Surgery for ventricular tachycardia originating from the left ventricular summit

Abstract

Open in new tabDownload slide

OBJECTIVES

Ventricular tachycardia (VT) originating from the left ventricular summit region, the most superior region of the left ventricle surrounded by the major coronary arteries and veins, is frequently refractory to pharmacological therapies and endocardial and epicardial catheter ablation.

METHODS

Eleven patients with an age from 31 to 79 (median 56) years old, underwent map-guided surgery for left ventricular summit VT. All patients had undergone 1–5 unsuccessful sessions of catheter ablation for incessant VT, preoperatively. Five patients had suffered VT storm and 1 had a history of cardiopulmonary resuscitation. Four patients had implanted with a defibrillator. Epicardium to endocardium transmural cryothermia was applied at the VT origin determined by intraoperative epicardial mapping with electro-anatomical mapping system. Harmonic scalpel was used to remove the epicardial fat and cryothermia was applied directly to the myocardium, avoiding thermal or mechanical injuries to the coronary vessels. Additional endocardial cryothermia at the VT origin was performed by a cryoprobe introduced into the left ventricular cavity through an aortotomy.

RESULTS

There was no surgical mortality or long-term mortality related to VT during a median follow-up period of 60 months (interquartile range: 34–82). Five-year freedom from preoperatively documented left ventricular summit VT and non-documented VT was 91% and 73%, respectively. All the patients with postoperative VT underwent successful catheter ablation. Other patients were free from VT during the follow-up period.

CONCLUSIONS

Epicardial to endocardial transmural cryothermia at the VT origin guided by intraoperative electro-anatomical mapping with a close collaboration with electrophysiologists was crucial in successful surgery for left ventricular summit VT.

INTRODUCTION

Left ventricular (LV) summit ventricular tachycardia (VT) is a rare form of life-threatening VT that originates from the most superior region of the LV, surrounded by the left anterior descending coronary artery (LAD), left circumflex coronary artery and the anterior interventricular vein (AIV) (Fig. 1) [1–4]. Catheter ablation of LV summit VT remains challenging, even with advanced technologies and techniques including a combined endocardial and epicardial approach [4–6]. Main branches of the coronary arteries and veins course through the LV summit region, therefore, there is a risk of mechanical or thermal injury to the adjacent coronary vessels by epicardial ablation [6, 7]. The myocardium is covered by thick fat tissue at the LV summit region, and this anatomical condition hampers effective epicardial ablation. It is essential to cause transmural cell death at the originating VT site to eradicate VT because intramural VT origin or reentry is thought to be the mechanism in LV summit VT [8, 9]. This can be difficult if the myocardium is thick, such as in the superior LV summit and in hypertrophic cardiomyopathies, and also with isolated endocardial or epicardial catheter ablation alone.

Surgical treatment also remains challenging and has been rarely performed [2] because of the inaccessible location and electrophysiological complexity of the VT. The tasks are at hand to precisely locate an invisible site for ablation and to identify an optimal surgical approach to safely create a reliable transmural cell death at the VT origin. We successfully performed a map-guided surgery in 11 patients with LV summit VT. We retrospectively review our experience and describe the surgical technique for the VT originating from this complicated region.

METHODS

This is a retrospective study to examine whether a surgical approach and technique is safe and effective for suppressing incessant and intractable VT and for preventing arrhythmic death in patients with LV summit VT, which is defined as VT that originates from the most superior region of the LV, surrounded by LAD, left circumflex coronary artery and anterior interventricular vein by catheter-based electrophysiological studies [1–4]. We reviewed the findings of the intraoperative mapping and postoperative recurrence of preoperatively documented LV summit VT as well as the emergence of preoperatively non-documented VT. Complications and early- and long-term mortalities after surgery were also examined.

Ethics

The surgical procedure was approved by the Japanese Ministry of Health, Labour, and Welfare. The intraoperative electrophysiological study, including the cardiac mapping study, and the study protocol using the patient information were approved by the institutional review board of Nippon Medical School. All the patients provided a written informed consent for the intraoperative mapping study and surgery.

Intraoperative mapping

All the patients, except those in whom the electrode manipulation on the epicardium was not feasible or safe, underwent an intraoperative mapping study using an electro-anatomical mapping system (CARTO System, Biosense Webster, Diamond Bar, CA). The setting and method of the intraoperative electro-anatomical mapping have been described previously [10]. Accurate characterization of QRS morphologies identical to those of the clinical VT is crucial for precise localization of VT origin for successful surgery of LV summit VT. However, not all the precordial leads of electrocardiogram (ECG) are properly recorded during open-chest procedures and QRS morphologies of the induced VT or the paced rhythm are assessed only by the limb leads and the limited precordial leads. After the chest was opened, the QRS morphologies during sinus rhythm, paced rhythm or induced VT were carefully verified before the electrogram acquisition to distinguish preoperatively documented VT from non-documented VT. First, a voltage map was constructed on a three-dimensional figure of the heart during sinus rhythm. Then, programmed stimulation was performed by a programmable stimulator or implanted implantable cardioverter defibrillator (ICD) to induce VT and an activation map of induced VT was constructed (Fig. 2). Patients were placed on normothermic cardiopulmonary bypass to maintain haemodynamics during sustained VT. Preoperatively constructed voltage or activation maps were displayed on a computer, and the regions of interest were precisely mapped intraoperatively. Preoperative and intraoperative maps were compared and the three-dimensional distribution of the substrates for VT was examined.

Surgical procedures

Map-guided surgical procedures were performed through a median sternotomy in all patients. After the sites of origin of VT were determined by intraoperative mapping and the heart was arrested by infusing a high-potassium and high-magnesium blood cardioplegia into the aortic root, the epicardial fat tissue covering over the VT origin was dissected and removed by using an ultrasonic scalpel to apply cryothermia directly on the myocardium (Fig. 3). The coronary vessels were carefully dissected and snared with silastic tapes, so that cryoprobe could be safely placed under the coronary vessels avoiding thermal or mechanical injuries to the vessels. Nitrous oxide was used for cryothermia as refrigerant to freeze the myocardium at -60 degrees Celsius or lower for 2 min.

Outcome evaluation and postoperative electrophysiological study

The patients were continuously monitored during the postoperative hospitalization, including continuous monitoring of the ECG. Complications or morbidities related to surgery or intraoperative mapping were recorded. All the patients underwent a catheter-based electrophysiologic study or a programmed ventricular stimulation using an implanted ICD before hospital discharge. QRS morphologies of the induced VT were compared with those of preoperatively documented clinical VT. VT with significantly different QRS morphologies from preoperatively documented VT was defined as preoperatively non-documented VT and analysed separately. Catheter-based electrophysiologic study was performed in the patients in whom sustained VT occurred postoperatively or sustained VT was inducible by the programmed stimulation by the ICD. Catheter ablation was performed in patients with inducible sustained VT and an ICD was implanted in a patient with a high risk of sudden cardiac death.

After hospital discharge, the patients were followed by the referral cardiologists. The patient survival, major morbidities, episode of sustained VT or VF, and ICD discharges were surveyed by asking the cardiologists or by directly interviewing the patient. The median follow-up period was 60 months (interquartile range: 34–82).

Statistical analyses

Continuous values were expressed as means [standard deviation (SD)]. Incidences of recurrence of preoperatively documented LV summit VT and emergence of preoperatively non-documented VT after surgery were examined. All the statistical analyses were performed by JMP version 12.1 (SAS). P < 0.05 was regarded as indicating a statistical significance.

RESULTS

Patients

There were 11 patients who underwent surgery for LV summit VT between March 2002 and February 2020. The patient characteristics are shown in Table 1. All the patients were referred to our hospital for ventricular tachyarrhythmias refractory to pharmacological and catheter therapies. Six patients (55%) had VT storm that is defined as ≥3 separate episodes of sustained VT within 24 h. One patient was resuscitated from ventricular tachyarrhythmias. One patient had incessant ventricular premature contraction, resulting in syncope and heart failure.

Underlying heart disease was hypertrophic cardiomyopathy (HCM) in 9 patients (82%), including dilated phase of HCM in 2 patients, and idiopathic in 2 patients. One patient had a mechanical valve implanted at the mitral position 18 years prior to VT surgery. The average LV ejection fraction was 49% (SD: 11), including 5 patients with LVEF <50%. The mean wall thickness of the interventricular septum, the LV posterior wall, and at the LV summit region was 11 mm (SD: 2), 10 mm (SD: 2) and 15 mm (SD: 2), respectively, measured by transthoracic echocardiography.

One to 2 different QRS morphologies of clinical VT were documented in each patient, including 2 patients with 2 morphologies. All the patients had received 1–4 anti-arrhythmic drugs, such as beta-blockers in 9 patients, amiodarone in 7, sotalol in 5, sodium channel blockers in 3 and multi-channel blockers in 2. Five patients (45%) had an ICD implanted preoperatively.

All the patients had undergone 1–5 sessions of radiofrequency catheter ablation before surgery, including epicardial ablation in 4 patients (36%) and ethanol injection into a branch of the coronary artery in 1 patient. Two patients had a complication of catheter ablation: one with an injury to the right coronary artery during epicardial ablation, requiring open-chest repair at the referral hospital, and the other with a narrowing of the left main coronary artery associated with endocardial ablation at the LV outflow.

Six patients (55%) underwent a repeated electrophysiological study at our hospital prior to surgery for precise localization of the site of VT origin. VT was induced by programmed stimulation and the QRS morphology was compared with those of the clinically documented VT. The site of the origin of VT was defined at the endocardial or epicardial catheter position where the earliest activation occurred in the QRS. In the patients in whom stable VT was not inducible or the haemodynamics during VT were intolerable, the site of VT origin was determined by pace mapping by comparing the QRS morphologies during rapid pacing from an endocardial or epicardial site to those of the clinically documented VT. While the tip of the catheter electrode was placed at the endocardial or epicardial VT origin, coronary angiograms were taken to locate the VT origin with a relation to the coronary arteries.

Findings of intraoperative mapping

One patient who was expected to have an epicardial adhesion due to the previous mitral valve replacement underwent an endocardial and epicardial catheter-based mapping study through a small subxiphoid pericardial window at the hybrid operating room 1 week prior to surgery and did not undergo intraoperative mapping study. The remaining patients underwent an intraoperative mapping study. Sites of origin of clinically documented VT were determined by preoperative and intraoperative mapping studies in all patients. The location of the epicardial origin of VT is shown in Fig. 1. All the sites of VT origin were located within the region of LV summit that was covered with thick epicardial fat, or adjacent to the major coronary vessels. Two patients exhibited 2 separated sites of VT origin with different QRS morphologies. One patient exhibited the earliest epicardial activation of the VT at the superior LV between the right ventricular outflow and the bifurcation of the left coronary artery behind the pulmonary artery trunk.

Remarks in surgical procedures

Dissection of the adhesion between the pericardium and epicardium was required in 5 patients: 4 patients with previous epicardial ablation and 1 with previous mitral valve surgery. In 1 patient in whom the site of origin of VT was located at the LV outflow behind the pulmonary artery trunk, the pulmonary artery trunk was transversely incised to apply cryothermia at the VT origin located near the left main coronary artery.

In addition to the epicardial ablation, endocardial cryothermia was performed to create transmural cell death at the VT origin in all patients. In 10 patients (91%), the ascending aorta was incised and a cryoprobe was introduced into the LV chamber across the aortic valve through an aortotomy and cryothermia was applied on the LV endocardium (Fig. 4). The site for the endocardial ablation was directed by a needle punctured perpendicularly at the epicardial VT origin. In 1 patient, in whom the site of origin of VT was located at lateral LV close to the mitral valve annulus, an LV endocardial cryothermia was performed by a cryoprobe introduced into the LV cavity across the mitral valve through a right-sided left atriotomy. The cryothermia was applied on the LV endocardium behind the posterior mitral valve leaflet in such a way as to connect the cryolesion to the mitral valve posterior annulus to prevent peri-mitral valve reentrant activations postoperatively.

Four patients underwent additional procedures: coronary artery bypass grafting in 2 patients, including 1 for the stenosis of the main trunk of the left coronary artery—likely due to the thermal effect of preoperative radiofrequency catheter ablation, mitral valve repair in 1 and the left atrial appendage closure in 1.

Surgical outcome

There were no operative death or complications associated with intraoperative mapping, cryoablation, or manipulation of the coronary vessels. During the median follow-up period of 60 months (interquartile range: 34–82), 2 patients died of heart failure at 11 and 40 months after surgery, respectively. Both the patients had cardiomyopathy with an LV ejection fraction of <40%, preoperatively. Another patient died of myelodysplastic syndrome at 23 months postoperatively. There were no patients who died of ventricular tachyarrhythmias.

Impact of surgery on ventricular tachycardia

Preoperatively documented LV summit VT recurred in 1 patient (9%) 1 month after surgery. The VT was suppressed by anti-arrhythmic drugs and catheter ablation was performed 7 months postoperatively. There was no other patient who presented with the recurrence of preoperatively documented LV summit VT during the follow-up period. However, preoperatively non-documented VT emerged in 2 patients 5 days and 11 months after surgery, respectively. There was another patient in whom VT was detected by the implanted ICD 6 months postoperatively. Characteristics of the VT were not determined in this patient because the ECG was not recorded. Five-year freedom from preoperatively documented LV summit VT and non-documented VT was 91% and 73%, respectively.

Radiofrequency catheter ablation was performed in 2 patients. One patient underwent catheter ablation for recurrent LV summit VT at 7 months after surgery. The LV epicardium was successfully ablated via the great cardiac vein through the coronary sinus. The other patient underwent catheter ablation on the 5th postoperative day for intractable non-documented VT storm with frequent ICD discharges. The origins of the VT are located at the right and left ventricular outflow of the interventricular septum and 3 different ventricular sites, but not the LV summit region. VT was successfully controlled by the catheter ablation.

A subcutaneous ICD was implanted in 1 patient 19 days after surgery because of a history of resuscitated cardiopulmonary arrest preoperatively, even though VT was not inducible in the postoperative electrophysiologic study. One patient who developed preoperatively non-documented VT 11 months postoperatively has progressively deteriorated cardiac function and is currently under consideration for heart transplantation. The remaining patients have been free from VT and ICD shocks during a follow-up period of 64 (SD: 40) months.

DISCUSSION

Our experience demonstrated that transmural cryothermia guided by intraoperative mapping provided a satisfactory outcome for patients with LV summit VT. All the patients were at risk of arrhythmic death due to incessant ventricular tachyarrhythmias even with intensive and substantial medical treatments and endocardial or epicardial catheter ablation, before referring to our hospital for definitive therapy. More than half of the patients suffered VT storm or electrical storm of implanted ICD. There was no surgical mortality or long-term mortality related to VT. There were 4 patients who presented with VT postoperatively: 1 with an early recurrence of preoperatively documented LV summit VT and 3 with an emergence of preoperatively non-documented VT. There were no patients who had a recurrence of preoperatively documented LV summit VT in mid- and long-term postoperative periods. All these patients were successfully treated by additional catheter ablation and other medical therapies.

Localization of the site for effective ablation is crucial in surgery for VT. The target for surgical ablation is not visible from the epicardium, because the epicardium at the LV summit region is covered by thick fat tissue and the mechanism of VT is intramural reentry associated with intramyocardial patchy scar in the hypertrophied myocardium [12, 13]. A close collaboration with electrophysiologists, particularly through preoperative and intraoperative mapping studies, is essential for three-dimensional recognition of the location and extent of the substrate of VT [10, 11].

Although intraoperative mapping provides definitive information of ventricular tachyarrhythmias in a real-time base, there are several limitations in clinical use. First, target arrhythmias for ablation are not necessarily induced intraoperatively, because of the effects of general anaesthesia, decreased sympathetic nerve tone, hypothermia and epicardial cooling and other intraoperative factors that can affect the inducibility of arrhythmias [14]. Second, the axis and morphology of QRS complex in ECG may alter during open-chest and cardiac repositioning, and those ECG characteristics are extremely important for defining whether induced VT is identical to preoperatively documented VT. Lastly, VT origin that is localized and displayed on computer-generated cardiac models should be transformed to visible media with anatomical landmarks, such as the coronary arteries, as a guide for accurate surgical ablation.

There were 3 patients who presented with preoperatively non-documented VT, postoperatively. The emergence of preoperatively non-documented VT might not be related to incompleteness of surgery, but rather due to the progression of the pathological changes or polymorphic nature of the arrhythmias, which means that the VT originates in >1 site or from multiple exits from an intramural reentrant circuit. As intramyocardial patchy scar is the underlying substrate for intramural reentry in VT associated with HCM, three-dimensional localization of the scar should be important to prevent preoperatively non-documented VT. Late gadolinium enhancement in magnetic resonance imaging has been shown to detect a myocardial scar in patients with cardiomyopathy [15, 16]. Combining the substrate localization by magnetic resonance images with electrophysiological study may enable more curative ablation of preoperatively documented and preventive ablation of non-documented VT.

Since the substrates of LV summit VT distribute intramurally in the hypertrophied myocardium, creation of a transmural cell death is mandatory for the complete suppression of this refractory and life-threating arrhythmia. Because the LV summit region is covered by thick fat tissue, the removal of epicardial fat is important for the direct thermal conduction of ablation energy to the myocardium. We safely dissected the epicardial fat tissue by using an ultrasonic scalpel, avoiding injuries to the coronary vessels. Endocardial cryothermia of the LV at the VT origin was performed by a cryoprobe introduced across the aortic or mitral valve in all patients. An injection needle was used to guide the endocardial cryothermia just underneath the epicardial origin of VT. Hybrid surgery guided by simultaneous epicardial and endocardial intraoperative mapping for three-dimensional understanding of the VT substrate, followed by surgical epicardial cryoablation and catheter-based LV endocardial ablation on a beating heart may be an effective and less-invasive approach for LV summit VT.

Two patients died of heart failure at the long-term postoperative period. Neither of the patients underwent postoperative coronary angiography; therefore, the possibility of the coronary artery injury caused by epicardial cryothermia cannot be completely ruled out as the mechanism for myocardial ischaemia and heart failure. A canine study of a direct cryothermia of the LAD showed no occlusion but showed an intimal hyperplasia of the coronary artery 6 months after cryothermia [17]. As we performed in the present study, the coronary arteries should be carefully dissected by using a harmonic scalpel and snared with silastic tapes to avoid any injury to the coronary arteries by epicardial cryothermia.

Limitations

This is a descriptive report of a surgical approach and technique performed at a single institute in only 11 patients. Additional evidence with a greater number of patients in multiple institutes and comparative trials with other therapeutic modalities, such as simultaneous epicardial and endocardia ablation with catheters in a randomized fashion, is required to examine the usefulness of the approach and techniques described in the present study. Because the LV summit VT is a rare form of complex and life-threatening arrhythmia, this may be difficult to verify; however, additional studies are truly warranted.

CONCLUSION

Surgical cryoablation guided by intraoperative mapping provided a satisfactory outcome in patients with LV summit VT. Intraoperative mapping for precise localization of VT origin, followed by epicardial cryoablation directly of the myocardium and additional cryoablation of the LV endocardium for a transmural cell death at the VT origin, was essential in surgery for this unique and refractory VT.

ACKNOWLEDGEMENTS

We are deeply grateful to Prof. Steven Kirk for language proofreading and Ms. Hitomi Horiguchi for preparing illustrations.

Conflict of interest: none declared.

DATA AVAILABILITY

The data underlying this article cannot be shared publicly due to the privacy of patients who underwent a surgery and participated in the study. The data will be shared on reasonable request to the corresponding author.

Author contributions

Takashi Nitta: Conceptualization; Investigation; Methodology; Supervision; Writing – original draft; Writing – review & editing. Shun-ichiro Sakamoto: Methodology; Validation; Visualization. Hiroshige Murata: Data curation; Investigation; Methodology. Kenji Suzuki: Validation; Visualization. Naoki Yamada: Validation; Visualization. Yuki Iwasaki: Data curation; Methodology; Validation. Yosuke Ishii: Supervision.

Reviewer information

European Journal of Cardio-Thoracic Surgery thanks W Brent Keeling, Ulrich Otto von Oppell, Olga N. Kislitsina and the other, anonymous reviewer(s) for their contribution to the peer review process of this article.

REFERENCES

ABBREVIATIONS

ABBREVIATIONS
 
• ECG

  Electrocardiogram

 
• HCM

  Hypertrophic cardiomyopathy

 
• ICD

  Implantable cardioverter defibrillator

 
• LAD

  Left anterior descending coronary artery

 
• LV

  Left ventricle

 
• SD

  Standard deviation

 
• VT

  Ventricular tachycardia