Cryothermic ablation in comparison with radiofrequency ablation
The current standard for percutaneous intracardiac lesion generation is radiofrequency ablation (RFA). RFA creates lesions of sufficient size to reach most reentrant pathways and irritable spots and therefore affects a cure of the resultant tachycardias. However because RFA causes significant endothelial disruption and blood clot formation it may cause adverse events such as stroke (1). It has limited efficacy in areas of low blood flow such as in the coronary sinus or between trabeculae. In such areas there is little to no cooling of the electrode by blood flow. Therefore electrode temperatures quickly rise and coagulate blood. Because RFA systems are typically temperature controlled, the circuit automatically limits power so that too small a lesion is created to reach the target and cure the patient (2). Tissue heating with RFA is greatest below the endocardial surface, so even though electrode temperature is held at 60 degrees Celsius e.g., temperatures below the encardium may reach 100 degrees Celsius causing steam formation. If the steam is breaches the epicardium, cardiac perforation and tamponade may ensue (3). The risk of perforation also increases because RFA denatures the extracellular tissue matrix (the structural backbone of the heart) weakening ablated tissue such that even the pressure from the catheter may be sufficient to perforate the heart (4). Cardiac excitability is reversibly impaired at tissue temperatures approximating 45 degrees Celsius, whereas irreversible injury occurs at 50 degrees (5). Unfortunately, it is not possible to hold electrode temperature down to see whether the effect will be desirable or not without making a permanent lesion. Therefore when RF energy is being delivered close to important structures such as the atrioventricular node (AVN), errant ablation can occur producing complete heart block (CHB) (6, 7) and then a permanent pacemaker (PPM) must be placed. RFA produces somewhat inhomogeneous lesions that can cause other arrhythmias (8). Finally, the stability of the RF catheter on the endocardial surface may be problematic in some cases making it very difficult to ablate successfully.
Cryoablation (CrA) of cardiac tissue was developed initially for open-heart surgical applications and applied by a hand-held probe to either the endocardium or the epicardium. A refrigerant was pumped under high pressure through a lumen to a reservoir in the tip of the probe where it evaporated cooling the tip. The gaseous refrigerant was then vacuumed via another lumen and collected externally. CrA was found in animal models to produce homogeneous lesions that were nonarrhythmogenic, and extracellular tissue architecture was preserved as was tissue tensile strength minimizing the risk of cardiac rupture (4). Reversible cooling of tissue to assess effect (“ice mapping”) before a lesion was produced was feasible and exploited such that surgical endocardial CrA close to the AVN was effective in curing patients with AV node reentry tachycardia (AVNRT) without producing heart block (9).
The application of a cryogenic probe to a tissue surface results in the formation of a well demarcated hemispherical block of frozen tissue, or “ice ball”, that can have either oblate or prolate geometry (10). Cells within the ice ball become irreversibly damaged and ultimately are replaced by fibrotic tissue. The mechanism of cell death has been the subject of some debate. Upon initial thawing, tissue evaluated by light microscopy appears to be intact, suggesting that subsequent edema and ischemia lead to cell death. However, electron microscopic data demonstrate early acute irreversible changes including fatal changes in subcellular organelle structure and mitochondrial destruction. These changes correlate with the presence of ice crystals observed during tissue freezing.
The progress to a stable lesion can be categorized into three phases: 1) freeze/thaw phase 2) hemorrhagic and inflammatory phase and 3) replacement fibrosis phase.
In the freeze/thaw phase, the application of cryogenic temperatures results in the formation of intracellular and extracellular ice crystals that vary in size and location depending on the tissue type, the proximity to the cryogenic probe, and on the presence or absence of blood flow during application. In perfused muscle, crystals within the ice ball tend to be irregularly distributed within muscle fibers and vary in size and shape. In contradistinction, ice crystals at the periphery of the ice ball tend to be extracellular. Ice crystals in excised (non-perfused) muscle are more regular in size and distribution suggesting that the irregularity noted in vivo is consequent to regional differences in the microcirculation (10). The mechanical effect of ice crystal formation per se is unclear. Ice crystals do not penetrate the cell membrane but cause compression and distortion of adjacent cytoplasmic components and nuclei (10,11). Smaller ice crystals in muscle tissue are associated with loss of structural detail and loss of intracellular organelles, whereas larger ice crystals compress adjacent structures with preservation of myofilament and mitochondrial architecture. Membrane integrity is preserved up to two minutes after thawing, supporting the view that the ice crystals themselves do not cause mechanical disruption.
Immediately following thaw in skeletal muscle frozen for 1 minute at -70°C, the earliest changes occur in myofilaments and mitochondria (10). The Z and I lines lose linearity and sometimes disappear altogether. The mitochondria appear enlarged, have decreased matrix density and cristae appear disrupted, although the outer mitochondrial membranes appear intact. Within 10 minutes of thaw, the sarcoplasmic reticulum appears more obvious due to the formation of distending vesicles within them, suggesting early disruption of intracellular transport mechanisms. There is a striking depletion of glycogen stores by this stage, in dramatic contrast to the effects of ischemia on skeletal muscle, in which glycogen depletion occurs only after 8 to 11 hours. Within one hour of thaw, glycogen stores are entirely depleted and there is further loss of myofilament structure. By two hours there are only remnants of normal myofibril structure; however, mitochondrial outer membranes and some crystal structures remain intact. Subsequent changes up to 10 hours following thaw are most notable for progressive mitochondrial damage. In vitro analysis of isolated frozen/thawed mitochondria has shown that irreversible damage to mitochondria occurs as a result of increased membrane permeability during the thawing phase (12). Loss of mitochondrial membrane integrity results in oxidation of endogenous pyridine nucleotides and subsequent membrane lipid peroxidation and enzyme hydrolysis. The latter result in disruption of the electron transport chain such that, despite evidence of late membrane repair, mitochondria become irreversibly de-energized (13).
Similar to the effects of cryogenic thermal skeletal muscle damage, the application of a cryogenic probe to atrial and ventricular myocardium results in the formation of an elliptical hemispheroid lesion (14). Within 30 minutes of thawing, the myocytes appear swollen and the myofilaments are extremely stretched. Mitochondria in the ice ball appear swollen and contain small densities within their matrix. Changes are less marked toward the periphery (15,16).
The hemorrhagic and inflammatory phase of myocardial damage the application of cryogenic temperatures to tissue is characterized by the development of hemorrhage (14), edema, and inflammation (16) (coagulation necrosis) that are evident within 48 hours following thaw. At one week following thaw the periphery of the lesions are sharply demarcated by inflammatory infiltrate (macrophages, lymphocytes, fibroblasts), fibrin and collagen stranding, and capillary ingrowth (16). Additionally, a few areas of hemorrhage are still evident and occasional foci of dystrophic calcification are seen. Harrison et al. (17) reported similar findings at one week following canine atrioventricular nodal (AVN) CrA. The lesions showed necrotic myocardial cells and conduction fibers, a polymorphonuclear infiltrate and marked hemorrhage in the periphery of the lesion. No changes in blood vessel walls were noted at this point, and one dog had thrombus associated with the hemorrhagic focus.
In the third and final replacement fibrosis phase, the cryolesion at 2 to 4 weeks of age consists largely of dense collagen and fat infiltration, with many small blood vessels present at the periphery. Lesions are composed of fibroblasts, capillaries and a moderate degree of fibrosis. At one month the lesion is marked by dense fibrosis (17). By 12 weeks lesions appear small and fibrotic, with a normal distribution of small and large blood vessels.
The size of the lesion created by cryogenic temperatures is determined by a variety of factors, including temperature, the size of the probe employed, and the duration and number of freeze/thaw cycles to which the tissue is subjected. For a given duration of exposure, lower temperatures generate progressively larger lesions. Within 5 minutes at a given temperature, however, the lesion size plateaus (11, 18, 19). Repetitive freeze/thaw cycles enlarge cryolesions beyond those obtained with prolonged freezing at a given temperature (20-22). The rate of conduction of the cold front increases with repetitive exposure, suggesting a progressive increase in thermal conductivity of tissue. Although the basis for this is not well understood, it may relate to a change in basic cellular structure and/or changes in the local microperfusion environment during the thaw cycle (11). Markovitz etal. demonstrated that varying the tank pressure when utilizing nitrous oxide-cooled cryogenic probes directly correlated with lesion size, with maximal lesion sizes being obtained at tank pressures greater than 700 PSI (23) They also demonstrated that cold potassium cardioplegia allowed for the formation of larger lesions at a given temperature and duration of cryogenic treatment. The subsequent use of cryogenic probes cooled by liquid nitrogen allowed for significantly lower temperatures and hence the generation of larger lesions for a given duration of exposure (24).
Vascular effects of cryoablation
The effect of cryogenic temperatures on the microcirculation is characterized by endothelial cell damage, platelet aggregation and flow stasis, occlusion and ultimately recanalization (10). Within 30 minutes of thaw endothelial cells appear swollen and protrude into the lumen of the vessel. At one hour defects in endothelial cell junctions are apparent. Carbon particles injected into thawing hamster tongue blood vessels migrate from the vessel lumen into the lumen wall but not the surrounding tissue (10). By two hours, prominent gaps form between endothelial cells. Post-capillary venules have similar albeit more marked changes during this phase of damage. Subsequently platelet plugs become associated with endothelial gaps, which lead ultimately to stasis and occlusion of vessels within the previously frozen tissue. Flow studies demonstrate recovery of flow immediately following thaw, dilatation and increased flow up to one hour following freezing, with cessation of flow occurring around 5 hours after thaw. Similarly, Brown et al. have demonstrated endothelial cell disruption, macromolecular leakage, and complete vascular stasis in the 2-hour period following surgery with a cryogenic probe of rat cremasteric muscle (25). At one week following thaw in epicardial blood vessels intimal thickening due to the presence of smooth muscle cells becomes evident (16). Similar changes have also been seen in cryogenic injury to rabbit aorta (26), in medium-sized arteries and arterioles 6 weeks following cryoablation of the AVN (27), and is not unlikely to see this with catheter-induced injury of coronary and carotid arteries (neo-intimal proliferation). While blood flow within the cryolesion itself is reduced, blood flow within tissue immediately adjacent to the cryolesion is unperturbed. This latter finding has been demonstrated in ventricular myocardium (28).
Histologic effects of cryoablation on the AV junction
The effects of cryogenic temperatures on cardiac conduction tissue have been well characterized in animal models (17, 24, 27) and in one autopsy series of four patients who died at varying times within one year of receiving CrA therapy of the AVN (29). Gillette et al. induced AVB in Hanford miniature swine utilizing a transvenous cryocatheter and sacrificed the animals that had persistent CHB one hour following the procedure. The ablation sites showed extensive hemorrhage and coagulation necrosis of myofibrils sharply demarcated from adjacent unaffected tissue (30). Fujino etal. extended these findings to characterization of tissue obtained 6 weeks following transvenous catheter CrA of the AVN (27). By this time the lesions were dense, fibrotic and homogenous. Harrison etal. induced CHB using cryogenic probes in a canine open-thoracotomy study and sacrificed the animals at 1, 2, 4, 8 and 12 weeks (17). The changes observed were similar to those already described, with necrosis of the conduction tissue and ultimately replacement by fibrotic tissue.
Human autopsy studies are consistent with the changes observed in animal models. In one patient with dilated cardiomyopathy who died suddenly 8 days following AVN surgical CrA, the area surrounding the AVN showed hemorrhage (29). The AV nodal artery had intramural fibrinoid necrosis. The conduction tissue itself showed necrosis and hemorrhage. The cryolesions from another patient who had undergone AVN CrA two months before death demonstrated discrete moderate-to-severe fibrosis and intimal narrowing of the AVN artery. At autopsy one year after cryogenic therapy in a third patient, the penetrating and branching portions of the His bundle (HB) showed marked fibrotic changes. In each case, the observed lesions were discrete and sharply delimited. None of the deaths was thought to be related to cryosurgery (CrS) (29).
Proarrhythmic potential of cryoablation lesions
Cryolesions, as discussed above, are discrete and composed largely of collagenous tissue. Cryolesions have sharp, well-demarcated margins with preserved blood flow (28). In contrast to coronary artery disease-associated scar, cryolesions exhibit low arrhythmogenic potential, as has been demonstrated in canine models (31-33). Holman et al. evaluated local electrical potentials generated prior to and immediately following the creation of ventricular cryolesions (31). Unipolar electrical potentials were measured from points spanning the ventricular myocardium using plunge electrodes. Cryolesions were generated by the application of a cryogenic probe cooled to -60°C with expanding nitrous oxide for 2 minutes. There was a proportionate decrease in local electrogram amplitude the closer the measuring electrode was to the developing cryolesion. The observed decrease in amplitude could reflect epicardial ice insulation or functional inhibition of myocardial electrical potential, a point that was not distinguished by the authors. A greater than 70% decrease in absolute amplitude from control unipolar potentials was predictive of cell death, as determined by subsequent histologic analysis 48 hrs or 2 weeks following the experiment. These findings were extended by Klein et al., who evaluated epicardial and transmural mapping after the production of cryolesions in canine myocardium, and also examined the arrhythmogenic potential of these lesions (32). They found that epicardial electrograms recorded directly above acute epicardial or intramural cryolesions showed amplitude loss, while those adjacent to the lesions were unaffected. This observation persisted 4 weeks after the acute lesion formation, clearly demonstrating sustained loss of myocardial potential following cryogenic temperatures. By studying the relative location of plunge electrodes microscopically with respect to cryolesions and surrounding normal tissue, this group observed normal electrical activation in histologically preserved tissue immediately adjacent to the cryogenic scar. A similar study performed on eight dogs by Hunt et al. corroborated these results (33). In contrast to the above studies where lesions were made in normal tissue, Reek and colleagues studied the feasibility of cryoablation in a sheep model of healed myocardial infarction. Three sheep with 5 different ventricular tachycardias (VT) were cryoablated with 2 of the 3 rendered noninducible after cryoablation (34).
A useful property of cryogenic temperature is its ability to reversibly block electrical conduction at less severe temperatures, a phenomenon referred to as cryotermination (35) or cryomapping (CrM). Analysis of canine ventricular myocardium exposed to temperatures decreasing from 37°C to -30°C demonstrated progressive slowing of conduction to the point of complete block (36). Cooling prolongs the local effective refractory period (ERP) causing conduction delay and block (37). These effects are short-lived, and therefore permit the mapping of focal tachycardias or tachycardia circuits by reversibly interrupting them. CrM has been used in the characterization and treatment of supraventricular and ventricular tachyarrhythmias in humans.
- Zhou L, Keane D, et al. Thrombogenic complications of cardiac radiofrequency catheter ablation. Journal of Cardiovascular Electrophysiology. 1999; 10:611-620.
- Morady F. Catheter Ablation of Supraventricular Arrhythmias: State of the Art. Pacing and Clinical Electrophysiology. 2004; 27:125-142.
- Haines D. The biophysics and pathophysiology of lesion formation during radiofrequency catheter ablation. In Zipes D and Jaliffe J, eds: Cardiac Electrophysiology- From Cell to Bedside. Edition 4. Saunders Publishing Company, 2004, pp. 1018-1027.
- Klein G, Harrison L, Ideker R et.al. Reaction of the myocardium to cryosurgery: electrophysiology and arrhythmogenic potential. Circulation. 1979; 59:364-372.
- Nath S, Lynch C, Whayne J, and Haines D. Cellular electrophysiological effects of hyperthermia on isolated guinea pig papillary muscle. Implications for catheter ablation. Circulation. 1993; 88:1826-1831.
- Hindricks G. on behalf of the multicenter European radiofrequency survey (MERFS) investigators of the working group on arrhythmias of the European society of cardiology. Incidence of complete atrioventricular block following attempted radiofrequency catheter modification of the atrioventricular node in 880 patients. European Heart Journal. 1996; 17:82-88.
- Calkins H, Yong P, Miller J. et.al. Catheter ablation of accessory pathways, atrioventricular nodal reentrant tachycardia, and the atrioventricular junction: final results of a prospective, multicenter clinical trial. Circulation. 1999; 99:262-270.
- Kimman G, Bogaard , Van Hemel N, et.al. Ten-year follow-up after radiofrequency catheter ablation for atrioventricular nodal reentrant tachycardia in the early days forever cured, or a source for new arrhythmias? Pacing and Clinical Electrophysiology. 2005; 28:1302-1309.
- Cox J, Holman W, Cain M. Cryosurgical treatment of atrioventricular node reentrant tachycardia. Circulation. 1987;76:1329-1336.
- Whittaker, D.K., Mechanisms of Tissue Destruction Following Cryosurgery. 1984. Annals of the Royal College of Surgeons of England, 66(5): p. 313-318.
- Gill, W., J. Fraser, and D.C. Carter, Repeated freeze-thaw cycles in cryosurgery. Nature, 1968. 219(5152): p. 410-3.
- Tsvetkov, T., et al., Functional Changes in Mitochondrial Properties As a Result of Their Membrane Cryodestruction 2. Influence of Freezing and Thawing On Atp Complex Activity of Intact Liver Mitochondria. 1985. Cryobiology, 22(2): p. 111-118.
- Petrenko, A., A mechanism of latent cryoinjury and reparation of mitochondria. Cryobiology, 1992. 29(1): p. 144-52.
- Holman, W., et al., Cardiac cryosurgery: Effects of myocardial temperature on cryolesion size. Surgery, 1983. 93(2): p. 268-272.
- Iida, S., T. Misaki, and T. Iwa, The histological effects of cryocoagulation on the myocardium and coronary arteries. Jpn J Surg, 1989. 19(3): p. 319-25.
- Mikat, E.M., et al., Reaction of the myocardium and coronary arteries to cryosurgery. Lab Invest, 1977. 37(6): p. 632-41.
- Harrison, L., et al., Cryosurgical ablation of the A-V node-His bundle: a new method for producing A-V block. Circulation, 1977. 55(3): p. 463-70.
- Mazur, M., Causes of injury in frozen and thawed cells. Fed Proc, 1965. 24: p. 5175-5182.
- Mazur, P., Cryobiology: the freezing of biological systems. Science, 1970. 168(934): p. 939-49.
- McGrath, J.J. Low Temperature Injury Processes. in Advances in Bioheat and Mass Transfer: Microscale Analysis of Thermal Injury Processes, Instrumentation, Modeling, and Clinical Applications. 1993: ASME.
- Stewart, G.J., et al., Hepatic cryotherapy: double-freeze cycles achieve greater hepatocellular injury in man. Cryobiology, 1995. 32(3): p. 215-9.
- Gage, A.A., et al., Effect of varying freezing and thawing rates in experimental cryosurgery. Cryobiology, 1985. 22(2): p. 175-82.
- Klein, G.J., et al., Cryosurgical ablation of the atrioventricular node-His bundle: long-term follow-up and properties of the junctional pacemaker. Circulation, 1980. 61(1): p. 8-15.
- Ghalili, K., et al., Comparison of left ventricular cryolesions created by liquid nitrogen and nitrous oxide. J Am Coll Cardiol, 1992. 20(6): p. 1425-9.
- Brown, N.J., et al., The effect of cryotherapy on the cremaster muscle microcirculation in vivo. 1994. British Journal of Cancer, 69(4): p. 706-710.
- Taylor, C., D. Baldwin, and G. Hass, Localized arteriosclerotic lesions induced in the aorta of the juvenile rabbit by freezing. Arch Pathol, 1950. 49: p. 623.
- Fujino, H., et al., Histologic study of chronic catheter cryoablation of atrioventricular conduction in swine. Am Heart J, 1993. 125(6): p. 1632-7.
- Holman, W.L., et al., Cardiac cryosurgery: regional myocardial blood flow of ventricular cryolesions. J Surg Res, 1986. 41(5): p. 524-8.
- Ohkawa, S., et al., Anatomic effects of cryoablation of the atrioventricular conduction system. Circulation, 1982. 65(6): p. 1155-62.
- Gillette, P.C., et al., Transvenous cryoablation of the bundle of His. Pacing Clin Electrophysiol, 1991. 14(4 Pt 1): p. 504-10.
- Holman, W.L., et al., Ventricular cryosurgery: short-term effects on intramural electrophysiology. Ann Thorac Surg, 1983. 35(4): p. 386-93.
- Klein, G.J., et al., Reaction of the myocardium to cryosurgery: electrophysiology and arrhythmogenic potential. Circulation, 1979. 59(2): p. 364-72.
- 33. Hunt, G.B., et al., Comparison of early and late dimensions and arrhythmogenicity of cryolesions in the normothermic canine heart. J Thorac Cardiovasc Surg, 1989. 97(2): p. 313-8.
- Reek S, Geller C, Schildhaus H, et.al. Feasibility of catheter cryoablation in normal ventricular myocardium and healed myocardial infarction. Pacing and Clinical Electrophysiology. 2004; 27:1530-1539.
- Gallagher, J.D., et al., Cryothermal mapping of recurrent ventricular tachycardia in man. Circulation, 1985. 71(4): p. 732-9.
- Gessman, L., et al., Localization and mechanism of ventricular tachycardia by ice mapping 1 week after the onset of myocardial infarction in dogs. Circulation, 1983. 68(3): p. 657-666.
- Wallace, A. and R. Mignone, Physiologic evidence concerning the reentry hypothesis of ectopic beats. Am Heart J, 1966. 72: p. 60-70.