vascularized over time, surviving for the animals’ lifetimes. In both studies, atrium–graft and graft–ventricle electrical conduction had to cross discontinuities in myocardial tissue, epicardium, or scars.
The principal possibility of passive cardiac action potential propagation through non-excitable mesenchymal cells had been known from cell culture studies for 45 years (6). Its presence was subsequently documented in patients with electrical propagation across atrio-atrial (7) and atrioventricular (8) lesions after cardiac surgery. The underlying electrotonic coupling of heterotypic heart muscle and mesenchymal cells, identified in vivo 10 years ago (9), is an important, yet often overlooked, facet of cardiac electrophysiological integration that deserves further attention (for a recent update on this field, see ).
Cardiac conduction pathway repair using engineered biological conduits has significant clinical translation potential. Use of human pluripotent stem cell–derived cardiomyocytes and fibroblasts for patient-specific electrical grafts is an obvious avenue. This option would add significantly to the current range of cell injection–based approaches for cardiac tissue repair (10), taking the field from the delivery of more locally acting agents to provision of targeted long-distance pathways.
Technological developments to grow human pluripotent stem cell–derived cardiomyocytes in aligned tissue strands have begun (11), although describing them as biowires understates their potential importance. Conduction in “wires” is passive and characterized by progressive signal attenuation, as would be observed with action potential propagation through tissue constructs made entirely from non-excitable cells. These would bridge relatively short distances only (up to 300 μm in vitro ), and they are likely to underlie coupling across the graft–myocardium interface. The key advantage of having cardiac myocytes as an integral component of a conduit, even if they themselves were not to form a continuous conductive chain, is that they act like action potential “repeater stations” (12), sustaining electrical signal fidelity from input to output of the graft.
Several hurdles must be overcome on the way to clinical application. These include:
• Automaticity. The ideal electrical conduit will have no or very low spontaneous excitation rates. This can presumably be achieved by selecting suitable cells for the graft.
• Connectivity. Grafts would ideally couple well at their ends but remain insulated along much of their progression. This pattern seems well supported by the approach presented by Cingolani et al. (1); they used a fixation technique that (in contrast to surgical integration) evades excess deposition of electrically insulating acellular connective tissue at contact sites, whereas leaving the intermediate conduit free from direct association with the myocardium.
• Source–sink relations. Linked to connectivity and directionality, it will be necessary to control input and output loads at the atrial and ventricular ends of the conduit. This may be achievable using advanced tissue engineering.
• Directionality. Avoidance of retrograde conduction may be attainable by controlling source–sink relations and connexin coupling patterns at both graft ends. This may benefit from quantitative computational modeling to aid design of tissue implants and their connections to the heart.
• Location. Relative ease of epicardial access should be balanced with the advantages of endocardial activation. For the proof of principle reported here, the former was understandably an overriding concern, but further research will be needed to ascertain the best placement of conduits for univentricular or potentially biventricular “pacing.”
• Durability. Cingolani et al. (1) observed effective atrioventricular coupling 3 days after conduit implantation. During this period, an alternative mechanistic explanation of atrioventricular synchronization may apply (Figure 1).
• Stability (mechanical). Constructs, in particular if attached mechanically only at their ends, must withstand their own and the heart’s contraction, as well as any hemodynamic and respiration-induced changes in the relative topology of atrial and ventricular attachment sites. This may be aided by suitably balancing muscle and nonmuscle construct constituents (as explored by the investigators) and by carefully scaling graft lengths to attachment loci.
• Stability (metabolic). Electrical conduits in the heart are likely to require vascularization, in particular if the myocytes within them mature. This at least is the message from trabeculae carneae, which, if larger than 4 to 8 myocytes per cross-section (constructs used here exceed the corresponding dimension by about an order of magnitude), regularly have their own blood supply with a capillary-to-myocyte ratio of ∼0.5 in the rat (13).
• Stability (immunological). To prevent rejection, the present study (1) explanted hearts for ex vivo experiments after 3 days. This concern may be avoided using individualized stem cell–derived grafts and biodegradable structural support materials for the initial conduit generation.
In spite of this long list, the vision of being able to patch up not only the plumbing of the heart using autologous replacement vessels, but also its wiring using individualized electrical conduits, is inspiring. Of course, electrical rewiring will not solve all forms of conduction block. Furthermore, adding an alternative activation pathway could be counterproductive, in particular for intermittent “functional” disturbances. Additional studies, involving larger numbers of animals and perhaps more traditional atrioventricular block models, are needed to systematically address the concerns listed previously, but an encouraging start has been made.
Only the future can tell whether this vision will materialize. Looking to lessons from the past, it is instructive to revisit the work of Starzl et al. (14) who, in 1963, tried to treat complete atrioventricular conduction block in dogs by autologous sinoatrial node transplantation into the right ventricle. Although they saw no evidence of a change in idioventricular rhythms after nodal transplant in any animal (0 of 9), dogs in the control series were effectively treated (2 of 2), if temporarily. In these animals, an atrioventricular sling was constructed (Figure 1) that afforded mechanically mediated atrioventricular synchronization, whereby the atrial “tug” triggered ventricular excitation. Pacing, on the basis of mechanoelectric feedback (15), lasted 2 to 4 days, at which time autopsy established that the atrial sling had become attenuated, rendering mechanical coupling ineffective. It would be interesting to explore whether similar mechanisms may have contributed to the 3-day observations reported here.
In conclusion, the paper by Cingolani et al. (1) serves as a reminder that “convention” is not “wisdom” and that the most intriguing observations tend to arise from venturing outside the box. Their study certainly does that.
The author expresses his gratitude to Dr. T. Alexander Quinn for helpful comments on the manuscript.
↵∗ Editorials published in the Journal of the American College of Cardiology reflect the views of the authors and do not necessarily represent the views of JACC or the American College of Cardiology.
Research in the author’s laboratory is supported by the European Research Council’s Advanced Grant (CardioNECT), the British Heart Foundation, and the Magdi Yacoub Institute. Dr. Kohl has reported that he has no relationships relevant to the contents of this paper to disclose.
- American College of Cardiology Foundation
- Cingolani E.,
- Ionta V.,
- Cheng K.,
- Giacomello A.,
- Cho H.C.,
- Marbán E.
- Kleber A.G.,
- Rudy Y.
- Gaudesius G.,
- Miragoli M.,
- Thomas S.P.,
- et al.
- Camelliti P.,
- Green C.R.,
- LeGrice I.,
- et al.
- Goo S.,
- Joshi P.,
- Sands G.,
- et al.