Median sternotomy has been a standard method of approach for open heart surgeries for many decades since it provides a definitive view of the surgical field.1 However, as the patients' social life improves, and as the competitions among medical institutions become intensive, lesser postoperative pain, shorter hospital stay, and better cosmetic results are being considered, and many minimally invasive surgical methods are being developed to satisfy the patients' variable demands.2-4 The less invasive surgical procedures have been growing in popularity in the past decade, and incisions have become increasingly smaller as endoscopic techniques have been developed.

The purpose of this study is to report our overall experience with minimally invasive mitral valve surgery using parasternal or lower-sternotomy approach, and compare the results of the two methods.

The summary of operative techniques, and patients' intraoperative data are shown in Table 5 and 6. For all patients, the mean length (cm) of skin incision was 10.3 ± 1.3 (range, 8 to 13). Mean operation time, cardiopulmonary bypass time, and aortic cross clamp time were 267.9 ± 57.3 (range, 107 to 425) minutes, 107.6 ± 26.1 (range, 60 to 212) minutes, and 73.1 ± 28.7 (range, 40 to 260) minutes, respectively. A mean amount of postoperative mediastinal bleeding on the first postoperative day was 438.6 ± 371.9 (range, 80 to 2080) mL. Mean duration of mechanical ventilator care, ICU stay, and hospital stay were 12.3 ± 4.9 (range, 4 to 28) hours, 35.3 ± 17.4 (range 17 to 96) hours, and 10.1 ± 6.6 (range 5 to 60 ) days. Postoperative NYHA functional class improved to 1.1 ± 0.4 (range 1 to 3). For those patients who had mitral valve repair, the grade of mitral regurgitation significantly decreased from 3.7 ± 0.46, preoperatively, to 0.6 ± 1.22, postoperatively. Eight patients (9%) required homologous blood transfusion, and mean amount of blood transfused were 2.5 ± 1.07 (range, 2 to 5) units. There was no hospital death, and early complications (5.8%) included superficial wound infection (2), re-exploration for bleeding (2), and prolonged ventilation due to pneumonia (1). There were no peripheral vascular complications related to femoral arterial cannulation. Mean follow up duration was 55.0 ± 25.7 (range 5 to 87) months, and follow up rate was 92%. Late complications occurred in 3 patients (3.5%). One patient had cerebral infarction that later died of infarct-related cerebral hemorrhage, and 2 patients who have had mitral valve repair had reoperations 51, and 63 months later because of significant residual mitral regurgitations. Two patients (2.3%) died of cerebral infarction, and multiple myeloma, respectively during follow up period.

The comparative results of the two different surgical approaches are shown in Table 6. Preoperative patients' profiles were not significantly different between two groups (Table 1). Mean wound length (9.21 ± 1.10 cm vs. 11.24 ± 0.82 cm, p < 0.05) was significantly shorter and mean operation time (294.74 ± 59.41 min vs 259.36 ±54.36 min, p = 0.03) was significantly longer in parasternal group. Mean cardiopulmonary bypass time (113.38 ± 30.39 min vs 105.78 ± 24.6 min, p = 0.31), and mean aortic cross clamp time (81.33 ± 28.69 min vs 70.42 ± 28.48, p = 0.14) were also longer in parasternal group, but not statistically different between two groups. Mean mechanical ventilation time (10.42 ± 4.36 hours. vs 12.9 ± 5.0 hours, p = 0.04) was significantly shorter, and mean amount of chest tube drainage was less in parasternal group with no significant difference. Mean follow duration was significantly longer in parasternal group, since we performed parasternal approach in the earlier period of this study.

Although cardiac surgery has been performed through a median sternotomy, traditionally, the left-sided approach had existed for cardiac procedures since 1910, when Alexis Carrel5 proposed it for a coronary artery bypass from the thoracic aorta, and in the 1960s, mitral valvuloplasties were commonly approached from the left side.6,7 In 1990s, the minimally invasive coronary bypass surgery began, stimulated by the increasing use of minimally invasive surgical techniques for all branches of surgery, the last being cardiac surgery.8 And in 1995, several groups began to explore the use of parasternal and ministernotomy incisions for aortic and mitral valve replacement and repair.2,4,9 To date, surgeons at the Cleveland Clinic have performed more than 2500 minimally invasive valve operations by combining direct vision with an upper hemisternotomy and modified perfusion techniques with less than a 1.0% mortality, reduced transfusions (< 10%), and earlier hospital discharge.1

The advantages of minimally invasive surgery are reducing surgical trauma, increasing patient's cosmetic satisfaction, less homologous blood transfusion, shorter hospital stay and cost, and faster recovery and returned to work.2-4,8,10,11 With these potential advantages, we started minimally invasive mitral valve operations in 1996, starting with right paramedian, or mini-thoracotomy incisions, and since 2000, we used lower sternotomy incisions, more often.

In our experience, right paramedian or thoracotomy incision brought about some difficulties in exposing the structures around the aorta, placing an arterial cannula directly into the ascending aorta, and cross-clamping the aorta. Thus most of the cases, another small incision on right or left inguinal area were necessary, and more patients complained of postoperative wound pain than the patients with lower sternotomy. Besides, this approach has more demerits including the difficulties to convert to midsternotomy, the possibility to injure the right internal mammary artery, and the possible development of asymmetric breasts.12 However, for female patients, the satisfaction for cosmetic result was superior, and most of the wound pains were well controlled with oral analgesics. Furthermore, the results of this study revealed that this patient group had more advantages including shorter wound length, and mechanical ventilator support required. On the other hands, lower ministernotomy provided a better view especially around the aorta, and in most of the cases, direct ascending aortic cannulations were feasible, and shorter operation time required. Since the two different methods of approaches have merits and demerits, now we provide the selected patients the informations about both methods, and decide which approach to perform, according to the patients' preference.

The overall results of this study showed low perioperative morbidity and mortality, and the late results also showed very low mortality, and reoperation rate. However, the mean hospital stay was slightly longer than that of other institutions, because of 3 patients operated for infective endocarditis. These patients had to stay longer for prolonged use of intravenous antibiotics, postoperatively (4 to 6 weeks).

Although minimally invasive techniques have been rapidly evolved, there are some elements limiting the extensive use of these techniques; 1) difficulties of recognizing the structures owing to the limited surgical view. 2) in patients with peripheral atherosclerotic disease with no access for femoral arterial or venous cannulations, 3) difficulties to decompress the heart during cardiopulmonary bypass, 4) difficulties to apply defibrillators, 5) difficulties for deairing, 6) limitation of surgical field for multiple valve operations, and 7) in patients with previous open heart surgery.13 Thus careful patients selection may be mandatory.