|
|
||||||||
Division of Thoracic Surgery, Massachusetts General Hospital, Boston, Massachusetts
Accepted for publication April 16, 2009.
* Address correspondence to Dr Lanuti, 55 Fruit St, Blake 1570, Boston, MA 01748 (Email: mlanuti{at}partners.org).
Presented at the Poster Session of the Forty-fifth Annual Meeting of The Society of Thoracic Surgeons, San Francisco, CA, Jan 26–28, 2009.
| Abstract |
|---|
|
|
|---|
Methods: Patients undergoing SVC resection from 1997 to 2007 for surgical management of benign and invasive neoplasms were retrospectively reviewed.
Results: We identified 19 patients requiring SVC resection. Malignant disease was resected in 17: lung cancer in 9 and mediastinal malignancy in 8. Two patients (10%) with benign processes required reconstruction for chronic SVC syndrome. Ringed Gore-Tex conduit (W. L. Gore and Associates, Flagstaff, AZ) was used for 12 reconstructions (63%) of the SVC, and 7 patients underwent primary closure or autologous pericardial patch repair. Preoperative chemoradiotherapy was administered to 9 patients (53%). There was one perioperative death (5%). Major postoperative morbidities included atrial fibrillation in 5, stroke in 2, respiratory failure in 3, myocardial infarction in 1, and Horner syndrome in 1. Median survival for the entire cohort was 45.5 months (range, 0.2 to 147 months), with a mean follow-up of 45.8 months. Five-year survival probability was 30% for patients with resected lung cancer and 56% for patients with resected anterior mediastinal malignancies.
Conclusions: Resection and reconstruction may be safely performed in selected patients for benign and malignant obstruction or infiltration of the SVC. Survival and intermediate-term patency after tubular grafting of the SVC are acceptable.
| Introduction |
|---|
|
|
|---|
With regard to lung cancer, the results of tumors involving the SVC are far better than one would predict with stage IIIB NSCLC, with a historical survival of 6% to 8% [7]. The proposed 2009 revision (7th edition) of the T N M classification of lung cancer by the International Association for the Study of Lung Cancer (IASLC) has taken into account the potential resectability and different prognosis of T4 N0-1 M0 and transferred these tumors to stage IIIA [8].
The feasibility and low morbidity of extended resection for pulmonary or mediastinal tumors infiltrating the SVC are well documented, but the selection of optimal candidates for aggressive resection is less well defined. Although some prognostic factors have been reported [9], ambiguity remains. Completeness of resection, extent of SVC involvement, and the presence of mediastinal lymph node metastases are all regarded as important in the evaluation of these select patients. There is little contention about SVC resection for anterior mediastinal tumors, but SVC resection in the context of NSCLC with nodal metastasis remains controversial. This study evaluates the outcomes of a heterogeneous group of patients undergoing en bloc SVC resection as part of an R0 resection of malignant disease of the lung and anterior mediastinum.
| Material and Methods |
|---|
|
|
|---|
Patients
The study population included consecutive patients undergoing SVC resection and reconstruction for treatment of benign or malignant disease on the Thoracic Surgery Service at the Massachusetts General Hospital between January 1997 and December 2007. Retrospective medical record reviews were performed. The specific objective of the study was to evaluate the outcomes of radical resection for disease involving the SVC. The study included all patients who underwent SVC resection for pulmonary, mediastinal, or thyroid malignancies, as well as 2 patients with SVC syndrome from obstructive benign disease.
No patient required extracorporeal bypass to achieve resection. The study was inclusive of all surgical approaches that provided suitable access, including median sternotomy, cervical sternotomy, and thoracotomy. Demographic, preoperative, intraoperative, and outcome measures were recorded. Data were obtained from the medical records, including office records, anesthesia records, and in-hospital records.
Preoperative assessment included clinical examination, diagnostic imaging, pulmonary function, quantitative ventilation-perfusion scan (for pneumonectomy), bronchoscopy, and mediastinoscopy in most patients with NSCLC. Aside from diagnostic chest computed tomography, metastatic survey was performed with brain magnetic resonance imaging (MRI), bone scan, and when available beginning in about 2000, whole-body positron emission tomography (PET). Echocardiography was routinely used to fully evaluate biventricular function and rule out severe tricuspid regurgitation.
Operative Repair
The degree of SVC infiltration dictated the type of resection. Primary repair was achieved in 5 patients when less than 50% of caval circumference was resected. Suture repair was conducted with running polypropylene suture, usually by tangential placement of a partial occlusion clamp. Larger defects were repaired in 2 patients using an autologous pericardial patch (Fig 1A and B). In 12 patients in whom SVC infiltration encompassed more than 50% of the vessel, the SVC was replaced with a ringed polytetrafluoroethylene (PTFE) graft (median size, 14 mm; sometimes reconstructed using technique in Fig 1C).
|
All patients received intravenous heparin sodium (50 to 100 U/kg) 5 to 10 minutes before vascular occlusion. After intraoperative heparinization, the partial thromboplastin time (PTT) was allowed to normalize without pharmacologic reversal. Patients were selectively transitioned to an antiplatelet agent or Coumadin (Bristol-Myers Squibb, Princeton, NJ) at the surgeon's discretion. Most patients with a PTFE vascular reconstruction received anticoagulation therapy for a minimum of 3 months.
Statistical Analysis
Survival curves were constructed using the Kaplan-Meier method and compared with the log-rank test. Survival analyses were calculated with SPSS 14.0 software (SPSS Inc, Chicago, Ill). Comparisons of survival were performed for patients undergoing neoadjuvant therapy, nodal status in resected NSCLC, and tumor location.
| Results |
|---|
|
|
|---|
|
|
|
|
|
Induction chemoradiotherapy was completed in 3 of 9 patients (33%) with NSCLC and in 6 of 8 (75%) with anterior mediastinal malignancy. Adjuvant chemotherapy was administered to 5 of 9 patients (56%) with resected lung cancer. There was no significant survival difference between patients receiving induction chemoradiotherapy and those undergoing primary resection, followed by adjuvant therapy.
Partial SVC resection was performed in 7 of 19 patients (37%), and an autologous pericardial patch was used for reconstruction in 2. Ringed PTFE grafts (size range, 8 to 18 mm) were used to reconstruct 12 SVC resections (63%), including ligation of a unilateral brachiocephalic vein in 5 patients. Symptomatic acute thrombosis of the SVC repair occurred in 2 patients (10%) early in the postoperative period (Table 1). One patient underwent SVC embolectomy within 24 hours with no further sequelae, and another patient with a resected intravascular SVC lipoma (primary SVC repair) was treated with anticoagulation, with resolution of upper extremity edema. Twelve patients (63%) were discharged with Coumadin therapy. Two patients had preoperative SVC thrombosis.
Bronchial stumps after pneumonectomy were selectively covered with vascularized tissue, including pericardial fat, omental transposition, or intercostal muscle. One postoperative death occurred within 30-days in a patient with anaplastic thyroid cancer who underwent thyroidectomy with right brachiocephalic and partial SVC resection. Major complications in the 19 patients were atrial fibrillation in 5 (26%), respiratory failure in 3 (16%), stroke in 2 (11%), early SVC thrombosis in 2 (11%), and 1 patient (5%) each with Horner syndrome, postoperative myocardial infarction, and acute respiratory distress syndrome.
SVC venous occlusion time associated with 12 circumferential resections was 50 minutes or less in all patients, with one exception. The need to replace the SVC was unanticipated preoperatively in one patient, and bleeding during dissection required SVC clamping for a prolonged period. This resulted in bilateral posterior ischemic optic neuropathy. The visual loss eventually improved over many months.
| Comment |
|---|
|
|
|---|
Although this study has a small sample size and is underpowered for any meaningful statistical analysis, the presence of nodal disease (N1-2) in resected NSCLC was significant (p = 0.04) for predicting poor long-term survival compared with node-negative disease (N0). This intuitive observation was corroborated in studies by Dartevelle and colleagues [3] in 1991, Suzuki and colleagues [4] in 2004, and Spaggiari and colleagues [5] in 2007. Spaggiari and colleagues reported a 5-year survival of 52% for N0 disease compared with 21% for pathologic N2 disease. In contrast to many of the publications examining extended resection for SVC invasive malignancy, Suzuki and colleagues [4] differentiated 5-year survival with SVC invasion by direct tumor extension (36%) from those with SVC invasion by metastatic lymph nodes (6.6%). This separation may be artificial when induction therapy obscures tissue boundaries as a treatment effect. This distinction was poorly characterized in our patient series, even in the absence of neoadjuvant therapy. Several authors [4, 13, 14] have observed no correlation between T N M nodal status and survival after SVC resection for lung cancer. The conflicting data speak to statistical weakness of small sample sizes.
The intraoperative management of SVC resection and reconstruction was relatively uniform in this clinical series. All circumferential SVC resections were performed with ringed PTFE (median size, 14; range, 8 to 18 mm) without a vascular shunt. Size was determined by extent of venous resection and the nature of the distal anastomosis to conform to native brachiocephalic vein or SVC. Other conduit options include tubularized bovine pericardium, spiral saphenous vein graft [15], and more recently, cryopreserved arterial allograft [16].
The overall thrombosis rate of PTFE reconstruction of the SVC has been reported to be 14% to 24% within 3 to 5 years [3, 5]. Most thromboses were observed within the first month of resection and reconstruction. The thrombotic risk is higher when SVC revascularization occurs in the presence of venous collaterals. We observed an overall SVC thrombosis rate of 10% (2 of 19) with a median follow-up of 45 months. Both patients experienced early thrombosis despite anticoagulation within 12 to 24 hours postoperatively. One patient underwent primary suture repair of the SVC and another patient underwent reconstruction with a 12-mm ringed PTFE conduit to the right brachiocephalic vein. The small conduit may have contributed to thrombosis. Although venous Doppler studies were not routine after 30 days, no late symptomatic SVC thrombosis was observed.
Coumadin was prescribed to 12 of 19 patients (63%) at discharge, with an average anticoagulation time of 3 to 6 months. Patients who underwent partial resection with suture or pericardial patch repair were often treated solely with enteric-coated aspirin (325 mg) for up to 3 months.
Our preference was for unilateral reconstruction of the brachiocephalic veins, with satisfactory long-term patency. In contrast, Shintani and colleagues [17] recommend bilateral brachiocephalic reconstruction due to a thrombosis rate exceeding 50% at 5 to 26 months, particularly with unilateral left venous reconstruction. All patients in the Shintani series received Coumadin anticoagulation for 6 months or longer.
Efforts were made to clamp the SVC proximal to the azygous vein to preserve collateral circulation and reduce cerebral edema. Clamping at the cavoatrial junction was routinely avoided to reduce the risk of sinoatrial node injury. Hemodynamic compromise during clamping was uncommon and less concerning in the 2 patients whose SVC was chronically stenosed or occluded before operation.
SVC resection and reconstruction was performed after intravenous volume expansion and the use of pharmacologic vasoconstrictors. Although the use of SVC shunts during complete caval occlusion has been described [18], we have not used them. One strategy to reduce caval occlusion time is to perform the distal anastomosis directly to the right atrial appendage by using a Satinsky clamp on the atrium. The proximal venous anastomosis can be performed last, thereby reducing central venous clamp time to that necessary for the proximal anastomosis.
In most cases, SVC resection and reconstruction was performed first, followed by right-sided bronchopulmonary resections, including pneumonectomy, carinal pneumonectomy, or resections with proximal pulmonary artery invasion. This sequence requires careful attention to avoid bacterial or tumor cell contamination of the prosthetic graft. In the setting of malignancy, an autotransfusion device is not recommended.
Venous clamp times were kept at less than 50 minutes, except in 1 patient. Another patient (50-minute SVC clamp time) sustained a transient right hemiparesis on postoperative day 2 while in atrial fibrillation due to an embolic middle cerebral artery occlusion. According to experimental data derived in 1989, 60 minutes of SVC occlusion was physiologically tolerated in a nonhuman primate model [19]; however, Dartevelle and colleagues [18] reported poor tolerance in humans when SVC clamping time exceeded 45 minutes.
Our series confirmed a 5-year survival of 30% and 56% for extended resection of lung and anterior mediastinal neoplasms, respectively, found to involve the SVC. This is congruent with previously published series [1–5, 12, 13, 14, 20] (Table 3). The extended resections were associated with acceptable morbidity and mortality. Our experience suggests that the need for SVC reconstruction should not be considered a contraindication for resection of a bronchopulmonary or mediastinal neoplasm in an otherwise potentially curable patient, provided a complete resection can be achieved.
|
| Acknowledgments |
|---|
|
|
|---|
| References |
|---|
|
|
|---|
Related Article
This article has been cited by other articles:
![]() |
A. M. Ciccone, F. Venuta, A. D'Andrilli, C. Andreetti, M. Ibrahim, T. De Giacomo, D. Massullo, and E. A. Rendina Long-term patency of the stapled bovine pericardial conduit for replacement of the superior vena cava Eur J Cardiothorac Surg, December 1, 2011; 40(6): 1487 - 1491. [Abstract] [Full Text] [PDF] |
||||
![]() |
N. Ohata, R. Nakahara, A. Ui, and H. Matsuguma Azygos flap as a novel method for superior vena cava reconstruction Interact CardioVasc Thorac Surg, November 1, 2010; 11(5): 519 - 521. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. Sekine, H. Suzuki, Y. Saitoh, H. Wada, and S. Yoshida Prosthetic Reconstruction of the Superior Vena Cava for Malignant Disease: Surgical Techniques and Outcomes Ann. Thorac. Surg., July 1, 2010; 90(1): 223 - 228. [Abstract] [Full Text] [PDF] |
||||
![]() |
L. Spaggiari Invited Commentary Ann. Thorac. Surg., August 1, 2009; 88(2): 397 - 398. [Full Text] [PDF] |
||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |