Case reports of steam pops using a dual-energy lattice-tip catheter during cavotricuspid isthmus ablation and review of the Manufacturer and User Facility Device Experience database Authors: Jad A Ballout, Kuldeep B Shah, Mahmoud Samy Ahmed, Brian D Williamson Published: 2026 Key Teaching Points •Although steam pops are less likely to occur with temperature-controlled catheters, they can still happen in the setting of lesion stacking and ablating in pouches even with the Sphere-9 catheter. Although steam pops are less likely to occur with temperature-controlled catheters, they can still happen in the setting of lesion stacking and ablating in pouches even with the Sphere-9 catheter. •Steam pops with the Sphere-9 catheter are difficult to predict because the system does not provide feedback on impedance and other traditional ablation parameters. Contact and anticipated contact force were similar between the lesions that caused the steam pops compared with those before them. Steam pops with the Sphere-9 catheter are difficult to predict because the system does not provide feedback on impedance and other traditional ablation parameters. Contact and anticipated contact force were similar between the lesions that caused the steam pops compared with those before them. •Using Sphere-9 in the pulsed field energy mode can in theory prevent steam pops although this should be used judiciously on the cavotricuspid isthmus to avoid coronary vasospasm. Using Sphere-9 in the pulsed field energy mode can in theory prevent steam pops although this should be used judiciously on the cavotricuspid isthmus to avoid coronary vasospasm. •1 limitation of Sphere-9 from this report is that temperature recording is from electrodes at an unclear distance from the point of maximum ablation energy delivery. 1 limitation of Sphere-9 from this report is that temperature recording is from electrodes at an unclear distance from the point of maximum ablation energy delivery. Introduction Since its commercial introduction in Europe in 2021 and first Food and Drug Administration (FDA) approval in the United States in 2023, there has been a major uptick in the use of pulsed field ablation (PFA) for atrial fibrillation (AF) ablation owing to ease of use, shorter ablation times, and superior safety profile compared with thermal ablation. Since its adoption, there have been no reported cases of permanent clinical phrenic nerve injury or pulmonary vein stenosis and few reported cases of esophageal injury or atrioesophageal fistula, specifically in the radiofrequency (RF) mode of the catheter, which is associated with mortality rates upward of 50%.1, 2, 3, 4 However, PFA can cause coronary vasospasm particularly on the cavotricuspid isthmus (CTI) occasionally developing ST-segment elevation and even ventricular fibrillation. High doses of intravenous nitroglycerin may be needed during PFA ablation on the CTI to prevent coronary vasospasm.5, 6, 7, 8 However, with that said, the long-term effects of PFA on the coronary arteries remain largely unknown with some early studies suggesting long-term coronary stenosis.9 A spherical lattice-tip dual-energy monopolar ablation catheter (Sphere-9™, Medtronic, Minneapolis, MN) that was recently FDA approved allows alternating between RF and pulsed field energy.10 Using RF on the CTI avoids the severe vasospasm but comes with traditional complications that can be seen with RF energy. Here, we present 2 case reports of steam pops that occurred during CTI ablation using the Sphere-9 catheter and a review of 15 cases reported in the Manufacturer and User Facility Device Experience (MAUDE) database. Case Report Case 1 A 77-year-old female with a history of paroxysmal AF status post cryoballoon pulmonary vein isolation (PVI) in 2020, heart failure with improved ejection fraction secondary to nonischemic cardiomyopathy, left bundle branch block, and nonobstructive coronary artery disease presented to the electrophysiology laboratory for repeat AF ablation procedure in the setting of now recurrent and persistent AF, as well as CTI ablation for typical atrial flutter. The procedure was performed under general anesthesia. Three femoral venous accesses were obtained for the ablation sheath, the coronary sinus catheter, and the intracardiac echocardiogram catheter. Intravenous heparin was administered after access, and a heparin drip was started. 1 transseptal puncture was performed, and the transseptal sheath was advanced into the left atrium. The Sphere-9 dual-energy catheter and the Affera™ mapping system (Medtronic) were used for mapping and ablation. The baseline voltage map showed reconnection of the left pulmonary veins and low-voltage areas on the posterior wall. Redo PVI and posterior wall isolation with evidence of bidirectional block was performed with PFA. We then withdrew the sheath and catheter back into the right atrium and proceeded with CTI ablation. An intracardiac echocardiogram was used to visualize the Sphere-9 catheter position and contact. We decided to use RF to avoid coronary vasospasm. With the 11th lesion, near the inferior vena cava (IVC) side of the CTI, a steam pop was heard, and ablation was stopped. The 11th ablation lesion along with their ablation characteristics, including lesion duration, power (purple), temperature (orange), and impedance (green), is presented in Figure 1. The lesion had a truncated duration of 4.5 seconds (compared with the standard 5 seconds), and the graphs show a sudden rise in impedance, along with an ablation power that is higher than the previous ablation lesions, but without an associated rise in tissue temperature. The details of the lesion that resulted in a steam pop during case 1 compared with the averages of those of the previous lesions are presented in Table 1. The lesion was truncated at 4.5 seconds compared with 5 seconds. Maximum and mean temperatures were similar between the steam and nonsteam pop lesions. Maximum current, mean current, and total energy delivered were higher in the lesion that resulted in the steam pop. Finally, the average distance between RF lesions on the CTI was 3.95 mm. The intracardiac echocardiogram did not show any evidence of effusion. We monitored the patient for an additional 10 minutes, and no effusion developed. After ablation, bidirectional block was confirmed across the CTI. The patient was discharged in stable condition. Case 1 steam pop during CTI ablation. CTI ablation started on lesion #61 on the tricuspid annulus. Steam pop occurred on ablation lesion #71 (11th CTI RF ablation point, red arrow). Review of ablation graphs (red rectangle) shows a truncated ablation lesion at 4.5 seconds instead of the full 5 seconds, along with a sudden rise in impedance (green graph). In addition, power output (purple graph) was higher than the previous ablation points to maintain target surface tissue temperature (orange graph). CTI = cavotricuspid isthmus; RF = radiofrequency. Table 1Parameters of steam pop lesions vs the other CTI ablation lesionsParameterCase 1Case 2Steam pop lesionAverage for nonsteam pop lesionsSteam pop lesionAverage for nonsteam pop lesionsDuration (s)4.552.95Maximum temperature (C)73736471Mean temperature (C)60625460Maximum current (mA)3320305033513246Mean current (mA)3063263031412921Energy delivered (J)1648133310061496CTI = cavotricuspid isthmus. Parameters of steam pop lesions vs the other CTI ablation lesions CTI = cavotricuspid isthmus. Case 2 A 78-year-old male with persistent AF and typical atrial flutter presented for AF and flutter ablation. He also has a history of coronary artery disease and myocardial infarction status post coronary artery bypass surgery, ischemic cardiomyopathy with heart failure with reduced ejection fraction, hypertension, hyperlipidemia, and type 2 diabetes. The procedural steps were identical to those in the previous case including general anesthesia, vascular access, and heparin administration. The patient was in atrial flutter when he presented to the laboratory with a cycle length of 366 ms. We used a Sphere-9 catheter and the Affera mapping system for activation mapping and entrainment from the CTI, which confirmed CTI-dependent atrial flutter. We started ablation from the distal CTI using RF energy. Halfway through the CTI line, we encountered a pouch on the CTI, and during our 11th ablation lesion, a steam pop was heard, and the ablation was stopped. The intracardiac echocardiogram did not show any effusion throughout the remainder of the case. Given that the catheter was away from the coronary artery, we decided to finish our CTI line using PFA. After ablation, bidirectional block was confirmed across the CTI. Ablation lesions 3–11 on the CTI, which were performed using RF, and lesion 12, which was the first PFA lesion on the CTI, are presented in Figure 2. The 11th ablation point (point #71) had a truncated duration of 2.9 seconds, and the graphs show a sudden rise in impedance, along with an ablation power that is higher than the previous ablation lesions, but without an associated rise in tissue temperature (tissue temperature was lower than the previous RF point). The details of the lesion that resulted in a steam pop during case 2 compared with the averages of those of the previous lesions are presented in Table 1. The lesion was stopped early at 2.9 seconds. Maximum and mean temperatures were lower for the lesion that resulted in the steam pop. Maximum current and mean current were higher in the lesion that resulted in the steam pop. The total energy delivered was lower during the lesion that caused the steam pop. Finally, the average distance between RF lesions on the CTI was 3.27 mm. Case 2 steam pop during CTI ablation. CTI ablation started on the tricuspid annulus. Steam pop occurred on ablation lesion #11 (red arrow). Review of ablation graphs (red rectangle) shows a truncated ablation lesion at 2.9 seconds instead of the full 5 seconds, along with a sudden rise in impedance (green graph). In addition, power output (purple graph) was higher than the previous ablation points in the setting of a suboptimal rise in surface tissue temperature (orange graph). CTI = cavotricuspid isthmus. After CTI ablation, we obtained transseptal access to the left atrium and performed PVI and posterior wall isolation using PFA and confirmed bidirectional block from the veins and posterior wall. There were no further complications, and the procedure was concluded. Discussion The Sphere-9 catheter was originally developed as an RF-only ablation catheter. It has a 9 mm expandable lattice tip made of conductive nitinol, with a central irrigating nozzle. The catheter functions as a continuous conductive surface with the same voltage potential across the entire surface. It has an effective surface area of 275 mm2 and delivers 3600 mA of current, with a resultant current density of 13 mA/mm2. In comparison, traditional irrigated-tip ablation catheters have a 3.5–4 mm tip electrode with a surface area of 27–31 mm2. At 40 W of power, the delivered current is approximately 600 mA, resulting in a current density of around 20 mA/mm2.11 The Sphere-9 catheter can deliver more current to the tissue at a lower current density, thereby resulting in larger lesions with less tissue overheating. The sphere has 9 temperature sensors along with integrated microelectrodes. The temperature sensors that are in contact with the tissue record tissue temperature. Unlike conventional RF ablation catheters, the Sphere-9 ablation catheter has only 2 commercially available preset RF settings, and parameters such as target temperature, power, and lesion time limits cannot be adjusted outside these presets. The first RF setting is the “anterior setting” that is temperature controlled, with a target temperature of 73°C, a 5-second lesion delivery duration, and an 80% current limit of the maximum deliverable current of 3600 mA. This setting is what is currently being used for ablation on the CTI. The second RF ablation setting is the “mitral setting” that has the same target parameters except for a higher current target at 90% of maximum. During RF delivery, the normal saline irrigation rate goes from a baseline of 4 mL/min to 30 mL/min. Steam pops occur when the myocardium is heated above 100°C resulting in water evaporation and formation of a steam bubble intramurally, which expands with continued RF delivery with rising pressure that eventually results in eruption of the gas bubble with a resultant popping sound. This usually occurs toward the endocardial surface (heat-damaged surface) leading to a crater formation but can occasionally occur toward the epicardium resulting in perforation and tamponade. Factors associated with steam pops include higher contact force and rapid impedance drop during RF delivery. Irrigated catheters are potentially more likely to be associated with steam pops given that irrigation cools the catheter tip and myocardial surface, the most tissue heating occurs owing to conductive heating, and the recorded surface temperature will be lower than the maximal tissue temperature. In contrast, temperature-controlled catheters, such as the QDot catheter (Biosense Webster) and Sphere-9, are theoretically less likely to result in steam pop formation.12,13 In both of our cases, the ablation profiles of the lesions that resulted in a steam pop showed early termination of RF delivery (before 5 seconds) and sudden impedance rise, which is expected. However, the unexpected finding was the suboptimal rise in surface temperature compared with the previous lesions, which resulted in the system increasing and maintaining a high-power delivery in an attempt to improve tissue heating but resulting in steam pop formation. The first steam pop occurred at the IVC end of the CTI, whereas the other one occurred in the mid-CTI inside a pouch without a significant rise in recorded surface temperatures. A possible explanation for the observed findings is that the electrode with the most contact with the tissue that was actually recording surface temperature was slightly away from the area with the highest tissue heating. In the case of the first steam pop, the highest tissue temperature could have been at the IVC edge of the CTI, whereas the recording electrode could have been slightly posterior to it recording from the IVC. In the case of the second steam pop, the highest tissue temperature could have been inside the pouch while the electrode was recording from immediately outside it. Another possible factor that may have contributed to this was the stacking of ablation lesions on the CTI. This may have resulted in continued excessive heating of the subendocardium/midmyocardium owing to conductive heating in the overlapped lesions leading to steam pop. This is supported by data presented in Table 1. In the first case, the lesion that resulted in steam pop formation had similar maximum and mean recorded tissue temperatures to the previous lesions; however, the catheter needed to deliver more power (higher current and total energy) to reach the same surface tissue temperatures. This higher power coupled with lesion stacking likely resulted in excessive tissue heating below the surface leading to a steam pop. In the second case, the lesion that caused the steam pop had lower maximum and mean recorded tissue temperatures than the previous lesions, and more power was needed to reach the recorded temperatures. This suggests that most likely the site where tissue temperature was being recorded by the catheter was away from the site of highest tissue heating. The total energy delivered to the tissue during the second case was lower for the lesion that caused the steam pop given that the lesion was stopped at 2.9 seconds compared with 5 seconds during the previous lesions; however, the mean power (E/t) delivered to the tissue was higher with the lesion that caused the steam pop (368 W vs 299 W). Unlike traditional RF ablation, during high-power short-duration ablation that Sphere-9 uses, monitoring and reacting to rapid impedance drop or sudden rise in impedance may not be as feasible, and avoiding lesion stacking or excessive contact/contact force may be a more feasible approach. We conducted a review of the US FDA’s MAUDE database (an open, deidentified repository comprising both mandatory industry submissions and voluntary reports from clinicians and patients regarding adverse device events). The search criteria from the MAUDE database included “Medtronic” for the category of manufacturer, “Affera” for the category of brand name, and the date range from November 7, 2024, to September 1, 2025. The results were then exported to an Excel sheet, and the complications that included steam pops were included. Between September 1, 2024, and August 26, 2025, there were a total of 15 reported steam pops (Table 2). 3 of these occurred in unreported locations and did not have any associated complications. 6 occurred during CTI ablation; 1 was at the IVC end of the CTI (case 1 in Table 2), 1 noted lesion stacking (case 3 in Table 2), 1 noted absence of lesion stacking, and 1 noted that a “temporary injury” occurred without additional details (case 9 in Table 2). No further information is available on the remaining 2 cases. None of the CTI cases resulted in further complications or prolongation in hospital stay. 3 steam pops occurred in the left atrium: 1 during ablation on the mitral annulus, 1 during ablation on the left atrial appendage ridge, and 1 during ablation anterior to the right superior pulmonary vein. The latter (case 15 of Table 2) resulted in pericardial effusion and tamponade requiring pericardiocentesis and Impella placement for hemodynamic support. Finally, 3 occurred in the left ventricle without additional complications. 1 of the left ventricular events occurred during the 11th second of RF delivery, which is an investigational setting for ventricular tachycardia ablation and is not yet available commercially. The MAUDE database has several limitations, namely being self-initiated and as such subject to reporting bias and lacking the total number of cases performed (ie, no denominator), and as such, the frequency of a complication or adverse outcome cannot be calculated from the MAUDE database. However, it is still a valuable tool to report adverse outcomes, especially those that were not seen in clinical trials of a certain device to alert operators to possible complications. Table 2Steam pop cases with the Sphere-9 catheter reported in the MAUDE databaseCaseSteam pop locationDateSequelae1CTI near IVCDecember 12, 2024None2UnknownJanuary 6, 2025None3CTIFebruary 3, 2025None4CTIFebruary 24, 2025None5UnknownMarch 18, 2025None6CTIMarch 28, 2025None7Left ventricleApril 7, 2025None8Mitral valveApril 16, 2025None9CTIApril 29, 2025Temporary injury reported10UnknownMay 6, 2025None11Left ventricleMay 22, 2025None12Left ventricleMay 29, 2025None13Left atrial appendage ridgeJune 16, 2025None14CTIJuly 16, 2025None15Right superior pulmonary veinJuly 21, 2025Tamponade requiring pericardiocentesis and Impella placementCTI = cavotricuspid isthmus; IVC = inferior vena cava; MAUDE = Manufacturer and User Facility Device Experience. Steam pop cases with the Sphere-9 catheter reported in the MAUDE database CTI = cavotricuspid isthmus; IVC = inferior vena cava; MAUDE = Manufacturer and User Facility Device Experience. Conclusion Although steam pops are uncommon with temperature-controlled catheters such as the Sphere-9 catheter, they can still occur, and studies are required to determine which parameters can reliably be tracked to decrease the risk. Funding Sources This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors. Disclosures The authors have no conflicts of interest to disclose. References