Early and mid-term outcomes of minimally invasive mitral valve repair via right mini-thoracotomy: 5-year experience with 129 consecutive patients

This retrospective observational study was performed using prospectively collected data of 309 consecutive patients who underwent mitral valve repair procedures between January 2014 and January 2020. All patients were routinely examined using contrast enhanced CT and head and neck magnetic resonance imaging. For preoperative enhanced CT in MIMVr cases, images were obtained with the patient in an exhalation phase with arms down. Patient selection was determined mainly based on the decision of the attending surgeon, though for high-risk patients such as those with impaired left ventricular function, a heart team composed of cardiac surgeons and cardiologists was responsible for the decision. Patients with anatomical and/or physiological problems were evaluated with reference to advice from an appropriate specialist. Notably, cases with a chest wall deformity such as scoliosis or funnel chest were carefully considered. A total of 54 sternotomy procedures were performed, with that modality chosen because of very poor left ventricular ejection fraction (less than 30%), severe peripheral vascular disease [ankle-brachial index (ABI) 0.6 or lower], strong pleural adhesions (history of right lung surgery, radiation therapy, etc.), severe chronic obstructive pulmonary disease (forced expiratory volume during first second < 1 L), active endocarditis, or concomitant coronary artery disease requiring surgical intervention. History of dialysis or impaired renal function was not considered to be a contraindication for MIMVr. Indications for treating liver cirrhosis patients with MIMVr were the same as used for cases of normal cardiac surgery with a full sternotomy procedure.

MIMVr with robotic assistance using the da Vinci system (Intuitive Surgical Inc, Mountain View, CA) was instituted at our institution in June 2018, thus from that time patients were treated with that system and they were excluded from this study. Of 256 candidates for MIMVr via an RT, 255 (99.6%) underwent an effective repair procedure, while 1 (0.4%) required replacement after the repair attempt because of severe annulus and leaflet calcification. After excluding 92 who underwent MIMVr with robotic assistance with the da Vinci system and 22 with concomitant aortic valve replacement, 141 patients who underwent MIMVr via an RT approach were enrolled, of whom 129 (91%) also had late follow-up (longer than 6-month) results available.

The main outcomes investigated were early and late mortality, perioperative complications, and freedom from mitral regurgitation (MR) recurrence and reoperation. Early mortality was defined as death from any cause occurring within 30 days of the operation or before hospital discharge. Stroke was defined as any new focal or global neurological deficit lasting more than 24 h, and was confirmed whenever possible by neurologist and neuro-radiologist examinations. Postoperative stroke was diagnosed when evidence was found of a new neurologic deficit, with the presence of morphologic substrate confirmed by computed tomography or magnetic resonance imaging. Respiratory insufficiency included ventilation failure, and performance of reintubation or a tracheostomy.

All pre- and postoperative transthoracic echocardiography results were obtained using a quantitative method. In cases with more than moderate MR, a qualitative method was added and performed by trained cardiac echography technicians and cardiologists. All patients were examined from 1 to 4 weeks, and again 12 and 24 weeks postoperatively, then were contacted thereafter for follow-up data. All patients in the late phase underwent follow-up examinations every 24 weeks. The median follow-up time was 35 ± 15 months (interquartile range 9–69 months) and follow-up data were 91.4% complete. The severity of MR was graded based on recommendations from the European Society of Cardiology and European Association for Cardio-Thoracic Surgery. Recurrent MR was defined as moderate or severe using a four-point grading system (trivial, 1; mild, 2; moderate, 3; severe, 4).

Following general anesthesia with differential lung ventilation, the patient was placed in a 20° left lateral position with a pillow under the left thorax. A cardiopulmonary bypass (CPB) was instituted via the femoral arterial or axillary artery and femoral venous cannulation. We selected the femoral artery for retrograde perfusion and the axillary artery for antegrade perfusion. Axillary cannulation was selected if any of the following criteria were satisfied in the aorta or an iliac artery at any point: (1) thickness of the thrombosis is > 3 mm; (2) thrombosis is more than one-third of total circumference; and (3) calcification present in the total circumference. Femoral cannulation was selected in the remaining patients.

The tip of the venous cannula was positioned in the right atrium under transesophageal echocardiographic (TEE) guidance. A right lateral mini-thoracotomy, 5–6 cm in length, was performed in the 4th intercostal space, with the incision placed 2–3 cm lateral to the nipple in males and in the sub-mammary crease in females (Fig. 1a). For patients with a narrow chest, such as a funnel chest, the incision was placed more to the lateral side of the anterior axillary line (Fig. 1b). Especially for funnel chest cases, double venous cannulas are recommended for positioning the SVC and IVC to secure the surgical field of view. On the other hand, for obese patients in whom the mitral annulus could be compressed by the elevated diaphragm and the heart tended to be in a lying position, the incision was placed more to the cranial side and the approach side was considered to be the 3rd intercostal space (Fig. 1c). Using a small thoracic and soft tissue retractor, a transthoracic Cygnet aortic cross-clamp (Vitalitec Inc, Plymouth, MA, USA) was inserted through the same 4th intercostal space. Antegrade cold blood cardioplegia (6 °C) was administered directly into the aortic root. The first dose included a potassium concentration of 10 mEq/L with flow at 400 mL/min for a concentration of 15 mL/kg, then repeated doses every 40 min including a potassium concentration of 5 mEq/L with flow at 400 mL/min for the same dose given and then repeated within 40 min if necessary. Throughout the procedure, the surgical field was flooded with carbon dioxide through the soft tissue retractor. The left atrium was opened posterior to Sondergaard’s groove and a left atrial retractor was used to expose the mitral valve. Standard mitral valve repair techniques were employed [7]. Deairing was performed by filling the left atrium with saline during closure and via the cardioplegia puncture site on the ascending aorta. In the setting of degenerative mitral valve disease, a mitral valve repair was the first approach attempted for a degenerative mitral valve, while mitral valve replacement was commonly used in cases with severe annulus and/or leaflet calcification.

All patients underwent accurate intraoperative transesophageal echocardiogram examinations performed by an anesthesiologist licensed by the Japanese Board of Perioperative Transesophageal Echocardiography (JB-POT) before and after weaning from the cardiopulmonary bypass machine. MR assessment was performed with a quantitative method. The policy of our institution states that in patients for whom a repair is attempted, the mitral valve should be replaced if (a) some degree of mitral regurgitation is shown in hydrostatic saline test results after several attempts, (b) the coaptation surface is not adequate to guarantee long durability, and (c) intraoperative echo findings show no more than mild MR.

This study was conducted following approval from the Institutional Review Board of Chiba-Nishi General Hospital, with a specific waiver granted for need of individual patient consent. Continuous data are expressed as mean ± standard deviation or median with interquartile range (IQR), and categorical data as percentages. Learning curves and trend lines were created using a two-dimensional scatter plots to show approximate curves for changes in operation, CPB, and aortic cross-clamp (ACC) times. Survival and time-to-event analyses for rates of reoperation, MACE events, and recurrent mitral regurgitation were assessed using the Kaplan–Meier method. The grade of residual or recurrent MR in each patient was compared with baseline preoperative measurements using Student’s t test for paired samples. All reported P values are two-sided and those < 0.05 were considered to indicate statistical significance. All statistical analyses of recorded data were performed using the Excel statistical software package (Ekuseru-Toukei 2010; Social Survey Research Information Co., Ltd., Tokyo, Japan).

Operative results are presented in Table 4. Rates of postoperative mortality and morbidity were low. There were no operative deaths (< 30 days or same hospital admission), with the median hospital stay 8 days (IQR 7–11) and ICU stay 2 days (IQR 2–3). A blood transfusion was required during hospitalization in 25 patients (19.4%). New-onset atrial fibrillation occurred in 20 (16%) of the 129 patients admitted for sinus rhythm. Reoperation for bleeding was performed in 2 (1.6%). There were no cerebral vascular accident incidents nor cases of acute renal failure requiring dialysis. 2 patients (1.6%) had pulmonary complications and a permanent pacemaker for complete heart block was required in an additional 2 (1.6%). Following the repair, residual MR at discharge was very low [no MR, n = 73 (56.6%); trivial, n = 54 (41.8%); mild, n = 2 (1.6%); moderate or severe MR, none].

Nine months was used to define the minimum mid-term follow-up period and clinical follow-up was achieved in 129 patients (91.4%), with a mean period of 35 months (range 9–69 months).

Overall, the 1-, 3-, and 5-year survival rates were 99.2 ± 0.01%, 98.1 ± 0.01%, and 88.8 ± 0.05%, respectively (Fig. 3). A total of five patients (3.9%) died during the follow-up period, though the cause of death could only be determined in 3, none cardiac related (pneumonia, 7 months; gastrointestinal bleeding, 25 months; lung cancer, 28 months; unknown, 8 and 18 months).

Following mitral valve repair, rates for freedom from MACE events after 1, 3, and 5 years were 100%, 98.9 ± 0.01%, and 95.6 ± 0.03%, respectively (Fig. 4).

For patients who underwent mitral valve repair, freedom from recurrent MR (grade > 2) after 1, 3, and 5 years was 99.2 ± 0.01%, 94.9 ± 0.03%, 94.9 ± 0.03%, respectively (Fig. 5).

Freedom from reoperation at 1, 3, and 5 years after mitral valve repair was 98.4 ± 0.01%, 98.4 ± 0.01%, and 98.4 ± 0.01%, respectively (Fig. 6). Two patients underwent a reoperation, 1 for a failed repair the next day of the first MIMVr and 1 for hemolytic anemia more than 1 month after the initial procedure, with the mitral valve successfully replaced under a sternotomy in both.