Biophysics and Pathophysiology of Radiofrequency Lesion Formation

Resistive Heating

The RF energy is a form of alternating electrical current that generates a lesion in the heart by electrical heating of the myocardium. A common form of RF ablation found in the medical environment is the electrocautery, which is used for tissue cutting and coagulation during surgical procedures. The goal of catheter ablation with RF energy is to transform electromagnetic energy into thermal energy in the tissue effectively and to destroy the arrhythmogenic tissues by heating them to a lethal temperature. The mode of tissue heating by RF energy is resistive (electrical) heating. As electrical current passes through a resistive medium, the voltage drops and heat is produced (similar to the heat that is created in an incandescent light bulb). The RF electrical current is typically delivered in a unipolar fashion with completion of the circuit through an indifferent electrode placed on the skin. Typically, an oscillation frequency of 500 to 750 kHz is selected. Lower frequencies are more likely to stimulate cardiac muscle and nerves, resulting in arrhythmias and pain sensation. Higher frequencies will result in tissue heating; however, in the megahertz range, the mode of energy transfer changes from electrical (resistive) heating to dielectric heating (as observed with microwave energy). With very high frequencies, conventional electrode catheters become less effective at transferring the electromagnetic energy to the tissue, and therefore complex and expensive catheter antenna designs must be used.

Resistive heat production within the tissue is proportional to the RF power density, and that, in turn, is proportional to the square of the current density ( Table 1.1 ). When RF energy is delivered in a unipolar fashion, the current distributes radially from the source. The current density decreases in proportion to the square of the distance from the RF electrode source. Thus direct resistive heating of the tissue decreases proportionally with the distance from the electrode to the fourth power ( Fig. 1.1 ). As a result, only the narrow rim of tissue in close contact with the catheter electrode (2–3 mm) is heated directly. All heating of deeper tissue layers occurs passively through heat conduction. If higher power levels are used, both the depth of direct resistive heating and the volume and radius of the virtual heat source will increase.

Thermal Conduction

Most of the tissue heating resulting in the formation of lesion during RF catheter ablation occurs as a result of thermal conduction from the direct resistive heat source. Transfer of heat through tissue follows basic thermodynamic principles and is represented by the bioheat transfer equation. Change in tissue temperature with increasing distance from the heat source is called the radial temperature gradient . At the onset of RF energy delivery, the temperature is very high at the source of heating and falls off rapidly over a short distance (see Fig. 1.1 and Video 1.1, Video 1.2 on Expert Consult). As time progresses, more thermal energy is transferred to deeper tissue layers by thermal conduction. The tissue temperature at any given distance from the heat source increases in a monoexponential fashion over time. Sites close to the heat source have a rapid rise in temperature (a short half-time of temperature rise), whereas sites remote from the source heat up more slowly. Eventually, the entire electrode–tissue system reaches steady state, meaning that the amount of energy entering the tissue at the thermal source equals the amount of energy that is being dissipated at the tissue margins beyond the lesion border. At steady state, the radial temperature gradient becomes constant. If RF power delivery is interrupted before steady state is achieved, tissue temperature will continue to rise in deeper tissue planes as a result of thermal conduction from more superficial layers heated to higher temperatures. In one study, the duration of continued rise in temperature at the lesion border zone after a 10-second RF energy delivery was 6 seconds. The temperature increased by an additional 3.4°C and remained above the temperature recorded at the termination of energy delivery for more than 18 seconds. This phenomenon, termed thermal latency, has important clinical implications because active ablation, with beneficial or adverse effects, will continue for a certain period despite cessation of RF current flow.

Because the mechanism of tissue injury in response to RF ablation is thermal, the final peak temperature at the border zone of the ablative lesion should be relatively constant. Experimental studies predict this temperature with hyperthermic ablation to be approximately 50°C, although alternative methods propose that this critical temperature may be higher. This is called the isotherm of irreversible tissue injury . The point at which the radial temperature gradient crosses the 50°C isothermal line defines the lesion radius in that dimension. One may predict the three-dimensional temperature gradients with thermodynamic modeling and finite-element analysis and by doing so can predict the anticipated lesion dimensions and geometry with the 50°C isotherm. In an idealized medium of uniform thermal conduction without convective heat loss, a number of relationships can be defined using boundary conditions when a steady-state radial temperature gradient is achieved. In this theoretical model, it is predicted that radial temperature gradient is inversely proportional to the distance from the heat source. The 50°C isotherm boundary (lesion radius) increases in distance from the source in direct proportion to the temperature at that source. It was predicted, then demonstrated experimentally, that in the absence of significant heat loss because of convective cooling, the lesion depth and diameter are best predicted by the electrode–tissue interface temperature. In the clinical setting, however, the opposing effects of convective cooling by circulating blood flow diminish the value of electrode-tip temperature monitoring to assess lesion size.

The idealized thermodynamic model of catheter ablation by tissue heating predicted, then demonstrated, that the radius of the lesion is directly proportional to the radius of the heat source ( Fig. 1.2 ). When one considers the virtual heat source radius as the shell of direct resistive heating in tissue contiguous to the electrode, it is not surprising that larger electrode diameter, length, and contact area all result in a larger source radius and larger lesion size, and that this may result in enhanced procedural success. Higher power delivery not only increases the source temperature but also increases the radius of directly heated tissue (i.e., the heat source), thereby increasing lesion size in two ways. These theoretical means of increasing the efficacy of RF catheter ablation have been realized in the clinical setting with large-tip catheters and cooled-tip catheters.

The relationship of ablation catheter distance from the ablation target to the power requirements for clinical effect was tested in a Langendorff-perfused canine heart preparation. Catheter ablation of the right bundle branch was attempted at varying distances, and during the delivery of RF energy, power was increased in a stepwise fashion. The RF power required to block the right bundle branch conduction increased exponentially with increasing distance from the catheter. At a distance of 4 mm, most RF energy deliveries reached the threshold of impedance rise before block was achieved. When pulsatile flow was streamed past the ablation electrode, the power requirements to cause block increased fourfold. Thus the efficiency of heating diminished with cooling from circulating blood, and small increases in distances from the ablation target corresponded with large increases in ablation power requirements, emphasizing the importance of optimal targeting for successful catheter ablation.

Sudden Impedance Rise

While the peak tissue temperature increases during ablation, the temperature at greater tissue depths also increases. A very high source temperature therefore should theoretically yield a very deep 50°C isotherm temperature and, in turn, very large ablative lesions. Unfortunately, this process is prevented in the biological setting because of the formation of coagulum and char at the electrode–tissue interface when temperatures exceed 100°C. At 100°C, blood literally begins to boil. This can be observed in the clinical setting with generation of showers of microbubbles if tissue heating is excessive. As the blood and tissue in contact with the electrode catheter desiccate, the residue of denatured proteins adheres to the electrode surface. These substances are electrically insulating and result in a smaller electrode surface area available for electrical conduction. In turn, the same magnitude of power is concentrated over a smaller surface area, and the power density increases. With higher power density, the heat production increases, and more coagulum is formed. Thus in a positive-feedback fashion, the electrode becomes completely encased in coagulum within 1 to 2 seconds. In a study testing ablation with a 2-mm-tip electrode in vitro and in vivo, a measured temperature of at least 100°C correlated closely with a sudden rise in electrical impedance ( Fig. 1.3 ). All modern RF energy ablation systems have an automatic energy cutoff if a rapid rise in electrical impedance is observed. Some experimenters have described accumulation of soft thrombus when temperatures exceed 80°C. This is likely caused by blood protein denaturation and accumulation, but fortunately appears to be more of a laboratory phenomenon than one observed in the clinical setting. When high temperatures and sudden rises in electrical impedance are observed, there is concern about the accumulation of char and coagulum, with the subsequent risk of char embolism. Reports of asymptomatic cerebral embolic lesions on diffusion weighted imaging–magnetic resonance imaging (MRI) images after atrial fibrillation ablation highlight the clinical significance of microembolism. Anticoagulation and antiplatelet therapies have been proposed as preventative measures, but meticulous sheath management avoidance of excessive heating at the electrode–tissue interface remains the best strategy to avoid this risk.

Convective Cooling

The major thermodynamic factor opposing the transfer of thermal energy to tissue is convective cooling. Convection is the process in which heat is rapidly distributed through a medium by active mixing of that medium. In the case of RF catheter ablation, the heat is produced by resistive heating and transferred to deeper layers by thermal conduction. Simultaneously, the heat is conducted back into the circulating blood pool and metal electrode tip. Because the blood is moving rapidly past the electrode and over the endocardial surface, and because water (the main constituent of blood) has a high heat capacity, a large amount of the heat produced at the site of ablation can be carried away by the blood. Convective cooling is such an important factor that it dominates the thermodynamics of catheter ablation. Efficiency of energy coupling to the tissue can be as low as 10%, depending on electrode size, catheter stability, and position relative to intracavitary blood flow. Unstable, sliding catheter contact results in significant tip cooling and decreased efficiency of tissue heating. This is most often observed with ablation along the tricuspid or mitral valve annuli, or on the left pulmonary vein ridge.

Paradoxically, the convective cooling phenomenon has been used to increase lesion size. As noted earlier, maximal power delivery during RF ablation is limited by boiling of blood and coagulum formation at the electrode tip. However, if the tip is cooled, a higher magnitude of power may be delivered without a sudden rise in electrical impedance. The higher magnitude of power increases the depth of direct resistive heating and, in turn, increases the radius of the effective heat source. In addition, higher temperatures are achieved 3 to 4 mm below the surface, and the entire radial temperature curve is shifted to a higher temperature over greater tissue depths. The result is a greater 50°C isotherm radius and a greater depth and diameter of the lesion. Nakagawa and coworkers demonstrated this phenomenon in a blood-superfused exposed thigh muscle preparation. In this study, intramural tissue temperatures 3.5 mm from the surface averaged 95°C with an irrigated-tip catheter despite a mean electrode–tissue interface temperature of 69°C. Lesion depths were 9.9 mm compared with 6.1 mm in a comparison group of temperature-feedback power control delivery with no electrode irrigation ( Fig. 1.4 ). An important finding of this study was that six of 75 lesions had a sudden rise in electrical impedance associated with an audible pop. In these cases, the intramural temperature exceeded 100°C, resulting in sudden steam formation and a steam pop. The clinical concern about pop lesions is that sudden venting of steam to the endocardial or epicardial surface (or both) can potentially cause perforation and tamponade. Monitoring intramyocardial steam formation with near field ultrasound to terminate energy delivery before steam venting can occur has been proposed as a method of mitigating this risk.

The observation of increasing lesion size with ablation-tip cooling holds true only when the ablation is not power limited. If the level of power used is insufficient to overcome the heat lost by convection, the resulting tissue heating may be inadequate. In this case, convective cooling will dissipate a greater proportion of energy, and less of the available RF energy will be converted into tissue heat. The resulting lesion may be smaller than it would be if there was no convective cooling. As power is increased to a higher level, more energy will be converted into tissue heat, which results in larger lesions. If power is unlimited and temperature-feedback power control delivery is used, greater magnitudes of convective cooling will allow for higher power levels and very large lesions. Thus paradoxically in this situation, lesion size may be inversely related to the electrode–tissue interface temperature if the ablation is not power limited. However, if power level is fixed (most commercial RF generators limit power delivery to 50 W for use with standard catheters), lesion size increases in proportion to the electrode–tissue interface temperature even in the setting of significant convective cooling ( Fig. 1.5 ).

The magnitude of convective cooling that is achieved with irrigated catheters is relatively small compared to the circulating blood pool, but this cooling can occur in the highly localized region at the electrode–tissue interface. The main effect of irrigation is to prevent excessive surface heating, boiling, char formation, and impedance rise despite use of high power amplitude. High versus low irrigation rates do not affect the ablation lesion depth if power is constant. However, high irrigation rates do result in smaller lesion diameter on the endocardial surface. Thus if thin-walled tissues are being ablated (e.g., posterior left atrial wall), then using high irrigation flow rate will not alter risk of injury to collateral structures, but it will decrease ablation efficacy by reducing superficial lesion size.

Electrode-tip cooling can be achieved passively or actively. Passive tip cooling occurs when the circulating blood flow cools the mass of the ablation electrode and cools the electrode–tissue interface. This can be enhanced by using a large ablation electrode or by using an electrode material with high thermal conductivity, such as gold or diamond. Active tip cooling can be realized with a closed or open perfused-tip system. One design recirculates the saline through a return port, and the opposing design infuses the saline through weep holes in the electrode into the bloodstream. Both designs are effective and result in larger lesions and greater procedure efficacy than standard RF catheter ablation, although open irrigation is preferred because operators observe less tendency for formation of thrombus compared to closed systems. Present-day electrode geometries vary considerably, tip irrigation is used routinely, and standard catheters embed their thermocouples within the mass of the electrode tip, thereby providing evidence of tip cooling with irrigation, but no ability to detect tissue heating. A new generation of temperature sensing catheters offers to improve this scenario. Six miniature thermocouple sensors have been positioned immediately below the electrode surface and distributed around the tip of a force sensing catheter, significantly improving ability to predict catheter orientation and lesion depth.

Because the peak tissue temperature is shifted from the endocardial surface to deeper intramyocardial layers, there is a risk of excessive intramural heating and pop lesions. The challenge for the clinician lies with the fact that with varying degrees of convective cooling, there is no reliable method for monitoring whether tissue heating is inadequate, optimal, or excessive. Cooling at the electrode–tissue interface limits the value of temperature monitoring to prevent excess power delivery and steam pops. New technologies such as MR thermography or near-field ultrasound-guided ablation may allow the operator to visualize lesion formation real time during energy delivery and more precisely adjust power and magnitude of convective cooling to optimize lesion formation. Catheter ablation in the pericardial space is a unique condition. Because there is no circulating blood, there is no convective cooling whatsoever. Ablations performed with conventional RF catheters yield very small lesions. Perfused-electrode catheters are usually used in this setting to provide some surface cooling and allow ablation at higher powers. In particular, linear ablation tools designed for surgical ablation of atrial fibrillation from the epicardial surface require cooling to achieve transmural atrial lesions.

Catheter Contact Force and Orientation

It has long been appreciated that electrode–tissue contact force is an important factor in successful RF energy lesion formation. With increasing force, greater proportion of the electrode surface area is in contact with the tissue, and there is more efficient energy coupling. In addition, with thin-walled tissues, the endocardial surface is slightly depressed with increasing contact force and the tissue is somewhat thinner at the contact point. This increases the likelihood that the RF lesion will be transmural. In recent years, force-sensing catheters have been developed that use either fiber–optic or piezoelectric components attached to a flexible catheter tip that can accurately measure the force applied to the tip electrode of the catheter. Lesions created with higher contact force (>20 g–force) were larger and required lower powers than lesions created with lower contact force. Contact force catheters have also been used to determine that the average force needed for atrial perforation in a swine model was 175 g–force (range, 77–376 g–force). Force-sensing catheters have been tested extensively in the clinical setting. Successful catheter ablation has been associated with higher contact forces, and higher forces that have been applied over longer durations (the force-time integral). However, there is a concern that using greater contact force and longer durations during ablation of the posterior left atrial wall may have contributed to a higher risk of atrial-esophageal fistula.

Catheter orientation will affect lesion size and geometry. Perpendicular catheter orientation results in less electrode surface area in contact with the tissue and more surface area in contact with the circulating blood pool. Parallel catheter orientation provides more electrode–tissue contact. With unrestricted power delivery, the parallel orientation should produce the larger lesion. In perfused-tip catheters, parallel orientation also results in more active tissue cooling and smaller lesion sizes than a perpendicular orientation. The resultant interplay among active cooling, passive cooling, and power availability or limitation determines whether the lesions will be larger or smaller in these varying conditions. If perfused-tip catheters are positioned in a parallel orientation with greater tissue cooling, the lesions are smaller in vitro because of diminished efficiency of energy delivery. The effects of catheter orientation are less important with 4- or 5-mm-tip catheters but become more dominant when 8- or 10-mm tips are used.

Electrical Current Distribution

Catheter ablation depends on the passage of RF electrical current through tissues. As tissue contact improves, the impedance of the RF electrical circuit decreases since there is lower impedance at an electrode–tissue interface than at an electrode-blood interface. A strong correlation is observed between effective lesion formation and rate of impedance fall during energy delivery because hotter tissue has a lower impedance than cooler tissue. When electrode–tissue interface temperature monitoring is unreliable because of high-magnitude convective cooling, an observed impedance drop is a useful sign that tissue heating is occurring. An initial impedance drop greater than 10 Ω is an indicator of good catheter contact as assessed by force-sensing catheters. With the progressive fall in impedance during ablation, the delivered current increases along with tissue heating. If no impedance drop is observed, catheter repositioning is warranted.

The magnitude of tissue heating is determined by the current density; in turn, the distribution of RF field around the electrodes in unipolar, bipolar, or phased RF energy delivery will determine the distribution of tissue heating. If energy is delivered in a unipolar fashion in an isotropic medium from a spherical electrode to an indifferent electrode with infinite surface area, current density around the electrode should be entirely uniform. As geometries and tissue properties change, heating becomes nonuniform. Standard 4- or 5-mm electrode tips are small enough so that heating around the tip is fairly evenly distributed, even with varying tip contact angle to the tissue. One study of ablation with a nonirrigated catheter equipped with multiple surface mounted thermistors showed that the single thermistor located at the tip of the 4-mm electrode accurately represented the maximum recorded temperature from all thermistors 96% of the time, and failed to predict sudden impedance rise in only one of 17 of cases where that occurred.

It has been proposed that power distribution during RF ablation can be modulated by altering the tonicity of the irrigation solution. RF power is dissipated throughout the entire circuit, including the catheter, catheter-tissue interface, myocardium, tissue between the heart and the dispersive electrode, skin-dispersive electrode interface, wire conductors, and RF generator electronics. If more power is channeled through this tissue rather than directly into the blood pool, lesion size should be larger. This was accomplished by Nguyen et al. by replacing 0.9% saline with 0.45% saline as the irrigation liquid. The hypotonic solution produced a higher impedance environment in the blood pool around the electrode and resulted in a greater proportion of RF current passing directly into the tissue. This resulted in 60% larger lesion volumes in experimental testing in vivo. This may be a useful strategy for ablation of deep intramyocardial targets.

Fat is distributed widely on the epicardium of the heart and reduces both electrical and thermal conductivity. Epicardial ablation over fat will result in minimal ablation of the underlying myocardium. Conversely, ablation of tissue insulated by fat outside of the ablation target will produce an insulating effect, with higher temperatures for longer durations after cessation of energy delivery.

Bipolar ablation uses a second active ablation electrode rather than a dispersive electrode, and the current flows between the two electrodes with heating occurring at both electrodes. For symmetrical ablation at both active electrodes of the bipole, it is important that the electrodes are similar in size, because heating is proportional to power density, which is a function of both power and electrode surface area. With bipolar delivery, the electrical field is densest between the electrodes, and so some additional volume heating may be achieved in the intervening tissues if the interelectrode spacing is close. If energy is delivered in a bipolar mode between contiguous electrodes on a catheter positioned parallel to the tissue, there may be improved lesion formation between the electrodes, but lesion depth will be less than that achieved with multipolar ablation in the unipolar mode. It is possible to deliver both unipolar RF energy to multiple poles and bipolar energy between poles simultaneously by altering the phase of RF signal between the two electrodes, or by using a duty cycle to alternate from unipolar to bipolar. The blended unipolar–bipolar RF lesion from contiguous electrodes will be deeper than a pure bipolar ablation but more continuous than a pure unipolar ablation.

Dispersive Electrode

For unipolar RF energy delivery, the power dissipated in the complete electrical circuit is proportional to the impedance and voltage drop for each component of the series circuit. The impedances of the ablation system generator electronics and transmission lines are low relative to the impedance of the tissue interposed between the catheter and the dispersive electrode, so most of the energy dissipation occurs within the body. The site of greatest impedance, voltage drop, and power dissipation is at the electrode–tissue interface. However, most power is consumed with electrical conduction through the body and blood pool and into the dispersive electrode. In fact, only a fraction of the total delivered power is actually deposited in the myocardial tissue ( Fig. 1.6 ). The return path of current to the indifferent electrode will certainly affect the current density close to that indifferent electrode, but its placement anterior versus posterior, and high versus low on the torso, has only a small effect on the distribution of RF current field lines within millimeters of the electrode. Therefore lesion geometry should not be affected greatly by dispersive electrode placement. However, the proportion of RF energy contributing to lesion formation will be reduced if a greater proportion of that energy is dissipated in a long return pathway to the dispersive electrode. When the ablation is power limited, it is advantageous to minimize the proportion of energy that is dissipated along the current pathway at sites other than the electrode–tissue interface to achieve the greatest magnitude of tissue heating and the largest lesion. In an experiment that tested placement of the dispersive electrode directly opposite to the ablation electrode versus at a more remote site, lesion depth increased by 26% with optimal placement. Vigorous skin preparation to minimize impedance at the skin interface with the dispersive electrode, closer placement of the dispersive electrode to the heart, and use of multiple dispersive electrodes to increase skin contact area will all increase tissue heating in a power-limited energy delivery. Nath and associates reported that in the setting of system impedance higher than 100 Ω, adding a second dispersive electrode increased the peak electrode-tip temperature during clinical catheter ablation ( Fig. 1.7 ).

The dispersive electrode has a large surface area relative to that of the ablation electrode so that the power density at the skin surface is uniformly low. As a consequence, there is minimal skin heating during RF catheter ablation. However, if high powers are used, and/or the contact surface area of the dispersive electrode is reduced (e.g., a partially detached electrode), excess power density with resultant skin heating can occur. Case reports of serious skin burns from the dispersive electrode emphasize this point. Sequential activation of two ground pads to allow intermittent cooling of each pad results in lower skin temperatures during high-power delivery.

Edge Effect

Electrical field lines are not entirely uniform around the tip of a unipolar ablation electrode. The distribution of field lines from an electrode source is affected by changes in electrode geometry. At points of geometric transition, the field lines become more concentrated. This so-called edge effect can result in significant nonuniformity of heating around electrodes. The less symmetrical the electrode design (such as the ones found with long electrodes), the greater the degree of nonuniform heating. McRury and coworkers tested ablation with 12.5-mm length electrodes and found that a centrally placed temperature sensor significantly underestimated the peak electrode–tissue interface temperature. Finite-element analysis demonstrated a concentration of electrical current at each edge of the electrode ( Fig. 1.8 ). When dual thermocouples were placed on the edge of the electrode, the risk of coagulum formation and impedance rise was significantly reduced during ablation testing in vivo.

Gross Pathology and Histopathology of the Ablative Lesion

The endocardial surface in contact with the ablation catheter shows pallor and sometimes a small depression caused by volume loss of the acute lesion. If excessive power has been applied, there may be visible coagulum or char adherent to the ablation site. On sectioning the acute lesion produced by RF energy, a central zone of pallor and tissue desiccation characterizes its gross appearance. There is volume loss, and the lesion frequently has a teardrop shape with a narrower lesion width immediately subendocardially and a wider width 2 to 3 mm below the endocardial surface. This is because of surface convective cooling by the endocardial blood flow. Immediately outside the pale central zone is a band of hemorrhagic tissue. Beyond that border, the tissue appears relatively normal. The acute lesion border, as assessed by vital staining, correlates with the border between the hemorrhagic and normal tissue ( Fig. 1.9 ). The histologic appearance of the lesion is consistent with coagulation necrosis. There are contraction bands in the sarcomeres, nuclear pyknosis, and basophilic stippling consistent with intracellular calcium overload.

The temperature at the border zone of an acute hyperthermic lesion assessed by vital staining with nitro blue tetrazolium was observed to be 52°C to 55°C in one study, and 60°C in another. However, it is likely that the actual isotherm of irreversible thermal injury occurs at a lower temperature boundary outside the lesion boundary, and that it cannot be identified acutely. Coagulation necrosis is a manifestation of thermal inactivation of the contractile and cytoskeletal proteins in the cell. Changes in the appearance of vital stains are caused by loss of enzyme activity, as is the case with nitro blue tetrazolium or triphenyl tetrazolium chloride staining and dehydrogenase activity. Therefore the acute assessment of the lesion border represents the border of thermal inactivation of various proteins, but the ultimate viability of the cell may depend on the integrity of more thermally sensitive organelles such as the plasma membrane (see later discussion). In the clinical setting, recorded temperature does correlate with response to ablation. In patients with manifest Wolff–Parkinson–White syndrome, the reversible accessory pathway conduction block with a nonirrigated catheter was observed at a mean electrode temperature of 50°C ± 8°C, whereas the permanent block occurred at a temperature of 62°C ± 15°C. In a study of electrode-tip temperature monitoring during atrioventricular junctional ablation, an accelerated junctional rhythm was observed at a mean temperature of 51°C ± 4°C. Permanent complete heart block was observed at ablation temperatures of 60°C ± 7°C. Because the targeted tissue was likely millimeters below the endocardial surface, the temperatures recorded by the catheter were expected to be higher than those achieved intramurally at the critical site of ablation.

The subacute pathology of the RF lesion is similar to what is observed with other types of injury. Although the appearance of typical coagulation necrosis persists, the lesion border becomes more sharply demarcated with infiltration of mononuclear inflammatory cells. A layer of fibrin adheres to the lesion surface, coating the area of endothelial injury. After 4 to 5 days, the transition zone at the lesion border is lost, and the border between the RF lesion and surrounding tissue becomes sharply demarcated. The changes in the transition zone within the first hours and days after ablation likely account for the phenomena of early arrhythmia recurrence (injury with recovery) or delayed cure (progressive injury caused by the secondary inflammatory response). The coagulation necrosis in the body of the lesion shows early evidence of fatty infiltration. By 8 weeks after the ablation procedure, the necrotic zone is replaced with fatty tissue, cartilage, and fibrosis and can be surrounded by chronic inflammation. The chronic RF ablative lesion evolves into a uniform scar. Like any fibrotic scar, there is significant contraction of the scar with healing. Relatively large and wide acute linear lesions have the final gross appearance of narrow lines of glistening scar when examined 6 months after the ablation procedure. The uniformity of the healed lesion accounts for the absence of any proarrhythmic effect of RF catheter ablation, unless multiple lesions with gaps are made. A group of patients who underwent pulmonary vein ablation and had clinical recurrence had full-thickness pulmonary vein antral biopsies at the time of a follow-up surgical maze procedure. Fifty percent of those specimens showed viable myocardium with or without scar on histopathologic analysis, explaining the reestablishment of pulmonary vein conduction after the acute catheter procedure.

Radiofrequency Lesion Ultrastructure

The ultrastructural appearance of the acute RF lesion offers some insight into the mechanism of tissue injury at the lesion border zone. In cases of experimental RF ablation in vivo, ventricular myocardium was examined in a band 3 mm from the edge of the acute pathologic lesion as defined by vital staining ( Fig. 1.10 ). It showed marked disruption in cellular architecture characterized by dissolution of lipid membranes and inactivation of structural proteins. The plasma membranes were severely disrupted or missing. There was extravasation of erythrocytes and complete absence of basement membrane. The mitochondria showed marked distortion of architecture with swollen and discontinuous cristae membranes. The sarcomeres were extended with loss of myofilament structure or were severely contracted. The T tubules and sarcoplasmic reticulum were either absent or severely disrupted. Gap junctions were also either severely distorted or absent. Thus despite the fact that the tissue examined was outside of the border of the acute pathologic lesion, the changes were profound enough to conclude that some progression of necrosis would occur within this border zone. The band of tissue 3 to 6 mm from the edge of the pathologic lesion was examined, and it showed significant ultrastructural abnormalities, but not as severe as those described closer to the lesion core. Severe abnormalities of the plasma membrane were still present, but gap junctions and mitochondria were mainly intact. The sarcomeres were variable in appearance, with some being relatively normal and some partially contracted. Although ultrastructural disarray was observed in the 3- to 6-mm zone, the myocytes appeared to be viable and would likely recover from the injury.