Quick links
Quick links
J Chest Surg 2025; 58(1): 21-30
Published online January 5, 2025 https://doi.org/10.5090/jcs.24.080
Copyright © Journal of Chest Surgery.
De Qing Görtzen, B.Sc., Fleur Sampon , M.D., Naomi Timmermans
, M.D., Joost Ter Woorst
, M.D., Ph.D., Ferdi Akca
, M.D., Ph.D.
Department of Cardiothoracic Surgery, Catharina Hospital, Eindhoven, The Netherlands
Correspondence to:Ferdi Akca
Tel 31-40-239-8680
E-mail ferdi.akca@catharinaziekenhuis.nl
ORCID
https://orcid.org/0000-0002-1748-3235
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Commentary: J Chest Surg. 2025;58(1):31-33 https://doi.org/10.5090/jcs.24.110
Background: This study presents an overview of our technique and the perioperative outcomes for the first 100 patients who underwent minimally invasive endoscopic-assisted off-pump multivessel bypass grafting (endoscopic coronary artery bypass [endo-CAB]) at the Catharina Hospital in Eindhoven.
Methods: The first 100 patients undergoing multivessel endo-CAB from May 2022 to March 2024 were included in this retrospective, single-center, observational study (N=100). The study encompassed both elective and urgent surgical revascularization. In all cases, endoscopic-assisted harvesting of the internal mammary artery, radial artery, or saphenous vein was performed, followed by beating-heart anastomoses through a mini-thoracotomy.
Results: A total of 226 distal anastomoses were performed, utilizing 102 left internal mammary arteries, 80 radial arteries, 30 right internal mammary arteries, and 14 saphenous veins. On average, each patient had 2.3 anastomoses. A Y graft configuration was employed in 78 patients, in-situ bilateral internal mammary artery inflow in 19 patients, and a proximal aortic graft in 3 patients. Four patients underwent concurrent arrhythmia surgery. Eleven patients received hybrid revascularization. There was 1 conversion to sternotomy (1%) and 3 instances where cardiopulmonary bypass was required (3%). The median operation time was 3.3 hours (interquartile range, 3.0–3.7 hours), and the median hospital stay was 4.0 days (interquartile range, 3–4 days). The in-hospital mortality rate was 1%.
Conclusion: Multivessel off-pump endo-CAB surgery can be safely performed with endoscopic-assisted conduit harvesting. Combining the benefits of a minimally invasive and anaortic approach may improve perioperative outcomes for patients requiring surgical revascularization. Further studies are necessary to establish the role of this technique in routine coronary surgery.
Keywords: Endoscopic coronary artery bypass grafting, Off-pump coronary bypass grafting, Minimally invasive cardiac surgery, Minimally invasive direct coronary artery bypass
The standard of care for managing advanced coronary artery disease is coronary artery bypass grafting (CABG). As the global burden of cardiovascular disease continues to rise, CABG remains the most commonly performed procedure in cardiac surgery [1]. CABG is conventionally performed through a full sternotomy on a non-beating heart, utilizing cardiopulmonary bypass (CPB). However, off-pump coronary artery bypass grafting (OPCAB) has become increasingly popular due to its lower risk of cerebrovascular accidents. This benefit arises because the technique avoids manipulation of the aorta, a method referred to as the “anaortic approach” [2-4].
Minimally invasive coronary surgery is a rapidly evolving and technically demanding field, especially for multivessel revascularization. Using minimally invasive techniques, a median sternotomy is avoided in favor of a mini- thoracotomy or total endoscopic techniques [5]. Various methods have been developed for both conduit harvesting—including direct vision, thoracoscopic-assisted, or robotic-assisted mammary artery harvest—and graft deployment, which can be performed off-pump, on a beating heart, or on an arrested heart [6-12]. The ideal revascularization strategy would avoid sternotomy entirely and achieve complete revascularization using an anaortic approach, thus providing a truly minimally invasive long-term solution for the patient.
We aimed to share our experience with minimally invasive, multivessel, endoscopic-assisted, off-pump coronary surgery (endoscopic coronary artery bypass [endo-CAB]) and provide a comprehensive overview of the perioperative data and outcomes for the first 100 patients undergoing this procedure at the Catharina Hospital in Eindhoven.
The institutional review board and the ethical committee of Medical Research Ethics Committees United approved the study and waived the requirement for informed consent (registration no., W23.029).
From May 2022 to March 2024, 100 patients undergoing multivessel endo-CAB were included in this retrospective, single-center, observational study at Catharina Hospital in Eindhoven. All procedures were performed by a dedicated minimally invasive off-pump coronary surgeon. Patient baseline characteristics were obtained from the Department of Cardiothoracic Surgery’s database. Our aim was to provide a comprehensive overview of our initial experience and procedural parameters, focusing on the first 100 patients treated with multivessel endo-CAB. The patients had either double (n=87) or triple (n=13) vessel coronary artery disease. For 11 patients, a hybrid revascularization strategy was determined by the heart team, which included a percutaneous intervention alongside the multivessel endo-CAB procedure.
All patients were evaluated by a heart team, which included a cardiac surgeon and an interventional cardiologist, to assess the need for surgical revascularization. Patients diagnosed with multivessel disease and deemed suitable for either elective or urgent surgical revascularization were considered eligible for endo-CAB. Initially, during the implementation phase of endo-CAB, patients with a lower body mass index (BMI) and double vessel disease were specifically chosen for this procedure. Morbid obesity and a salvage situation involving hemodynamic instability were established as definitive exclusion criteria.
All patients underwent the standard preoperative workup required for any coronary surgery procedure, which included coronary angiography, chest radiography, electrocardiogram, transthoracic echocardiogram, bilateral blood pressure measurements, and blood panel laboratory tests. Following the preoperative workup, we determined whether a patient was eligible for standalone multiple-vessel endo- CAB, hybrid revascularization, or concomitant arrhythmia surgery. Routine preoperative pulmonary function tests or computed tomography scans were not included in our preoperative workup.
The procedure is performed under general anesthesia with the patient in a supine position. To improve surgical access, a pillow is placed under the left scapula, elevating the left hemithorax. Standard arterial and central venous lines, along with an endotracheal tube, are positioned before the procedure begins. A selective endobronchial blocker (Teleflex, Wayne, PA, USA) is inserted through the endotracheal tube to enable unilateral ventilation during the grafting process. Additionally, the groin is marked and draped in preparation for potential femoral cannulation if hemodynamic instability occurs. If the radial artery is planned to be harvested endoscopically at the same time, the right arm is positioned at a 90° angle from the thorax. Furthermore, transesophageal echocardiography is utilized to evaluate left ventricular and valve function.
Three small incisions are made on the left side of the chest: 1 at the third intercostal space at the anterior axillary line, and 2 more at the second and fourth intercostal spaces, positioned more anteriorly. Through these incisions, 5-mm endoscopic ports are inserted, and capnothorax is established at a pressure of 8 to 12 mm Hg (Fig. 1A). A 5-mm, 0°, 2-dimensional camera (Karl Storz GmbH, Tuttlingen, Germany) is then introduced. Using the other ports, standard long-shafted video-assisted thoracic surgery instruments and a Ligasure Maryland device (Medtronic, Dublin, Ireland) are also introduced. This setup enables harvesting of both the left internal mammary artery (LIMA) and the right internal mammary artery (RIMA).
Before harvesting the mammary artery, the pericardium is initially opened anterior to the left phrenic nerve to identify and assess coronary targets. A clip is placed near the grafting site on the left anterior descending artery (LAD) to facilitate easy identification through the small thoracotomy. When harvesting the bilateral mammary arteries (BIMAs), both the anterior mediastinum and the right pleural space are opened. The RIMA is harvested first, following the opening of the endothoracic fascia. Utilizing long-shafted video-assisted thoracic surgery instruments and the Ligasure device, the RIMA is harvested along with its adjacent veins in a semi-skeletonized fashion. The RIMA can be harvested proximally beyond the subclavian vein and distally up to its bifurcation. Chest wall deformities, such as pectus excavatum, can complicate harvesting the RIMA from the left side. A right thoracoscopic approach or opting for a radial artery can mitigate this issue. Once the RIMA is harvested to the desired length, the LIMA is then harvested.
Thoracoscopically, the phrenic nerve can be easily identified. The LIMA is harvested from the proximal area up to the level of the phrenic nerve and extends distally to its bifurcation. Any side branches of the LIMA are ligated using the Ligasure device (Fig. 1C). Once the LIMA has been fully dissected, heparin is administered intravenously to achieve an activated clotting time of over 300 seconds. The distal portion of the LIMA is secured with titanium clips (Endo-Clip; Medtronic). The LIMA is then sutured to the pericardium using a 5-0 polypropylene stitch. After dividing the LIMA, the location for the mini-thoracotomy is marked. Intrathoracic pressure is relieved to accurately determine the location of the coronary target without the influence of artificial capnothorax pressure. A transthoracic needle is used to puncture the intercostal space above the primary coronary target, ensuring an optimal placement for the mini-thoracotomy (Fig. 1B). The precise location may vary depending on the patient’s specific anatomy.
When a radial artery is used as a second arterial conduit, the LIMA and radial artery are harvested simultaneously to achieve complete arterial revascularization. Endoscopic radial artery harvesting was performed following the methods described by Reyes et al. [13] and Connolly et al. [14], utilizing an endoscopic vessel harvesting system (Vasoview Hemopro; Getinge, Gothenburg, Sweden). Similarly, endoscopic techniques were employed for saphenous vein harvesting.
Once all conduits have been harvested and the mini-thoracotomy site has been marked, the endoscopic ports are removed, and unilateral ventilation is initiated. The 5–6 cm incision is made at the previously marked location, between the third and fifth intercostal spaces. For grafting the LAD and right descending posterior, we generally adopt a more medial approach. If grafting of lateral branches is required, a more lateral access must be established. In female patients, the incision is preferably made in the inframammary area for aesthetic reasons. When the mini-thoracotomy is correctly positioned, the LIMA will be attached to the pericardium directly beneath the incision. Fig. 2A illustrates the LIMA as visible through the thoracotomy.
During multivessel procedures, various graft configurations may be employed. These include both BIMA in situ grafts and Y- or T-graft configurations using a free-RIMA or radial artery. Fig. 2A–C illustrate a T-graft comprising the LIMA and radial artery performed through a mini- thoracotomy. If a saphenous vein is utilized, the proximal anastomosis is created on the ascending aorta using techniques previously described by Issa and Ruel [10] and McGinn et al. [11]. Traction sutures are applied to the right lateral pericardium, and a sponge is positioned adjacent to the aorta. The Octopus Nuvo (Octopus Nuvo Tissue Stabiliser; Medtronic) is employed to depress the pulmonary artery, increasing maneuverability.
For cardiac positioning, we use the epicardial stabilizer and the heart positioning device (Starfish Evo; Medtronic), as previously described by Davierwala et al. [9], Issa and Ruel [10], and Kikuchi and Endo [15]. Depending on the coronary targets, we either use both devices or only the epicardial stabilizer. The heart is positioned with wet sponges, and the stabilizers are placed as shown in Fig. 2D. For the LAD and diagonal branches, only the epicardial stabilizer is used, as depicted in Fig. 3A. For the anterolateral, lateral, inferolateral, and inferior anastomoses, the armless Starfish Evo is also utilized, as illustrated in Fig. 3B and C. The armless Starfish Evo is employed to elevate the apex, facilitating access to the coronary targets, and a silk 1-0 suture is tied around the Starfish to assist in positioning. For the inferior wall, we elevate the apex of the heart with wet sponges and position the armless Starfish towards the apex. We often loosen our pericardial sutures, as they may limit visibility and maneuverability of the heart. We pull the Starfish cranially to stabilize the position for exposing the right posterior descending coronary artery. Subsequently, we position the epicardial stabilizer to secure a stable field for the coronary anastomosis. Once the heart is positioned, we ensure cardiac stability before continuing to manipulate the coronary artery. In cases of hemodynamic instability, we first elevate the legs to increase venous inflow and reposition the heart to stabilize it. If hemodynamic instability or ventricular arrhythmias persist, femoral CPB support is initiated. When using CPB, anastomoses are performed on the beating heart using similar techniques.
Once all coronary targets had been grafted, protamine was administered. The flow through the graft was measured using transit-time flow measurement (TTFM) to ascertain graft function (Medistim Vascular TTFM; Medistim, Oslo, Norway). After confirming that the graft function was sufficient, a 28F chest tube was inserted into the left pleural cavity through 1 of the 3 incisions previously made for the LIMA harvest. Ropivacaine was administered at 3 intercostal levels, and the skin was then closed (Fig. 3D).
Patients were considered for concomitant arrhythmia surgery if they presented with symptomatic paroxysmal or persistent atrial fibrillation (AF) and had a left atrial volume index of less than 70 mL/m2. A detailed account of our criteria for patient selection and the techniques used in concomitant thoracoscopic arrhythmia surgery has been previously published [16]. For patients with a high BMI, the primary surgical objective was to achieve minimally invasive coronary revascularization. If limited intrathoracic space increased the risk of complications during concomitant pulmonary vein isolation, only the left atrial appendage (LAA) clip (AtriClip; Atricure, Mason, OH, USA) would be applied to mitigate the risk of stroke.
All patients were transferred to the intensive care unit (ICU) after the procedure. Patients with radial artery grafts were administered calcium antagonists postoperatively to prevent graft spasms. Paracetamol, oral opioids, and intravenous nonsteroidal anti-inflammatory drugs were utilized for postoperative pain management. Following discharge, patients attended a follow-up appointment at the outpatient clinic 14 days later, where they underwent chest radiography and laboratory blood tests.
The data was analyzed using JASP ver. 0.18.1.0 (Universiteit van Amsterdam, Amsterdam, Netherlands) to obtain descriptives, boxplots, and Q-Q plots, identifying outliers for each continuous variable. Categorical variables are presented as frequency and percentage, while continuous variables are shown as mean and standard deviation if they are normally distributed, and as median and interquartile range (IQR) if they are not, based on the Q-Q plot analysis. RStudio ver. 4.4.0 (R Core Team, Boston, MA, USA) was utilized to visualize data and compute the Pearson correlation coefficient.
The patient demographics, procedural parameters, and perioperative characteristics are displayed in Table 1. Most of the patients were men (90.0%) and the mean age was 65.5±9.1 years. The mean BMI was 26.7±3.3 kg/m2, and the median body surface area was 2.00 m2 (IQR, 1.91–2.10 m2). Seventeen percent of the patients (17.0%) had diabetes, 7 patients had peripheral artery disease (7%), and 5 patients had AF (5%). The median Euroscore II for the population was 1.00 (IQR, 0.80–1.61).
Table 1. Demographics and perioperative characteristics of the endoscopic coronary artery bypass group (N=100)
Characteristic | Value |
---|---|
Sex (men) | 90 (90.0) |
Age (yr) | 65.5±9.1 |
Height (cm) | 175 (171–180) |
Weight (kg) | 82.3±12.8 |
Body mass index (kg/m2) | 26.7±3.3 |
Body surface area (m2) | 2.00 (1.91–2.10) |
Diabetes | 17 (17.0) |
Peripheral vascular disease | 7 (7.0) |
Atrial fibrillation | 5 (5.0) |
Previous cardiac surgery | 0 |
Recent myocardial infarction | 28 (28.0) |
Prior percutaneous coronary intervention | 33 (33.0) |
Left ventricular ejection fraction | |
Good (<50%) | 78 (78.0) |
Moderate (31%–50%) | 18 (18.0) |
Poor (21%–30%) | 3 (3.0) |
Very poor (<20%) | 1 (1.0) |
Chronic obstructive pulmonary disease | 4 (4.0) |
Pulmonary hypertension | 0 |
Cerebrovascular accident | 6 (6.0) |
Renal impairment | 5 (5.0) |
Dialysis | 0 |
Level of urgency | |
Elective | 55 (55.0) |
Urgent | 44 (44.0) |
Emergency | 1 (1.0) |
Euroscore II | 1.00 (0.80–1.61) |
Preoperative hemoglobin (mmol/L) | 9.1 (8.3–9.6) |
Double vessel disease | 87 (87.0) |
Triple vessel disease | 13 (13.0) |
Concomitant arrhythmia surgery | 4 (4.0) |
Isolated LAA clip | 3 (3.0) |
Thoracoscopic AF ablation including LAA clip | 1 (1.0) |
No. of distal anastomoses | |
2 | 75 (75.0) |
3 | 24 (24.0) |
4 | 1 (1.0) |
No. of anastomoses with left IMA | |
1 | 96 (96.0) |
2 | 3 (3.0) |
No. of anastomoses with right IMA | |
1 | 26 (26.0) |
2 | 2 (2.0) |
No. of anastomoses with radial artery | |
1 | 42 (42.0) |
2 | 19 (19.0) |
No. of anastomoses with SVG | |
1 | 6 (6.0) |
2 | 4 (4.0) |
Y-graft configuration | 78 (78.0) |
Bilateral internal mammary artery in situ | 19 (19.0) |
Proximal aorta | 3 (3.0) |
Hybrid revascularization | 11 (11.0) |
PCI, circumflex artery | 5 (5.0) |
PCI, right coronary artery | 6 (6.0) |
Conversion to sternotomy | 1 (1.0) |
CPB-assisted | 3 (3.0) |
Operation time (hr) | 3.3 (3.0–3.7) |
Values are presented as frequency (%) for categorical variables; for continuous variables, they are shown as mean±standard deviation if normally distributed, or as median (interquartile range) if non-normally distributed.
LAA, left atrial appendage; AF, atrial fibrillation; IMA, internal mammary artery; SVG, saphenous vein graft; PCI, percutaneous coronary intervention; CPB cardiopulmonary bypass.
Four patients underwent concomitant arrhythmia surgery. Three underwent isolated LAA clip placement and 1 patient underwent thoracoscopic AF ablation, including an LAA clip. The patients received an average of 2.3 distal anastomoses. The total number of grafts per conduit is plotted in Fig. 4A. The first choice of conduit was the LIMA with 102 anastomoses, while the radial artery was used for 80 anastomoses, the RIMA for 30 anastomoses, and the saphenous vein for 14 anastomoses. A Y-graft configuration was used in 78 patients (78%), the BIMAs in situ were utilized in 19 patients (19%), and a proximal aorta graft was placed in 3 cases (3%) (Fig. 4B). Coronary anastomoses were performed on the anterior (n=100), anterolateral (n=42), lateral (n=65), inferolateral (n=5), and inferior wall (n=14) (Fig. 4C, Table 2). In our population, 11 patients underwent hybrid revascularization (11%). In 1 patient, a conversion to sternotomy was performed due to limited visibility (1%), and 3 cases needed CPB support (3%). In 1 patient, this was planned preoperatively, while in 2 patients, CPB was used due to hemodynamic instability. The median operation time was 3.3 hours (IQR, 3.0–3.7 hours). The Pearson correlation coefficient was calculated to illustrate a possible learning curve in Fig. 5. The Pearson correlation was calculated to be r(98)=-0.109 (p=0.279).
Table 2. Graft locations in the endoscopic coronary artery bypass group
Location | No. |
---|---|
Anterior | 100 |
Anterolateral | 42 |
Lateral | 65 |
Inferolateral | 5 |
Inferior | 14 |
Table 3 describes the postoperative characteristics and complications. The total ventilation time was 7.6 hours (IQR, 6.6–9.3 hours), and the total postoperative blood loss was 350 mL (IQR, 200–550 mL). The median stay in the ICU was 1.0 days (IQR, 0.5–1 days), and the total postoperative hospital stay was 4 days (IQR, 3–4 days). In-hospital mortality occurred in 1 patient (1%) due to shock caused by a postoperative stress response due to previously undiagnosed cancer metastases. One patient required postoperative re-exploration for bleeding (1%), and 1 patient had postoperative ischemia requiring revision of the bypass graft (1%). One patient had a pulmonary embolism that required oral anticoagulation. There was no occurrence of stroke. During hospital admission, 7 patients developed new-onset AF that required oral anticoagulation (7%).
Table 3. Postoperative characteristics of the endoscopic coronary artery bypass group (N=100)
Variable | Value |
---|---|
Total ventilation time (hr) | 7.6 (6.6–9.3) |
Postoperative blood loss (mL) | 350 (200–550) |
Intensive care unit admission time (day) | 1.0 (0.5–1) |
Total hospital stay (day) | 4 (3–4) |
In hospital mortality | 1 (1.0) |
Intra-aortic balloon pump | 0 |
Extracorporeal membrane oxygenation | 0 |
CK-MB max (ng/mL) | 25.5 (19.0–35.3) |
Discharge hemoglobin (mmol/L) | 7.7 (6.8–8.3) |
Transfusion | 6 (6.0) |
Re-exploration for bleeding | 1 (1.0) |
Ischemia postoperative | 1 (1.0) |
Graft revision | 1 (1.0) |
Acute percutaneous coronary intervention | 0 |
Cerebrovascular accident | 0 |
Tamponade during initial admission | 0 |
Tamponade after discharge | 1 (1.0) |
Pulmonary embolism | 1 (1.0) |
Wound infection | 0 |
New-onset atrial fibrillation | 7 (7.0) |
Hospital-acquired pneumonia | 3 (3.0) |
Thoracentesis | 3 (3.0) |
30-day readmission | 2 (2.0) |
Target vessel reintervention | 3 (3.0) |
Values are presented as number (%) for categorical variables and median (interquartile range) for continuous variables if non-normally distributed.
CK-MB, creatine kinase-myoglobin binding.
During follow-up, 1 patient developed late tamponade (1%) that required pericardiocentesis and 3 patients had left-sided pleural effusion that required thoracentesis (3%). Two patients were readmitted within 30 days after discharge for symptomatic postoperative anemia (n=1) and dyspnea due to pleural effusion requiring thoracentesis (n=1). Three patients received unexpected target vessel reintervention (3%) during follow-up.
In this study, we described the outcomes of our first 100 endoscopic-assisted, off-pump, multivessel coronary bypass grafting procedures. All patients underwent endoscopic-assisted graft harvesting and received coronary bypasses via a mini-thoracotomy. On average, each patient had 2.3 distal anastomoses. Eleven patients underwent hybrid revascularization. There was 1 conversion to sternotomy and 3 instances where CPB was required. The median operation time was 3.3 hours, and the in-hospital mortality rate was 1%. Our findings indicate that multivessel off-pump endo-CAB can be performed safely with favorable postoperative outcomes.
Multivessel endo-CAB is a technically demanding procedure with a steep learning curve. It utilizes conventional off-pump coronary techniques. This approach differs technically from bypass grafting performed on an arrested heart [17]. Patients with advanced atherosclerosis of the aorta, chronic lung disease, or kidney problems may benefit from this approach. Other advantages include reduced risks of postoperative inflammation, infection, and arrhythmia [18]. Safely performing OPCAB requires specific surgical training and dexterity to achieve complete familiarity [19,20]. As Shaefi et al. [2] previously described, maintaining hemodynamic stability and minimizing coronary ischemia are critical considerations. We believe that to safely perform multivessel endo-CAB as described above, a surgeon must be comfortable with sternotomy off-pump techniques, ensuring adequate exposure and stabilization for the precise creation of anastomoses on a beating heart.
We further discuss the challenges associated with the thoracoscopic harvest of the mammary artery. Our group has previously detailed an extensive examination of the learning curve associated with thoracoscopic harvesting. In our initial 80 cases, we observed a significant reduction in LIMA harvesting time, from an average of over 100 minutes to approximately 60 minutes [7,21]. Vassiliades was the first to demonstrate the feasibility of thoracoscopic BIMA harvest with favorable outcomes [22]. Starting the procedure with endoscopic assistance offers several advantages. First, it allows for the opening of the pericardium to identify coronary targets and potentially alter the graft configuration. Second, it facilitates the precise identification of the optimal location for the mini-thoracotomy. Finally, it enables the performance of concurrent thoracoscopic arrhythmia surgery, thus maintaining the principles of minimally invasive surgery while effectively addressing multiple comorbidities [22].
Our second most frequently used conduit is the radial artery, which can be harvested through minimally invasive techniques [14]. Although the RIMA is an excellent graft conduit option, it is less often utilized due to the slightly more challenging harvesting process from the left side of the thorax, leading us to prefer the endoscopic radial artery approach [23]. In this study, we resorted to venous grafts only when arterial conduits were not available, or when technical difficulties or patient comorbidity were present. In almost all cases, we preferred the Y configuration or BIMA in situ as the graft configuration, to provide an all- arterial, anaortic approach. The benefits of this approach have been described by Puskas et al. [3].
This procedure involves both surgical and anesthesiological learning curves. Close collaboration with the anesthesiologist is crucial during heart manipulation, similar to what is required in sternotomy off-pump revascularization. Factors such as table positioning, the administration of inotropes or vasopressors, and the sequencing of grafts are critical for ensuring the safety of the procedure. Occasionally, the heart may need to be positioned slightly differently to maintain stable hemodynamics during the coronary anastomosis, which necessitates effective communication with the anesthesia team.
Ideally, the endo-CAB procedure is performed off-pump and without manipulating the aorta, termed “anaortic.” If the surgeon lacks experience with off-pump beating heart surgery, they can decide preoperatively to use a pump. Therefore, pump-assisted endo-CAB may be an attractive option for those with limited experience, as it allows the surgeon to become acquainted with the minimally invasive aspects of the procedure.
The study is a single-center retrospective analysis. Additionally, our follow-up period was brief, and the angiographic evidence is confined to cases of hybrid revascularization and target vessel reinterventions. Nevertheless, we can tentatively conclude that for selected patients, multivessel endo-CAB offers favorable perioperative outcomes, with target vessel reintervention rates comparable to those of sternotomy techniques.
Multivessel off-pump endo-CAB can be safely performed using thoracoscopic IMA harvesting techniques. We believe that combining the benefits of minimally invasive surgery with an anaortic approach could enhance perioperative outcomes for patients undergoing coronary surgery. Mastery of conventional off-pump techniques is crucial to successfully perform this procedure. In selected patients, multivessel endo-CAB yields favorable procedural outcomes.
Authors contributions
Conceptualization: DQG, FA. Data curation: DQG, FA. Formal analysis: DQG, FA. Methodology: DQG, FA. Software: DQG, FA. Writing original draft: DQG, FA. Writing review & editing: DQG, FS, NT, JTW, FA. Approval of final manuscript: all authors.
Conflict of interest
Joost Ter Woorst is a proctor for OPCAB at Medtronic. Ferdi Akca is a proctor for Endo-CAB and OPCAB at Medtronic. Except for that, no potential conflict of interest relevant to this article was reported.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Acknowledgments
Data will be made available by the corresponding author upon reasonable request.
J Chest Surg 2025; 58(1): 21-30
Published online January 5, 2025 https://doi.org/10.5090/jcs.24.080
Copyright © Journal of Chest Surgery.
De Qing Görtzen, B.Sc., Fleur Sampon , M.D., Naomi Timmermans
, M.D., Joost Ter Woorst
, M.D., Ph.D., Ferdi Akca
, M.D., Ph.D.
Department of Cardiothoracic Surgery, Catharina Hospital, Eindhoven, The Netherlands
Correspondence to:Ferdi Akca
Tel 31-40-239-8680
E-mail ferdi.akca@catharinaziekenhuis.nl
ORCID
https://orcid.org/0000-0002-1748-3235
This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/licenses/by-nc/4.0) which permits unrestricted non-commercial use, distribution, and reproduction in any medium, provided the original work is properly cited.
Commentary: J Chest Surg. 2025;58(1):31-33 https://doi.org/10.5090/jcs.24.110
Background: This study presents an overview of our technique and the perioperative outcomes for the first 100 patients who underwent minimally invasive endoscopic-assisted off-pump multivessel bypass grafting (endoscopic coronary artery bypass [endo-CAB]) at the Catharina Hospital in Eindhoven.
Methods: The first 100 patients undergoing multivessel endo-CAB from May 2022 to March 2024 were included in this retrospective, single-center, observational study (N=100). The study encompassed both elective and urgent surgical revascularization. In all cases, endoscopic-assisted harvesting of the internal mammary artery, radial artery, or saphenous vein was performed, followed by beating-heart anastomoses through a mini-thoracotomy.
Results: A total of 226 distal anastomoses were performed, utilizing 102 left internal mammary arteries, 80 radial arteries, 30 right internal mammary arteries, and 14 saphenous veins. On average, each patient had 2.3 anastomoses. A Y graft configuration was employed in 78 patients, in-situ bilateral internal mammary artery inflow in 19 patients, and a proximal aortic graft in 3 patients. Four patients underwent concurrent arrhythmia surgery. Eleven patients received hybrid revascularization. There was 1 conversion to sternotomy (1%) and 3 instances where cardiopulmonary bypass was required (3%). The median operation time was 3.3 hours (interquartile range, 3.0–3.7 hours), and the median hospital stay was 4.0 days (interquartile range, 3–4 days). The in-hospital mortality rate was 1%.
Conclusion: Multivessel off-pump endo-CAB surgery can be safely performed with endoscopic-assisted conduit harvesting. Combining the benefits of a minimally invasive and anaortic approach may improve perioperative outcomes for patients requiring surgical revascularization. Further studies are necessary to establish the role of this technique in routine coronary surgery.
Keywords: Endoscopic coronary artery bypass grafting, Off-pump coronary bypass grafting, Minimally invasive cardiac surgery, Minimally invasive direct coronary artery bypass
The standard of care for managing advanced coronary artery disease is coronary artery bypass grafting (CABG). As the global burden of cardiovascular disease continues to rise, CABG remains the most commonly performed procedure in cardiac surgery [1]. CABG is conventionally performed through a full sternotomy on a non-beating heart, utilizing cardiopulmonary bypass (CPB). However, off-pump coronary artery bypass grafting (OPCAB) has become increasingly popular due to its lower risk of cerebrovascular accidents. This benefit arises because the technique avoids manipulation of the aorta, a method referred to as the “anaortic approach” [2-4].
Minimally invasive coronary surgery is a rapidly evolving and technically demanding field, especially for multivessel revascularization. Using minimally invasive techniques, a median sternotomy is avoided in favor of a mini- thoracotomy or total endoscopic techniques [5]. Various methods have been developed for both conduit harvesting—including direct vision, thoracoscopic-assisted, or robotic-assisted mammary artery harvest—and graft deployment, which can be performed off-pump, on a beating heart, or on an arrested heart [6-12]. The ideal revascularization strategy would avoid sternotomy entirely and achieve complete revascularization using an anaortic approach, thus providing a truly minimally invasive long-term solution for the patient.
We aimed to share our experience with minimally invasive, multivessel, endoscopic-assisted, off-pump coronary surgery (endoscopic coronary artery bypass [endo-CAB]) and provide a comprehensive overview of the perioperative data and outcomes for the first 100 patients undergoing this procedure at the Catharina Hospital in Eindhoven.
The institutional review board and the ethical committee of Medical Research Ethics Committees United approved the study and waived the requirement for informed consent (registration no., W23.029).
From May 2022 to March 2024, 100 patients undergoing multivessel endo-CAB were included in this retrospective, single-center, observational study at Catharina Hospital in Eindhoven. All procedures were performed by a dedicated minimally invasive off-pump coronary surgeon. Patient baseline characteristics were obtained from the Department of Cardiothoracic Surgery’s database. Our aim was to provide a comprehensive overview of our initial experience and procedural parameters, focusing on the first 100 patients treated with multivessel endo-CAB. The patients had either double (n=87) or triple (n=13) vessel coronary artery disease. For 11 patients, a hybrid revascularization strategy was determined by the heart team, which included a percutaneous intervention alongside the multivessel endo-CAB procedure.
All patients were evaluated by a heart team, which included a cardiac surgeon and an interventional cardiologist, to assess the need for surgical revascularization. Patients diagnosed with multivessel disease and deemed suitable for either elective or urgent surgical revascularization were considered eligible for endo-CAB. Initially, during the implementation phase of endo-CAB, patients with a lower body mass index (BMI) and double vessel disease were specifically chosen for this procedure. Morbid obesity and a salvage situation involving hemodynamic instability were established as definitive exclusion criteria.
All patients underwent the standard preoperative workup required for any coronary surgery procedure, which included coronary angiography, chest radiography, electrocardiogram, transthoracic echocardiogram, bilateral blood pressure measurements, and blood panel laboratory tests. Following the preoperative workup, we determined whether a patient was eligible for standalone multiple-vessel endo- CAB, hybrid revascularization, or concomitant arrhythmia surgery. Routine preoperative pulmonary function tests or computed tomography scans were not included in our preoperative workup.
The procedure is performed under general anesthesia with the patient in a supine position. To improve surgical access, a pillow is placed under the left scapula, elevating the left hemithorax. Standard arterial and central venous lines, along with an endotracheal tube, are positioned before the procedure begins. A selective endobronchial blocker (Teleflex, Wayne, PA, USA) is inserted through the endotracheal tube to enable unilateral ventilation during the grafting process. Additionally, the groin is marked and draped in preparation for potential femoral cannulation if hemodynamic instability occurs. If the radial artery is planned to be harvested endoscopically at the same time, the right arm is positioned at a 90° angle from the thorax. Furthermore, transesophageal echocardiography is utilized to evaluate left ventricular and valve function.
Three small incisions are made on the left side of the chest: 1 at the third intercostal space at the anterior axillary line, and 2 more at the second and fourth intercostal spaces, positioned more anteriorly. Through these incisions, 5-mm endoscopic ports are inserted, and capnothorax is established at a pressure of 8 to 12 mm Hg (Fig. 1A). A 5-mm, 0°, 2-dimensional camera (Karl Storz GmbH, Tuttlingen, Germany) is then introduced. Using the other ports, standard long-shafted video-assisted thoracic surgery instruments and a Ligasure Maryland device (Medtronic, Dublin, Ireland) are also introduced. This setup enables harvesting of both the left internal mammary artery (LIMA) and the right internal mammary artery (RIMA).
Before harvesting the mammary artery, the pericardium is initially opened anterior to the left phrenic nerve to identify and assess coronary targets. A clip is placed near the grafting site on the left anterior descending artery (LAD) to facilitate easy identification through the small thoracotomy. When harvesting the bilateral mammary arteries (BIMAs), both the anterior mediastinum and the right pleural space are opened. The RIMA is harvested first, following the opening of the endothoracic fascia. Utilizing long-shafted video-assisted thoracic surgery instruments and the Ligasure device, the RIMA is harvested along with its adjacent veins in a semi-skeletonized fashion. The RIMA can be harvested proximally beyond the subclavian vein and distally up to its bifurcation. Chest wall deformities, such as pectus excavatum, can complicate harvesting the RIMA from the left side. A right thoracoscopic approach or opting for a radial artery can mitigate this issue. Once the RIMA is harvested to the desired length, the LIMA is then harvested.
Thoracoscopically, the phrenic nerve can be easily identified. The LIMA is harvested from the proximal area up to the level of the phrenic nerve and extends distally to its bifurcation. Any side branches of the LIMA are ligated using the Ligasure device (Fig. 1C). Once the LIMA has been fully dissected, heparin is administered intravenously to achieve an activated clotting time of over 300 seconds. The distal portion of the LIMA is secured with titanium clips (Endo-Clip; Medtronic). The LIMA is then sutured to the pericardium using a 5-0 polypropylene stitch. After dividing the LIMA, the location for the mini-thoracotomy is marked. Intrathoracic pressure is relieved to accurately determine the location of the coronary target without the influence of artificial capnothorax pressure. A transthoracic needle is used to puncture the intercostal space above the primary coronary target, ensuring an optimal placement for the mini-thoracotomy (Fig. 1B). The precise location may vary depending on the patient’s specific anatomy.
When a radial artery is used as a second arterial conduit, the LIMA and radial artery are harvested simultaneously to achieve complete arterial revascularization. Endoscopic radial artery harvesting was performed following the methods described by Reyes et al. [13] and Connolly et al. [14], utilizing an endoscopic vessel harvesting system (Vasoview Hemopro; Getinge, Gothenburg, Sweden). Similarly, endoscopic techniques were employed for saphenous vein harvesting.
Once all conduits have been harvested and the mini-thoracotomy site has been marked, the endoscopic ports are removed, and unilateral ventilation is initiated. The 5–6 cm incision is made at the previously marked location, between the third and fifth intercostal spaces. For grafting the LAD and right descending posterior, we generally adopt a more medial approach. If grafting of lateral branches is required, a more lateral access must be established. In female patients, the incision is preferably made in the inframammary area for aesthetic reasons. When the mini-thoracotomy is correctly positioned, the LIMA will be attached to the pericardium directly beneath the incision. Fig. 2A illustrates the LIMA as visible through the thoracotomy.
During multivessel procedures, various graft configurations may be employed. These include both BIMA in situ grafts and Y- or T-graft configurations using a free-RIMA or radial artery. Fig. 2A–C illustrate a T-graft comprising the LIMA and radial artery performed through a mini- thoracotomy. If a saphenous vein is utilized, the proximal anastomosis is created on the ascending aorta using techniques previously described by Issa and Ruel [10] and McGinn et al. [11]. Traction sutures are applied to the right lateral pericardium, and a sponge is positioned adjacent to the aorta. The Octopus Nuvo (Octopus Nuvo Tissue Stabiliser; Medtronic) is employed to depress the pulmonary artery, increasing maneuverability.
For cardiac positioning, we use the epicardial stabilizer and the heart positioning device (Starfish Evo; Medtronic), as previously described by Davierwala et al. [9], Issa and Ruel [10], and Kikuchi and Endo [15]. Depending on the coronary targets, we either use both devices or only the epicardial stabilizer. The heart is positioned with wet sponges, and the stabilizers are placed as shown in Fig. 2D. For the LAD and diagonal branches, only the epicardial stabilizer is used, as depicted in Fig. 3A. For the anterolateral, lateral, inferolateral, and inferior anastomoses, the armless Starfish Evo is also utilized, as illustrated in Fig. 3B and C. The armless Starfish Evo is employed to elevate the apex, facilitating access to the coronary targets, and a silk 1-0 suture is tied around the Starfish to assist in positioning. For the inferior wall, we elevate the apex of the heart with wet sponges and position the armless Starfish towards the apex. We often loosen our pericardial sutures, as they may limit visibility and maneuverability of the heart. We pull the Starfish cranially to stabilize the position for exposing the right posterior descending coronary artery. Subsequently, we position the epicardial stabilizer to secure a stable field for the coronary anastomosis. Once the heart is positioned, we ensure cardiac stability before continuing to manipulate the coronary artery. In cases of hemodynamic instability, we first elevate the legs to increase venous inflow and reposition the heart to stabilize it. If hemodynamic instability or ventricular arrhythmias persist, femoral CPB support is initiated. When using CPB, anastomoses are performed on the beating heart using similar techniques.
Once all coronary targets had been grafted, protamine was administered. The flow through the graft was measured using transit-time flow measurement (TTFM) to ascertain graft function (Medistim Vascular TTFM; Medistim, Oslo, Norway). After confirming that the graft function was sufficient, a 28F chest tube was inserted into the left pleural cavity through 1 of the 3 incisions previously made for the LIMA harvest. Ropivacaine was administered at 3 intercostal levels, and the skin was then closed (Fig. 3D).
Patients were considered for concomitant arrhythmia surgery if they presented with symptomatic paroxysmal or persistent atrial fibrillation (AF) and had a left atrial volume index of less than 70 mL/m2. A detailed account of our criteria for patient selection and the techniques used in concomitant thoracoscopic arrhythmia surgery has been previously published [16]. For patients with a high BMI, the primary surgical objective was to achieve minimally invasive coronary revascularization. If limited intrathoracic space increased the risk of complications during concomitant pulmonary vein isolation, only the left atrial appendage (LAA) clip (AtriClip; Atricure, Mason, OH, USA) would be applied to mitigate the risk of stroke.
All patients were transferred to the intensive care unit (ICU) after the procedure. Patients with radial artery grafts were administered calcium antagonists postoperatively to prevent graft spasms. Paracetamol, oral opioids, and intravenous nonsteroidal anti-inflammatory drugs were utilized for postoperative pain management. Following discharge, patients attended a follow-up appointment at the outpatient clinic 14 days later, where they underwent chest radiography and laboratory blood tests.
The data was analyzed using JASP ver. 0.18.1.0 (Universiteit van Amsterdam, Amsterdam, Netherlands) to obtain descriptives, boxplots, and Q-Q plots, identifying outliers for each continuous variable. Categorical variables are presented as frequency and percentage, while continuous variables are shown as mean and standard deviation if they are normally distributed, and as median and interquartile range (IQR) if they are not, based on the Q-Q plot analysis. RStudio ver. 4.4.0 (R Core Team, Boston, MA, USA) was utilized to visualize data and compute the Pearson correlation coefficient.
The patient demographics, procedural parameters, and perioperative characteristics are displayed in Table 1. Most of the patients were men (90.0%) and the mean age was 65.5±9.1 years. The mean BMI was 26.7±3.3 kg/m2, and the median body surface area was 2.00 m2 (IQR, 1.91–2.10 m2). Seventeen percent of the patients (17.0%) had diabetes, 7 patients had peripheral artery disease (7%), and 5 patients had AF (5%). The median Euroscore II for the population was 1.00 (IQR, 0.80–1.61).
Table 1 . Demographics and perioperative characteristics of the endoscopic coronary artery bypass group (N=100).
Characteristic | Value |
---|---|
Sex (men) | 90 (90.0) |
Age (yr) | 65.5±9.1 |
Height (cm) | 175 (171–180) |
Weight (kg) | 82.3±12.8 |
Body mass index (kg/m2) | 26.7±3.3 |
Body surface area (m2) | 2.00 (1.91–2.10) |
Diabetes | 17 (17.0) |
Peripheral vascular disease | 7 (7.0) |
Atrial fibrillation | 5 (5.0) |
Previous cardiac surgery | 0 |
Recent myocardial infarction | 28 (28.0) |
Prior percutaneous coronary intervention | 33 (33.0) |
Left ventricular ejection fraction | |
Good (<50%) | 78 (78.0) |
Moderate (31%–50%) | 18 (18.0) |
Poor (21%–30%) | 3 (3.0) |
Very poor (<20%) | 1 (1.0) |
Chronic obstructive pulmonary disease | 4 (4.0) |
Pulmonary hypertension | 0 |
Cerebrovascular accident | 6 (6.0) |
Renal impairment | 5 (5.0) |
Dialysis | 0 |
Level of urgency | |
Elective | 55 (55.0) |
Urgent | 44 (44.0) |
Emergency | 1 (1.0) |
Euroscore II | 1.00 (0.80–1.61) |
Preoperative hemoglobin (mmol/L) | 9.1 (8.3–9.6) |
Double vessel disease | 87 (87.0) |
Triple vessel disease | 13 (13.0) |
Concomitant arrhythmia surgery | 4 (4.0) |
Isolated LAA clip | 3 (3.0) |
Thoracoscopic AF ablation including LAA clip | 1 (1.0) |
No. of distal anastomoses | |
2 | 75 (75.0) |
3 | 24 (24.0) |
4 | 1 (1.0) |
No. of anastomoses with left IMA | |
1 | 96 (96.0) |
2 | 3 (3.0) |
No. of anastomoses with right IMA | |
1 | 26 (26.0) |
2 | 2 (2.0) |
No. of anastomoses with radial artery | |
1 | 42 (42.0) |
2 | 19 (19.0) |
No. of anastomoses with SVG | |
1 | 6 (6.0) |
2 | 4 (4.0) |
Y-graft configuration | 78 (78.0) |
Bilateral internal mammary artery in situ | 19 (19.0) |
Proximal aorta | 3 (3.0) |
Hybrid revascularization | 11 (11.0) |
PCI, circumflex artery | 5 (5.0) |
PCI, right coronary artery | 6 (6.0) |
Conversion to sternotomy | 1 (1.0) |
CPB-assisted | 3 (3.0) |
Operation time (hr) | 3.3 (3.0–3.7) |
Values are presented as frequency (%) for categorical variables; for continuous variables, they are shown as mean±standard deviation if normally distributed, or as median (interquartile range) if non-normally distributed..
LAA, left atrial appendage; AF, atrial fibrillation; IMA, internal mammary artery; SVG, saphenous vein graft; PCI, percutaneous coronary intervention; CPB cardiopulmonary bypass..
Four patients underwent concomitant arrhythmia surgery. Three underwent isolated LAA clip placement and 1 patient underwent thoracoscopic AF ablation, including an LAA clip. The patients received an average of 2.3 distal anastomoses. The total number of grafts per conduit is plotted in Fig. 4A. The first choice of conduit was the LIMA with 102 anastomoses, while the radial artery was used for 80 anastomoses, the RIMA for 30 anastomoses, and the saphenous vein for 14 anastomoses. A Y-graft configuration was used in 78 patients (78%), the BIMAs in situ were utilized in 19 patients (19%), and a proximal aorta graft was placed in 3 cases (3%) (Fig. 4B). Coronary anastomoses were performed on the anterior (n=100), anterolateral (n=42), lateral (n=65), inferolateral (n=5), and inferior wall (n=14) (Fig. 4C, Table 2). In our population, 11 patients underwent hybrid revascularization (11%). In 1 patient, a conversion to sternotomy was performed due to limited visibility (1%), and 3 cases needed CPB support (3%). In 1 patient, this was planned preoperatively, while in 2 patients, CPB was used due to hemodynamic instability. The median operation time was 3.3 hours (IQR, 3.0–3.7 hours). The Pearson correlation coefficient was calculated to illustrate a possible learning curve in Fig. 5. The Pearson correlation was calculated to be r(98)=-0.109 (p=0.279).
Table 2 . Graft locations in the endoscopic coronary artery bypass group.
Location | No. |
---|---|
Anterior | 100 |
Anterolateral | 42 |
Lateral | 65 |
Inferolateral | 5 |
Inferior | 14 |
Table 3 describes the postoperative characteristics and complications. The total ventilation time was 7.6 hours (IQR, 6.6–9.3 hours), and the total postoperative blood loss was 350 mL (IQR, 200–550 mL). The median stay in the ICU was 1.0 days (IQR, 0.5–1 days), and the total postoperative hospital stay was 4 days (IQR, 3–4 days). In-hospital mortality occurred in 1 patient (1%) due to shock caused by a postoperative stress response due to previously undiagnosed cancer metastases. One patient required postoperative re-exploration for bleeding (1%), and 1 patient had postoperative ischemia requiring revision of the bypass graft (1%). One patient had a pulmonary embolism that required oral anticoagulation. There was no occurrence of stroke. During hospital admission, 7 patients developed new-onset AF that required oral anticoagulation (7%).
Table 3 . Postoperative characteristics of the endoscopic coronary artery bypass group (N=100).
Variable | Value |
---|---|
Total ventilation time (hr) | 7.6 (6.6–9.3) |
Postoperative blood loss (mL) | 350 (200–550) |
Intensive care unit admission time (day) | 1.0 (0.5–1) |
Total hospital stay (day) | 4 (3–4) |
In hospital mortality | 1 (1.0) |
Intra-aortic balloon pump | 0 |
Extracorporeal membrane oxygenation | 0 |
CK-MB max (ng/mL) | 25.5 (19.0–35.3) |
Discharge hemoglobin (mmol/L) | 7.7 (6.8–8.3) |
Transfusion | 6 (6.0) |
Re-exploration for bleeding | 1 (1.0) |
Ischemia postoperative | 1 (1.0) |
Graft revision | 1 (1.0) |
Acute percutaneous coronary intervention | 0 |
Cerebrovascular accident | 0 |
Tamponade during initial admission | 0 |
Tamponade after discharge | 1 (1.0) |
Pulmonary embolism | 1 (1.0) |
Wound infection | 0 |
New-onset atrial fibrillation | 7 (7.0) |
Hospital-acquired pneumonia | 3 (3.0) |
Thoracentesis | 3 (3.0) |
30-day readmission | 2 (2.0) |
Target vessel reintervention | 3 (3.0) |
Values are presented as number (%) for categorical variables and median (interquartile range) for continuous variables if non-normally distributed..
CK-MB, creatine kinase-myoglobin binding..
During follow-up, 1 patient developed late tamponade (1%) that required pericardiocentesis and 3 patients had left-sided pleural effusion that required thoracentesis (3%). Two patients were readmitted within 30 days after discharge for symptomatic postoperative anemia (n=1) and dyspnea due to pleural effusion requiring thoracentesis (n=1). Three patients received unexpected target vessel reintervention (3%) during follow-up.
In this study, we described the outcomes of our first 100 endoscopic-assisted, off-pump, multivessel coronary bypass grafting procedures. All patients underwent endoscopic-assisted graft harvesting and received coronary bypasses via a mini-thoracotomy. On average, each patient had 2.3 distal anastomoses. Eleven patients underwent hybrid revascularization. There was 1 conversion to sternotomy and 3 instances where CPB was required. The median operation time was 3.3 hours, and the in-hospital mortality rate was 1%. Our findings indicate that multivessel off-pump endo-CAB can be performed safely with favorable postoperative outcomes.
Multivessel endo-CAB is a technically demanding procedure with a steep learning curve. It utilizes conventional off-pump coronary techniques. This approach differs technically from bypass grafting performed on an arrested heart [17]. Patients with advanced atherosclerosis of the aorta, chronic lung disease, or kidney problems may benefit from this approach. Other advantages include reduced risks of postoperative inflammation, infection, and arrhythmia [18]. Safely performing OPCAB requires specific surgical training and dexterity to achieve complete familiarity [19,20]. As Shaefi et al. [2] previously described, maintaining hemodynamic stability and minimizing coronary ischemia are critical considerations. We believe that to safely perform multivessel endo-CAB as described above, a surgeon must be comfortable with sternotomy off-pump techniques, ensuring adequate exposure and stabilization for the precise creation of anastomoses on a beating heart.
We further discuss the challenges associated with the thoracoscopic harvest of the mammary artery. Our group has previously detailed an extensive examination of the learning curve associated with thoracoscopic harvesting. In our initial 80 cases, we observed a significant reduction in LIMA harvesting time, from an average of over 100 minutes to approximately 60 minutes [7,21]. Vassiliades was the first to demonstrate the feasibility of thoracoscopic BIMA harvest with favorable outcomes [22]. Starting the procedure with endoscopic assistance offers several advantages. First, it allows for the opening of the pericardium to identify coronary targets and potentially alter the graft configuration. Second, it facilitates the precise identification of the optimal location for the mini-thoracotomy. Finally, it enables the performance of concurrent thoracoscopic arrhythmia surgery, thus maintaining the principles of minimally invasive surgery while effectively addressing multiple comorbidities [22].
Our second most frequently used conduit is the radial artery, which can be harvested through minimally invasive techniques [14]. Although the RIMA is an excellent graft conduit option, it is less often utilized due to the slightly more challenging harvesting process from the left side of the thorax, leading us to prefer the endoscopic radial artery approach [23]. In this study, we resorted to venous grafts only when arterial conduits were not available, or when technical difficulties or patient comorbidity were present. In almost all cases, we preferred the Y configuration or BIMA in situ as the graft configuration, to provide an all- arterial, anaortic approach. The benefits of this approach have been described by Puskas et al. [3].
This procedure involves both surgical and anesthesiological learning curves. Close collaboration with the anesthesiologist is crucial during heart manipulation, similar to what is required in sternotomy off-pump revascularization. Factors such as table positioning, the administration of inotropes or vasopressors, and the sequencing of grafts are critical for ensuring the safety of the procedure. Occasionally, the heart may need to be positioned slightly differently to maintain stable hemodynamics during the coronary anastomosis, which necessitates effective communication with the anesthesia team.
Ideally, the endo-CAB procedure is performed off-pump and without manipulating the aorta, termed “anaortic.” If the surgeon lacks experience with off-pump beating heart surgery, they can decide preoperatively to use a pump. Therefore, pump-assisted endo-CAB may be an attractive option for those with limited experience, as it allows the surgeon to become acquainted with the minimally invasive aspects of the procedure.
The study is a single-center retrospective analysis. Additionally, our follow-up period was brief, and the angiographic evidence is confined to cases of hybrid revascularization and target vessel reinterventions. Nevertheless, we can tentatively conclude that for selected patients, multivessel endo-CAB offers favorable perioperative outcomes, with target vessel reintervention rates comparable to those of sternotomy techniques.
Multivessel off-pump endo-CAB can be safely performed using thoracoscopic IMA harvesting techniques. We believe that combining the benefits of minimally invasive surgery with an anaortic approach could enhance perioperative outcomes for patients undergoing coronary surgery. Mastery of conventional off-pump techniques is crucial to successfully perform this procedure. In selected patients, multivessel endo-CAB yields favorable procedural outcomes.
Authors contributions
Conceptualization: DQG, FA. Data curation: DQG, FA. Formal analysis: DQG, FA. Methodology: DQG, FA. Software: DQG, FA. Writing original draft: DQG, FA. Writing review & editing: DQG, FS, NT, JTW, FA. Approval of final manuscript: all authors.
Conflict of interest
Joost Ter Woorst is a proctor for OPCAB at Medtronic. Ferdi Akca is a proctor for Endo-CAB and OPCAB at Medtronic. Except for that, no potential conflict of interest relevant to this article was reported.
Funding
This research did not receive any specific grant from funding agencies in the public, commercial, or not-for-profit sectors.
Acknowledgments
Data will be made available by the corresponding author upon reasonable request.
Table 1 . Demographics and perioperative characteristics of the endoscopic coronary artery bypass group (N=100).
Characteristic | Value |
---|---|
Sex (men) | 90 (90.0) |
Age (yr) | 65.5±9.1 |
Height (cm) | 175 (171–180) |
Weight (kg) | 82.3±12.8 |
Body mass index (kg/m2) | 26.7±3.3 |
Body surface area (m2) | 2.00 (1.91–2.10) |
Diabetes | 17 (17.0) |
Peripheral vascular disease | 7 (7.0) |
Atrial fibrillation | 5 (5.0) |
Previous cardiac surgery | 0 |
Recent myocardial infarction | 28 (28.0) |
Prior percutaneous coronary intervention | 33 (33.0) |
Left ventricular ejection fraction | |
Good (<50%) | 78 (78.0) |
Moderate (31%–50%) | 18 (18.0) |
Poor (21%–30%) | 3 (3.0) |
Very poor (<20%) | 1 (1.0) |
Chronic obstructive pulmonary disease | 4 (4.0) |
Pulmonary hypertension | 0 |
Cerebrovascular accident | 6 (6.0) |
Renal impairment | 5 (5.0) |
Dialysis | 0 |
Level of urgency | |
Elective | 55 (55.0) |
Urgent | 44 (44.0) |
Emergency | 1 (1.0) |
Euroscore II | 1.00 (0.80–1.61) |
Preoperative hemoglobin (mmol/L) | 9.1 (8.3–9.6) |
Double vessel disease | 87 (87.0) |
Triple vessel disease | 13 (13.0) |
Concomitant arrhythmia surgery | 4 (4.0) |
Isolated LAA clip | 3 (3.0) |
Thoracoscopic AF ablation including LAA clip | 1 (1.0) |
No. of distal anastomoses | |
2 | 75 (75.0) |
3 | 24 (24.0) |
4 | 1 (1.0) |
No. of anastomoses with left IMA | |
1 | 96 (96.0) |
2 | 3 (3.0) |
No. of anastomoses with right IMA | |
1 | 26 (26.0) |
2 | 2 (2.0) |
No. of anastomoses with radial artery | |
1 | 42 (42.0) |
2 | 19 (19.0) |
No. of anastomoses with SVG | |
1 | 6 (6.0) |
2 | 4 (4.0) |
Y-graft configuration | 78 (78.0) |
Bilateral internal mammary artery in situ | 19 (19.0) |
Proximal aorta | 3 (3.0) |
Hybrid revascularization | 11 (11.0) |
PCI, circumflex artery | 5 (5.0) |
PCI, right coronary artery | 6 (6.0) |
Conversion to sternotomy | 1 (1.0) |
CPB-assisted | 3 (3.0) |
Operation time (hr) | 3.3 (3.0–3.7) |
Values are presented as frequency (%) for categorical variables; for continuous variables, they are shown as mean±standard deviation if normally distributed, or as median (interquartile range) if non-normally distributed..
LAA, left atrial appendage; AF, atrial fibrillation; IMA, internal mammary artery; SVG, saphenous vein graft; PCI, percutaneous coronary intervention; CPB cardiopulmonary bypass..
Table 2 . Graft locations in the endoscopic coronary artery bypass group.
Location | No. |
---|---|
Anterior | 100 |
Anterolateral | 42 |
Lateral | 65 |
Inferolateral | 5 |
Inferior | 14 |
Table 3 . Postoperative characteristics of the endoscopic coronary artery bypass group (N=100).
Variable | Value |
---|---|
Total ventilation time (hr) | 7.6 (6.6–9.3) |
Postoperative blood loss (mL) | 350 (200–550) |
Intensive care unit admission time (day) | 1.0 (0.5–1) |
Total hospital stay (day) | 4 (3–4) |
In hospital mortality | 1 (1.0) |
Intra-aortic balloon pump | 0 |
Extracorporeal membrane oxygenation | 0 |
CK-MB max (ng/mL) | 25.5 (19.0–35.3) |
Discharge hemoglobin (mmol/L) | 7.7 (6.8–8.3) |
Transfusion | 6 (6.0) |
Re-exploration for bleeding | 1 (1.0) |
Ischemia postoperative | 1 (1.0) |
Graft revision | 1 (1.0) |
Acute percutaneous coronary intervention | 0 |
Cerebrovascular accident | 0 |
Tamponade during initial admission | 0 |
Tamponade after discharge | 1 (1.0) |
Pulmonary embolism | 1 (1.0) |
Wound infection | 0 |
New-onset atrial fibrillation | 7 (7.0) |
Hospital-acquired pneumonia | 3 (3.0) |
Thoracentesis | 3 (3.0) |
30-day readmission | 2 (2.0) |
Target vessel reintervention | 3 (3.0) |
Values are presented as number (%) for categorical variables and median (interquartile range) for continuous variables if non-normally distributed..
CK-MB, creatine kinase-myoglobin binding..