Comparison of Implantability, Early Post-Operative Valve Function and Structural Valve Deterioration between the Carpentier-Edward Perimount Magna Valve and St. Jude Medical Trifecta Valve / 日本心臓血管外科学会雑誌

Purpose : Easy and safe implantability, good post-operative valve function and good long-term durability are required for any bioprosthetic valve implanted in aortic position. The Carpentier Edwards Perimount Magna valve (Magna) was introduced in 2009 and the St. Jude Medical Trifecta valve (Trifecta) was introduced in 2012 to our institution. In this study, we compared implantability, early post-operative valve function and structural valve deterioration (SVD) between these two valves. Patients and Methods : Between January 2009 and December 2019, Magna or Trifecta were electively implanted for 254 patients (Magna 151 patients and Trifecta 103 patients) and these patients were included in this study. Implantability was evaluated by occurrence of intraoperative valve dysfunction. Early post-operative valve function was evaluated by mean pressure gradient (m-PG) and indexed aortic valve area (AVAI) by ultrasonography performed 10 days after surgery. The relationship between indexed bioprosthetic valve orifice area calculated from internal diameter (GOAI) and AVAI was evaluated. If there was a significant relationship between GOAI and AVAI, maximum body surface area (BSA) to obtain AVAI≥0.85 cm2/m2 was estimated from 99% reliable interval of regression line. Results : Age, gender, and BSA did not differ between the two groups. There was no intraoperative valve dysfunction in Magna ; however we experienced one patient with severe aortic regurgitation due to stent distortion by the aortic wall during surgery with the 25 mm Trifecta valve. For this patient, Trifecta was replaced with Magna intra-operatively. In the 19 mm valve, AVAI was significantly larger (1.12±0.27 cm2/m2 vs. 0.88±0.21 cm2/m2, p<0.001) and m-PG was significantly lower (8.7±2.7 mmHg vs. 17.2±6.3 mmHg, p<0.001) in Trifecta. The frequency of AVAI<0.85 cm2/m2 (24% vs. 49%, p=0.036) and the frequency of m-PG≥20 mmHg (0% vs. 26%, p=0.006) were significantly less in Trifecta. There was significant relationship between GOAI and AVAI in both valves. Maximum BSA to obtain AVAI ≥0.85 cm2/m2 was estimated as 1.35 m2 in Magna and 1.50 m2 in Trifecta. In the 21 mm valve, AVAI was significantly larger (1.14±0.23 cm2/m2 vs. 0.92±0.22 cm2/m2, p<0.001) and m-PG was significantly lower (7.8±3.2 mmHg vs. 14.6±4.7 mmHg, p<0.001) in Trifecta. The frequency of AVAI<0.85 cm2/m2 was significantly less in Trifecta (11% vs. 42%, p=0.002) ; however, the frequency of m-PG≥20 mmHg did not differ significantly. There was a significant relationship between GOAI and AVAI in Magna and Trifecta. Maximum BSA to obtain AVAI ≥0.85 cm2/m2 was estimated as 1.49 m2 in Magna and 1.70 m2 in Trifecta. In the 23 and 25 mm valves, AVAI was significantly larger and m-PG was significantly lower in Trifecta. However neither the frequency of AVAI<0.85 cm2/m2 nor m-PG≥20 mmHg differed between the two valves. There was one early (27 months after surgery) SVD due to leaflet tear in Trifecta and two SVDs due to leaflet calcification more than 10 years after surgery in Magna. Conclusion : For Trifecta implantation, valve size selection seemed to be important and larger valves should be avoided with narrow ST junctions. Selection of 19 and 21 mm Magna valves should be limited for the patient with a BSA less than 1.35 and 1.49 m2 respectively. In Trifecta, early SVD might occur and careful follow-up is necessary.

Preventive Effect of Tolvaptan on Pleural Effusion after Cardiac Valvular Surgery / 日本心臓血管外科学会雑誌

Background : Post-operative fluid management after cardiac valvular surgery is very important. In our institute, carperitide 0.0125 γ was started during surgery and oral furosemide 20-40 mg/day and spironolactone 25 mg/day were started at post-operative day (POD) 1 as the standard therapy. Tolvaptan, vasopressin V2 receptor antagonist, was started when fluid retention such as pleural effusion occurred. With this strategy, the frequency of pleural drainage was more than 40%. Therefore we changed our standard therapy in February 2018. In this new standard therapy, carperitide (0.0125 γ) was started and maintained until oral intake became possible and tolvaptan 7.5 mg was started with furosemide 20 mg and spironolactone 25 mg as oral medicine usually at POD 1. In this study, whether tolvaptan prevents pleural effusion or not after cardiac surgery was examined. Subjects and Methods : Sixty-four patients were operated during February 2017 and December 2018 were included in this study. Thirty-two patients operated in the period until January 2018 served as control group and were compared with 32 patients for whom tolvaptan was started on POD 1 (tolvaptan group). Results : There was no significant difference between two groups for background, operative procedure, operation time, cardiopulmonary bypass time, aortic cross clamp time and fluid balance during procedure. Tolvaptan was given to all patients in the tolvaptan group and in 22% of patients in the control group. Oral furosemide dose (tolvaptan group 21±5 mg/day, control group 31±20 mg/day, p＝0.0112), and the frequency of patients with intravenous furosemide administration (tolvaptan group 9%, control group 44%, p＝0.0038) were significantly less in tolvaptan group. In the tolvaptan group, intravenous furosemide administrated only once in all patients, whereas the frequency of intravenous furosemide administration was 1-32 times, average 6.6 times in control group. Tolvaptan was stopped within 1 week because of too much urination in two patients and the elevation of liver enzyme in two patients without any adverse effects. Post-operative urination volume until POD 5 did not differ. In both groups, body weight increased at POD 1 and 2 and returned to pre-operative weight at POD 3. Pleural effusion was significantly less in the tolvaptan group at POD 3 (tolvaptan group : none 66%, small amount 22%, moderate amount 3%, drain tube inserted 9%, control group : none 16%, small amount 34%, moderate amount 13%, drain tube inserted 38%, p＝0.0003), at POD 7 (tolvaptan group : none 72%, small amount 28%, vs., control group : none 47%, small amount 19%, moderate amount 22%, drain tube inserted 13%, p＝0.0041) and at discharge (tolvaptan group : none 94%, small amount 6%, vs., control group : none 69%, small amount 22%, moderate amount 9%, p＝0.0301). The frequency of pleural drainage was also less in the tolvaptan group (tolvaptan group 9.4%, control group 44%, p＝0.0038). Conclusion : After cardiac valvular surgery, tolvaptan started at POD 1 is very effective to reduce the frequency of pleural effusion and pleural drainage, and careful checking for too much urination and the elevation of liver enzymes is mandatory.

Brachiocephalic Artery Cannulation for Patients with Diseased Ascending Aorta / 日本心臓血管外科学会雑誌

<p><b>Objective</b> : The ascending aortic cannulation (Ao-C) is the routine procedure for cardiopulmonary bypass (CPB) in our hospital. However, for patients with diseased ascending aorta, such as severely calcified aorta, dissected or aneurysmal aorta, we used brachiocephalic artery (BCA) cannulation. The effectiveness and simplicity of BCA cannulation was evaluated. <b>Methods</b> : For patients with diseased ascending aorta, BCA was cannulated when the diameter of BCA is larger than 10 mm and is free from calcification, since January 2013. There were 62 patients who underwent aortic valve replacement (AVR) for aortic valve stenosis and BCA cannulation was applied for 11 patients. Standard Ao-C was used for remaining 51 patients. There were 44 patients with dissected or aneurysmal ascending aorta and BCA cannulation was applied for 7 patients, axillary artery perfusion was used for 15 patients and standard Ao-C was used for 22 patients. Consciousness level at the time of awaking from general anesthesia and any complication related with BCA cannulation was evaluated for the effectiveness. Simplicity was evaluated by the time required to establish CPB after skin incision. <b>Results</b> : In AVR patients, there was 1 patient with delayed consciousness level recovery with BCA cannulation and this patient was found to have cerebral infarction by CT. Intraoperative aortic dissection, probably due to BCA cannulation was observed in 1 patient, very old fragile and long period steroid user. In diseased ascending aorta patients, no patient suffered neurological accident nor any complication due to cannulation. In AVR patients, the time required to establish CPB after skin incision was 51+/-9 min in BCA cannulation and 47+/-10 min in Ao-C patients (<i>p</i>=0.34). In diseased ascending aorta patients, the time required to establish CPB after skin incision was 49+/-49 min in BCA cannulation and 51+/-16 min (<i>p</i>=0.82). <b>Conclusion</b> : BCA cannulation is a very simple and safe technique to establish CPB for patients with diseased ascending aorta. However great care should be taken, and BCA cannulation should be avoided for the long term steroid users or patients with connective tissue disease.</p>

Right Atrial Approach for Surgical Repair of a Posterior Postinfarction Ventricular Septal Defect / 日本心臓血管外科学会雑誌

We performed transatrial repair of postinfarction posterior ventricular septal defect (VSP) in a 69-year-old man who was transferred to our hospital with a diagnosis of posterior acute myocardial infarction and VSP. Coronary angiogram revealed total occlusion of the right coronary artery at #3 and 75% stenosis of the left circumflex artery at #13. UCG revealed the ventricular septal defect on the posterior ventricular septum without LV wall motion abnormality. Surgical repair was planned around 3-4 weeks later because his hemodynamic state was stable without inotropes nor IABP support upon arrival. Under general anesthesia, standard median sternotomy was performed and cardiopulmonary bypass was established with the ascending aorta and bicaval cannulation. Cardiac arrest was achieved with antegrade cold crystalloid cardioplegic solution and an oblique right atrial incision was made. The VSP was visualized via the tricuspid valve. The location of VSP was confirmed with saline injection from the LA vent line. VSP was closed with two patches, consist of a Teflon felt and a bovine pericardial patch, from the left and right ventricle side with six 4-0 polypropylene mattress sutures. Also coronary artery bypass for LCx was performed with a saphenous vein graft. The postoperative course was uneventful. There was no residual ventricular septal shunt and LV function was normal by UCG. Right atrial approach for surgical repair seemed to be useful for posterior VSP.

The Effectiveness of Left Side Pericardiotomy in Off-Pump Coronary Artery Bypass Grafting / 日本心臓血管外科学会雑誌

In off-pump coronary artery bypass grafting (OPCAB), adequate exposure under stable hemodynamic condition is mandatory. We introduced left side pericardiotomy to expose the left anterior descending artery without lifting up the ventricle in 2008. With this pericardiotomy approach, the exposure of the circumflex and right coronary artery territory became easier and OPCAB with left side pericardiotomy was compared with OPCAB with midline pericardiotomy. From 2004 to 2011, 194 elective first time coronary artery bypass grafting (CABG) were performed in our hospital. Before 2008, 62 patients underwent OPCAB with midline pericardiotomy which constituted 61% of the CABG in that period. After 2008, the pericardium was dissected on the left side and a small pericadiotomy was made on the left side of the main pulmonary artery. This incision was then extended to the apex. With this pericardiotomy, only two patients underwent CABG with cardiopulmonary bypass (one patient with 15% left ventricle ejection fraction and one more patient who developed acute coronary syndrome during anesthesia induction). Thus 91 out of 93 patients underwent OPCAB (98%) (Group L). In Group L, old myocardial infarction and unstable angina patients were frequent. The frequency of the patients with left ventricular ejection fraction less than 40% tended to be more in Group L. The operation time was significantly shorter in Group L (Group M 305±71 min, Group L 223±54, <i>p</i><0.0001) and the number of distal anastomoses number was significantly more in Group L (Group M 2.3±0.7, Group L 2.8±1.0, <i>p</i><0.0001). Blood pressure during left circumflex coronary artery and right coronary artery anastomosis was significantly higher in Group L, and even continuous dopamine infusion requirement was significantly less in Group L (92% in Group M, 13% in Group L, <i>p</i><0.001) among the patients with left ventricle ejection fraction less than 60%. There was only 1 hospital death in Group M. Postoperative maximum CK-MB was significantly lower in Group L (Group M 48±107 IU/<i>l</i>, Group L 13±16 IU/<i>l</i>, <i>p</i>=0.005) and the patients with CK-MB more than 12 IU/<i>l </i>was significantly frequent in Group M (Group M 73%, Group L 33%, <i>p</i><0.0001). Postoperative ICU and hospital stay period was significantly shorter in Group L (ICU stay : Group M 3.4±2.3 days, Group L 2.0±1.4 days, <i>p</i><0.0001, hospital stay : Group M 27±21 days, Group L 16±7 days, <i>p</i><0.0001). The patency of the graft to the left anterior descending artery did not differ significantly (Group M 94%, Group L 99%), however the patencies of the grafts to left circumflex artery and right coronary artery were significantly better in Group L (left circumflex artery : Group M 75%, Group L 98%, <i>p</i>=0.001, right coronary artery : Group M 81%, Group L 98%, <i>p</i>=0.014). Left side pericardiotomy seemed to be useful because OPCAB with left side pericardiotomy yielded shorter operation time, less myocardial enzyme release, improved postoperative recovery and better patency of graft to the left circumflex and right coronary artery.

A Case of Thoracic Endovascular Aortic Repair with Fenestrated Stentgraft for Ulcer-like Projection of the Proximal Anastomosis after Total Arch Replacement for Acute Type A Aortic Dissection after Open Heart Surgery / 日本心臓血管外科学会雑誌

We report a case of thoracic endovascular aortic repair (TEVAR) with a fenestrated stent graft for ulcer-like projection (ULP) of the proximal anastomosis after total arch replacement (TAR) for acute type A aortic dissection (DAA). A 73-year-old woman with a history of surgical resection of a left atrial myxoma in January 2009 underwent TAR for DAA in November 2011. The contrast enhanced CT (CE-CT) 72 days after TAR revealed two ULPs anterior and posterior to the proximal anastomosis. Surgical repair would be difficult because of the history of cardiac and aortic surgery, therefore TEVAR with a fenestrated stent graft was performed. The postoperative course was uneventful and she was discharged on the 8th postoperative day. The CE-CT 3 months after TEVAR showed almost completely thrombosed ULPs. Endovascular repair with fenestrated stent graft for the proximal anastomotic ULP can be a useful and effective treatment.

Effect of a Renal Protection Protocol on the Renal Function after Endovascular Aortic Aneurysm Repair / 日本心臓血管外科学会雑誌

Endovascular aortic aneurysm repair using stent graft (SG) for both thoracic and abdominal aortic aneurysms (SG therapy) rapidly became widespread in Japan because of its relatively low invasiveness. Pre- and postoperative contrast enhanced CT are mandatory in SG therapy and angiography is required during SG therapy. Therefore contrast induced nephropathy (CIN) might occur after SG therapy. In our hospital, a renal protection protocol (oral <i>N</i>-acetylcysteine, perioperative normal saline infusion and bicarbonate infusion during SG therapy) was introduced in June 2010. In this report, the effect of the renal protection protocol on renal function after SG therapy was evaluated. During May 2008 and March 2012, 229 patients underwent SG therapy in our hospital. Serum creatinine (CRTN) was higher than 1.5 mg/dl and estimated glomerular filtration rate (eGFR) was less than 50 ml/min/1.73 m² in 26 patients. In these 26 patients, the renal protection protocol was applied in 15 patients (group P) and group P was compared with the 11 patients without renal protection protocol (group N). Also the relationship between CIN occurrence and preoperative renal function was evaluated in 192 patients who did not receive the renal protection protocol. CIN was defined as more than 25% or 0.5 mg/dl increase of CRTN based on the European Guidelines. As renal protection protocol, <i>N</i>-acetylcysteine (600 mg) was given 4 times every 12 h. Normal saline infusion was started on the evening of the day before surgery at the rate of 50 ml/h and was continued until 1h before surgery. Sodium bicarbonate solution (151 mEq/<i>l</i>) was started 1 h before surgery at the rate of 180 ml/h and the infusion rate was decreased to 60 ml/h during surgery. After surgery, 1,000 ml of normal saline was given at a rate of 60 ml/h. In group N, CRTN increased 1 and 3 days after SG therapy and returned to baseline level 6 days after SG therapy. On the other hand, CRTN was lower than baseline after SG therapy in group P. At 3 days after SG therapy, the percent change of CRTN component with baseline level was significantly lower in group P (14.5±19.1% in group N, -3.7±15.8% in group P, <i>p</i>=0.014). CIN occurrence tended to be more in group N (45% in group N, 7% in group P, <i>p</i>=0.054). Among the 192 patients without the renal protection protocol, CIN occurred in 16 patients (29.1%) out of 55 patients with preoperative CRTN≥1.0 mg/dl and eGFR≤50 ml/min/1.73 m², however CIN occurred in only 1 patient (0.7%) among 137 patients with preoperative renal function out of this range (<i>p</i><0.001). Renal protection protocol seemed to be effective to prevent CIN after SG therapy. Renal protection might be useful for patients with a CRTN≧1.0 mg/dl and eGFR≦50 ml/min/1.73 m².

Surgeon-Modified Zenith Stent Graft System for Endovascular Repair of Abdominal Aortic Aneurysm with Short Proximal Neck / 日本心臓血管外科学会雑誌

Endovascular repair for abdominal aortic aneurysm (EVAR) has become widespread in Japan because of its low invasiveness. However adequate proximal neck length is required for EVAR. Unfortunately the surgical mortality of para-renal aortic aneurysm cases has been higher than that of infrarenal aortic aneurysm cases, especially in high-risk patients. A manufacture-modified fenestrated Zenith stent graft system has already been developed, however this new device is not yet available in Japan. Furthermore this device could not be used in an emergency situation because it takes 2-3 weeks for preparation. Therefore we introduced a surgeon-modified fenestrated Zenith stent graft (fenestrated Zenith) system in December 2010 for patients with a proximal neck length of 5-10 mm. The fenestrated Zenith was not indicated if the supra-renal angle and proximal neck angle exceeded 35°. From May 2007 to February 2012, abdominal aortic aneurysms (AAA) with a short neck were repaired with fenestrated Zenith in 11 high-risk patients (group Fene), and AAAs with a proximal neck length of more than 15 mm were repaired with a standard Zenith in 43 patients (group IFU). There were two ruptured AAA in the Fene group. Proximal neck length was significantly shorter in the Fene group (5.5±1.4 mm in the Fene group, 26.4±9.5 mm in the IFU group, <i>p</i><0.0001) and proximal neck angle was significantly less in the Fene group (20±13° in the Fene group, 36±18° in the IFU group, <i>p</i>=0.008). The Zenith stentgraft system was deployed successfully in all patients. The frequency of type Ia endoleak detected by angiography after stent graft deployment and balloon attachment did not differ significantly (36% in the Fene group 26% in the IFU group, <i>p</i>=0.475) and the frequency of Palmaz stent requirement for type Ia endoleak which persisted after 10 min of additional balloon attachment also did not differ significantly (27% in Fene group, 9% in IFU group). All fenestrated renal arteries were shown to be patent by angiography. There was no hospital death despite 2 cases of ruptured AAA, nor were these major complications in either group. Serum creatinine levels at 1, 3, 6 and 30 days after EVAR did not differ significantly between the 2 groups. In 9 out of 11 patients, only type II endoleaks were detected and aneurysm shrinkage tended to be more in Fene group (9.9±5.7 mm in Fene group, 5.4±6.1 mm in IFU group, <i>p</i>=0.062) on enhanced CT 6 months after EVAR. Also all fenestrated renal arteries were patent in these 9 patients. The surgeon-modified fenestrated Zenith system seemed to be effective for AAA patients with short proximal necks, but long term follow up is mandatory.

Endovascular Abdominal Aortic Aneurysm Repair with on Excluder for Proximal Neck Angulation / 日本心臓血管外科学会雑誌

The results of endovascular abdominal aortic aneurysm repair (EVAR) for severe neck angulation with an Excluder were evaluated. We performed EVAR in 51 patients, using an Excluder, from September 2007 to September 2011. The angle between proximal neck and the aneurysm (Angle) was less than 61° in 31 patients (Group I), 61-90° in 13 patients (Group II) and more than 90° in 7 patients (Group III). In Groups I and II, the angled proximal neck was straightened with a stiff guide wire and a Trunkipsilateral device was deployed slowly (aortic modification technique). In Group III, the device modification technique was applied. In this technique, the stiff guide wire was inserted in the aortic root. The Trunkipsilateral device was bent to the contra lateral limb side and was inserted into the aorta. The stiff guide wire was pushed in with a fulcrum at the aortic valve. This procedure resulted in bending of the wire and the trunk-ipsilteral device became parallel to the proximal neck. The renal artery position was confirmed on angiographys and the main body was deployed slowly. We performed angiography after planned device deployment to evaluated Type Ia endoleak, and if it was observed, an additional procedure such as Aortic Extender or Palmaz XL stent deployment was performed and the Type Ia endoleak was evaluated during the procedure by completion angiography. The Angle change was measured by enhanced CT at 7 days and 6 month after EVAR. The Angle were 97-137° in Group III. The frequency of Type Ia endoleak after planned device deployment (35% in Group I, 55% in Group II and 17% in Group III), additional procedure for Type Ia endoleak (29% in Group I, 23% in Group II and 14% in Group III) and Type Ia endoleak by completion angiography (0% in Group I, 8% in Group II and 14% in Group III) did not differ significantly between the 3 groups. When Group I was sub divided into those with Angle less than 41° (Group Ia, 15 patients) and those with an Angle from 41 to 60° (Group Ib, 16 patients), Type Ia endoleak after planned device deployment (18% in Group Ia, 63% in Group Ib) was significantly more frequent in Group Ib and the additional procedure for Type Ia endoleak (7% in Group Ia, 50% in Group Ib) was more frequent in Group Ib. The Angle significantly decreased 7 days after EVAR and did not change thereafter in all 3 groups. EVAR with an Excluder for severe neck angulation was feasible by device modification with the bending technique. This technique might be useful for patients with an Angle of more than 41°.

Endovascular Therapy for Aortic Emergency / 日本心臓血管外科学会雑誌

Aortic emergencies are surgically challenging and the mortality rate remains high. Since December 2003, we have performed endovascular treatment with a stent graft (EVT) in 15 cases of aortic emergency, including 8 cases of aortic rupture or traumatic aortic disruption, 1 case of traumatic iliac artery disruption, 3 cases of aortobronchial fistula (ABF), and 3 cases of aortoenteric fistula (AEF). In 9 cases of aortic rupture and traumatic aortic or iliac disruption, 1 patient died due to traumatic cerebral hemorrhage, but the remaining 8 patients were discharged without complications. While hemoptysis was resolved in all 3 patients with ABF, 1 patient with primary ABF died due to pneumonia, and 1 patient with secondary ABF died due to multiple organ failure. Furthermore, 1 patient with primary AEF progressed well without any evidence of postprocedural stent graft infection. In 2 patients with secondary AEF, both required secondary surgical graft excision, and 1 of these died due to the recurrence of infection. EVT has shown good results in hemostasis for aortic emergency. However, if a source of infection persists, secondary surgical intervention is required in some cases. Commercially available endovascular stent graft make it possible to treat tortuous segments of the thoracic aorta and the abdominal aorta, therefore we expect there to be more patients with aortic emergencies who require EVT.

Endovascular Treatment for Ruptured Infected Descending Thoracic Aortic Aneurysm / 日本心臓血管外科学会雑誌

Endovascular treatment with a hand-made stent graft was performed successfully for a ruptured infected descending thoracic aortic aneurysm in the patient under chronic hemodialysis. The patient was a 77-year-old man and hemodialysis was continued with a double lumen catheter because of internal shunt occlusion since November 2005. He developed MRSA sepsis around January 2006 and was treated with antibiotics. Left pleural effusion was drained on March 1st, 2006. Initially left effusion from the left side was fluently bloody, however it became bloody the next day. Enhanced CT revealed the descending thoracic aortic aneurysm and left hemo-thorax. He was referred to our service, and was transportedly via ambulance. Home made stent graft was deployed under local anesthesia 2h after the arrival. His hemodynamic status became stable and left pleural hematoma was removed by video assisted thoracic surgery 2 days after endovascular treatment. His postoperative course was uneventful and antibiotics were continued 4 weeks. He is doing well without any recurrence of infection 23 month after the surgery. Careful follow up is mandatory, and endovascular treatment would be useful for the ruptured infected aortic aneurysm.

Comparison of the Impact of Skeletonization and Semi-Skeletonization by Harmonic Scalpel on the Left Internal Thoracic Artery Graft Flow / 日本心臓血管外科学会雑誌

Intraoperative graft flow, measured by a BF 1001 CardioMed Flowmeter and the postoperative graft diameter of the left internal thoracic artery, measured by coronary angiogram were compared in the semi-skeletonization method group (23 patients) and the skeletonization method group (29 patients). There was no significant difference between 2 groups in terms of age, gender, body surface area, diabetes mellitus, LAD lesion, preoperative ejection fraction, operation time, cardiopulmonary bypass time and aortic cross-clamp time. Graft flow was significantly larger in the skeletonization method group than in the semi-skeletonization method group (50.4±21.7ml/min vs 36.9±12.8ml/min, <i>p</i>=0.019). However graft diameter did not differ significantly between the 2 methods (2.46±0.44mm in the skeletonization method group and 2.38±0.42mm in the semi-skeletonization method, <i>p</i>=0.991). These results suggested that left internal thoracic artery could be used longer by either 2 methods than by pedicled harvesting technique, however skeletonization method caused less spasm during graft harvesting than the semi-skeletonization method.