Minimally invasive surgical treatment of coronary artery multivessel disease

Minimally Invasive Surgical Treatment of Coronary Artery Multivessel Disease Vassilios Gulielmos, MD, Michael Knaut, MD, Romuald Cichon, MD, Michael Brandt, Thorsten Jost, MD, Klaus Matschke, MD, and Stephan Schu¨ler, MD Cardiovascular Institute, University Hospital Dresden, Dresden, Germany

Background. If coronary artery multivessel disease is the target of a minimally invasive procedure, either median sternotomy or cardiopulmonary bypass can be avoided. Methods. We used an alternate technique instead of minithoracotomy and cardiopulmonary bypass to treat 102 patients (82 men, 20 women; age range, 39 to 82 years; median, 61.0 6 8.9 years) for coronary artery singlevessel, double-vessel, or multivessel disease between November 1996 and January 1998. Twenty-nine patients (22 men, 7 women; age range, 46 to 78 years; median, 69.0 6 8.4 years), who were in a high-risk group for the development of perioperative complications because of the use of cardiopulmonary bypass, received median sternotomy and a beating heart procedure using the

Octopus stabilizing technique. The left anterior descending coronary artery was the target vessel in all patients except for 1, in whom the left internal mammary artery was used. Results. There was no intraoperative death in either series. In the beating heart group (Octopus) 2 patients died on postoperative day 31 and 35, respectively, of postoperative pneumonia. Conclusions. Both techniques present safe alternative procedures to conventional coronary artery bypass grafting in patients with coronary artery multivessel disease.

M

Material and Methods

inimally or less invasive surgical techniques have been developed for the treatment of coronary artery disease [1–3] to minimize surgical trauma in cardiac surgery as in general surgery, avoiding median sternotomy, or cardiopulmonary bypass (CPB), or even both [4 – 6]. In March 1996 we introduced the Port-Access (Heartport, Redwood City, CA) technique at our institution as an initial trial in Europe; this technique preserves the safety standards of cardiac operations with CPB and cardioplegic arrest. Because of our protocol the method was used only for coronary artery single-vessel disease (CASVD). On the basis of the experience gained using the PortAccess technique we developed a surgical technique for the treatment of coronary artery multivessel disease (CAMVD) avoiding median sternotomy and using CPB. Patients with impaired left ventricular function or other high-risk factors for the use of CPB such as impaired renal or lung function or heavily calcified aorta were excluded from this series and received a beating heart procedure using the Octopus stabilizing system [6].

Presented at “Facts and Myths of Minimally Invasive Cardiac Surgery: Current Trends in Thoracic Surgery IV,” New Orleans, LA, Jan 24, 1998. Address reprint requests to Dr Schu¨ler, Cardiovascular Institute, University Hospital Dresden, Fetscherstrasse 76, D-01307 Dresden, Germany.

© 1998 by The Society of Thoracic Surgeons Published by Elsevier Science Inc

(Ann Thorac Surg 1998;66:1018 –21) © 1998 by The Society of Thoracic Surgeons

Dresden Technique Between November 1996 and January 1998, 102 patients (82 men and 20 women; age range, 39 and 82 years; median 61.0 6 8.9 years) with CASVD, coronary artery double-vessel disease (CADVD), and CAMVD were evaluated for this technique. The exclusion criteria used were calcified aorta, obesity (body mass index . 130%), and impaired left ventricular function (ejection fraction ,0.35). Clinical staging revealed 15.8% of the patients to be in Canadian Cardiovascular Society stage 1, 46.4% in stage 2, 35.4% in stage 3, and 2.4% in stage 4; 39.0% of the patients were in New York Heart Association (NYHA) functional class I, 46.3% in class II, and 14.7% in class III. Fifty-four patients suffered from CADVD or coronary artery triple-vessel disease (CATVD). The pattern of the coronary morphology is shown in Table 1. Hospital institutional review board approval and informed consent from all patients were obtained. At operation the patient was placed in supine position with a rubber cushion under the left shoulder and the left arm attached to the body dorsally to the posterior axillary line. For the surgical gate in most cases the third or the second intercostal space was chosen. A 6- to 9-cm skin incision at the level of the chosen intercostal space was made and the sternal edge of the upper and the lower rib were divided but not removed. The use of a small wound retractor enabled harvesting of the left internal mammary artery (LIMA) as a pedicle up to the first rib and down to the fifth or sixth rib. The pericardium was 0003-4975/98/$19.00 PII S0003-4975(98)00710-3

Ann Thorac Surg 1998;66:1018 –21

MINIMALLY INVASIVE GULIELMOS ET AL MINIMALLY INVASIVE SURGICAL TREATMENT OF CAD

1019

Table 1. Coronary Lesions, Coronary Vessels Grafted and Used Conduits Surgical Technique

Coronary Lesions

No. of Patients

Coronary Vessels Grafted and Conduits Used

No. of Patients

Dresden technique (n 5 102)

LAD LAD and DB LAD and IB LAD and LCX LAD, DB and LCX Main stem Main stem and RCA LAD and RCA LAD, DB, and RCA LAD, RCA and IB LAD, LCX and RCA LAD, RCA, LCX and IB LAD muscle bridge

25 22 2 11 5 7 1 7 5 1 12 3 1

Octopus technique (n 5 29)

LAD LAD and LCX LAD, LCX, and RCA RCA LAD and RCA

13 1 8 1 6

LL VG (LAD) LL 1 RCA(VG) LL 1 RCA(VG) 1 DB(VG) LL 1 MB(VG) LL 1 MB(VG) 1 PDA(VG) LL 1 IB(VG) LL 1 IB(VG) 1 MB(VG) LL 1 DB(VG) LL 1 LCX(VG) LL 1 RCA(VG) 1 MB(VG) LL 1 MB(VG) 1 DB(VG) LIMA-FG(LAD) 1 RIMA-T graft (IB) LIMA-FG(LAD) 1 LCX(VG) LIMA-FG(LAD) 1 DB(VG) Ostioplasty MBD (LAD) LL LIMA (DB) RCA (VG) RCA (VG) 1 IB (VG) LIMA (PDA)

30 1 7 1 12 1 5 1 30 1 4 3 1 1 1 1 1 25 1 1 1 1

DB 5 diagonal branch of LAD; FG 5 free graft; IB 5 intermediate branch; LAD 5 left arterior descending coronary artery; (right) internal mammary artery; LCX 5 left circumflex artery; LL 5 LIMA (LAD); MB 5 marginal branch of LCX; descending artery; RCA 5 right coronary artery; VG 5 vein graft.

opened longitudinally. At the same time saphenous vein segments were harvested by another surgeon if necessary. The femoral vein was cannulated for venous drainage for CPB. The ascending aorta was cannulated and a conventional aortic clamp was used for external crossclamping. Antegrade cold crystalloid cardioplegia was applied through the ascending aorta. During cardioplegic arrest, aortic root venting was made and the relaxed heart was rotated for exposure of coronary arteries of the left or the right coronary system, or both. End-to-side anastomoses were performed between vein grafts and the coronary arteries in a standard fashion (Fig 1) followed by anastomosis of the LIMA to the left anterior descending coronary artery (LAD). Before removal of the aortic clamp, antegrade deairing was done through the ascending aorta. Proximal anastomoses were performed with a conventional side-biting clamp. The patient was weaned from CPB and cannulas were removed from the aorta and the femoral vein. The sternal edges of the third and fourth rib were attached to the sternum using two steel wires and both ribs were approximated to each other with a 1-mm strong suture (poly-p-dioxanon). Two chest tubes were left in place, and the chest incision was closed in layers. Left internal mammary artery bypass to LAD and vein grafts to other coronary arteries if necessary were performed in all patients except for 3. Of those 3 patients, 1 received a dissection of a LIMA muscle bridge, 1 an ostioplasty, and another a vein graft to the LAD. The patterns of the coronary vessels grafted and the conduits

LIMA (RIMA) 5 left PDA 5 posterior

used are listed in Table 1. Perioperative data included total duration of operation, LIMA harvesting time, duration of CPB, and postoperative ventilation, intensive care unit stay, and hospitalization (Table 2).

Fig 1. Anastomosis between a vein graft (A) and a diagonal branch of the left anterior descending artery (B) performed with the Dresden technique.

1020

MINIMALLY INVASIVE GULIELMOS ET AL MINIMALLY INVASIVE SURGICAL TREATMENT OF CAD

Table 2. Resultsa Variable Operation time LIMA CPB Cross-clamp Ventilation ICU stay Hospital stay a

Dresden Technique (n 5 102)

Octopus Technique (n 5 29)

256 6 59.1 min 43.0 6 18.5 min 91.0 6 34.1 min 49.5 6 20.2 min 260 6 147.7 min 1 6 0.4 days 6 6 1.4 days

153.0 6 43.1 min ... ... ... 240 6 672 min 1 6 3.8 days 6 6 3.9 days

All values are mean 6 standard error of the mean.

CPB 5 cardiopulmonary bypass; 5 left internal mammary artery.

ICU 5 intensive care unit;

LIMA

Octopus Technique Twenty-nine patients with CASVD, CADVD, or CAMVD were evaluated from July 1997 to January 1998 for beating heart procedures using median sternotomy. The age of the 22 male and 7 female patients ranged between 46 and 78 years (median, 69.0 6 8.4 years). Inclusion criteria were impaired left ventricular ejection fraction, impaired renal or lung function, or heavily calcified aorta to avoid CPB-related complications. Clinical staging revealed 1 patient (3.4%) to be in Canadian Cardiovascular Society stage 1, 12 patients (41.4%) to be in stage 2, 14 patients (48.3%) to be in stage 3, and 2 patients (6.9%) to be in stage 4; 2 patients (6.9%) were in NYHA class I, 11 patients (37.9%) in NYHA class II, and 16 patients (55.2%) in NYHA class III. Fifteen patients suffered from severe CADVD or CATVD. The pattern of the coronary morphology is shown in Table 1. A single-lumen ventilation tube was used for general anesthesia. At operation the patient was placed in supine position and conventional median sternotomy was performed. The LIMA and saphenous vein grafts were harvested conventionally. With a conventional chest retractor, the pericardium was opened longitudinally. After exposure of the right coronary artery, two vessel loops were used for occlusion of the coronary artery. The vessel was dissected and the anastomosis was performed in a standard fashion using the Octopus stabilizing system [6] (Fig 2). Before performing the next anastomosis we attached the proximal end of this vein graft to the ascending aorta with a side-biting clamp. The marginal branch of the circumflex artery, the intermediate branch, or the diagonal branch could be grafted, followed by LIMA to LAD. Heparin was antagonized with protamine and hemostasis was made. One chest tube was left in the left pleural cavity. Two substernal tubes were left in place, and the sternum was closed with wires. All patients except for 3 received a LIMA bypass graft to the LAD. The patterns of the coronary vessels grafted and the conduits used are listed in Table 1. Perioperative data such as total duration of operation, postoperative ventilation, intensive care unit stay, and hospital stay were monitored. Because of the study protocol all patients had a follow-up including physical examinations, 12-lead electrocardiogram, and roentgenograms on postoperative day 2.

Ann Thorac Surg 1998;66:1018 –21

In addition, patients were followed up 12 weeks after the operation.

Results Dresden Technique All patients who underwent this operation are alive at present and well up to 14 months postoperatively. There was no perioperative infarction or necessity for inotropic support. One patient needed reexploration because of bleeding. In 1 patient the procedure had to be converted to median sternotomy because of an undiagnosed heavily calcified ascending aorta. Further postoperative complications were wound infections in 4 patients and lymphatic fistulas in 4 patients from our initial series using dissection of the groin for venous cannulation. Table 2 shows duration of operation, duration of LIMA harvesting, intensive care unit stay, and hospitalization of all the procedures. The 12-week follow-up completed in 53 patients (52.5%) revealed angina pectoris in 2 patients (3.8%). These 2 patients underwent coronary angiography showing stenosis of the LAD anastomosis in the first patient and a new stenosis in the LAD 2 cm distal to the LIMA anastomosis in the second patient. Both patients underwent an uncomplicated percutaneous transluminal coronary angioplasty. Thirty-three patients (62.4%) were in NYHA class I, 19 patients (35.8%) were in class II, and 1 patient (1.8%) was in class III; 51 patients (96.2%) were in Canadian Cardiovascular Society stage 1, 1 patient (1.9%) was in stage 2, and 1 patient (1.9%) was in stage 3. The 3-month control stress electrocardiogram was completed in 53 patients (52.5%) and showed no evidence of myocardial ischemia. Pain assessment revealed pain in 1

Fig 2. The suction paddle-handles (A, B) are connected with two rigid arms to the operating table. The coronary vessel is occluded using two vessel loops (C, D).

Ann Thorac Surg 1998;66:1018 –21

patient (1.8%), who claimed to have some pain during body exercise.

Octopus Technique Atrial fibrillation developed in 1 patient during the anastomosis between a vein graft and the right coronary artery. After the anastomosis was already performed, defibrillation was necessary. No patient needed inotropic support, and all patients left the operating theater in sinus rhythm without signs of ischemia. Two multimorbid patients died of pneumonia on postoperative day 31 and postoperative day 35. In another patient with an impaired left ventricular ejection fraction, postoperative low cardiac output developed and necessitated treatment with the intraaortic balloon pump and low inotropic support for 5 days. This patient was weaned from mechanical and drug support and is alive and well today. Table 2 shows duration of operation, intensive care unit stay, and hospitalization. The 12-week follow-up completed in 16 patients (55.2%) revealed no angina pectoris. Of the patients receiving the 3-month follow-up 9 patients (56.3%) were in Canadian Cardiovacular Society stage 1, 6 patients (37.5%) were in stage 2, and 1 patient (6.2%) was in stage 3; 7 patients (43.8%) were in NYHA class I, 7 patients (43.8%) were in class II, and 2 patients (12.4%) were in class III. Three-month control stress electrocardiogram was completed in 12 patients (41.4%) and showed no evidence of myocardial ischemia. Pain assessment revealed pain in 1 patient (6.2%), who claimed to have some pain during body exercise, combined with an irritation of the proximal wound end.

Comment Since the first minimally invasive bypass procedure using a LIMA bypass to the LAD avoiding median sternotomy and CPB was performed by Benetti and colleagues [1], a rapid development of minimally invasive techniques for the treatment of coronary artery disease was encountered [1–3, 6]. After introduction of minimally invasive coronary artery bypass grafting for CASVD, however, it became obvious that for the treatment of CAMVD median sternotomy and CPB can be avoided using alternative approaches [7]. When the Dresden technique is used with CPB, avoiding median sternotomy, even patients with multivessel disease can be treated safely with a very low morbidity [7]. This technique gives good access to all areas of the heart through a small chest incision because of CPB and cardioplegic arrest, and even complex procedures, such as bilateral internal mammary artery grafting and ostio-

MINIMALLY INVASIVE GULIELMOS ET AL MINIMALLY INVASIVE SURGICAL TREATMENT OF CAD

1021

plasty, can be performed. No vascular complications have been observed, as in other techniques [5, 8], caused by antegrade aortic cannulation. The Octopus stabilizing technique presents an adequate alternative for patients with an impaired left ventricular ejection fraction, impaired renal function, or other important risk factors for the use of CPB. This technique is applied at our institution through a median sternotomy to reach the LAD, the right coronary artery, and in some cases branches of the left circumflex artery as well. Results at our institution using the Octopus stabilizing system are not as good as those with the Dresden technique because of the patient selection; however, mortality in this group was not technique related, but due to the overall severity of impaired clinical status of the patients. For patients undergoing conventional coronary artery bypass grafting at our institution the mean hospital stay is 8 days, which means that when either of these two techniques was used hospitalization was shortened (see Table 2), thus illustrating a faster rehabilitation. Each technique reduces surgical trauma in a different way; therefore, they should be considered as complementary procedures for two patients groups with different morbidities.

References 1. Benetti FJ, Ballester C, Sani G, Boonstra P, Grandjean J. Video assisted coronary bypass surgery. J Cardiol Surg 1995;10: 620–5. 2. Stevens JH, Burdon TA, Peters WS, et al. Port-Access coronary artery bypass grafting: a proposed surgical method. J Thorac Cardiovasc Surg 1996;111:567–73. 3. Fonger J. Reoperative and alternative MICAB approaches: subxiphoid and lateral thoracotomy [Abstract]. World Congress on Minimally Invasive Cardiac Surgery, Paris, May 30 –31, 1997. 4. Calafiore AM, Di Giammarco G, Teodori G, et al. Left anterior descending coronary artery grafting via left anterior small thoracotomy without cardiopulmonary bypass. Ann Thorac Surg 1996;61:1658– 65. 5. Reichenspurner H, Gulielmos V, Daniel WG, Schueler S. Minimally invasive coronary artery bypass (CABS) with the safety of cardiopulmonary bypass and cardioplegic arrest. N Engl J Med 1997;336:67– 8. 6. Jansen E, Gruendemann P, Borst C, et al. Use of the Octopus method in off-pump single and multivessel CABG [Abstract]. World Congress on Minimally Invasive Cardiac Surgery, Paris, May 30 –31, 1997. 7. Gulielmos V, Knaut M, Dangel M, Wunderlich J, Schmidt V, Schueler S. A new minimally invasive surgical technique for the treatment of coronary artery multi vessel disease (CAD). Circulation 1997;96(Suppl 1):681. 8. Mohr FW, Falk V, Diegeler A, Walther T, Wendler R, Gummert J, Autschbach R. Videoassistierte minimal invasive Mitralklappenchirurgie. Zeitschr Kardiol 1997;86(Suppl 2):297.