Extended Operation for Thoracic Malignancies Invading Superior Vena Cava And/or Innominate Vein
Zhai Chun-bo,Yu Feng-lei,Wang Xiang,Liu Wen-liang,Hu Jian-guo,Lu Wei-dong,Zhang Ming
DOI: https://doi.org/10.1097/00029330-200707020-00015
2007-01-01
Abstract:Patients with thoracic malignancies invading the superior vena cava (SVC) and innominate vein that cause venous flow obstruction have a poor quality of life and short survival time. Dartevelle et al1 reported that the 3-year survival rates were 27% for patients with mediastinal malignancies and non-small cell lung cancer. Recently, some surgeons’ experience suggested that the results of extensive resection and reconstruction of the SVC and innominate vein were excellent.2–4 From May 2003 to December 2005, radical resection and reconstruction of the SVC and/or innominate vein were performed in a group of patients for mediastinal or lung tumors invading the superior vena cava and/or innominate vein. We reported our experience in this retrospective study to explore the operative method, the optimal vascular graft, and the benefits of extended resection and reconstruction of SVC and/or innominate vein for invasive thoracic tumors. CLINICAL DATA Patients Twenty-one patients (11 men and 10 women) who had undergone resection of the SVC and/or innominate vein were included in the study. The median age of the patients was 39 years (age range, 15–71 years). The primary symptoms were edema (52.4%, 11/21), nonproductive cough (38.1%, 8/21) and chest wall intumescence (9.5%, 2/21). The pathological diagnosis were malignant thymoma (52.4%, 11/21), lung adenocarcinoma (19.0%, 4/21), malignant lymphoma (Hodgkin's disease, 4.8%, 1/21 and non-Hodgkin's lymphoma, 14.3%, 3/21), postoperative breast cancer (5.0%, 1/21) and teratoma (5.0%, 1/21). The initial several patients received centesis and mediastinoscope biopsy, however, the masses were so giant that it was difficult to acquire a definite pathological diagnosis by centesis and mediastinoscope biopsy. So all but one patient did not receive preoperative radiotherapy or chemotherapy and the rest of patients did not continue to perform preoperative pathological biopsy. Information on vessels and nerves invaded The left innominate vein was involveded in four cases. The SVC and innominate vein were involveded in 13 cases. The SVC, innominate vein and subclavian artery were involved in one case. The SVC, innominate vein and artery were involved in three cases. One side of phrenic nerve was involved in 11 cases. The double phrenic nerves were involved in one case. One side of vagus nerve was involved in seven cases. The pulmonary lobes were invaded in 12 cases. The chest walls were invaded in two cases. Auxiliary examination Chest plain film and enhancement computed tomography (CT) were parts of patient's routine examinations. Eleven cases were examined by magnetic resonance imaging (MRI); 3 were examined by transthoracic color Doppler transonogram; 1 by phlebography and 4 by bronchofibroscope (BFS). Procedure The routine median sternotomy was used, however, the incision may have been extended toward the ipsilateral neck of the tumor and the ipsilateral clavicle may have been cut off if necessary. Central venous pressure (CVP) was monitored by left internal jugular vein and blood pressure (BP) was monitored by left radial artery. Transfusion was performed by a vein of lower extremity. For cases receiving resection of half of the innominate vein only, anastomosis was between the innominate vein and the right cardiac atrium or SVC. For cases receiving resection of the SVC and innominate vein, the vascular grafts were implanted to connect the right innominate vein and SVC when the connection between the left innominate vein and right cardiac atrium had been finished. For cases receiving replacement of bifurcated vascular graft, anastomosis between one branch of the bifurcated vascular graft and the left innominate vein was performed followed by anastomosis between the SVC and bifurcated vascular graft. Finally, an anastomosis between the right innominate vein and the other branch of the bifurcated vascular graft was performed. For cases receiving replacement of a self-made bifurcated vascular graft, one straight Gore-Tex vascular graft with external support was transected along about 3/4 of the perimeter. Grafts were fixed in place by 5/0 Prolene suture (Prolene®BLUE Polypropylene Monofilament Non-Adsorbable Suture, Ethicon Inc., USA) at the included angle of the incision to avoid continuous extension of the incision. Next, the anastomosis between the left innominate vein and half of the vascular graft was performed, then, the anastomosis between the superior vena cava and the self-made bifurcated vascular graft was performed. Lastly, the anastomosis between the right innominate vein and the other half of the vascular graft was performed. Data on the number of variables, including demographics symptoms, signs and any preoperative therapy were collected. Tumor histological features and operative details were also collected. All complications were noted, along with overall survival duration. All data were handled by SPSS13.0. Survival curves were calculated using Kplan-Meier method. The statistical test to compare survival was the Log rank test. A P<0.05 was considered statistically significant. No intraoperative death was observed, however, the cyanosis of the face and edema of the bulbar conjunctive in different degrees, were found in each case. The average amount of blood transfused was 2243.5 ml (ranging 350–5800 ml). The average operative time was 462.6 minutes (ranging 290–800 minutes). The average staying days in the intensive care unit (ICU) were (4.0±1.8) days. The average days depending on a ventilator were (1.7±1.5) days. Sequelae of nervous system was not found. Eleven of sixteen patients who underwent operation with curative intent had complete resection (R0), while the rest of five patients had residual microscopic disease only (R1 resection). Of the five patients who underwent operation with incomplete resection, three patients had residual microscopic disease (R1 resection), while two patients had residual macroscopic disease (R2 resection) and were marked with a titanic clip. One of the two patients who were marked with titanic clips died 15 days after the operation. The overall classification of resection of the tumor from venous structures was R0 in 11 (52.4%), R1 in 8 (38.1%), and R2 in 2 (9.5%). One of the twelve patients who received resection of phrenic nerve received resection of both phrenic nerves and fixation of the diaphragm. Seven patients received resection of vagus nerve of the wounded side. Twelve patients received a pulmonary lobectomy. Two patients received a reconstruction of the chest wall. The vascular grafts included autologous pericardium (3 of 21, Fig. 1), bovine jugular vein (laboratory of department of Cardio-Thoracic Surgery, Second Xiangya Hospital, Central South University; 2 of 21, Fig. 1), Gore-Tex graft (FEP Ringed Gore-tex® Vascular Graft; 12 of 21; or Bifurcated Gore-tex® Stretch Vascular Graft, W.L. Gore & Associates, Arizona; 3 of 21, Fig. 2), self-made bifurcated vascular graft by Gore-tex straight vascular graft (1 of 21, Fig. 3). One patient received resection of artery media and the defect was wraped up by artificial vascular flake.Fig. 1.: The primed bovine jugular vein(↑) on the left of the blood vessel forceps. The blue-black on the right of the blood vessel forceps was autologous pericardial vascular graft(⇑).Fig. 2.: Superior vena cava and innominate veins were replaced by Gore-tex bifurcated stretch vascular graft(↑). Aorta(⇑) was encased by vascular graft after the resection of arterial adventitia and media.Fig. 3.: Self-made bifurcated vascular graft by one straight Gore -Tex vascular graft with external support.The postoperative complications included pneumonia (1), acute respiratory distress (1), wound infection (3) and hoarseness (3). The median follow-up duration was 20 months. The median survival was (28.0±5.3) months. The median survival duration has not yet been reached for the patients who underwent complete resection and was (13.0±8.7) months for the patients who underwent incomplete resection. The difference in survival between complete resection and incomplete resection was statistically significant (P<0.001). The overall 1- and 2-year survival rate was 71% and 33%. The 1-year survival rates of patients with non-small cell lung cancer, thymoma and malignant lymphoma was 0, 83% and 100% respectively. The overall difference in survival between lung cancer, thymoma and lymphoma patients was statistically significant (P=0.001, Fig. 4). The survival comparison between thymoma and lung cancer was statistically significant (P=0.001). The survival comparison between lymphoma and lung cancer was statistically significant (P=0.021). The survival comparison between thymoma and lymphoma was not statistically significant (P=0.658).Fig 4.: The survival curves of the lung cancer, lymphoma and thymoma.Eighteen patients received postoperative chemotherapy and/or radiotherapy. Artificial graft occlusion occurred in two of the four patients who received postoperative phlebography. One occlusion of an artificial graft occurred one month after the operation and the other occlusion of an autologous pericardial graft occurred nine months after the operation. Neither of them had serious consequences because of the established collateral circulation. The cross sectional area of the bovine jugular vein of one patient was confirmed by follow-up computed tomography observations to gradually decrease at the third, sixth, ninth and twelfth month after the operation. DISCUSSION Extended resection and reconstruction of the SVC and/or innominate vein for thoracic malignances was not a new procedure. Especially, the resection of non-small lung cancer invading the SVC had been performed by many surgeons5–7 and the results were exciting. Thus, a radically surgical approach for resection of these tumors can be justified in a selected group of patients. Patient selection and surgical contraindications The patients reviewed in this study were a selected group. The decision to proceed with surgery was made on an individual basis after compositive assessment. Factors addressed included biologic features of the tumor (i.e. the tumor's aggressiveness, reaction to chemotherapy and resectability), symptomatology and physiologic status of the patient. The contraindications of surgery include: (1) Metastasis found beyond the thorax. (2) Tumor has invaded significantly intrathoracic organs or tissure, such as the bilateral phrenic nerve, bilateral vagus, one side of recurrent laryngeal nerve or implantated extensively on the heart and/or diaphragma, or tumor invading the aorta which leads to aortic dissecting aneurysm. However, regional intrusion of the trunk of the aorta and pulmonary artery, and compression of the trachea and/or esophagus are not surgical contraindications. (3) Patients who have a combination of serious damage to their circulatory system, respiratory system, hematological system and the functions of liver and kidney. Vascular graft selection The vascular grafts included autologous pericardium, primed bovine jugular vein, Gore-tex graft with removeable external support and bifurcated stretch vascular graft. The follow-up results indicated that the biological graft of one case had a trend to become gradually stenotic, biological grafts were not the preferred vascular substitute except for emergency operations. The self-made bifurcated stretch vascular or product of bifurcated stretch vascular can recover the blood flow of the SVC and innominate vein based on anatomy and function. So, both of them are preferred. Ambi-operative treatment Steady and safe anesthesia was required for the success of the operation. Drugs were transported by vein of lower extremity. Urethral catheterization, measurement of blood pressure by the radial artery puncture, and measurement of central venous pressure by the internal jugular vein puncture were the routine procedures. Heparin (0.5 mg/kg) was dripped during the operation and the first day after the operation to prevent thrombogenesis. From the beginning of the second day, warfarin was used to prevent thrombogenesis in stead of Heparin. Prothrombin time (PT) was controlled between 18 to 22 seconds. Warfarin was used to prevent thrombogenesis for 3 months for patients with biologic vascular grafts, and forever for patients with artificial vascular grafts. Hemostatics was forbidden to inject through veins of the upper extremities. Intraoperative management Sternotomy can sufficiently expose the anterior mediastinum, root of the neck, right thorax and left upper lobes of the lung. This incision is suitable for resection of anterior mediastinal malignances, reconstruction of the SVC and innominate vein, resection of right lung cancer and left mediastinal lymph nodes, dissociation of the root of great vessels, control of bleeding and construction of a bypass. A thoroughgoing dissociation of the malignance and stopping bleeding elaboratively were also very important. The dissection of the tumor may be considered when the thoroughgoing dissociation can not be achieved. The arterial adventitia and/or media suffering from invasion can be resected when systolic blood pressure was controlled to 60–70 mmHg. To prevent rupture and aneurysm the defect should be reinforced by an artificial vascular flake. During the procedure of anastomosis heparin solution was used to irrigate regions of stoma and graft to prevent thrombosis. Doty et al8 suggested the interrupted time should be less than 45 minutes. Some surgeons9 also suggested that the central venous pressure should be lower than 45 mmHg. We agreed with them and exsanguination was performed when the central venous pressure exceeded 45 mmHg. The intraoperative injury of the phrenic nerve and recurrent laryngeal nerve should be prevented. If the phrenic nerve and recurrent laryngeal have been invaded, the contralateral nerve should be protected from injury. Resection of the bilateral phrenic nerve, vagus and sometimes one side of recurrent laryngeal nerve will produce serious problem of respiratory efficiency. One patient with resection of the bilateral phrenic nerve stayed in the intensive care unit for 7 days and still complained little dyspnea now. Postoperative care and management These patients experienced a large wound and long operative times, so their hemodynamics should be monitored rigorously. The time on a ventilator should be extended properly. Postoperative key treatment included supplementation with colloid, limitation of crystalloid solution, analgesia, as well as coughing up phlegm. Patients should receive chemotherapy and/or radiotherapy as soon as possible when they have recovered from the operation. Outcome In conclusion, the review of extended resection and reconstruction of SVC and/or innominate vein showed that no morbidity or mobility rates increased and the procedure was safe. Patients suffering from complete clearance of their tumor had a better survival rate and quality of life. Although only a small group of patients were studied, it showed that long term survival was achieved at completion of surgery although the survival of lung cancer was lower than the survival seen for other thoracic malignaces. For patients with lung cancer, the injury of radical operation was giant while the effect was poor, and this kind of operation was not suitable for them. In addition, we found that the bifurcated vascular graft should be preferred. A controversial question is whether this operation is suitable for patients with primary mediastinal malignant lymphoma. Most scholars suggested that lymphoma is systematic disease and chemotherapy or radiotherapy were the major therapy approach. We agreed that those patients who were either insensitive to chemotherapy or radiotherapy or their SVC or trachea was oppressed by tumor should receive the operation10. However, it is difficult to a definite diagnosis, sometimes. So preoperative chemotherapy or radiotherapy can not be performed. Acknowledgement: We thank Dr. ZHOU Wen-wu for his help in English translation.