Life-table analysis of patency of primary vascular reconstruction.
Life-table analysis of tumor-free survival.
Combined actuarial analysis of tumor-free survival and primary patency of vascular reconstruction.
DiPerna CA, Bowdish ME, Weaver FA, Bremner RM, Jabbour N, Skinner D, Menendez LR, Hood DB, Rowe VL, Katz S, Kohl R. Concomitant Vascular Procedures for Malignancies With Vascular Invasion. Arch Surg. 2002;137(8):901-907. doi:10.1001/archsurg.137.8.901
Contemporary reconstructive vascular techniques can be safely used to permit resection of tumors invading major vascular structures.
Review of vascular surgery registry between January 1, 1987, and December 31, 2001.
An academic medical center and affiliated institutions.
Forty-nine patients (37 males and 12 females) aged 15 through 80 years (mean age, 55 years) who required concomitant vascular resection and reconstruction to allow complete tumor resection.
Main Outcome Measures
Early (<30 days) morbidity and mortality, late (>30 days) vascular morbidity and mortality, primary patency of the vascular reconstruction, and tumor-free survival.
Aortic resection with graft reconstruction was performed in 20 patients (41.7%) and inferior vena cava resection with reconstruction in 6 patients (12.5%). Five patients (10.4%) had both the aorta and inferior vena cava resected and reconstructed. Iliac, femoral, or popliteal reconstructions were performed in 15 patients (31.3%). Portal vein reconstruction was performed to permit resection of pancreatic neoplasms in 8 patients (16.7%). Resection and reconstruction of either a brachiocephalic vessel or superior vena cava was performed in 4 patients. Thirty-day mortality was 2.1%, as 1 patient died of a myocardial infarction following tumor resection with vascular reconstruction. Overall 30-day morbidity was 12.2%. Early vascular morbidity included bleeding from an arterial anastomosis and a compartment syndrome requiring fasciotomy. Primary patency of the vascular reconstructions at 24 months was 90% and tumor-free survival was 70%. Thirty-one patients (63%) were alive, without tumor recurrence and with a patent vascular reconstruction at 24 months. No patient died or lost a limb due to occlusion of the vascular reconstruction.
Contemporary reconstructive vascular procedures permit resection of tumors that involve major vascular structures with acceptable early and late morbidity and mortality.
OVER THE past few decades the postoperative morbidity and mortality following a major oncologic resection have steadily declined.1,2 With this improvement in surgical outcomes, many surgeons have liberalized the indications for resection of selected tumors to include those with invasion of critical vascular structures that require concomitant vascular reconstruction.3,4 Although several surgeons have reported favorable outcomes for such an approach in selected lower extremity sarcomas, head and neck neoplasms, pancreatic neoplasms, and testicular carcinomas, others have raised concern that postoperative vascular morbidity and mortality are excessive.1- 10 In addition, little is known concerning the late vascular morbidity of reconstructions performed at the time of tumor resection, including the incidence of occlusion of the vascular reconstruction or graft as well as associated limb loss or death due to such vascular events. To investigate these issues, a review of our institutional experience with these extended oncologic procedures was performed to (1) examine the perioperative morbidity and mortality of an en bloc tumor excision requiring a major vascular resection and reconstruction, (2) define patient and tumor-free survival following such procedures, and (3) assess the early and late vascular morbidity and mortality of the major vascular resection and reconstruction and its effect on survival.
Patients at University of Southern California (USC)–affiliated hospitals (USC University Hospital, Norris Cancer Institute, Los Angeles County–USC Medical Center) between January 1, 1987, and December 31, 2001, who required a concomitant resection of a major vascular structure and vascular reconstruction in an attempt to achieve complete tumor resection were identified by query of the USC Vascular Registry. Those patients who underwent vascular resection or ligation, vascular intervention for injury to vessels during an oncologic procedure, resection of noncritical or nonaxial vascular structures that did not require reconstruction, or whose vascular procedure was performed for reasons other than tumor resection, that is, concomitant abdominal aortic or iliac aneurysms, were excluded.
The vascular resections and reconstructions were performed using standard operative techniques. Control of involved vascular structures proximal and distal to the tumor mass was accomplished prior to en bloc resection of the tumor with involved vasculature. Patients received systemic heparin sodium prior to resection of the tumor specimen and division of the involved vascular structures. Vascular reconstructions were tailored to the vessels involved and used a variety of vascular grafts.
Medical records were reviewed to delineate the operative procedure, vascular graft used, pathology of the tumor, and perioperative and postoperative morbidity and mortality. Late survival and tumor recurrence were determined and characterized by review of clinical records and pathology reports. Patency of the vascular reconstruction was determined by review of follow-up duplex sonograms, computed tomographic scans, and clinical examinations. The main outcome measures were early (<30 days) morbidity and mortality, late (>30 days) vascular morbidity and mortality, primary patency of the vascular reconstruction, and tumor-free survival.
A 2-tailed Fisher exact test and 95% confidence intervals were used to evaluate the risk of postoperative complications for the following variables: age, sex, any comorbid medical condition, graft type, aortic and/or inferior vena cava procedure, tumor location, and resection margins positive for tumor and/or nodal and/or distant metastases present in the resected specimen. Actuarial estimates of tumor-free survival and primary patency of the vascular reconstruction were derived using life-table methods. For these analyses, imaging or clinical evidence of occlusion of the reconstruction or hemodynamically significant stenosis of a graft that required repair was considered a failure of the vascular reconstruction. Evidence of tumor recurrence or death by any cause was used to determine tumor-free survival. The SAS statistical software (The SAS System, Release 8.2; SAS Institute Inc, Cary, NC) was used for statistical analysis.
Forty-nine patients (37 males and 12 females) with ages ranging from 15 to 80 years (mean age, 55 years) underwent an oncologic procedure that required a concomitant vascular resection and reconstruction due to invasion or adherence of the neoplasm to a major vascular structure. Medical cormorbid conditions included hypertension in 8 patients (16%), diabetes mellitus in 3 patients (6.0%), known coronary artery disease in 3 patients (6.0%), and chronic obstructive pulmonary disease in 2 patients (4.1%). The pathology of the tumor resected included the abdominopelvic (34 patients), extremity (11 patients), thoracic (3 patients), and neck (1 patient) malignancies listed in Table 1. The oncologic procedure was the initial attempt at tumor resection in 42 patients and performed for recurrence of a previously resected neoplasm in 7 patients.
Vascular reconstructions included infrarenal (18 patients), juxtarenal (1 patient), or suprarenal (1 patient) aortic resections with prosthetic graft reconstruction in 20 patients (41.7%). Inferior vena cava resection with ringed polytetrafluoroethylene [PTFE] graft reconstruction was used in 6 patients (12.5%). Five patients (10.4%) had both the infrarenal aorta and inferior vena cava resected and reconstructed. Iliac or lower extremity reconstructions were performed in 15 patients (31.3%) using both autogenous and prosthetic grafts. To permit resection of a pancreatic neoplasm, portal vein reconstruction by primary anastomosis or autogenous graft was performed in 8 patients (16.7%). Vascular resection and reconstruction for tumor invasion of the carotid (interposition saphenous vein graft), subclavian (PTFE), superior vena cava (superior vena cava to internal jugular vein), and innominate artery (PTFE) were each performed on 1 occasion.
Pathologic examination of the resected specimen documented incomplete tumor resection with positive margins in 2 patients, positive nodal metastases in the surrounding node-bearing soft tissues in 3 patients, and distant metastatic disease that was resected along with the primary neoplasm in 5 patients. Tumor type of the distant metastatic disease was testicular (3 patients), fallopian (1 patient), and carcinoid (1 patient).
Overall 30-day mortality was 2.1% as 1 patient died on postoperative day 5 due to a myocardial infarction that followed resection of the infrarenal aorta, vena cava, and a renal clear cell adenocarcinoma. Thirty-day morbidity was 12.2% and included 2 wound infections, 1 hematoma requiring drainage, and 1 episode of combined sepsis, adult respiratory distress syndrome, and a chylothorax. Two complications occurred related to the vascular procedures performed. In 1 patient, significant bleeding on the third postoperative day from an anastomosis of a femoropopliteal bypass graft necessitated urgent exploratory surgery and repair. A second patient developed a compartment syndrome that required a fasciotomy after en bloc resection of a large, bulky testicular neoplasm that had invaded both the aorta and inferior vena cava. Ultimately, the patient had a residual peroneal nerve palsy requiring the long-term use of an orthotic device. The odds ratio for the risk of postoperative morbidity or mortality were greater than 2.5 in those patients with positive margins or nodal and/or metastatic disease at the time of the operation (Table 2). The odds ratios were less than 2.0 for age, any comorbid medical condition, sex, graft type, aortic and/or caval reconstruction, and tumor location.
Mean follow-up was 19.1 months (median follow-up, 24 months). Late vascular morbidity occurred in 5 patients (10%). One patient experienced thrombosis of a saphenous vein graft used to reconstruct the superficial femoral artery 4 months after resection of an extremity sarcoma. This required a femoropopliteal bypass that remains patent at 3 months. Another patient following resection of an extremity sarcoma developed at 2 months a long segment stenosis in the saphenous vein used to reconstruct the superficial femoral artery. A graft revision of interposition vein was successfully performed. In another patient an iliofemoral reconstruction with PTFE occluded 8 months after resection of a recurrent transitional cell carcinoma. A femoral-femoral ringed PTFE bypass was performed that remained patent for an additional 26 months. A fourth patient experienced the occlusion of a limb of an aortobiiliac graft placed 10 years earlier for resection of a testicular tumor. Patency of the graft was reestablished by thrombolysis. The limb remains patent 14 months later. The final patient developed significant lower extremity edema and pain when a ringed PTFE graft used to reconstruct the iliofemoral vein thrombosed 12 months after resection of a chondrosarcoma. Concomitantly, the patient developed a local recurrence of the chondrosarcoma that required hemipelvectomy.
Follow-up duplex sonograms (13 patients), contrast-enhanced computed tomographic scans (26 patients), or follow-up clinical examinations (9 patients)were used in 48 patients (98.0%) to document patency of the vascular reconstruction. One patient who was lost to follow-up after the immediate postoperative period is excluded from this analysis. For the 9 patients who underwent only a clinical examination, patency of the vascular reconstruction was inferred by absence on clinical evaluation of extremity edema, critical limb ischemia, and absent or diminished extremity pulses or abdominal pain. Actuarial estimates of primary patency at 12 and 24 months were 93% and 90%, respectively (Figure 1).
Eleven patients (22.9%) developed locoregional recurrence of the resected neoplasm during the course of follow-up. Recurrent tumors included 5 extremity-acetabular sarcomas, 3 testicular carcinomas, 1 fallopian adenocarcinoma, 1 extremity melanoma, and 1 pancreatic adenocarcinoma. Recurrence of testicular carcinoma (3 patients), fallopian adenocarcinoma (1 patient), osteogenic carcinoma (1 patient), leiomyosarcoma (1 patient), pancreatic adenocarcinoma (1 patient), and melanoma (1 patient) led to the deaths of 8 patients. Actuarial life-table estimates of tumor-free survival at 12 months was 95% and at 24 months was 70% (Figure 2). At 24 months, 31 (63%) of the 49 patients were alive, free of recurrent tumor with primary patency of the vascular reconstruction (Figure 3). This was not statistically different from tumor-free survival alone at the same time point (P = .32).
Current vascular techniques permit reconstruction of major vascular structures with a high degree of reliability and durability.1- 8 Despite the recognized success of vascular reconstructions, particularly for major abdominal and thoracic vessels, many surgeons still consider tumor invasion of critical vascular structures a relative contraindication to tumor removal.1,11 The current study was performed to determine the short-term and long-term vascular morbidity and mortality associated with resection of neoplasms that are adherent to critical, major vascular structures.
The surgical literature includes many studies that describe the outcomes for tumors requiring a concomitant major vascular resection and reconstruction. Kelly et al6 described 6 patients with large testicular tumors that required en bloc aortic resection and reconstruction for complete tumor resection with no associated vascular-related complications. Vena cava reconstruction for a variety of retroperitoneal tumors was shown by Bower et al12 to be associated with minimal early vascular morbidity and a graft patency of greater than 80% at 24 months. Pedrazzoli et al1 reviewed 22 studies, encompassing 841 patients with pancreatic cancer who required portal vein resection and reconstruction at the time of pancreaticoduodenectomy. Their analysis demonstrated that the addition of a portal vein resection and reconstruction added little to the morbidity and mortality of pancreaticoduodenectomy. Wright et al3 reported only 1 major carotid complication in 20 patients undergoing oncologic resections that required carotid resection and reconstruction. Koperna et al5 found that vascular morbidity was 21% in 13 patients who underwent resections of lower extremity sarcomas that required vascular resection and reconstruction. However, late graft complications were few and limb salvage was possible in patients who otherwise would have required amputation. In contrast, to these favorable reports, Bianchi et al7 found morbidity and mortality to be significant and substantially increased in patients undergoing a variety of oncologic and concomitant vascular procedures at their institution. However, this report includes not only patients undergoing elective resection of tumors involving vascular structures, but also patients who required a "rescue" vascular procedure to address either a postoperative vascular complication or an unplanned injury to a critical vascular structure at the time of tumor resection.
Similar to the report by Bianchi et al,7 the current study analyzes a single institutional experience in which several vascular techniques were applied to permit resection of a diverse group of neoplasms. Unlike the report by Bianchi et al,7 our study includes only patients undergoing an elective resection and is not concerned with those in whom a vascular procedure was performed to manage an intraoperative or postoperative vascular complication. We found in this defined patient population that vascular-related morbidity and mortality were uncommon and overall early complications were comparable to that reported for resection of the tumor alone. Furthermore, analysis of several preoperative and operative variables failed to document a relative increase in complications for aortic or vena caval procedures as well as the various grafts used. Evidence of positive margins and/or nodal or metastatic disease was the only variable with an odds ratio greater than 2.5 that, although not statistically significant, is suggestive of a relative increase in the risk of postoperative complications when compared with the other variables considered.
Another potential concern of any vascular reconstruction is late failure with compromise to vital structures or limb loss. Late vascular morbidity did occur with failure of 4 vascular reconstructions within the first 12 months and a graft limb thrombosis occurring at 10 years. Of the 5 late vascular complications, 4 were successfully managed by placement of another vascular graft or endovascular intervention. The remaining patient had tumor recurrence in temporal proximity to the graft occlusion. Hemipelvectomy for oncologic indications accounted for the patient's limb loss. Overall, patency of the vascular reconstructions was 90% at 24 months. But more importantly, early or late vascular complications were not responsible for any observed limb loss or death.
The primary value of an oncologic procedure can be measured in tumor-free survival. Overall tumor-free survival was 70% at 24 months and for selected tumors either similar or somewhat different than that reported in the literature. With a follow-up of 1 month to 16 years, Spitz et al13 reported 52.6% of patients were disease free after resection of bulky testicular tumors that required concomitant caval resection. In the present series, disease-free survival for patients with testicular tumors was 77% (3 patients). All 3 deaths were due to tumor recurrence including 1 in a patient with a radioresistant seminoma (<2% of seminomas). In reports by Koperna et al5 and Karakousis et al,8 local recurrence rates for lower extremity sarcomas were found to be 6% and 0%, respectively. Koperna et al5 noted a mortality rate of 30.8% due to development of metastatic disease. Of the 11 lower extremity sarcomas, 3 (27.3%)recurred in this study, and 3 patients (27.3%) died of distant metastatic disease. The collective review of Pedrazzolli et al1 of patients undergoing portal vein resection and reconstruction at time of pancreaticoduodenectomy documented a 5-year survival of 6%. Of the 7 patients in the present study only 1 patient (20.8%) had developed recurrent disease at the time of this article. This unexpected result may be because of the highly select nature of these 7 patients, since they represent less than 10% of patients subjected to pancreaticoduodenectomy at our institution over the past 4 years. Also, the addition of portal vein resection during pancreaticoduodenectomy has only recently been introduced at our institution, resulting in limited patient follow-up. The expectation is that with longer follow-up, the incidence of recurrent disease will increase.14- 16 If this fails to occur, the potential factors that may be responsible for improved survival will need to be further investigated.
When performing a vascular resection and reconstruction as part of an oncologic procedure, the additional late morbidity that may occur with failure of the reconstruction can be analyzed by constructing a life table where tumor recurrence or failure of a vascular reconstruction is a primary end point. Whereas tumor-free survival was 70% at 24 months the addition of failure of the vascular reconstruction decreased the favorable outcome to 63%. Since this small decrease failed to achieve statistical significance, it adds further support to the conclusion that the additional morbidity and mortality of the vascular procedure have little influence on long-term outcome. This conclusion, however, should be viewed in the context of the limitations of the present study. First, as for most studies of this nature, the sample size is small and the method is retrospective, which makes any statistical analysis less certain. Second, in 9 patients who did not have an objective imaging study to document patency of the vascular reconstruction, clinical examination alone was the determinant. However, since all vessels included in the study were major axial vessels that would cause morbidity usually not missed by an experienced observer, it is unlikely that many vascular failures went unrecognized.
Invasion of major vascular structures has been considered by some a barrier to the removal of large neoplasms. The results reported herein support the contrary contention that the need for resection and reconstruction of a major vascular structure should not prohibit the resection of any given tumor. The study demonstrates that most major vascular reconstructions performed in this setting have a high degree of success, do not appreciably increase overall early morbidity and mortality, and have little influence on long-term outcome. Although late graft occlusions can occur, most can be managed successfully and do not result in death or limb loss. In essence, long-term patient outcome is not determined by the need to perform a concomitant vascular procedure but rather by the biological behavior of the resected malignancy.
This paper was presented at the 73rd Annual Meeting of the Pacific Coast Surgical Association, Las Vegas, Nev, February 18, 2002, and is published after peer review and revision. The discussion is based on the originally submitted manuscript and not the revised manuscript.
We acknowledge the following for their contributions: Linda S. Chan, PhD, USC–Los Angeles County Medical Center, and Earl H. Leonard, Childrens Hospital of Los Angeles, for statistical assistance, as well as Jenny Moser, RN, Hillary Gibbons, PA, and Rita Poquette, RN, for assistance with patient follow-up.
Corresponding author: Fred A. Weaver, MD, Department of Surgery, Division of Vascular Surgery, Keck School of Medicine, University of Southern California, 1510 San Pablo St, Suite 514, Los Angeles, CA 90033 (e-mail: email@example.com).
Jeffrey L. Ballard, MD, Loma Linda, Calif: Dr Bowdish and colleagues from USC have presented a heterogeneous group of 49 oncologic patients who had major vascular reconstruction in conjunction with attempted curative tumor resection. Excluded patients were those having vascular resection or ligation without reconstruction, vascular injury during tumor resection, and resection of minor blood vessels. Main outcome measures are quite reasonable and statistical analysis is straightforward. I jump-started my discussion by reviewing a paper that we presented at the 70th Annual Meeting of the Pacific Coast Surgical Association meeting in Los Cabo, Mexico titled, "Vascular Reconstruction and Major Resection for Malignancy."
In that presentation, which covered the period 1980 to 1998, it was concluded that long-term patient survival was actually poor when resections for carcinoma were associated with major vessel infiltration or when a major oncologic resection required an emergent vascular procedure. In contradistinction, Dr Bowdish and colleagues have demonstrated that contemporary reconstructive vascular procedures can facilitate the extent of tumor resectability without a significant increase in early or late morbidity or mortality. Despite the complicated nature of these procedures, the overall 30-day morbidity in this 14-year retrospective review was only 14%. Tumor-free survival at 12 and 24 months was 95 and 70%, respectively, by life-table analysis and remarkably, 63% of patients were alive, free of recurrent disease with a patent vascular reconstruction at 24 months. These data are not only encouraging but also demonstrate the potential outcome when diverse surgical specialties combine to offer complete patient care.
Interestingly, our papers discuss a similar topic but have decidedly different conclusions. I wonder if this may be related to the type of tumor that was treated in each series. My impression from briefly reviewing the oncologic literature is that completely resected urologic malignancies and soft tissue sarcomas have a generally better outcome than head and neck malignancies that tend to have aggressive biology and notoriously poor long-term survival. Your series had 1 neck malignancy compared with 10 or 26% in our series. Do you think this may explain the difference in the survival curves? It is also notable that your life table demonstrates a steep drop off in tumor-free survival from 12 to 24 months and that mean patient follow-up was only 19 months. In our series, mean follow-up for surviving patients was almost twice as long. If you rethink your data extrapolated out another year or 2 do you still paint as rosy a conclusion?
I have 2 final questions and a comment. One, do your surgical oncologists seek formal advice from you prior to tumor resection or are you "curbsided" in the OR on your way to an elective AAA (abdominal aortic aneurysemectomy)repair? After our publication, we have received numerous preoperative consultations, which allows us to see the patient before surgery, review all pertinent scans and studies and to completely plan both procedures. Two, based on your experience with such a diverse group of patients, procedures, and malignancies do you have any advice to impart to us regarding when it would be appropriate to retreat from these potential surgical landmines and have you learned when to concede victory to the tumor?
My comment relates to quality of life notwithstanding the virtues of a patent vascular graft or a live patient. However, it seems that this may be the most important, nonstudied aspect of major vascular reconstruction and resection for malignancy. Prospective analysis of life quality before and after complex vascular or oncologic procedures may provide essential data that could be used by treating physicians to preoperatively stratify outcome in these difficult clinical situations.
Theodore X. O'Connell, MD, Los Angeles, Calif: Obviously, from a vascular surgery aspect, this is a real tour de force, but as an oncologic surgeon I want to know how the patients are really going to benefit long term. How many patients are going to be really cured? As I look at the paper, it looks like only about 25% of the patients have long-term control and possible cure of the tumor. It's a real mixed bag of tumor types. Which are the patients that are cured? What kind of tumors are cured? What are their characteristics, and which tumors are not cured? When should I ask my vascular surgeon to help me, in a sarcoma, in a pancreatic cancer, etc? When do the patients really benefit from this type of operation?
Frederick W. Grannis, Jr, MD, Duarte, Calif: I am fortunate to work in an institution where our surgical oncologists do involve the vascular surgeon early in the process. I believe that it is also important is to involve the plastic surgeon early in the process, particularly for complex head and neck and groin reconstructions. Specifically with respect to venous anastomosis, which I have found to be the most difficult part of management in limb salvage, of the lower and upper extremities, how do surgeons at USC manage venous reconstruction in grafting? Do they use anticoagulant therapy long-term postoperatively? Do they use AV (arteriovenous) fistulas to try and keep those venous grafts open?
James J. Peck, MD, Portland, Ore: My question has to do with the quality of life. In the extremity lesions, how often was the nerve involved and, if the nerve was involved, did that preclude a vascular reconstruction?
John MacFarlane, MD, Vancouver, British Columbia: I enjoyed this paper very much. I was curious to know if the authors have any information on the actual involvement of the vascular structures with tumor. It's long been my contention that many solid tumors do not actually invade vessels but they constrict them. If there is information of this type in the paper, is there a difference in the outcome?
John Vetto, MD, Portland: I wonder if the authors would comment on the role of radiation therapy, either with the anastomosis or instead of it, specifically brachytherapy.
Dr Weaver: First of all I would like to thank Dr Ballard for his discussion and comments. With regards to tumor biology, I do think that this is an important difference between our study and Dr Ballard's and also the emergent nature of many of the vascular interventions in Dr Ballard's paper is another important difference.
We had a large number of testicular tumors that have a very favorable outcome and afflict a young patient population. We had also a large cohort of extremity sarcomas that is predominantly a young patient population as well, with reasonable outcomes. Most of these procedures, although we scanned our registry for the past 15 years, were done more recently, and consequently we do not have long-term follow-up on many of the patients. I think the group that is the most interesting is the pancreatic adenocarcinomas. Eight pancreatic resections, 7 for adenocarcinoma, along with portal vein reconstruction were done, and we have complete follow-up on all patients. With a mean follow-up of 13 months, we only have 1 recurrence. In a collective review by Pedrazzolli et al, the 5-year survival for that group was only 6%, so clearly longer follow-up probably will show a drop off in survival.
With regard to curbside consults, in the very initial phases of this series we did experience a little bit of a curbside approach, but this has changed and in most cases we work hand-in-hand with the oncologic surgeon with regards to preoperative planning. I think that has really made a difference with regard to our outcomes.
When do you concede victory to the tumor? We leave that up to the oncologic surgeon. The message of this paper is that vascular reconstruction should not preclude the resection of a neoplasm, the decision to resect a particular neoplasm is really in the hands of the oncologic surgeon and is dependent on the tumor biology. That is really the determining factor.
Dr O'Connell, concerning tumor types, where this is this advantageous, the testicular tumors and extremity sarcomas respond favorably and this has been not only our experience but that of the literature as well. I already mentioned the pancreatic tumors. The head and neck carcinomas are a mixed bag, and we have not had a lot of experience with those neoplasms. Our head and neck surgeons have not been enthusiastic about using vascular reconstructions in those particular cases.
With regard to venous reconstruction, we do not use AV fistulas. We will anticoagulate those patients undergoing venous reconstructions. We use ringed PTFE grafts at the femoral and caval levels. If we are at the popliteal level, we will use vein to reconstruct the popliteal vein. Anticoagulant therapy continued for the caval grafts for about 6 months. The same is true for the lower extremity PTFE grafts although we do not have many of those in this series.
Dr Peck asked about nerve involvement. Nerve involvement is a relative contraindication to resection. If there is nerve involvement, then we would not advocate proceeding with such a major procedure. There are occasional patients who absolutely will refuse to have their limb removed, and in those instances the oncologic surgeon has had us do reconstructions even though the leg is not functional. The wisdom of that approach certainly is questionable.
With regards to the involvement or invasion of vascular structures, unfortunately, I do not have any pathologic information with regards to that particular component. It would be interesting to look at that aspect. We certainly have the ability and the tissue blocks do exist. With the testicular carcinomas, for which we have the largest experience and the largest vascular structures, my impression is that it is more of an adherence rather than an invasion by the tumor into the wall of the vessel, but it has so firmly adhered to the wall that there is no way that you can take the tumor out without resecting the vessel.
With regards to radiation treatment or radiotherapy, that again is a decision that is the oncologic surgeons'. It does not preclude us from performing a vascular reconstruction if a patient that has had a previously irradiated tumor removed. The plastic surgeons come into play predominantly in head and neck neoplasms but also a great deal in the extremity reconstructions. Usually it is the rectus flaps that are used to cover the femoral triangle when large tumors are removed in that area and at least 50% of sarcoma resections in that area have required some form of adjunctive plastic surgical procedure as well. I would like to thank all of the discussants and, in particular, Dr Ballard for his discussion and the opportunity to contribute to the meeting.