Feasibility, safety, and short-term outcome of totally thoracoscopic mitral valve procedure

Accumulating knowledge of the structure, function, and pathology of the mitral valve (MV) has led to favourable surgical results in MV procedures. Advances in imaging and surgical instruments have allowed surgeons to perform less invasive sternum-sparing MV surgery [1]. Several kinds of minimally invasive access for MV procedures in adult patients have been popularly introduced into clinical practice. Lower hemisternotomy was used as a popular form of a minimally invasive MV procedure [2]. Right anterolateral minithoracotomy for MV surgery was a safe approach with very low operative mortality comparable to standard median sternotomy [3, 4]. MV surgery through a right minithoracotomy did not result in increased morbidity, mortality or procedural duration; therefore, long-term survival outcomes were satisfactory [5, 6].

Endeavours to reduce surgical trauma, hasten patient recovery, improve cosmetics, and increase patient satisfaction continued to further promote minimally invasive procedures. Additional minimally invasive surgical approaches, such as total video-assisted thoracoscopy or robotic assistance, have also been applied to repair congenital heart defects to minimize surgical trauma and improve cosmetic results [7‐9]. Apart from comparable merits in cosmetics to robotic surgery, an approach using totally thoracoscopy assistance to repair MV without complicated hardware equipment has a remarkable advantage [10].

However, the application of totally video-assisted thoracoscopic procedures on MV has been very limited in clinical practice. It was demonstrated that adverse events, including intraoperative conversion to sternotomy, re-exploration for bleeding, and valve-related reoperation within the same hospital stay, occurred with a predicted 25% rate at the beginning; this trend remained sluggish with declines in conjunction with increased operation training [11]. In this study, we reported the two-year experience of our department regarding the totally endoscopic procedure on MV performed by a single experienced surgeon (Z.XS.).

The totally video-assisted thoracoscopic procedure for MV disease is technically challenging, and its application is currently restricted to a handful of experienced operators because it entails the surgeon overcoming a lengthy learning curve [14]. In our department, the technique was performed on 51 cases by one experienced surgeon since the end of 2014. The result was technically feasible, required no transition to median sternotomy, and had a low rate of adverse events.

The totally video-assisted thoracoscopic procedure has been reported as a safe and effective method for repairing simple congenital heart defects with a rapid recovery of life quality and lower pain levels compared with both minithoracotomy and median sternotomy [15]. Meanwhile, lower systemic inflammatory reactions and myocardial damage were also observed after ASD repair with the method [16].

Compared with conventional median sternotomy, the sternum-exempt procedure could not only preserve the integrity of the osseous thoracic wall but also save time on the formidable task of haemostasis as in median sternotomy. Compared with minithoracotomy, which requires a 6–8 cm skin incision in the right inframammary groove to create a small anterolateral “working port” [17], the totally video-assisted thoracoscopic MV procedure had an advantage in terms of minimal invasiveness. The totally thoracoscopic MV procedure, which offers fewer surgical incisions and improved cosmetics, was more attractive to patients than other approaches. The sparing of a rigid incision retractor rendered less injury to chest wall integrity. Nonetheless, high technical demands have impeded the popularization of minimally invasive endoscopic MV surgery.

It is too challenging to widely introduce this procedure because it is still difficult to perform for a large majority of surgeons. The survey of surgeons who were experienced in minimally invasive MV surgery showed that 90 % of the respondents believed that more than 20 cases were required to gain familiarity with the procedure [18]. Our department introduced a skilful cardiac surgeon who was well versed in performing the MV procedure with the guidance of a video-assisted thoracoscope instead of port-access minithoracotomy and robotic assistance. He had already carried out nearly 100 cases of MV repair/replacement from 2012 to 2014.

There were some disadvantages and pitfalls in terms of the characteristics of the totally video-assisted thoracoscopic MV procedure. Compared with the three-dimension vision under the totally direct open procedure, the manipulation zone lacked a stereoscopic experience. The field of vision during the operation was narrowed down by enlarged thoracoscope so that the surgeon had increased difficulty during manipulation. Each knotting entailed crossing of the sutures outside of chest cavity and then pulling down the knots with the assistance of a knot pusher.

There were some technical limitations in its clinical application. One of the key elements for procedural success was CPB establishment through femoral artery-vein bypass. Patients had a low body weight with smaller femoral blood vessels, which might hinder the insertion of venous or arterial cannulas for optimal CPB. As Table 1 indicates, the great majority of MV patients in our region were female and rheumatic, short in statue, with a BMI ranging from 18.1 to 30.8. Six patients received a 16F femoral artery cannula, and only one case underwent bilateral femoral artery cannulation due to hypoperfusion. The SVC was completely cannulated with a 16F cannula, which was inserted through the RJV during the initial phase of the programme. Due to the scar effect on the neck and the associated risk of penetrating into the pleural cavity, the plan was terminated after performing the first 11 cases in the middle of 2015. Instead, the operator completed the catheterization of the SVC directly through port 1 under the view of thoracoscopy. The femoral vein was successfully cannulated with a 28F cannula for IVC drainage. This satisfactory perfusion result could be due to the lower morbidity rate of peripheral artery disease in this population. Elective preoperative coronary angiography was often performed on the right radial artery, which avoided possible intima injury to the femoral artery.

The second element was heart arrest and cardiac protection. In our study, a lengthened cardioplegia perfusion cannula and transthoracic aortic clamp were introduced conventionally and smoothly. It was reported that intraluminal aortic clamping was applied by means of a sophisticated device consisting of a three-lumen catheter named an Endoclamp, which enabled simultaneous occlusion of the aorta, antegrade delivery of cardioplegia, and venting through the aortic root [19]. The minimally invasive approach should be excluded in patients with aortic regurgitation (AR) greater than grade I because of the risks of inadequate cardioplegia delivery and left ventricular distension [12]. No evidence in our study demonstrated that retrograde perfusion per se during minimally invasive MV surgery increased the incidence of cerebrovascular accidents [20].

The requirement of rethoracotomy for bleeding represented a major drawback of minimally invasive MV surgery because the entire operative field could not be directly visualized. In this study, there was one case undergoing transition to mini-thoracotomy for bleeding at the site of temporary pacemaker wire implantation, but no reexploration for bleeding happened after chest wall closure. This excellent outcome was attributable to overcoming the learning curve and innovative measure by the operator, who fixed the pacemaker wire with Teflon pressed with a titanium clip at both ends [21]. It was also routine for the operator to use a thoracoscope to inspect the inner thoracic wall thoroughly before closing the incision. It had been recorded that the rate of reexplorations for bleeding fell; it was reduced from 8.2 to 1.9% after 300 operations [11]. In contrast, we observed one case in which that patient died from prosthesis thrombosis because of insufficient anticoagulation. Due to fear of bleeding on the condition of thrombocytopenia and continuous excessive drainage, we did not regularly prescribe warfarin for prosthesis anticoagulation. The subsequent MV prosthesis replacement re-do procedure was ineffective, and the patient died from coagulation complications soon after the procedure. Granted that only two patients suffered reoperation after ACC release, the resultant operative quality of totally video-assisted thoracoscopic MV surgery could be on par with the conventional median sternotomy approach.

Regarding feasibility, chest wall deformity was restricted to perform the totally video-assisted thoracoscopic procedure to some extent. Due to case limitation and careful patient selection, we were not confronted with any cases of serious chest deformity or secondary port-access video-assisted thoracoscopic procedures. Nevertheless, the mere presence of uncorrected congenital chest wall deformities should not deter surgeons from minimally invasive cardiac surgery [22]. Loose pleural adhesion did not restrict the feasibility unless indivisible adhesion existed. We also experienced several cases of painstaking synechiotomy and then completed the MV procedure under the guidance of thoracoscope. It had been shown that the outcomes in redo-port access surgery after previous port access surgery were favourable [23]. In our population, two cases required re-do surgery of a previous MV procedure were successfully completed in patients who underwent percutaneous balloon mitral valvuloplasty and MVR at median sternotomy. To some extent, the procedure from right thoracotomy access should be the optimal choice for re-do MV patients because post-sternum adhesion is avoided. The LA blade retractor used to pull up the LA anterior wall was held by a shaft that penetrated into chest cavity rather than through the main incision [24, 25]. This approach could minimize the size of the main incision, which merely allowed the prosthesis to enter into the port smoothly. Meanwhile, the additional penetrating point for the shaft did not increase the cosmetic burden.

Concomitant TVP was not associated with a risk-adjusted increase in mortality, regardless of TR severity. A more liberal approach to TVP at the time of MVR might be justified when long-term benefits are thought to outweigh incremental short-term morbidity risk [26]. In our study, we completed MV procedures associated with TVP in 28 cases. The operator also performed the Warden procedure in one case who suffered from ASD combined with supracardiac partial anomalous right upper pulmonary venous connection to SVC. A giant self-pericardium patch was added to the RA as an atrial septum to separate the total pulmonary venous connection into the LA [27].

We completed 51 cases of MV procedures totally by thoracoscopic visual guidance. The aetiology of these MV diseases was comprised of mitral regurgitation, rheumatic mitral disorders, congenital heart defect or left atrial myxoma. Through three incisions that were 2–4 cm length and one penetrating point in the right chest wall, mitral valve replacement or repair was successfully administered by an experienced surgeon. The additional procedures included LA thrombus removal, Cox-maze radiofrequency, ASD repair, and TV annuloplasty. Apart from its minimal invasiveness characteristics and wide range of clinical applications, the MV procedure under the assistance of thoracoscopic visual guidance and TEE surveillance was also a safe and effective operation approach. In total, the thoracoscopic procedure for conventional MV diseases is worth being advocated broadly, and more experienced surgeons should be trained on this procedure due to its high-demand in terms of manipulation skills and three-dimensional conception. In summary, the totally thoracoscopic procedure for MV disease by an experienced operator is feasible, safe, effective, and merits widespread adoption.