Minimally invasive mitral valve surgery (MIS) refers to a collection of techniques and operation-specific technologies, all of which are designed to lessen surgical trauma and improve clinical outcomes. In the last 20 years, enhanced visualization and instrumentation as well as modified perfusion and aortic occlusion methods have propelled the development of MIS operations. Cohn and Cosgrove (1996) first modified cardiopulmonary bypass techniques to enable safe, effective, minimally invasive aortic and mitral valve surgery.^1–3 Concurrently, innovative port-access methods, using endo-aortic balloon occlusion, were proven to be effective.^4,5 Thereafter, video-assisted and robotic methods were developed and applied effectively by several surgical groups.

Despite acceptance today, many surgeons initially were very critical of performing any complex valve operations through small incisions, owing to safety risks and the possibility of inferior results.^6,7 Since then many institutions worldwide have published excellent MIS comparative outcomes with traditional surgery. These showed definite clinical advantages, which included decreased blood loss and less transfusions as well as minimal postoperative care and pain. Collectively, these translated into shorter hospital stays, faster return to normal activities, less use of rehabilitation resources, and overall healthcare cost savings. Each of these advantages has been a driver for the continued development and expansion of MIS operations.

Today, replacing and repairing cardiac valves through small incisions is a standard practice. The combination of alternative sternal and thoracotomy approaches, new aortic clamping techniques, modified cardiopulmonary bypass circuits and cannulas, shafted-instruments, and three previous generations of robotic systems, all have set the stage for current minimally invasive and robotic mitral surgery, which has gained acceptance at many centers worldwide. Large institutional series as well as several meta-analyses have confirmed that MIS mitral valve surgery is safe and effective for most patients.^8–14 Although overall mitral repair rates in the United States have increased from 51% in 2000 to 62% in 2010, adoption of minimally invasive operations lags behind at less than 30% of repair cases.^15–17 In comparison, in 2015 nearly 50% of over 5000 mitral valve repairs in Germany were done minimally invasively.¹⁸ At our center, mitral repair rates approach 100% for both MIS and robotic operations in selected patients with degenerative mitral disease.

With better understanding of the natural history of degenerative mitral valve disease, more asymptomatic patients now are being referred for repairs. Recent 2014 ACC/AHA (United States) and/or 2012 ESC/EACTS (European) Heart Valve Guidelines suggested that these patients should be referred only to high volume centers, having a heart team approach with a 95% chance of a repair and less than 1% operative mortality.^19,20 These same “high bar” standards now have been set for minimally invasive repairs. To this end, surgeons who plan to embark on this trek should have significant experience in repairing valves through a sternotomy as well as have a significant patient volume, a unified heart team, and full support from the operating room and hospital administration.¹⁶

Hopefully, this chapter will serve both as an educational tool and helpful guide for those who plan to do, or currently are doing, minimally invasive mitral and tricuspid valve surgery. Today, most surgeons have reserved the hemisternotomy for aortic valve surgery. The majority of minimally invasive mitral and tricuspid repairs are done either through a right minithoracotomy or even less invasive port access methods. Thus, this chapter will relate only to direct-vision, video-assisted (endoscopic), and robot-assisted MIS mitral/tricuspid operations performed via right thorax access.

The First Operations

T. Lauder Brunton first suggested that the mitral valve stenosis could be treated surgically.²¹ In May of 1923, Elliott Cutler performed a mitral valvulotomy in an 11-year-old girl using a tenotomy knife. Thereafter, she lived for 4.5 years. His next six patients died, and he abandoned the procedure.²² Sir Henry Souttar performed a successful digital commissurotomy in May of 1925. His patient recovered uneventfully but he was never referred another patient for this operation.²³ Despite these early repair attempts, mitral surgery remained dormant until 1948 to 1949 when Horace Smithy, Charles Bailey Dwight Harken, and Russell Brock independently revitalized the technique of commissurotomy.^24–27 There were early successes by Lillehei and McGoon to repair degenerative mitral valves.^28,29 Other surgeons developed various annuloplasty techniques.^30,31 Some of these operations were quite successful but the development of a mitral valve replacement prosthesis eclipsed further development of repair techniques.^32,33

Traditional Repair Success

The concept of repairing mitral valves was revived and developed by Carpentier and Duran.^34,35 Carpentier’s landmark address entitled The French Correction (1983) made surgeons realize that most insufficient degenerative valves could be repaired and that the benefits exceeded those of a replacement.³⁶ His techniques were proven to yield excellent results when done through a median sternotomy. Many subsequent clinical series confirmed these benefits and repair became a standard of care in Europe and the United States, albeit slowly adopted.^37–40 The perfection of mitral repair operations done through a sternotomy laid the foundation for minimally invasive techniques to be the next wave of advancement.

Minimally Invasive Mitral Valve Surgery

The idea of MIS mitral surgery emanated during the “Heartport Era” with the development of balloon aortic occlusion devices, long-shafted instruments, and port cardiac access.^4,5 These innovations facilitated MIS greatly and encouraged surgeons to expand the development of even less invasive operations. Cohn and Cosgrove among others showed modified less invasive sternal incisions safe and efficacious when replacing and repairing mitral valves.^1–3 Later, propensity matched patient series confirmed that repair results were similar to sternotomy-based operations.¹²

The concept of operating inside the heart using video-assisted or endoscopic vision is not new. In November of 1923, Duff Allen and Evarts Graham planned to use a cardioscope to visualize a commissurotomy in a 31-year-old woman but were unsuccessful.⁴¹ Harken experimented with intracardiac visualization techniques in 1943.⁴² In 1958, Sakakibara predicted that valve operations could be done using endoscopic secondary vision.^43,44 Kaneko (1995) used video assistance through a sternotomy to aid in mitral repairs and commissurotomies.⁴⁵

In February of 1996, Carpentier performed the first videoscopic minimally invasive mitral repair, which was done via a right minithoracotomy.⁴⁶ Our group performed the first camera-directed minimally invasive mitral replacement 2 months later.^47–49 Mohr, Reichenspurner, Vanermen, Hargrove, and Chitwood expanded the influence of video-directed minimally invasive mitral surgery by publishing large mitral repair series with excellent results.^8–10,14

Robot-Assisted Mitral Valve Surgery

In 1998, Carpentier accomplished the first robotic mitral repair using a daVinci robot prototype.⁵⁰ A week later, Mohr completed seven successful robot-assisted mitral repairs with this system.⁵¹ These operations proved that mitral repairs could be done safely using robotic telepresence alone. In 2000, Grossi performed a mitral valve leaflet repair using the Zeus robot.⁵² In 1999, our group purchased the first commercial daVinci Surgical System (Intuitive Surgical, Inc., Sunnyvale, CA) in the United States and then developed additional instruments and repair techniques in our robotic laboratory. In May of 2002, under an FDA safety and efficacy clinical trial, we performed the first complete robotic mitral repair with leaflet resection and a band annuloplasty using the daVinci system.⁵³ This device was FDA approved for intracardiac surgery in 2002 after two clinical trials.⁵⁴ Our inaugural series of 300 patients confirmed the safety and efficacy of robot-assisted mitral valve repair.⁵⁵

Dedicated robotic mitral valve repair referral programs have shown outcomes similar to operations done either through a sternotomy, hemisternotomy, minithoracotomy, or port only access.^56–60 The device, instrument, and maintenance costs have been challenged; however, surgeons at the Mayo and Cleveland Clinics have shown that optimized robotic care paths can render economic parity with other incisional approaches.^61,62

In previous book chapters, we have compared a surgeon’s progression in minimally invasive valve surgery skills akin to a “Mount Everest trek.” Although this analogy may be less apropos today, it does bespeak the advantages of an advancing pathway on which one can accommodate anywhere along this surgical escarpment. Embarking from a conventional median sternotomy-based operation or “base camp,” surgeons can advance progressively toward less invasiveness through experience and methodological acclimatization. In this schema, entry levels of technical complexity are mastered before advancing past small-incision, direct-vision approaches (Level 1) toward more complex video-assisted/directed procedures (Level 2 or 3), and finally to completely robot-assisted operations (Level 4). With the evolution of technology and advancing surgical expertise, most established mitral repair surgeons are able to attain “comfort zones” along this trek.

This chapter shows that patient selection, preoperative screening, patient positioning, the operative setup, perfusion, and operative management are generally the same for all of our MIS and robot-assisted operations. This standardization should help surgeons move more rapidly through the progression of MIS operations. Major differences between direct vision and video-assisted MIS operations relate mainly to the size of the minithoracotomy. In robot-assisted operations, changes relate to instrument arm placement, instruments, and telemanipulation methods. Mitral and tricuspid valve repair techniques have been standardized for all three operative approaches.

Table 40-1 shows what we consider to be an ideal mitral valve operation. Today, we have been able to achieve most of these factors: however, others remain for the future. Clearly, the die has been cast (alea iacta est) regarding the absolute requirement for patient safety and optimal long-lasting repairs, no matter what technique is employed. Patients who are selected for minimally invasive mitral surgery should have the same indications as outlined in either the 2014 ACC/AHA (United States) or 2012 ESC/EACTS (European) Guidelines.^19,20 At our institution, all patients with either degenerative or functional mitral insufficiency are considered for a videoscopic or robotic MIS. Asymptomatic patients are selected based on the IIa recommendations in the recent ACC/AHA guidelines.¹⁹ Patients with significant comorbidities, and those requiring multivessel coronary revascularization, an aortic valve replacement or have a significantly dilated ascending aorta should be operated upon through a sternotomy. Operative risks, age, fragility, and mitral pathology complexity all should be considered when selecting these patients. All patients should be informed of alternative approaches, including a traditional sternotomy, and they should understand that there is a small possibility of requiring a sternal conversion. Absolute and relative contraindications to robotic and MIS mitral surgery are listed in Table 40-2. Some relative contraindications can be managed by selecting alternate methods for perfusion and myocardial protection (eg, axillary artery cannulation and hypothermic ventricular fibrillation). For single vessel coronary disease, preoperative coronary stenting has obviated the need to avoid consideration for a mitral/tricuspid MIS repair.

Candidates for any MIS or robot-assisted mitral operation should be screened carefully for peripheral vascular and coronary artery disease as well as pulmonary maladies. In most patients, either a computed tomographic angiogram and/or contrast coronary angiogram should be performed. Suspect patients should undergo tomographic and/or ultrasound screening for aortoiliac, carotid, and general aortic disease. Peripheral perfusion and aortic endoballoon occlusion should be avoided in patients with severe aortic atherosclerosis or a dilated ascending aorta. Pulmonary function tests should be done in heavy smokers and those having symptoms of obstructive disease. A detailed transthoracic echocardiogram should be done to define general valve pathology, ventricular function, and the presence of pulmonary hypertension. If the latter is present, a right heart catheterization may be indicated. In patients with complex disease, we perform a preoperative 3D transesophageal echocardiogram (TEE).

We always develop an intraoperative “blueprint” to plan mitral and tricuspid repairs Table 40-3. After patients are anesthetized, the direction of each jet (leak) is mapped with both 2D and 3D TEE studies. The mobility and prolapse level is determined for each restricted or prolapsing segment. Leaflet segments (P₁-P₃, A₁-A₃) are measured. Next, the planar angle between the aortic and mitral valve annulus is determined. Finally, the annular diameter, outflow tract septal thickness, and coaptation point to septal (C Sept) distances are measured. Operative planning includes the avoidance of systolic anterior motion of the anterior leaflet (SAM). Thus, we pay special attention to the length of A₂ and P₁-P₃ heights (annulus to coapting edge). Table 40-4 shows the anatomic, operative, and dynamic features that can contribute to postoperative SAM. Preventative structural repair measures that may avoid this complication include: (1) implantation of a large enough ring/band, (2) reduction of the posterior leaflet height to 15 mm or less, and (3) achievement of an optimal leaflet coaptation surface (8–10 mm). The length of A₂ guides us for annuloplasty band size selection. Finally, a 3D valve model is constructed from these measurements (Fig. 40-1 A–C). During the valve reconstruction, this model as well as other imaging studies can be visualized in the daVinci operating console using the Tile Pro software (Intuitive Surgical, Inc., Sunnyvale, CA). To determine the need for an adjunctive tricuspid valve repair, we quantitate the regurgitant volume and measure the annular size. We now include a tricuspid annuloplasty repair in all patients that have even moderate insufficiency with annular dilatation of more than 4 cm.

As mentioned before, the operative organization and management are very similar for minithoracotomy or port access (MIS) video-assisted, direct vision, and robot-assisted mitral repairs (RMVP) and replacements. The Atlas of Robotic Cardiac Surgery details anesthesia management, instrument setup, and conduct of these and other operations at several well-known centers.⁶³ Moreover, Cardiopulmonary Bypass and Mechanical Support: Principles and Practice provides precise details of our method of cardiopulmonary perfusion (CPB) for all of these operations.⁶⁴ The use of endo-balloon aortic occlusion requires special attention and is detailed in these texts.

Patient Position

To help widen the right intercostal spaces (ICSs), which helps to prevent rib spreading, the right chest should be elevated by 30° using a midthorax towel roll. The right arm should be “sling” positioned inferior to the posterior axillary line. Some surgeons prefer to position the flexed right arm above the head using different forms of stabilization. At this time, anterior and posterior Zoll defibrillator pads should be placed to subtend the heart. To monitor adequate limb perfusion during CPB, we place Invos System (Somanetics, Inc., Troy, MI) oxygen saturation monitoring patches on each thigh. Standard skin preparation and draping must provide wide access to the right chest, sternum, and both groins. To facilitate transthoracic aortic clamp placement, the right chest should be prepared sterilely as far as the posterior axillary line.

Anesthesia Management

Although some surgeons prefer not to isolate the right lung, we believe that unilateral ventilation can be done safely and facilitates inspection for bleeding at the end of the operation. In the past, we deflated the lung first and then opened the pericardium with long instruments before establishing CPB. However, we now begin CPB before deflating the right lung and avoid manipulating it while heparinized. To isolate the right lung, either a double-lumen endotracheal tube or a right endobronchial blocker is used. In all of our minimally invasive cardiac surgery operations, we monitor closely cardiac dynamics and function as well as leg and brain perfusion. Forehead patches are applied to monitor bispectral index (BIS, Covidien, Inc., Boulder, CO), which provides an index of cerebral perfusion and anesthetic levels.

A left radial artery catheter is placed for systemic blood pressure monitoring. If intra-aortic balloon occlusion (clamping) is planned, arterial pressures should be measured continuously in both arms. This is done to ensure that the endoballoon has not inadvertently occluded the innominate artery. Thereafter, a right internal jugular vein catheter is inserted for drug infusions and placement of a flow directed Swan–Ganz pulmonary artery catheter (Edwards Lifesciences, Irvine, CA). Using the “double stick” technique, a thin-walled (15-Fr or 17-Fr) Bio-Medicus (Medtronic, Inc., St. Paul, MN) right internal jugular venous drainage cannula is inserted under echocardiographic guidance (Fig. 40-2). Some surgeons use a single femoral venous return cannula that is passed through the right atrium into the superior vena cava (SVC). However, to assure a dry surgical field throughout these operations, we have found that dual cannulation is more reliable. Lastly, the 3D transesophageal echo probe is positioned and detailed studies are done.

Working Incision

Today, most minimally invasive and robotic mitral/tricuspid valve operations are performed through a right fourth ICS minithoracotomy. Current visualization choices include a direct view through the incision, 2D endoscopic, or 3D robotic. However, the new Aesculap 3D EinsteinVision system (E Braun, Inc., Tuttlingen, Germany) has been compared favorably to daVinci robotic 3D visualization.⁶⁵ The size of the working incision usually depends on the visualization and operative methods chosen by the surgeon.

For both direct vision and 2D endoscopic MIS, a 4 to 5 cm minithoracotomy is made in the fourth ICS near the anterior axillary line. We suggest a slightly larger incision for surgeons just beginning MIS mitral surgery. For robotic operations, a smaller working incision (2 to 3 cm) is placed in the same topographic chest region. In preference to using a rib-spreading retractor, we now prefer a flexible Alexis soft tissue wound protector (Applied Medical, Inc., Rancho Santa Margarita, CA). In the past, we used the Perivue Soft Tissue retractor (Edwards Lifesciences, Inc., Irvine, CA) for nonrib spreading access. We have found that this minimizes postoperative pain and provides good working-incision exposure for either direct vision, video-assisted, or robotic minimally invasive mitral/tricuspid valve surgery. Other robotic surgeons have found that true port-access (1 to 2 cm) can be used without compromising either the operation or safety.^57–60

Cannulation and Cardiopulmonary Perfusion

All perfusion cannulas are placed under echocardiographic guidance using the Seldinger guide-wire technique. Through a 2-cm oblique groin incision adventitial 4-0 polypropylene oval (longitudinal) purse-string sutures are placed in both femoral vessels near the inguinal ligament. The right femoral artery is cannulated with either a 17 or 19-Fr Bio-Medicus cannula (Medtronic, Inc., Minneapolis, MN). For inferior vena caval drainage, either a 22-Fr (single stage) or a 23/25-Fr (dual stage) RAP femoral venous cannula (LivaNova, Inc., Arvada, CO) is passed into the right atrium (Fig. 40-3). In corpulent patients, we tunnel cannulas through the upper thigh subcutaneous tissue. This allows coaxial dilators and cannulas to have safer passage into vessels at a 30° to 45° angle. Vacuum-assisted venous drainage is used in all of our operations. Presently, we use Sorin S5^® heart-lung machine, modified for use with a venous bag reservoir or V-Bag (Circulatory Technology, Inc., Oyster Bay, NY) and equipped with magnetic drive for Revolution centrifugal pump suction (LivaNova USA, Inc., Arvada, CO) (Fig. 40-4). Figure 40-5 Illustrates our current perfusion circuit for MIS and robotic heart surgery.

Cardiac Air Removal

Meticulous cardiac air removal is particularly important in minimally invasive valve operations. Difficulty exists in manipulating and deairing the cardiac apex, as it cannot be elevated. Also, with a right anterolateral minithoracotomy, air tends to be retained along the dorsally oriented ventricular septum and in the right pulmonary veins. Moreover, this position places the right coronary ostium in the most vulnerable position to entrap air. Continuous carbon-dioxide (CO₂) insufflation has been particularly helpful in minimizing intracardiac air and should be begun before cardiac chambers are opened. Carbon dioxide is much more soluble in blood than air and displaces it very effectively. We infuse CO₂ continuously (4-5 L/min) into the thorax, and prior to cross clamp release, ventilate both lungs vigorously to draw the gas deep into all pulmonary veins. After atriotomy closure and following cross-clamp release, suction is applied to the aortic root vent. One can compress the right coronary artery origin during early ejection. Constant transesophageal echocardiographic monitoring is essential to assure adequate air removal before weaning from cardiopulmonary bypass.

Alternate Arterial Cannulation Techniques

Ileo-femoral and/or aortic atherosclerosis may preclude safe retrograde perfusion. To cannulate the ascending aorta directly, it is important to place two concentric pledgeted purse strings near the innominate artery origin. We use either a Biomedicus (Medtronic, Inc., Minneapolis, MN) guide-wire directed cannula or a 23-Fr Straight Shot device (Edwards Lifesciences, Irvine, CA), which is passed through the chest wall via a 10-mm trocar. If it is not feasible to cannulate the ascending aorta directly, we use the right axillary artery for antegrade perfusion. The artery is exposed through an infra-clavicular incision. An 8-mm GelSoft knitted graft (Vascutek, Terumo, Ann Arbor, MI) is sewn end-to-side to the axillary artery with 5-0 polypropylene suture. Thereafter, the graft is connected to the bypass circuit using either an appropriate size arterial cannula or a three-eighth inch pump tubing connector. Terumo also makes a PTFE graft that is annealed directly to a perfusion cannula (Fig. 40-6 A and B). Alternatively, the axillary artery can be cannulated directly; however, distal arm perfusion should be monitored while the cannula is in place.

Aortic Occlusion

For aortic occlusion, we use a transthoracic cross clamp Scanlan International, Inc., St. Paul, MN as it has been proven to be safe, reliable, economic, and simple to apply (Fig. 40-7A and B). However, a number of surgeons use the IntraClude endoballoon (Edwards Lifesciences, Irvine, CA) for aortic occlusion (Fig. 40-8). This technique has a steeper learning curve than using the clamp. The balloon position must be precise and remain stable in the ascending aorta. There is a potential for either innominate artery occlusion or intraventricular displacement. Therefore, it is essential echo monitor the balloon position throughout the operation. Also, introduction through the specialized femoral arterial cannula can limit limb perfusion. In this circumstance, the endoballoon catheter should be reinserted through the other femoral artery. Despite these concerns, this method can provide effective aortic occlusion with a simultaneous route for delivering antegrade cardioplegia and venting air.