Feasibility of concomitant exclusion of left atrial appendage during novel transapical off-pump beating heart mitral valve repair

Abstract

The AtriClip device enables the safe and reproducible epicardial clipping of the left atrial appendage. Transapical off-pump beating heart mitral valve repair using NeoChord DS100 Artificial Chordae Delivery System has matured and become more standardized. We aim to evaluate the feasibility of combining NeoChord repair and left atrial appendage exclusion in a single procedure through the same minithoracotomy in patients with mitral valve prolapse and atrial fibrillation. From 2018 to 2019, seven patients with severe mitral regurgitation and atrial fibrillation underwent transesophageal echocardiography-guided transapical off-pump mitral valve repair with the novel NeoChord DS 1000 system and concomitant left atrial appendage exclusion using the AtriClip Pro II device. Both procedures were performed via left mini-thoracotomy. The AtriClip device was applied after the NeoChord repair was done. All seven patients had less than moderate mitral regurgitation after the NeoChord repair and successful left atrial appendage occlusion. There were no device or procedure-related complications. Clinical follow-up revealed significant symptomatic improvement, and no cardiovascular complications were reported. Transesophageal echocardiography at 6–12 months post-procedure showed stable left atrial appendage occlusion with no residual flow between the left atrium and the left atrial appendage and a stump of less than 5 mm. Beating heart epicardial clipping of the left atrial appendage using AtriClip concomitant with transapical mitral valve repair using Neochord DS 1000 system is a feasible and safe treatment option in mitral valve prolapse and atrial fibrillation in patients with limited indications. However, its safety needs to be confirmed in a larger series of patients.

Supplementary Information

The online version contains supplementary material available at 10.1007/s10047-023-01383-2.

Introduction

Degenerative mitral regurgitation is the third most common cardiovascular disease in the US and Europe, accounting for one-third of all mitral regurgitation cases. It is a progressive condition that, if left untreated, can lead to congestive heart failure and death. While no medical therapy is curative, most cases are treated with surgery, preferably mitral valve repair. Transesophageal echocardiography (TEE)-guided transapical off-pump beating heart chordal implantation with NeoChord DS 1000 Artificial Chordae Delivery System is a minimally invasive option gaining popularity. It has been performed on more than 1200 patients worldwide [1]. Its safety and efficacy in repairing mitral valve prolapse, even in high-risk patients, have been demonstrated.

Atrial fibrillation (AF) is a frequent complication or consequence in patients with mitral valve diseases [2]. Because of the short and long-term risks associated with AF, surgical ablation for AF is recommended at the time of concomitant mitral operations (Class I, Level A) [3]. However, the actual rate of surgical ablation for AF in patients undergoing cardiac surgery is low at 40% [3]. Part of the reason might be due to potential increased surgical risk and questionable effectiveness, particularly in patients with large left atrium and long-standing persistent AF.

Since the left atrial appendage (LAA) has been recognized as the source of embolic material in patients with AF, surgical LAA occlusion is typically undertaken as a sole concomitant procedure if the Maze procedure is deemed unsuitable. Because it has been shown to be an effective alternative to anticoagulation, LAA occlusion could help to reduce long-term stroke risk and anticoagulation-related complications [4].

While the surgical exclusion of LAA by internal or external sutures or stapling devices has been criticized for its safety and effectiveness in achieving the goal [5], epicardial LAA occlusion with the AtriClip device has emerged as a safe and effective technique for occluding the LAA successfully [6, 7]. The clip is now routinely used in conjunction with sternotomy or minimally invasive open-heart surgery.

Because both off-pump beating heart mitral valve repair and LAA clip application are performed using a small left thoracotomy, both procedures can theoretically be performed in the same setting. Also, especially if the patient is in AF, it maximizes the benefits of a less invasive technique. The goal of this paper was to present the first case series that used this combined approach.

Methods

This series retrospectively reviewed the first seven patients who underwent mitral valve repair with NeoChord DS 1000 Artificial Chordae Delivery System (NeoChord, Inc., St. Louis Park, MN) with concomitant LAA closure with AtriClip (AtriCure, Mason, OH, USA). The study was approved on 18/7/2018 by the local ethics committee (No. UW 18-337), and written consent for the combined procedure was obtained from all patients.

Patient selection

All the patients referred for mitral valve surgery were screened according to their clinical and echocardiogram data. Patients who were potential candidates for off-pump beating heart mitral valve repair underwent a preoperative TEE to assess their suitability for NeoChord repair. All the patients had a guideline-based indication for surgical correction of mitral regurgitation (MR) [8].

The morphology of the mitral valve was assessed by TEE prior to surgery, according to the methodology reported previously [9]. The length of the leaflets and the annulus diameter were measured in two (2D) and three dimensional (3D) views of the prolapsing segments. According to the mitral valve morphology, anatomical types were defined as previously described: ‘Type A’, isolated central posterior leaflet prolapse/flail; ‘Type B’, posterior multisegment prolapse/flail and ‘Type C’, anterior, bileaflet or paracommissural disease with or without leaflet and annular calcifications [10].

Apart from the area of prolapse, the leaflet-to-annulus ratio or index (LTI) was used to predict procedural success. LTI is defined as the ratio between the sum of anterior and posterior leaflet lengths over antero-posterior length, and if greater than 1.2, the coaptation is expected to be adequate after correction of the prolapse by the NeoChord repair [11].

Even though the selection criteria for NeoChord repair were mainly based on the presence of favourable anatomy, it was reserved for older and higher-risk patients due to the lack of long-term results.

Patients were excluded from the study if they had previously undergone mitral valve surgery or any other cardiac procedures within the previous three months, were in an emergency situation or failed to provide informed consent.

After the patient agreed to the NeoChord procedure, those diagnosed with paroxysmal AF or persistent AF were offered the option of concomitant LAA closure. During the discussion, they were informed that the concomitant procedure would only be carried out if the intra-operative TEE revealed no left atrial clot, an exclusion criterion for the LAA closure procedure.

Operative technique

The NeoChord repair was performed as described previously [12–14]. Briefly, the patient was put under general anaesthesia and intubated with a single-lumen endotracheal tube without lung isolation and monitored as usual. A perfusionist with a primed cardiopulmonary circuit was on standby in the operating room. Before draping, hand-held transthoracic echocardiography was used to determine the location of the apex, and a 3 to 5 cm left thoracotomy incision was made. The pericardium was then opened, and a soft tissue retractor was inserted directly into the pericardium as a wound retractor, eliminating the need for rib-spreading. The apical entry site was identified by digital palpation and TEE, posterior and lateral to the true apex (Fig. 1). Heparin was given to achieve an activated clotting time of 250 s. Two ‘U’ purse-string sutures with 2/0 Prolene (Ethicon, Somerville, NJ, USA) pledgets were applied at the entry site. The NeoChord DS 1000 system was prepared and introduced into the left ventricular cavity of the beating heart. Under real-time 2D/3D TEE guidance, jaws of the NeoChord system were opened, the prolapsing leaflet segment was grasped, and Gore-Tex CV-4 neochordae (Gore-Tex; W.L. Gore & Associates Inc., Flagstaff, AZ, USA) were implanted on it. The procedure was repeated until a sufficient number of neochordae had been implanted to support the prolapsing segment. We aimed to achieve ‘over-correction’; usually, a trivial to mild residual posterior directing MR would be left. Procedural success was defined as less than moderate MR at the end of the procedure.

Fig. 1.

NeoChord procedure. A TEE finger test. X-plane image of left ventricle showing finger probing (red arrows) postero-lateral to the true apex; B purse-string sutures at the entry site; C blue arrow showing completed NeoChord procedure with 3 fixed artificial chordae at the entry site. Red arrow showing camera port inserted at 3rd intercostal space, anterior axillary line; D TEE X-plane assessment of mitral valve showing no prolapse and residual mitral regurgitation

The LAA was managed once the NeoChord procedure was completed and the apical entrance site was closed. A 5 mm port was inserted in the 3^rd intercostal space near the anterior axillary line (Fig. 2) directly through the chest wall and pericardium during temporary respiratory apnea of about 5 s. The superior margin of LAA was identified using a 30-degree endoscope passed through this port. The AtriClip device was inserted through the pre-existing left thoracotomy wound (Fig. 3). The base of LAA was visualized and measured using an AtriClip sizer to select the appropriate device length (35, 40, 45 or 50 mm). The AtriClip Pro 1 clip was utilized in the first two cases, while the rest of the patients received the AtriClip Pro 2 clip. By adjusting the device knobs and utilizing minimally invasive instruments, the clip was put near the base of LAA. After the clip was temporarily clamped, TEE and endoscopic assessment were used to confirm its proper placement, which was defined by previous reports as a residual stump of less than 10 mm and the absence of flow between the left atrium and the LAA with color Doppler [4]. If not, the clip was opened and repositioned. Once the satisfactory position was achieved, the clip was released. The deploying system was withdrawn, and the minithoracotomy was closed in a routine fashion.

Fig. 2.

Setup of left atrial appendage closure. A Atricure Pro 2 clip-direction adjusted according to the LAA before insertion; B the clip inserted through the thoracotomy wound (red arrow) used for the NeoChord procedure; C the camera port used for chest tube insertion after the procedure

Fig. 3.

AtriClip closure of the left atrial appendage. A Sizing of LAA; B application of LAA clip; C AtriClip deployed; D TEE assessment of the LAA: blue arrows—TEE X-plane images of LAA before closure, orange arrows-TEE X-plane images showing closed LAA and red bracket-assessment of the residual stump length

All patients were monitored in the intensive care unit on the first post-operative day. They were discharged after the removal of the chest tube and echocardiogram.

At 6 months, 1-year, and 2-year post-procedure, all patients had routine clinical follow-ups, and echocardiography was performed to assess the mitral valve. TEE assessment of the LAA status was usually not done until three months after the surgery. However, due to COVID-19, it was further delayed for some patients.

Statistical analysis

Demographic categorical variables were expressed as percentages, while quantitative variables were expressed as the median and interquartile range (IQR) or mean ± SD (minimum and maximum). SPSS statistical software was used for this purpose (IBM SPSS Statistics, version 24.0. Armonk, NY, USA).

Results

Patient demographic and operative data are summarized in Table 1. The mean age of this series of patients was 74.3 ± 5.9 years with a mean EuroScore II of 2.69% ± 1.8, a CHA2DS2-VASc Score of 4 ± 0.58 and a HAS-BLED Score of 2.14 ± 0.69. HAS-BLED Score evaluates bleeding risk in patients with atrial fibrillation. Five out of seven patients (71.4%) were in persistent atrial fibrillation and taking an anticoagulant, and the duration of AF was 17.3 ± 13.7 months. All patients had severe mitral regurgitation with preserved left ventricular function (Table 2).

Table 1.

Patient characteristics

Patient no.	1	2	3	4	5	6	7
Age (Years)	77	76	71	64	82	78	72
Gender	F	M	F	F	F	M	F
Hypertension	Yes	Yes	No	Yes	No	Yes	Yes
Dyslipidemia	Yes	Yes	No	Yes	No	Yes	No
Diabetes	No	No	No	No	No	Yes	No
Renal impairment	No	No	No	No	No	Yes	No
Other comorbidities	Bronchiectasis		Ca breast	Thyrotoxicosis	Ca stomach		Ca cervix
NYHA functional class	2	2	3	3	3	2	2
EuroScore II (%)	5.25	1.41	1.51	1.25	5.27	2.6	1.57
CHA2DS2-VASc Score	5	4	3	4	4	4	4
HAS-BLED Score	3	3	2	2	1	2	2
Pre-operative AF type	Persistent	Persistent	Persistent	Persistent	Persistent	Paroxysmal	Paroxysmal
Duration of AF (Months)	33	21.5	1.5	4.5	26	-	-
Pre-operative anticoagulation	Apixaban	Warfarin	Nil	Apixaban	Aspirin	Aspirin	Nil
Anatomical MV type	A	B	A	C	B	C	A
Leaflet involvement	P2	P2 & P3	P2	P2 w/annular calcification	P2 & P3	P3 close to commissure	P2 & P3
Operative and post-operative
Implanted neochords (n)	4	3	2	3	3	3	3
AtriClip size (mm)	40	45	40	40	35	40	45
Surgery time (Minutes)	231	188	125	137	140	92	165
Complications	Nil	Nil	Nil	PEF	Nil	Nil	Nil

AF, Atrial fibrillation; MV, mitral valve; NOAC, novel oral anticoagulants; NYHA, New York Heart Association; PEF, pericardial effusion

Table 2.

Echocardiography and transesophageal echocardiogram assessment data

Patient no.	1	2	3	4	5	6	7
Echocardiography
Follow-up (Months)	12.2	14.7	18	13.6	12	24.7	24.4
LVEF (%)
Pre-operative	50	55	60	60	62	60	65
Follow-up	60	60	70	70	75	65	65
Mitral regurgitation
Pre-operative	Severe	Severe	Severe	Severe	Severe	Severe	Severe
End of the procedure	Mild	Mild	Mild	Mild	Mild	Mild	Mild
Follow-up	Trivial	Trivial	Mild	Mild	Moderate	Trivial-mild	Mild
LAA status
Follow-up TEE (Months)	23	15	11	20	8	7	9
Occlusion	Successful	Successful	Successful	Successful	Successful	Successful	Successful
Flow to LAA	None	None	None	None	None	None	None
Max Stump depth (mm)	6.6	4.2	5.5	8	7.2	5	5.2
Change compared to intra-operative TEE	None	None	None	None	None	None	None

LVEF, left ventricular ejection fraction; LAA, left atrial appendage; TEE, transesophageal echocardiography

Operative outcomes

In all the patients, procedural success was achieved in both the NeoChord mitral valve repair and LAA exclusion (less than moderate residual MR and less than 1 cm LAA stump, respectively). On average, 3.0 ± 0.58 (Range, 2 and 4) neochordae were implanted. The mean procedural duration was 154 ± 45.4 min. There were no intraoperative deaths and no major cardiovascular complications in this series. A median hemoglobin decrease of 1.0 g% (IQR, 0.07–1.9) was observed. A minor complication of pericardial effusion necessitating pericardial tapping on the sixteenth post-operative day occurred in 1 patient.

Patients were transferred to the ward after a median stay of 29 h (IQR, 25–31) in the ICU and were discharged after a median hospital stay of 10 days (IQR, 8–15). All patients resumed their pre-operative anticoagulation regimen except one patient with persistent AF who was given warfarin.

Follow-up

Follow-up completeness was 100%. At the time of the latest follow-up, none of the patients had suffered a stroke or a transient ischemic attack. All the patients showed clinical improvement of at least 1 NYHA class (Table 3). Follow-up echocardiogram at 1 year and 2 years (in 2 patients) showed similar or reduced mitral regurgitation compared to post procedural assessment. Only 1 patient progressed from mild to moderate MR but remained asymptomatic (Table 2).

Table 3.

Follow-up data

Patient no.	1	2	3	4	5	6	7
Follow-up time (months)	30	28	23	22	20	18	16
NYHA functional class	1	1	2	1	2	1	1
Stroke	None	None	None	None	None	None	None
Anticoagulation	Aspirin	Aspirin	Nil	Aspirin	Aspirin	Aspirin	Aspirin
Rhythm	AF	AF	AF	AF	AF	SR	SR

TEE assessment in all the patients showed that the size of the LAA stump remained static since the LAA exclusion procedure (Table 2). At the follow-up TEE assessment, there was no flow between the left atrium and the LAA; the mean stump depth was 6 mm, which was comparable to the post-operative TEE. Following TEE evaluation, all patients' anticoagulation was switched to an antiplatelet drug at the subsequent follow-up.

Discussion

To our knowledge, this is the first case series reporting the LAA closure with the clip as an add-on procedure in patients undergoing off-pump beating heart mitral valve repair with NeoChord 1000 DS system. We demonstrated that performing the two procedures as a single combined surgery is safe, effective, and simple in patients with limited indications.

By far, the most prevalent cause of degenerative mitral regurgitation is mitral valve prolapse, also known as Carpentier type II lesion. Chordae reconstruction is a well-established repair procedure with excellent short and long-term results. The NeoChord system is a novel, minimally invasive technology for inserting artificial chordae into a beating heart mitral valve. Studies have shown it is a safe and effective alternative for correcting anatomically suitable prolapses [15].

AF is a common co-morbidity in patients with mitral valve disease. In our unpublished observation, AF was present in approximately 33.3% of the patients put up for mitral valve surgery at our center. Concomitant Maze procedure with LAA occlusion should be performed during mitral valve surgery as it carries a high success rate and can provide long-term benefits to the patients. Several societies and guidelines have classified this concomitant procedure during mitral valve surgery as a class I indication [3]. On the other hand, isolated LAA occlusion may be undertaken to minimize the risk of ischemic stroke in patients with high surgical risk or those in whom conversion to sinus rhythm seems very unlikely [16].

The importance of managing LAA during the time of other cardiac surgery in preventing strokes has been demonstrated by the LAAOS III trial [4]. Although the best protection is achieved by combining anticoagulation with LAA closure, the study showed a 40% reduction in long-term stroke rate with LAA closure alone [16]. Trials comparing anticoagulation and left atrial appendage occlusion add to the evidence favouring LAA closure. These trials described the non-inferiority of LAA occlusion to anti-coagulation in preventing stroke while avoiding the risk of major bleeding due to anti-coagulation medications. After extensive discussion, all our patients in this series decided to cease taking anticoagulants. This condition could reflect the preferences of the majority of patients and explain the low compliance rate for oral anticoagulants [16].

It is critical to have effective LAA closure to reap the potential benefits. The LAA could be closed using a variety of surgical techniques. Endocardial closure carried the highest failure rate, while stapler ligation might have the highest bleeding rate. Noncutting staplers can be used to exclude the LAA, although late lumen recanalization and bleeding along the staple lines can occur. Cutting staplers guarantee that the trabeculated part of the LAA is removed and pericardial buttressing stops bleeding along the staple line; nonetheless, the delicate tissue underneath the staple line is susceptible to tearing. If there is flow to the LAA or a big stump following surgical closure, the risk of stroke could be higher than if it is not touched at all [16]. We adopted the LAAO trial's criteria for successful LAA occlusion, defined as no residual flow to LAA and a residual stump of less than 1 cm [4]. Among all surgical techniques, the LAA clip has been shown to be the most effective. It can be applied regardless of LAA morphology or LA dilatation, or ostial size. Since it is placed on the epicardial surface, the risk of complications of thrombosis and infection seen with the percutaneous device is potentially reduced [17]. The safety and efficacy of AtriClip have been shown in the EXCLUDE trial, and the US food and drug administration has approved its use in patients during concomitant cardiac surgery. Successful exclusion has been reported in up to 98.4% of patients at 1 year with no device-related complications [6]. In a large multicenter study of over 200 patients who underwent exclusion with an AtriClip at the time of thoracoscopic ablation, the freedom from cerebrovascular events was 99.1% [18].

A few reports described the thoracoscopic application of the clip on a beating heart either as a stand-alone procedure or with catheter ablation [17, 19]. However, the stand-alone thoracoscopic application of the clip requires the working port to be placed more cranially and the pericardium to be opened more posteriorly to have better access to the left atrial dome and pulmonary veins. That meant additional wound and procedure risks to this group of older and higher-risk patients. Moreover, most of the thoracoscopic ablations were performed in non-valvular AF patients, and the efficiency in valvular AF was not well established. The NeoChord procedure combined with the LAA clip application is one step forward and more advantageous as both procedures share the same surgical access and can be performed in a single setting.

In this case series, we have highlighted a few technical aspects critical to achieving it safely and effectively. AtriClip was applied after the mitral pathology was corrected by the NeoChord. In the presence of severe mitral regurgitation, there is increased pressure in the left atrium, and LAA has higher and earlier systole filling [20, 21]. There is an increased risk of bleeding and rupture if the appendage is manipulated at this time, as LAA is fragile and distended. Second, using Atricure Pro 2 or a newer type of clip during such a combination operation setting would be easier. The control knobs help to direct and fine-tune the position and these versions of the clip have enhanced hoopless end effector that allows for smooth retrieval of the delivery system after applying the clip. The use of a 30-degree endoscope further aids the clip’s positioning by locating the LAA margin. The camera port would not cause extra trauma to the patient as it served as a chest tube insertion site at the end of the procedure.

All seven patients had successful LAA occlusion according to the intra-operative TEE criteria. Re-assessment TEE at least 6 months afterwards demonstrated stable results. This reinforced the findings of a previous study that showed optimal LAA occlusion evidenced by intraoperative TEE may obviate the need for follow-up imaging [22]. The pros and cons of anticoagulation were discussed with all patients with atrial fibrillation, but they all decided to take just aspirin.

It has been projected that beating heart transapical repair using the Neochord DS 1000 system will be more generalizable and reproducible because of its technical simplicity and shorter learning curve [14]. We have performed 40 cases of mitral repair using the Neochord DS 1000 system in selected patients and have had no failure cases thus far. We went a step further and demonstrated that Neochord DS 1000 system with concomitant LAA occlusion is likewise as simple and can offer further benefits in those with atrial fibrillation.

Conclusion

Our first experience proved the concept that concomitant LAA occlusion during the NeoChord procedure is technically feasible and safe. A larger study is needed to confirm that the device's deployment during the NeoChord procedure can be done quickly and securely without any device-related or procedure-related complications. Combining the NeoChord procedure with LAA clip occlusion offers greater benefits in mitral regurgitation and atrial fibrillation in patients with limited indications.

Supplementary Information

Below is the link to the electronic supplementary material.

Funding

There was no funding available for the study.

Data Availability

Data is available within the article and its supplementary information.

Declarations

Conflict of interest

The authors have no conflicts of interest to declare.