Incidence

Previous studies reported the incidence of cardiac perforation with catheter-based interventions as follows: 0.8% for all procedures, 1.5% to 4.7% for valvuloplasty, 0.5% to 0.8% for angioplasty-atherectomy, 0.01% for diagnostic catheterization, 0.1% to 0.2% for electrophysiological (EP) studies, 0.2% for SVT ablation, 0.4% to 2.7% for VT ablation, and 0.5% to 4.0% for left atrial (LA) ablation.^1,4,5

The incidence of procedure-related cardiac tamponade increases with the increasing number of catheter ablation procedures for the treatment of AF, which typically involves one or more atrial septal punctures, extensive intracardiac catheter manipulation and ablation within the thin-walled LA, and the need for high levels of systemic anticoagulation during the procedure.^2,6

Mechanism

Cardiac perforation can be caused by mechanical trauma secondary to catheter manipulation, myocardial tissue rupture due to radiofrequency (RF) ablation, or inadvertent injury during atrial septal puncture. Extensive catheter manipulation, especially in areas with thin walls such as the LA roof, right ventricular (RV) free wall, RV outflow tract (RVOT), RV apex, and atrial appendages, can cause mechanical tear of the chamber wall. Cardiac perforation can also occur away from the area of mapping and ablation, usually caused by penetration of an unattended RV catheter in the setting of rapid and vigorous heart action because of high-dose isoproterenol infusion. Inadvertent puncture of the right atrial (RA) free wall or lateral LA wall can occur during transseptal catheterization. High RF power output, especially when using large- or irrigated-tip ablation catheters, can increase the risk of cardiac perforation. An audible pop associated with an abrupt rise in impedance is heard in many patients who develop tamponade. Popping occurs because of tissue boiling causing myocardial rupture, and is increased by irrigated-tip ablation, high tissue-catheter interface flow, poor or unstable tissue contact, and high catheter tip temperature.^2,6

Detection

Prompt recognition and management of cardiac perforation are critical to prevent the development of cardiac tamponade and potentially life-threatening hemodynamic collapse. Although the presentation often is dramatic with abrupt hypotension, it can be insidious with a more gradual fall in blood pressure. The use of an arterial line that provides continuous blood pressure monitoring can help detect early hemodynamic compromise. Sinus tachycardia is common in the setting of cardiac tamponade that gradually develops. Nonetheless, the absence of tachycardia does not exclude cardiac tamponade; in fact sinus bradycardia can develop secondary to a vagal reflex in the setting of a rapidly developing cardiac tamponade. Any chest pain that persists beyond the completion of an ablation lesion, especially if associated with hypotension and diaphoresis, should alert the operator to the possible development of pericardial effusion.^2,6–8

Assessment of the cardiac silhouette fluoroscopically can provide the first clue, especially if a similar assessment was made at baseline. Decreased excursion of the lateral heart border on fluoroscopy in the left anterior oblique (LAO) projection, indicating accumulating pericardial effusion, usually can be seen well before a drop in blood pressure and prior to progression to cardiac tamponade. Some operators obtain a baseline LAO cine image at the onset of a procedure to serve as a reference for comparison during the procedure, followed by intermittent evaluation of the same fluoroscopic projection during the procedure.^2,6–8

Transthoracic echocardiography is the most definitive method for confirming the development of pericardial effusion. Intracardiac echocardiography (ICE), commonly used to guide transseptal puncture for LA procedures, also can facilitate the early detection of pericardial effusions before the emergence of tamponade physiology. Most of the pericardial space is not visualized from the RA imaging venue used to visualize the interatrial septum, and limited catheter rotation is usually required. Advancing the ICE catheter into the RV and rotating the transducer against the interventricular septum can readily identify pericardial effusions (Fig. 32-1).

FIGURE 32-1 Intracardiac echocardiography (ICE) imaging of pericardial effusion. The phased-array ICE transducer is positioned in the right ventricle, showing a short-axis view of the left ventricle (LV) and a moderate pericardial effusion (PE). IVS = interventricular septum.

Perforation without tamponade may occur during placement of diagnostic catheters or while accessing the LA by transseptal catheterization; detection at this time can prevent progression to tamponade but requires a high index of suspicion. Several clues should alert the operator to the possibility of cardiac perforation, including a ventricular catheter that reaches the edge of the cardiac silhouette, high pacing thresholds at sites despite normal electrograms and apparent good tissue contact on fluoroscopy, a right bundle branch block (RBBB) complex during “RV” pacing, and a catheter tip position discordant from where it should be (i.e., an RV catheter too far leftward with intended RV apical location; Fig. 32-2). With atrial septal puncture for accessing the left heart chambers, the aorta can be inadvertently entered; if only the needle enters the aorta and this is recognized before advancing the dilator and sheath, the needle can be withdrawn and the patient monitored for stability of vital signs and with echocardiography. The procedure can be continued if there is no accumulation of pericardial fluid after 15 to 30 minutes of monitoring. If the dilator and sheath have been advanced into the aortic root before the error is recognized, it is imperative that the sheath not be removed immediately, as this can result in immediate and perhaps irretrievable hemodynamic collapse due to intractable bleeding into the pericardial space.

FIGURE 32-2 Cardiac perforation during electrophysiological study. Surface ECG and intracardiac recordings from a patient who was undergoing electrophysiological study. Recordings show poor capture from a catheter intended for the right ventricular apex (RVA); captured complexes have a right bundle branch block pattern. These features are explained by the accompanying right anterior oblique fluoroscopic image showing that the RVA catheter (arrow) has perforated the RV anterior wall and is pacing the anterolateral left ventricular epicardium.

Management

The management of pericardial effusion is largely determined by its relative size and hemodynamic effect. Trivial pericardial effusions, if recognized early during the procedure, should be monitored continuously but do not warrant termination of the procedure. For larger effusions, the procedure should be terminated and anticoagulation, if administered, should be reversed. Protamine is used to reverse the effects of heparin, and activated factor VII, fresh frozen plasma, and vitamin K can be used in patients with therapeutic anticoagulation with warfarin. Intravenous fluids, vasopressors, and transfusion of blood products can be required, depending on the extent of the effusion and the severity of hemodynamic decompensation. Intravenous atropine can be of value in patients with increased vagal tone.^3,9

If perforation without tamponade is recognized, an echocardiogram should be obtained as a baseline. If the offending catheter is a standard 5 or 6 Fr shaft, it can often be withdrawn back into the heart while monitoring the echocardiogram for accumulation of pericardial fluid. Most often, there will be no bleeding into the pericardial space and the procedure can be continued. If systemic anticoagulation is planned for left-sided catheterization, the operator must decide how important it is to continue the procedure with the risk of bleeding into the pericardium. If a larger catheter or large vascular sheath has been inadvertently placed in the pericardial space, echocardiography should be obtained along with cardiothoracic surgical consultation. In many if not most cases, the sheath may be safely withdrawn into the heart without adverse consequences, but the team must be ready to transport the patient to an operating room for repair of a hole or tear in the wall of the affected heart chamber. In some cases, especially when the patient is fully anticoagulated, it may be prudent to transport the patient to the operating room and prepare for urgent sternotomy before removing the catheter/sheath so as to be able to rapidly enter the chest if hemodynamic collapse occurs.

Although pericardiocentesis typically is required for large pericardial effusions, smaller effusions manifesting signs of cardiac tamponade, or both, and can effectively restore hemodynamic function, it carries the potential risk of cardiac chamber laceration, inadvertent puncture of the RV, pneumothorax, and infection. In a subset of patients (especially older female patients with a small to moderate amount of pericardial effusion who are not anticoagulated), a conservative approach incorporating administration of intravenous fluids and vasopressors to address the hemodynamic consequences of cardiac tamponade was found in a recent report a reasonable initial strategy and could obviate the need for emergency pericardiocentesis.^2,7

Pericardiocentesis should be performed promptly when indicated, because there is usually a narrow therapeutic time window for intervention before critical hemodynamic compromise ensues. The procedure can be performed under fluoroscopic guidance (Video 22) . Echocardiography can also be used to guide pericardiocentesis (Video 23 ). When echocardiography is not readily available, fluoroscopy is usually adequate to guide a safe procedure, and reliance on transthoracic echocardiography should not lead to unnecessary delay in diagnosis or therapy. The technique of percutaneous pericardiocentesis is discussed in Chapter 27. In most patients, an indwelling catheter is required for a short interval after initial drainage to confirm that the bleeding has stopped and that no effusion is reaccumulating. In the event of persistence or rapid reaccumulation of the effusion, exploration by thoracic surgery can be required. Up to 13.3% of LA perforations during AF ablation require surgical closure. Autotransfusion of blood removed from the pericardial space can be of value in patients with persistent bleeding, and is best done using an autologous blood recovery system, because direct autotransfusion can result in a systemic inflammatory response.^3,6,9

Early pericarditis after cardiac perforation is common. In one report, 53.3% of such patients had persistent chest pain after effusion evacuation and removal of the pericardial catheter suggestive of pericardial inflammation. Nonsteroidal antiinflammatory agents are adequate in most patients. Additionally, intrapericardial steroids (triamcinolone, 20 mg) can help reduce pericardial inflammation. Subacute reaccumulation of pericardial fluid suggestive of postcardiac injury syndrome or inflammatory pericarditis can also occur, requiring repeat pericardiocentesis.^3,9

Mechanism

Potential sources of emboli include thrombus formation on high-profile wires, catheters, and sheaths inserted in the LA or LV, char formation at the tip of the ablation catheter or at the site of ablation, thrombi or air passing through a patent foramen ovale or transseptal puncture, and dislodgment of a preexistent LA or LV thrombus by catheter manipulation. Additionally, new LA appendage thrombi may develop in patients with persistent AF after planned or inadvertent conversion to sinus rhythm, especially in the setting of insufficient levels of anticoagulation before, during, or after ablation. Furthermore, endocardial disruption from the ablation lesions can potentially become a nidus for thrombus formation. Also, aortic atheroembolism can occur during retrograde transaortic LV access for ablation of VTs or atrioventricular (AV) bypass tracts (BTs), and it tends to be associated with difficulty in manipulation of catheters within a severely diseased aorta.¹⁰

Thromboembolic events typically occur within 24 hours of the ablation procedure, with the high-risk period extending for the first 2 weeks following ablation. Cerebral thromboembolism is most common, but emboli can also involve the coronary, abdominal, or peripheral vascular circulations. Although silent cerebral thromboembolism has been reported, its incidence and clinical significance are unknown.¹¹

Prevention

Prevention remains the best strategy in minimizing cerebrovascular events during left heart mapping and ablation, and this may be achieved by the following: (1) preprocedural transthoracic or transesophageal echocardiography (TEE) in patients with AF as well as those with ischemic VT with LV dysfunction; (2) aggressive intraprocedural anticoagulation, including early heparin administration (once vascular access is obtained), followed by continuous infusion to maintain the activated clotting time at greater than 300 seconds; (3) meticulous attention to sheath management, including constant infusion of heparinized saline and air filters; (4) minimizing char formation during lesion creation by regulating power delivery to prevent abrupt impedance rise; and (5) using ICE during AF ablation for early detection of intracardiac thrombi and accelerated bubble formation consistent with endocardial tissue disruption with RF application. Open-irrigation RF ablation or cryoablation, compared with standard 4- or 8-mm solid tip or closed irrigation electrodes, can potentially decrease the formation of char and thrombus at the tip of the ablation catheter.¹⁰ Administration of large doses of protamine on completion of the ablation procedure to reverse heparin abruptly can potentially promote thrombogenesis and warrants further evaluation to confirm its safety. Continuation of warfarin at a therapeutic level at the time of AF ablation can potentially be a better alternative to strategies that use bridging with heparin or enoxaparin, as it eliminates a period of inadequate anticoagulation immediately following the ablation procedure, a critical period for thromboembolic risk because of the inflammation and irritation inherently associated with ablation.^12,13

In patients with severe disease of the aorta, using a long vascular sheath to advance the ablation catheter directly into the LV can potentially reduce the risk of aortic atheroembolism.

Mechanism

Small amounts of air embolized in the venous circulation are generally broken up in the capillary bed and absorbed from the circulation without significant sequelae. Embolization of large volumes of air (>5 mL/kg) can cause severe complications (shock or cardiac arrest). However, complications have been reported with as little as 20 mL of air. On the other hand, embolization of as little as 2 or 3 mL of air into the arterial circulation can be fatal.

Paradoxical air embolization into the arterial circulation can occur through direct passage of air into the arterial system via anomalous structures such as an atrial or ventricular septal defect, a patent foramen ovale, or pulmonary arterial-venous malformations. Direct arterial air embolism is caused by the introduction of air into the LA via the transseptal sheath as well as via long vascular sheaths occasionally used for catheter stabilization during transaortic mapping and ablation in the LV.^3,9

With the patient in the supine position, large volumes of air rapidly entering the venous circulation typically lodge in the RVOT or pulmonary artery, and can cause RVOT obstruction, marked reduction of cardiac output, and potentially hemodynamic collapse. Also, air embolization can lead to serious inflammatory changes in the pulmonary vessels including direct endothelial damage and accumulation of platelets and fibrin. Air in the systemic circulation can induce ischemia by various mechanisms, such as obstruction of the blood flow, vasospasm, and thrombus formation because of platelet activation.

Clinical Presentation

Most cases of venous air embolism are subclinical and do not result in untoward outcomes, and even when symptomatic, they go unrecognized because of the nonspecific nature of clinical presentation that can mimic other cardiac, pulmonary, and neurological dysfunctions. Therefore, a high index of suspicion is necessary to establish the diagnosis.³

The outcome of venous air embolism is directly related to the amount of air and the rate at which it enters the vein. Spontaneously breathing patients can experience more serious consequences than those under controlled positive-pressure ventilation because they generate negative intrathoracic pressure during the respiratory cycle, facilitating air entrainment. Awake patients typically manifest shortness of breath, continuous coughing, chest pain, and a sense of “impending doom.” Jugular venous distention, hypotension, tachycardia, and electrocardiographic (ECG) signs of right heart strain (ST-T wave abnormalities) can be observed. Severe cases are characterized by cardiovascular collapse.^3,9

Arterial air emboli can distribute to almost any organ and can have devastating clinical sequelae. Direct cerebral air embolism can be associated with altered mental status, seizures, and focal neurological signs. A common presentation of air embolism during LA mapping and ablation is acute inferior ischemia, heart block, or both (Fig. 32-3). This reflects preferential downstream migration of air emboli into the right coronary artery.

FIGURE 32-3 Coronary air embolism. Leads II and V₁ are shown from a patient with a left-sided bypass tract who suffered air embolism during transseptal catheterization and catheter advancement into the left atrium. Snapshots of ECGs taken at the designated times show the baseline (10:33) normal findings, dramatic ST elevation immediately following transseptal catheterization (10:46), and progressive resolution over the next 25 minutes. Time scale and 10-mV calibration marks are shown.

Detection

Routine diagnostic modalities to identify air embolism in the terminal arterial circulations lack sensitivity, and diagnosis is typically based on the appropriate clinical scenario, with possible air identified in cardiac chambers. Air in the RV or pulmonary artery can be visualized on fluoroscopy as well as transthoracic echocardiography. TEE is currently the most sensitive monitoring device for venous air embolism. Prompt CT or magnetic resonance (MR) imaging obtained before the intravascular air is absorbed may show multiple serpiginous hypodensities representing air in the cerebral vasculature, with or without acute infarction, but imaging later in the course of the disease shows diffuse acute infarcts (Fig. 32-4).^3,9

FIGURE 32-4 Cerebral air embolism. A transverse section of a magnetic resonance image of the brain in a patient who sustained cerebral air embolism during ablation of atrial fibrillation, showing many areas of diffusion restriction scattered throughout both hemispheres, representing acute infarcts.

Prevention

The optimal management of air embolism is prevention. Careful sheath management, including constant infusion of heparinized saline and air filters, should be observed. Although air can be introduced through the infusion line, it can also occur with suction when catheters are removed. Therefore, whenever catheters are removed, they need to be withdrawn slowly to minimize suction effects and the fluid column within the sheath should be aspirated simultaneously. The sheath should then be aspirated and irrigated to ascertain that neither air nor blood has collected in the sheath.³ Importantly, the entire volume of the sheath should be aspirated after initial deployment as well as after each time a catheter is removed and reinserted, in order to ensure that a continuous column of fluid is present in the sheath and disallowing the possibility of trapped air that could otherwise be introduced when a catheter is advanced through the sheath. Testing to ascertain the sheath’s capacity prior to insertion is a good practice.

Management

Initially, it is essential to take all measures necessary to prevent further air embolization. For venous air embolism, placing the patient in the left lateral decubitus and Trendelenburg position helps air remain in the RA, where it will not contribute to an “air lock” in the RVOT. Additionally, direct extraction of air from the venous circulation if localized in the RA or RV can be attempted by aspiration from a central venous or pulmonary artery catheter. Air aspiration should be performed with the patient supine or in a Trendelenburg position while holding his or her breath at the end of inspiration or during a Valsalva maneuver.¹⁴

Administration of intravenous fluids and vasopressors can be necessary for hemodynamic support. Supplemental 100% oxygen therapy can reduce the size of the air embolus by increasing the rate of nitrogen absorption from air bubbles. Because gas bubbles are not buoyant enough to counteract arterial blood flow, these measures are more suitable for treating venous and RV air emboli.¹⁴

When cerebral air embolism is suspected, it is important to maximize cerebral perfusion by administration of fluids and supplemental oxygen. Hyperbaric oxygen therapy is regarded as the treatment of choice in patients with cerebral air embolism and prompt transfer to a hyperbaric oxygen therapy center should be considered. When started within a few hours, hyperbaric oxygen therapy can potentially compress the existing bubbles, speed resolution of bubbles by establishing a high diffusion gradient, improve oxygenation of ischemic tissues, and reduce endothelial thromboinflammatory injury.^3,11,14

Mechanism

The potential for acute coronary artery occlusion is a significant risk consideration with catheter ablation within the aortic cusps. Damage to the coronary arteries can result from catheter manipulation and ablation in the aortic root, secondary to RF energy delivery in close proximity to the ostium of the right or left coronary artery, as well as inadvertent catheter engagement of the left main coronary artery when ablating in the left coronary cusp.^17,18

It is also important to recognize the potential risk of coronary artery damage when ablating in the RVOT and pulmonary artery. The right coronary artery is typically 4 to 5 mm away from the proximal part of the RVOT near the free wall, and is separated by a variable amount of fat. Additionally, the cephalocaudal separation of the pulmonic and aortic valves results in close proximity of the pulmonic valve to the origin of the right coronary artery, and the left main coronary artery lies in immediate posterior proximity to the subvalvular RVOT near the pulmonic valve as well as the supravalvular pulmonary artery.^17,18

Because of the proximity of the right coronary artery to the cavotricuspid isthmus and the left circumflex artery to the lateral mitral annulus, coronary artery injury can potentially occur during ablation at mitral or tricuspid annuli.¹⁵ Transient inferior ST segment elevation and acute occlusion of the right coronary artery have been reported following ablation of the cavotricuspid isthmus, and left circumflex artery injury has been reported following ablation of left-sided AV BTs. Also, late stenosis of the right coronary artery has been observed in children with Ebstein anomaly of the tricuspid valve who undergo ablation of a right-sided AV BT.¹⁹

Percutaneous epicardial ablation probably poses a greater risk of coronary artery injury, severe spasm, or both, whereby the large epicardial coronary arteries can be in direct contact with the ablation catheter and movement of the ablation catheter in the pericardial space can occur during ablation endangering adjacent coronary arteries. The coronary sinus (CS) and its branches are also in close relation to the distal circumflex and the posterolateral branches of the right coronary artery, and ablation within the coronary venous system (for AV BTs, lateral mitral isthmus, and VT) can cause coronary injuries.^5,15

RF injury to coronary arteries can cause acute thrombosis or damage to the arterial wall. Arterial injury is inversely proportional to vessel diameter; there is little evidence of injury when vessel diameter exceeds 0.5 to 1.0 mm.¹⁵

Coronary artery spasm has been reported during mapping in the pericardial space and can occur as a result of ablation close to the artery. Extrinsic compression of a coronary artery can also result from edema caused by nearby ablation.

Clinical Presentation

Coronary arterial injury generally presents with acute coronary occlusion or spasm at the time of ablation associated with chest pain and ST elevation; nonetheless, delayed presentations can also occur.¹⁵

Importantly, RF ablation frequently (in 25% to 100% of cases) causes significant elevations in cardiac troponin levels that are not related to coronary arterial injury, with mean peak levels of 0.13 to 6 ng/mL for troponin I, and 0.20 to 2.41 ng/mL for troponin T. In contrast to acute coronary syndromes, troponin elevations observed following RF ablation typically peak early (2 to 8 hours versus 18 to 24 hours) and normalize early.²⁰ The levels of troponin elevation correlate with the number of RF lesions applied, the site of lesions (ventricular more than atrial more than annular), and the approach to the left side (transaortic more than transseptal). Ablation of focal lesions (atrioventricular nodal reentrant tachycardia [AVNRT], atrioventricular reentrant tachycardia [AVRT]) causes less frequent (in 25% to 88% of cases) elevation of troponin levels as compared with linear lesion ablation (typical and atypical atrial flutter [AFL], AF, and VT) where all patients are expected to have an increase in troponin levels. The prognostic significance of asymptomatic elevations of troponin I remains unclear.²⁰

Prevention

Several precautions are important to avoid injury to coronary arteries during percutaneous epicardial ablation, ablation via the coronary venous system, and ablation in the coronary cusps. Prior to ablation near or within a branch of the CS or in the pericardial space, direct visualization of the relation between the ablation site and adjacent coronary arteries must be obtained, usually by coronary angiography performed with the ablation catheter on the target site. An absolute safe distance between the ablation site and epicardial artery has not been defined; nonetheless, a distance of at least 5 mm between the ablation catheter and an epicardial artery is commonly accepted. It is also important to ensure that the catheter is not touching the vessel at any point of the cardiac cycle during angiography. Cryoablation appears to have less risk of coronary injury in animal models, but can still create occlusion and intimal damage when in close proximity, particularly to small vessels.⁵

When ablation in the aortic cusp is planned, it is preferable to insert a 5 Fr pigtail catheter into the aortic root and obtain aortic root angiography to visualize the aortic root and the ostia of the right and left coronary arteries.¹⁸ Selective angiography of the coronary arteries also can be performed with the ablation catheter positioned at the desired location so as to assess the anatomical relationships between these structures and the location of the ablation catheter. If the origin of VT is in the left coronary cusp, it is preferable to cannulate the left main coronary artery with a 5 Fr left Judkins catheter, which serves as a marker and for protection of the left main coronary artery in case of ablation catheter dislodgment during RF application. Alternatively, a combination of electroanatomical mapping and ICE imaging can be used to confirm anatomical location, catheter tip position and contact, and distance to coronary vasculature and to monitor RF delivery.¹⁷

To avoid injuring the coronary arteries during ablation of arrhythmias arising from the coronary cusps, it is preferable to start RF energy delivery at a low power output (15 W) and then increase to no more than 30 W to achieve a target temperature of approximately 50°C. Additionally, ablation should also be performed during continuous fluoroscopy to observe for catheter dislodgment, and energy delivery should be discontinued in case of even minimal dislodgment from the site showing the best mapping findings. RF application should also be stopped if the repetitive or sustained VT cannot be terminated after 10 seconds. Coronary angiography is often performed immediately after the ablation procedure to exclude coronary artery spasm, dissection, or thrombus.¹⁸

Atrioventricular Block

AV block is the most important complication of AVNRT ablation. AV block occurs in about 0.2% to 0.8% of slow pathway ablation using the posterior approach, generally occurs during RF delivery or within the first 24 hours postablation, and is almost always preceded by junctional ectopy with VA block. The level of block is usually in the atrioventricular node (AVN). Predictors of AV block include proximity of the anatomical ablation site to the compact AVN, occurrence of fast junctional tachycardia (cycle length <350 milliseconds) during RF application, occurrence of junctional rhythm with VA block, the number of RF applications (related to the amount of tissue damage), and significant worsening of anterograde AV conduction during the ablation procedure. The anterior approach for AVNRT ablation is associated with a higher risk of inadvertent complete AV block (approximately 10%, but ranging from 2% to 20%).²¹ During ablation of AVNRT, careful monitoring of AV conduction during RF application and prompt discontinuation of energy delivery on evidence of AV block are the keys to minimize the incidence of AV block. RF delivery should be immediately discontinued when the following occur: (1) the impedance rises suddenly (>10 Ω); (2) the PR interval (during normal sinus rhythm [NSR] or atrial pacing) prolongs; (3) AV block develops; (4) retrograde conduction block is observed during junctional ectopy; or (5) fast junctional tachycardia (tachycardia cycle length <350 milliseconds) occurs, which can herald imminent heart block.

AV block can also occur when ablating along the interventricular septum in the close vicinity of the compact AVN or His bundle (HB), including ablation of superoparaseptal and midseptal AV BTs or ATs as well as para-Hisian VTs. Additionally, because the penetrating HB lies at the membranous septum formed in part by the commissure between the right coronary and noncoronary cusps, ablation of VTs originating in this region can potentially damage the HB (Fig. 32-5).

FIGURE 32-5 Atrioventricular (AV) block complicating catheter ablation of premature ventricular complexes (PVCs) originating from the para-Hisian region. In A, note the presence of anterograde His potentials on the proximal and distal ablation electrodes (ABLp and ABLd, respectively) during sinus rhythm (blue arrows) and retrograde His potentials during PVCs (red arrows) at the site of earliest ventricular activation during PVCs. In B, radiofrequency energy delivery results in complete AV block secondary to injury to the His bundle.

When a His potential is detectable at the ablation catheter location, titrated RF energy output and RF application for a short duration should be used, coupled with overdrive atrial pacing to monitor AV conduction in case accelerated junctional rhythm occurs. Alternatively, cryoablation can be used in this region with its slightly better safety margin for AV conduction.^22,23

In some patients, mechanical trauma from the catheter induces temporary AV block. Also, catheter manipulation in the RV or LV can cause mechanical bundle branch block (often transient), which can potentially cause complete AV block in patients with preexisting block in the contralateral bundle branch. Furthermore, ablation of the right bundle branch in patients with bundle branch reentrant VT is commonly followed by complete AV block (in up to 30%) because of preexisting severe disease in the left bundle branch.

Macroreentrant Atrial Tachycardias

Whereas the occurrence of new arrhythmias following focal RF ablation has not been a clinical problem, linear ablation lesions, if incomplete, can promote reentry propagating through the gaps in the ablation lines. Even when conduction block across the ablation lines is verified, the ablation lines commonly performed in the LA for ablation of AF can still facilitate reentry by providing conduction obstacles and protected isthmuses with adjacent anatomical structures in the atrium.^24,25 In fact, LA macroreentry is a relatively common complication of catheter ablation of AF, occurring in up to 50% of cases. The incidence of ATs seems to be lower following segmental ostial pulmonary vein (PV) isolation (<5%) compared with circumferential PV isolation, and much higher following circumferential or linear LA ablation (>30%). Targeting complex fractionated atrial electrograms without linear lesions or PV isolation is associated with a moderate risk (8.3%) for post-ablation ATs. Following stepwise approaches of catheter ablation incorporating extensive lesions in the LA and RA to terminate persistent AF, ATs can be observed in more than 50% of patients.^26–28 Although ATs can occur at variable time intervals after AF ablation procedures, the most common timing appears to be 1 to 2 months postablation.^29,30

Ventricular Arrhythmias

Ventricular fibrillation has been reported in up to 6% of patients with chronic AF after AVN ablation when the ventricular pacing rate is lower than 70 beats/min. This complication can be minimized by postablation pacing for 3 months at a higher rate—that is, 90 beats/min. A possible mechanism for postablation ventricular arrhythmia is activation of the sympathetic nervous system and a prolongation of action potential duration.

Inappropriate Sinus Tachycardia

Inappropriate sinus tachycardia can occur in some patients after posteroseptal BT or AVN ablation, suggesting disruption of the parasympathetic input and/or sympathetic input into the sinus node and AVN.

Clinical Presentation

Phrenic nerve injury can be asymptomatic in about one-third of cases. The most frequent symptom is dyspnea. Other symptoms or clinical findings are cough or hiccup during ablation and the development of postablation pneumonia or pleural effusion. In asymptomatic patients, the diagnosis is made on the routine chest x-ray (Fig. 32-6) with hemidiaphragm paresis or paralysis (hemidiaphragm elevation with paradoxical movement). There is no active treatment known to aid phrenic nerve healing. Complete or partial recovery of diaphragmatic function can be observed in 66% and 17% of patients, respectively, sometimes not realized until several weeks or even months later.³⁶

FIGURE 32-6 Chest x-rays (posteroanterior views) showing right hemidiaphragmatic paralysis (arrowheads) caused by right phrenic nerve injury (A), and left hemidiaphragmatic paralysis caused by left phrenic nerve injury (B) during catheter ablation of atrial fibrillation.

Prevention

In areas at high risk of phrenic nerve injury (RA free wall, posteroseptal part of the SVC, inferoanterior aspect of the right PV ostium, and proximal LA appendage roof) that require ablation, high-output (10 mA) pacing should be performed before energy delivery. The ability to pace the phrenic nerve should prompt attempting to find a slightly different site or, if this is not possible, applying RF energy at low power and/or for short duration. This technique can be combined with electroanatomical mapping of the RA, SVC, LA, and right PVs to reconstruct the anatomical course of the right phrenic nerve. Even when the phrenic nerve cannot be paced, intermittent fluoroscopic visualization of the ipsilateral diaphragm movement should be performed during RF application at high-risk sites, and RF delivery should be terminated if diaphragmatic excursion decreases.

An alternative technique is to position a catheter in the SVC at a site where the phrenic nerve can be consistently captured, and pace the phrenic nerve during ablation. RF energy delivery should be stopped immediately if diaphragmatic contraction becomes less vigorous or ceases.

Early recognition of phrenic nerve injury during RF delivery (which can be indicated by the development of hiccup, cough, or decrease in diaphragmatic excursion during energy delivery) allows immediate interruption of the application prior to the onset of permanent injury, which is associated with the rapid recovery of phrenic nerve function. This is not feasible in patients undergoing general anesthesia.

Recently, diaphragmatic electromyography has been used to monitor phrenic nerve integrity during ablation. A progressive decline in compound motor action potential amplitude heralded phrenic nerve palsy, with a 30% decrease yielding the best predictive value, preceding hemidiaphragmatic paralysis by about 30 seconds. To record right diaphragmatic compound motor action potentials, two standard surface electrodes are connected to a central computerized electrophysiology workstation. The recording electrode is placed on the thorax 5 cm superior to the tip of the xiphoid process and the reference electrode is positioned along the right costal margin with a 16-cm interelectrode distance. Diaphragmatic bipolar electromyographic signals are recorded during continuous pacing of the right phrenic nerve, using a catheter in the SVC. The clinical application of this approach to prevent phrenic nerve palsy induced by cryoballoon ablation of the right superior PV is still in its preliminary stages, and further evaluation is required to validate and optimize this approach.^37,38

The use of a non-RF source of energy is unlikely to prevent this complication because phrenic nerve injury has been reported with ultrasound, laser, and cryotherapy. Furthermore, phrenic nerve injury occurs independently of the strategy of AF ablation used (PV isolation versus wide circumferential LA ablation).

During percutaneous epicardial ablation, proximity to the phrenic nerve can be detected by high-output pacing (typically at >10 mA) to detect diaphragmatic stimulation on fluoroscopy, allowing its course to be marked on a three-dimensional (3-D) map.³⁹ The nerve can branch over the LV, so simply noting sites at which phrenic nerve capture occurs does not necessarily delineate the course of all branches. It is important to recognize that detection by phrenic nerve capture is prevented by the use of paralytic agents during general anesthesia. However, these drugs are typically used during induction of anesthesia and their effects have dissipated by the time ablation is being performed. Catheter ablation should be avoided at sites adjacent to the phrenic nerve. Alternatively, moving the nerve away from the myocardium by injection of air into the pericardium or placement of a balloon catheter between the ablation target site and nerve can allow safe RF ablation. Alternative energy sources such as cryoenergy have been used to prevent phrenic nerve injury. Cryomapping uses temporary phrenic nerve injury to determine when to avoid full cryoablation. However, data on the success of cryoablation in the pericardial space are limited.^34,35

Incidence

PV stenosis is one of the most serious complications of AF ablation and has been reported with a wide range of incidence (0% to 42%), mainly because of different ablation strategies and differing methodology for estimating postablation PV ostia. The severity of PV stenosis is generally categorized as mild (<50%), moderate (50% to 70%), or severe (>70%). When moderate and mild lesions are included, the overall rate is as high as 15.5% (Fig. 32-7). However, the rate of severe PV stenosis has fallen to as low as 1.0% to 1.4% with increasing experience and improvements in technique, such as performing ablation outside the PVs and the use of ICE. Late progression of PV stenosis to severe stenosis or even complete occlusion can occur in up to 27% of patients (Fig. 32-8), particularly in the smaller PVs. Of note, stenosis seems to be more common in the left-sided PVs.¹¹

FIGURE 32-7 Pulmonary vein (PV) stenosis. A, CT image of the left atrium and PVs at baseline before segmental ostial PV isolation. B, CT image 3 months after the ablation procedure, showing moderate stenosis of the left superior pulmonary vein (LSPV; arrows) and mild stenosis of the left inferior pulmonary vein (LIPV). C, Angiogram (left anterior oblique view) of the LSPV, showing a moderate degree of stenosis (arrows). RIPV = right inferior pulmonary vein; RSPV = right superior pulmonary vein.

FIGURE 32-8 Pulmonary vein (PV) occlusion. Upper panels, CT image of the left atrium (LA) and PVs at baseline before segmental ostial PV isolation (A) and 5 months after ablation (B). Following ablation complete occlusion of both superior PVs is observed. Middle panels, Fluoroscopic (anteroposterior) views during angiography of the right superior pulmonary vein (RSPV) (C) and left superior pulmonary vein (LSPV) (D), showing severe stenosis-occlusion of both PVs at the ostia. PV angiography was performed with contrast injection via a catheter positioned in the pulmonary artery (arrowheads). Lower panels, Fluoroscopic (anteroposterior) views during stenting of the LSPV: E, Angioplasty balloon is inflated for stent deployment in the LSPV; F, angiogram of the left superior PV following stenting. LIPV = left inferior pulmonary vein; RIPV = right inferior pulmonary vein.

Circumferential LA ablation, with ablation limited to atrial tissue outside the PV orifice, seems to have lower rates of PV stenosis, whereas focal ablation inside the PV has a higher risk. Moreover, with PV isolation techniques, the use of an individual encircling lesion set that requires RF ablation between the ipsilateral PVs carries a higher risk of PV stenosis. With the focal ablation approach, PV stenosis can occur at a rate up to 33% to 42%. Although this represents an overestimation of the problem because increased TEE flow velocity was used to establish the diagnosis, it remains a major concern. With PV isolation techniques, the incidence of severe PV stenosis is approximately 1% to 5%. MR and spiral CT scanning have shown that although current PV isolation techniques are associated with minimal reduction of the PV ostial diameter, stenosis exceeding 50% can occur in up to 18% of cases, whereas very severe stenosis (>90%) is found in up to 2% of cases. Following circumferential LA ablation, the prevalence of PV stenosis likely ranges between 2% and 7%. In one report, a detectable reduction in PV diameter was present in 38% of PVs; however, moderate and severe PV stenosis was observed in 3.2% and 0.6% of PVs, respectively.^3,9

Mechanism

The mechanism of the development of PV stenosis remains unclear. Acute reduction of PV diameter (by approximately 30%) has been demonstrated during ablation. It was initially speculated that these changes are caused by edema and spasm. Thermal injury, induced by RF, has recently been shown to produce Ca²⁺-mediated irreversible tissue contracture occurring with the production of most if not all RF lesions, and therefore almost certainly contributes to stenosis. It is well known that thermal injury produces denaturation of extracellular proteins, especially collagen. Thermally induced collagen shrinkage has been well documented in animal and human studies. When near-circumferential lesions are produced within the tubular confines of the PVs, the consequences can be easily demonstrated as PV contracture. Thus, PV stenosis likely results from a combination of endothelial disruption with platelet activation and later neointimal proliferation, reversible edema, collagen denaturation, and shrinkage and thermal contracture. The occurrence and degree of PV stenosis correlate with the amount of energy delivered and lesion extension. The high incidence of mild PV stenosis likely reflects PV reverse remodeling rather than pathological PV stenosis. This hypothesis is based on the observation that all the mild PV stenoses were concentric rather than eccentric.^3,9

When severe stenosis or occlusion of a PV develops suddenly, gradual decline and then cessation of the arterial flow to the affected lung segment are observed. This is caused by a decline in the arteriovenous gradient as well as compression by the developing tissue edema. As a consequence, the involved alveoli are affected by the resulting ischemia and surrounding edema, leading to atelectasis, infarction, or susceptibility to infections. With the resulting alterations in pulmonary hemodynamics, redistribution of blood flow occurs with the opening of vascular channels or neovascularization in which tissue hypoxia is known to play a role. Hence, the venous drainage of the affected segment becomes mainly dependent on the ipsilateral veins draining the healthy lobes. If the ipsilateral vein(s) is also stenosed, the impedance to the pulmonary flow increases, adding to the hemodynamic burden. Symptoms attributable to PV stenosis are typically not seen unless the perfusion of the affected lobe falls below 20% or the perfusion of the entire lung on the affected side falls below 25%.⁴⁰

Clinical Presentation

PV stenosis after ablation is frequently asymptomatic, especially when a mild or moderate degree of PV stenosis is present or a single vein is involved. Severe PV stenosis can be associated with various respiratory symptoms that frequently mimic more common lung diseases, such as asthma, pneumonia, lung cancer, and pulmonary embolism. A high degree of suspicion is necessary to avoid performing misleading diagnostic procedures and to allow proper and prompt management. The onset of symptoms is usually several months after ablation. The initial manifestation is generally dyspnea on exertion, which typically evolves over the course of 1 to 3 months. Persistent cough is a common symptom. Pleuritic chest pain is a late symptom, and hemoptysis is uncommon. Both pleuritic chest pain and hemoptysis are likely related to complete vessel or branch occlusion.^3,9

It seems that symptom severity is related not only to the degree of stenosis but also to the number of PVs with stenosis. Symptoms may improve spontaneously over time in a significant percentage of patients. In one report, this improvement occurred in 50% of patients and was always related to improvement in the radiological abnormalities previously detected, although other hemodynamic compensatory mechanisms (e.g., development of collaterals) can also play a role.

Detection

To date, it remains unclear which is the best and most cost-effective noninvasive modality for detecting PV stenosis after ablation. TEE and CT scanning are most commonly used (Videos 24 and 25 ); however, no single imaging modality has been able to assess all relevant aspects of PV stenosis. Therefore, complementary methods are used (see Fig. 32-7). Although many investigators recommend routine follow-up CT or MR imaging for detection of asymptomatic PV stenosis 3 to 4 months after ablation of AF, it is unknown whether early diagnosis and treatment of asymptomatic PV stenosis provide any long-term advantage to the patient.⁴⁰ Nevertheless, it is recommended that follow-up PV imaging be performed to screen for PV stenosis during the initial experience of an AF ablation technique for quality control purposes.^3,9

Spiral CT and MR imaging of the PVs are readily available and can reveal the correct diagnosis; they are probably the most helpful in identifying the location and extent of the stenosis (see Fig. 32-7). However, the sensitivity and specificity of these imaging modalities need to be studied further. Of note, the diagnosis of PV occlusion cannot be made reliably using CT or MR imaging, and pulmonary artery wedge angiography is the only method for confirming the diagnosis.⁴¹

TEE provides adequate visualization of the superior PVs, but imaging of the right and left inferior PVs is inconsistent. Hence, TEE is not typically used as a first-line imaging study for routine screening. Nonetheless, TEE provides additional functional data regarding the significance of PV stenosis.⁴⁰ On TEE, PV stenosis is usually indicated by an increased maximal PV Doppler flow velocity or by the presence of turbulence and deformity of the flow signal, as defined by a minimal flow between systolic and diastolic peak flow of 60% of the mean of both peaks. ICE allows straightforward imaging of all PVs and provides a good view into the first 1 to 3 cm of the vein. Color flow contrast imaging is also useful for identifying the location of the orifice of the most tightly stenotic vessels.

Blood gas assessment and pulmonary function testing do not seem to be useful as screening measures for early PV stenosis. Radiographic findings in PV stenosis are nonspecific. Chest radiography often reveals parenchymal consolidation, pleural effusion, or both. Ventilation/perfusion scanning can be done to characterize the functional significance of the stenosis. In the presence of significant PV stenosis, perfusion abnormalities are usually marked and appear similar to the changes seen in pulmonary embolism.

Prevention

The best way to manage PV stenosis is to avoid it. PV stenosis is independently related to RF lesion location, size, and distribution and to baseline PV diameter . The prevalence of this complication has decreased because of various factors, including abandonment of in-vein ablation at the site of the AF focus, limiting ablation to the extraostial portion of the PV or PV antrum, use of advanced imaging techniques to guide catheter placement and RF application, reduction in target ablation temperature and energy output, and increased operator experience.¹¹

Cryothermal ablation has unique characteristics that may prove to be desirable when lesions are required within the PV. Specifically, there is less endothelial disruption, maintenance of extracellular collagen matrix without collagen denaturation, and no collagen contracture related to thermal effects. These characteristics may translate into a reduction in PV stenosis as compared with RF ablation.

Management

PV intervention is recommended in symptomatic patients. PV stenting appears to be associated with a lower restenosis rate compared with angioplasty (Video 26). ⁴² For asymptomatic patients with severe PV stenosis, some investigators have also suggested interventional therapy because lesions can progress and eventually lead to total occlusion, precluding dilation and potentially resulting in hemodynamic compromise (see Fig. 32-8).¹¹ Others have proposed regular clinical follow-up and intervention considered only when symptoms develop. However, it is important to understand that patients can become symptomatic only after the long-term sequelae of PV occlusion become manifest, at which time intervention may not be technically feasible or, even when successful, may not be associated with significant improvement in perfusion of the affected lung segments. It is also important to consider the age, functional capacity, and associated comorbidities as well as anatomical and technical factors pertaining to the stenotic vein when making a decision regarding interventional treatment in asymptomatic patients.¹¹

TEE or ICE can facilitate accurate stent placement by reliably identifying the orifice of the tightly stenotic vein. Unfortunately, both in-stent and in-segment restenosis can recur in up to 61% of patients, and repeat intervention is warranted in symptomatic patients. Stenosis within the body of the stent is likely caused by neointimal hyperplasia and fibrosis. Out-of-stent restenosis is usually observed at bifurcation points beyond the stent into the vessel, also suggesting the progression of PV pathology, despite stent placement. There is no experience to date with the use of drug-eluting stents in this setting, but such a high restenosis rate makes a strong case for the investigation of those devices and larger diameter stents.¹¹

Incidence

The true incidence of atrioesophageal fistula after catheter ablation is unknown, but has been estimated to be between 0.03% and 0.2%, although underreporting is likely.^43,44 Despite its rarity, atrioesophageal fistula remains a devastating complication associated with high mortality, accounting for 15.6% of cases with fatal outcome (the second leading cause of death post AF ablation, following cardiac tamponade).⁴⁵

On the other hand, esophageal mucosal changes consistent with thermal injury have been reported in up to 47% of the patients following AF ablation, and esophageal ulcerations confirmed by esophagogastroscopy or capsule endoscopy can be observed in 14% to 18%.^46,47

Mechanism

The precise mechanism of atrioesophageal fistula formation has not yet been determined. Proposed potential mechanisms include direct thermal injury of the esophageal wall in immediate proximity to the posterior LA wall, thermal injury of the arterial blood supply to the esophagus, and mechanical injury to the esophagus during periprocedural TEE as a contributing or predisposing factor.³

The esophagus is located at the center of the posterior mediastinum and runs in close proximity to the posterior LA wall for about 5 cm of its course. In that region, the esophagus is separated from the LA only by the pericardial sac (the oblique sinus), which insinuates itself between the openings of the right and left PVs. The LA posterior wall thickness is 2 to 4 mm and the thickness of the esophageal wall is 2 to 3 mm in cross section at the LA-esophageal area of contact. During RF catheter ablation, lesion depth, extension, and volume are related to the design of the ablation electrode and RF power delivered. Theoretically, esophageal damage can occur, even during LA ablation using a standard 4-mm-tip ablation catheter. However, cooled and large-tip ablation catheters induce extended lesions. Thus, delivering high-power RF applications on the LA posterior wall between the left and right PVs can potentially deeply damage the esophagus, resulting in atrioesophageal fistula formation.^48,49

The esophagus does not have a serosal layer and is fixed primarily in the pharynx and gastroesophageal junction. In the mediastinum, the esophageal position is dynamic because of peristalsis. In fact, the esophagus often shifts sideways by 2 cm or more in most patients undergoing catheter ablation for AF under conscious sedation. Additionally, the position of the esophagus in relation to the LA and PV demonstrates high variability. In many cases, the esophagus is very close to the ostia of the PVs and lies only a short distance from the LA wall. In most subjects (90%), the esophagus courses down beside the ostia of the left PVs with a mean distance of 10.1 mm to the left superior PV and 2.8 mm to the left inferior PV. Therefore, anatomical localization of the esophagus can be critical before or during AF ablation to prevent atrioesophageal fistula, especially as there is a need for transmural atrial lesions.

The risk of esophageal injury likely is enhanced by increasing the magnitude and duration of local tissue heating (i.e., the total ablation energy delivered to cardiac tissue near the esophagus), which is related to catheter tip size, contact pressure, catheter orientation, and power output and duration. In fact, all reported cases of atrioesophageal fistulas have in common attempts to perform deep RF lesions on the LA posterior wall. Additionally, ablation strategies incorporating extensive ablation in the LA posterior wall and the type of AF (persistent more than paroxysmal AF, possibly due to the additional linear ablation lesions) have been associated with an increased rate of esophageal ulcerations. A short LA-to-esophagus distance, the use of nasogastric tubes, and general anesthesia are other potential risk factors for esophageal injury during RF ablation.^49,50

Cryoballoon ablation commonly causes significant decreases in esophageal temperature, especially during cryoablation of the inferior PVs, resulting in esophageal ulcerations. Although no instances of atrioesophageal fistula formation after cryoballoon PV isolation have yet been reported, the total number of cryoballoon ablation cases performed is relatively small to reliably detect this rare complication. Furthermore, esophageal ulcerations have been observed following cryoballoon ablation at rates (17% of patients) similar to those seen with RF ablation. Although there is no definitive proof that esophageal ulcer formation is predictive of fistulas, it is reasonable to assume that esophageal ulcerations may represent the first step on the way to atrioesophageal fistula. Therefore, until further experience indicates otherwise, cryoballoon ablation should not be considered completely safe in regard to atrioesophageal fistula formation.⁵¹

Excessive esophageal heating, esophageal ulcerations, and atrioesophageal fistula formation have been observed post PV isolation using high-intensity focused ultrasound (HIFU). In a recent study, elevated esophageal temperature prompting cessation of HIFU energy delivery occurred in 9% of PVs. Despite use of the safety algorithm, the occurrence of esophageal thermal damage and lethal atrioesophageal fistula could not be prevented, occurring in 1 of 28 patients. This safety concern argues against the clinical use of HIFU in its current iteration for PV isolation.⁵²

Clinical Presentation

The clinical course of atrioesophageal fistula formation varies in abruptness of presentation and timing (ranging from 2 days to 4 weeks), and mostly is manifest after patient discharge from the hospital. The presenting symptoms can include hematemesis, sepsis, or air embolization and stroke. The leading symptom of esophageal perforation is high fever or severe chest or epigastric pain.

Leukocytosis is the earliest and most sensitive laboratory marker. Thoracic CT or MR imaging is the most valuable diagnostic examination (Fig. 32-9, Video 27 ). The dramatic neurological complications occur with a delay of at least a few hours after the first symptoms. Endoscopy is a diagnostic modality that should be avoided because insufflation of the esophagus with air can result in a devastating cerebrovascular accident and death secondary to a large air embolus. Immediate surgery (within a few hours after the first symptoms) can potentially prevent neurological complications and possibly result in a high survival rate, without residue. Delay of treatment seems to have devastating results.³

FIGURE 32-9 Atrioesophageal fistula postablation of atrial fibrillation. Multiple transverse sections of a chest computed tomography scan, revealing air bubbles (blue arrows) in the superior aspect of the left atrium (A-C) and at the level of the mitral valve (D). Note that air appears to originate from the anterior esophageal wall (B and C). Nasogastric tube is visualized in the esophagus (yellow arrow).

Prevention

Various strategies may be used to avoid the development of an atrioesophageal fistula, including the following: (1) assessment of esophageal position before and during ablation; (2) avoiding ablation in the vicinity of the esophagus; (3) mechanical displacement of the esophagus; (4) monitoring of the esophageal temperature during energy delivery; and (5) reduction of ablation energy power output and duration at sites close to the esophagus. However, because of the rarity of this complication, it remains unproven whether the use of these approaches lowers or eliminates the risk of esophageal perforation or fistula formation, and the optimal technique has not yet been determined. Therefore, it appears prudent to apply a combination of these preventive techniques during AF ablation. These strategies are discussed in detail in Chapter 11.^{48–50,53–55}

Clinical Presentation

In EP procedures, the patient tissue receiving the greatest radiation dose is the skin area of the back at the entrance point of the x-ray beam, and skin injuries are well-recognized complications for catheter ablation procedures. The threshold for transient erythema and epilation is 2 to 3 Gy, for acute skin injury is 2 to 8 Gy, and for chronic radiation injury is 10 Gy.⁵⁶ Obese patients, those with preexisting health conditions such as collagen vascular disease, diabetes mellitus, or ataxia telangiectasia, and those who received a high radiation dose from a previous procedure, can be at a greater risk for radiation skin damage.

Acute radiation injury (characterized by erythema with vesiculation, erosion, and pain) usually starts within days after exposure and persists up to several weeks (Fig. 32-10). Chronic radiation injury manifests months to years after exposure and presents clinically as permanent erythema, dermal atrophy, and ulceration.

FIGURE 32-10 Acute radiation-induced skin injury (erythema) 2 to 3 weeks following radiofrequency ablation of atrial fibrillation using biplane fluoroscopy. The affected areas have well-defined borders congruent with the entrance of the x-ray beam.

Acute radiation-induced skin injury can be misdiagnosed as contact dermatitis, viral or bacterial infection, or a spider bite, and a high index of suspicion is required. Fluoroscopy-induced injuries affect an area congruent to the entrance of the x-ray beam that typically has well-defined borders (see Fig. 32-10), which occur when the beam is not moved or resized during prolonged fluoroscopy over one site. It is important to alert patients who had undergone prolonged procedures or known to have received a high skin radiation dose to examine him- or herself about 2 to 3 weeks after the procedure to look for skin changes and to contact the interventionalist if any are observed.

Skin biopsy is usually not necessary and not recommended, especially given the potential risk of a nonhealing ulcer at the site of biopsy in the radiation-damaged skin area. Furthermore, the findings on skin biopsy generally are not pathognomonic for radiation change.

The estimated lifetime risk of a fatal malignancy after AF ablation using a modern low-frame pulsed fluoroscopy system is relatively low (0.15% for women and 0.21% for men in one report) and is higher than, although within the range of, previously reported risk to result from the ablation of standard types of supraventricular arrhythmias (0.03% to 0.26%). It has been estimated that for every 60 minutes of fluoroscopy, the mean total lifetime excess risk of a fatal malignancy is 0.03% to 0.065%.

Prevention

Increasing availability and familiarity with 3-D mapping systems should significantly reduce fluoroscopy time and the need for biplane fluoroscopy. The use of remote navigation systems also is likely to reduce fluoroscopy exposure of the patient and operator.⁵⁷ Additionally, judicious and intermittent use of fluoroscopy, that is, keeping the x-rays on for only a few seconds at a time, long enough to view the current catheter position, can reduce total fluoroscopic times considerably. The presence of grids in x-ray systems primarily increases the contrast and hence the image quality; however, they increase the dose to the patient and staff by a factor of two or more, and removal of the grid can help reduce the dose of radiation. Additionally, periodically rotating the fluoroscope at different angles helps spread the total radiation dose over a broader area of the patient’s skin so that no single region receives the entire dose. Other methods include adjustment of beam quality, using pulsed fluoroscopy, inserting appropriate metal filters (aluminum, copper, or other materials) into the beam at the collimator, and minimizing geometric magnification by keeping the image receptor close to the patient and the source away.