- Email: [email protected]
Mediastinoscopy in superior vena cava obstruction: analysis of 80 consecutive patients
Mediastinoscopy in Superior Vena Cava Obstruction: Analysis of 80 Consecutive Patients Tommaso Claudio Mineo, MD, Vincenzo Ambrogi, MD, Italo Nofroni, BS, and Chiara Pistolese, MD Department of Thoracic Surgery, Tor Vergata University, Rome, Italy
Background. Prejudices against mediastinoscopy in superior vena cava obstruction still remain. Hereby we analyze risk/benefit balance in a large series of patients. Methods. Eighty consecutive patients underwent cervical mediastinoscopy for caval obstruction, 51 after uncertain diagnosis obtained by lesser techniques, 17 after ineffective chemotherapy (n ⴝ 9) or radiotherapy (n ⴝ 8). In 12 patients we immediately performed mediastinoscopy as an urgent procedure. In addition the examination was combined with left anterior mediastinotomy (n ⴝ 7) for staging purposes. Results. No perioperative mortality was recorded. Five patients had significant bleeding, but only one required sternotomy. Definitive diagnosis was obtained in all
patients: 50 lung cancer, 17 lymphoma, 7 invasive thymoma, 3 postradiation fibrosis, 2 metastatic lymph nodes from renal carcinoma, and 1 fibrosing mediastinitis. Specific therapy had excellent effects in 71 patients, negligible in 7, and adverse in 2. Postmediastinoscopy brachial venous pressure had a mean significant decrease (p < 0.0001). Lung cancer was the sole variable significantly associated with unfavorable outcome (p < 0.0004). Conclusions. Mediastinoscopy should be routinely included after less invasive procedures in the diagnostic program because it is simple, low risk, and effective.
I
hospital at Tor Vergata University of Rome. In all patients the examination was conducted by the same surgeon (T.C.M.) so that the two series can be considered homogeneous and available for a unique evaluation. Patient population consisted of 59 men and 21 women ranging in age from 19 to 83 years (mean, 55.5 ⫾ 7.3 standard deviation). The experience was the result of 111 patients with superior vena caval obstruction who underwent preliminary evaluation in the two university hospitals, La Sapienza (n ⫽ 81) and Tor Vergata (n ⫽ 30). Information was retrieved from both operative notes and medical records. Clinical evolution of the disease was retrieved from the oncologist or general practitioner who treated and followed-up the patient. According to our retrospective analysis, caval syndrome was present as the first sign of the underlying disease in 76 of these patients (69%). In 34 patients the obstruction had a rapid onset. Major reason for presentation was always neck swelling. Other symptoms included moderate (n ⫽ 73) or severe dyspnea (n ⫽ 19), cough (n ⫽ 64), hemoptysis (n ⫽ 12), phrenic nerve paresis (n ⫽ 18), weight loss (n ⫽ 34), hoarseness (n ⫽ 15), and dysphagia (n ⫽ 11). No patient presented or developed major neurologic symptoms (eg, seizure, coma). In 85 patients, the clinical diagnosis of obstruction was confirmed by contrast studies, either venography (n ⫽ 31) and angio-computed tomographic scan (n ⫽ 54). Caval outflow was judged as type I according to the classification of Stanford and Doty [7] in 5 patients, type II in 17, type III in 25, and type IV in 38.
n 1959 Carlens [1] first introduced the mediastinoscopy procedure. It filled an important void by reducing the number of patients undergoing thoracotomy for unresectable bronchial carcinoma. Since then, mediastinoscopy has gained an increasing importance in the staging of lung cancer [2]. For many years it has been claimed that mediastinoscopy in patients with superior vena cava obstruction was an inappropriate procedure with a high mortality and morbidity rate [3]. Nevertheless, mediastinoscopy in caval obstruction has been advocated as the most reliable technique in obtaining rapid histologic diagnosis [4 – 6]. Besides, there is no evidence of such high mortality or morbidity rates in literature [6]. In spite of this statement, a common and silent prejudice against the examination still persists. We decided to review our experience with the aim of emphasizing the safety and the advantages of the mediastinoscopy in superior vena caval obstruction.
Patients and Methods Our review embraces 23 years (1974 to 1997) with a total of 80 patients who underwent cervical mediastinoscopy for superior vena caval obstruction in two departments of Thoracic Surgery in consecutive periods. The first 57 patients were observed at La Sapienza University of Rome until 1986 and the other 23 at our present teaching Accepted for publication Feb 2, 1999. Address reprint requests to Dr Mineo, Cattedra di Chirurgia Toracica, Ospedale S. Eugenio, P.le Umanesimo 10, 00144 Rome, Italy; e-mail: [email protected].
© 1999 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 1999;68:223– 6) © 1999 by The Society of Thoracic Surgeons
0003-4975/99/$20.00 PII S0003-4975(99)00455-5
224
MINEO ET AL MEDIASTINOSCOPY IN CAVAL OBSTRUCTION
As far as we know a preliminary bronchoscopic examination was performed in 92 patients, 80 of these in our department. Among this group histologic biopsies were possible in 16 patients, in the other cases diagnosis was attempted by bronchoalveolar lavage cytology. The examination was effective in establishing a diagnosis in 21 patients (26.2%). It was negative in 19 patients and nondiagnostic to start therapy in 40. Fine needle biopsy was done in 24 patients with negative bronchoscopy. Sufficient material to establish a correct histologic diagnosis was obtained only in four instances (16.7%). As a result of the initial screening, 31 patients were not considered for mediastinoscopy as a result of an already known diagnosis by lesser diagnostic examinations (n ⫽ 25) or they refused consent to further examinations (n ⫽ 6). Informed consent for mediastinoscopy was obtained in all other patients. We decided for immediate mediastinoscopy on an urgent basis as first diagnostic procedure in 12 patients. In 51, the mediastinoscopy was performed after less invasive diagnostic techniques (bronchoscopy and fine needle biopsy), which either failed to achieve sufficient sample for a correct diagnosis (n ⫽ 36) or provided negative results (n ⫽ 15). Another 17 patients had been studied elsewhere and referred to us for mediastinoscopy after initiating ineffective therapy (incomplete radiation and chemotherapeutic cycles [n ⫽ 9] or prebiopsy radiotherapy [n ⫽ 8]). Venous pressure values in the brachial district were available in 72 patients immediately before the induction and 6 hours after the procedure. It was calculated by the height (in centimeters) above the right atrium level reached by a column of water in a No. 12 catheter connected to the brachial venous system with the patient lying on supine decubitus. Mediastinoscopy was always conducted with the patient in a supine decubitus position with the head and truncus elevated. The operative field was widely prepared in the eventuality of an emergency thoracotomy. General anesthesia was preferred and the patient was always ventilated through a single lumen endotracheal tube. The operative risk was classified according to Bigsby and colleagues [8] in low (n ⫽ 47) and high risk (n ⫽ 33). Before commencing the procedure, the presence of enlarged scalene nodes was always researched and excluded. Transcervical mediastinoscopy in a superior vena cava obstruction did not substantially differ from the usual procedure, apart from a more careful dissection and accurate hemostasis. As usual, samples were sent for frozen section analysis. The examination ended as soon as the pathologist indicated that the tissue was suitable for histologic diagnosis. In 5 patients, the presence of enlarged prethyroid lymph nodes impeded access to the pretracheal space, but biopsies of these lymph nodes were diagnostic. Statistical evaluation did not include these patients. On the basis of computed tomographic scan findings, in 7 patients further biopsies were provided through a second access during the same operation by a left ante-
Ann Thorac Surg 1999;68:223– 6
rior mediastinotomy. Thus, we were able to reach concomitant left-sided masses, which were the primitive lung cancer (n ⫽ 2) and controlateral neoplastic subaortic lymph nodes (n ⫽ 2). In the last 3 patients anterior mediastinotomy was adopted after a preliminary response of fibrous tissue. This was done to achieve supplementary biopsies from other mediastinal areas. Dependency between survival and subcategories was tested by 2 test. Because of the abnormal distribution of the data, comparison between caval pressure values before and after the procedure was performed by the Wilcoxon test.
Results The procedure, recorded for the last series of 23 patients, had a mean duration of 26.45 minutes including the time for frozen section. No perioperative mortality was recorded. No airway obstruction and cardiovascular collapse at general anesthesia induction were observed. We only noted episodes of mild bronchospasm, which were controlled by steroid injection. Cardiac arrhythmias and temporary arterial hypotension were monitored and successfully treated. We had five cases of major bleeding. Histology in these patients was small cell carcinoma (n ⫽ 3), sclerosis after radiation therapy (n ⫽ 1), and invasive thymoma (n ⫽ 1). Mean venous pressure was 30.8 cm H2O (range, 26 to 38), which is less than the analogous value measured in the entire population at study; types of venous outflow detected in the patients who bled were II (n ⫽ 2), III (n ⫽ 2), and IV (n ⫽ 1). No apparent correlation between risk of bleeding and venous pressure or type of outflow was found. In 4 patients hemorrhage was provoked by tear lesions of the minor venous vessels and it was controlled by local compression and positioning metallic clips or oxidized regenerated cellulose. Only 1 patient required total median sternotomy through which resection of an invasive thymoma and grafting of a superior cava vein by a Gore-Tex (WL Gore & Associates Inc, Flagstaff, AZ) prosthesis were possible. Anesthetic recovery was rapid and all the patients were soon able to breathe spontaneously. The postoperative period was uneventful and wounds healed without major infection. In the last series of 23 patients, 14 procedures were performed on an outpatient basis discharging them after 6 hours of observation. Brachial venous pressure showed a significant decrease between before and after mediastinoscopy mean values by Wilcoxon test (Table 1). Thirty-four patients presented a subjective improvement of symptoms within a few hours after the operation even in the absence of specific therapy. Mean venous pressure diminution was more significant (7.1 versus 3.2 cm H2O; p ⬍ 0.0001) in this group of patients. Biopsies performed by transcervical mediastinoscopy established exact definitive tissue diagnosis in all patients. Fifty-three patients had lung cancer, squamous cell carcinoma in 16, adenocarcinoma in 11, large cell carcinoma in 4, and small cell carcinoma in 22. The
Ann Thorac Surg 1999;68:223– 6
MINEO ET AL MEDIASTINOSCOPY IN CAVAL OBSTRUCTION
225
Table 1. Mediastinoscopic Changes of Venous Pressure Categories Lung cancer Lymphoma Thymoma Type 1–2a Type 3 Type 4 Early improved Total a
No. of Patients
Mean Preoperative Pressure (cm H2O)
Mean Postoperative Pressure (cm H2O)
Significancy (Wilcoxon test)
46 15 6 15 24 33 34 72
32.6 ⫾ 8.29 33.3 ⫾ 8.94 30.0 ⫾ 6.26 24.3 ⫾ 4.83 31.2 ⫾ 6.60 37.1 ⫾ 7.89 33.6 ⫾ 9.34 32.6 ⫾ 8.37
27.4 ⫾ 8.51 30.1 ⫾ 7.49 26.3 ⫾ 6.88 22.1 ⫾ 5.17 26.8 ⫾ 6.22 30.7 ⫾ 9.37 26.5 ⫾ 9.32 27.7 ⫾ 8.29
p ⬍ 0.0001 p ⬍ 0.03 NS NS p ⬍ 0.0006 p ⬍ 0.0001 p ⬍ 0.0001 p ⬍ 0.0001
Type according to the Stanford and Doty [7] classification.
Mean ⫾ standard deviation values for peculiar categories and relative significancy at Wilcoxon test are shown.
majority of the tumors originated in the right upper lobe (n ⫽ 27) and most of these (n ⫽ 21) had contiguity with the cava vein. In the other patients, caval obstruction was provoked by mediastinal lymph nodal involvement. In 3 patients with lung cancer, the syndrome was secondary to sclerosis after radiation therapy and it was completely resolved by steroid administration. In another 17 patients obstruction was secondary to enlarged mediastinal lymph nodes due to lymphoma, 11 nonHodgkin’s and 6 Hodgkin’s. Seven patients had an invasive thymoma, 2 had metastatic lymph nodes from a renal carcinoma, and 1 had fibrosing mediastinitis. In the 12 patients who underwent urgent mediastinoscopy, we were able to start appropriate therapy within 24 hours after the operation. The majority of patients (n ⫽ 71) showed a dramatic improvement of clinical signs of caval obstruction and related symptoms within the first week of treatment. In 7 patients therapy produced only mild or nonsignificant effects and in 2 patients, worsening the general condition. All patients with lung carcinoma for whom we have information (n ⫽ 24) died with a median survival of 6 months. Death was attributable to generalized disease in almost all instances. Only 1 patient died for direct progression of the caval syndrome. At present, 5 of 12 lymphomatous and 4 of the 6 thymomatous patients have remission of disease. In 9 patients with lymphoma we were able to restage the neoplasm after chemotherapy by repeat mediastinoscopy. Once again the examination was easily performed without complications. The 2 test disclosed no significant association between mortality and any of the following factors: Bigsby’s high risk scale, Stanford and Doty III or IV types, caval pressure more than 40 cm H2O, rapid obstruction and postmediastinoscopy brachial venous pressure variation. On the other hand, lung cancer was significantly associated with unfavorable outcome ( p ⬍ 0.0004).
Comment In more than 90% of the patients, caval obstruction is attributable to a malignant disease [8, 9]. Bronchogenic carcinoma is reported in about 80% of these patients,
with lymphoma and metastatic disease comprising 15% and 5%, respectively [8, 9]. In our observations, superior vena cava obstruction arose with the following frequency: 1.73% of lung cancer, 2.77% of the nonHodgkin’s and 3.07% of the Hodgkin’s lymphomas, and 4.25% of the thymomas. Because of the risks related to general anesthetic and local dissection, mediastinoscopy was and still is considered potentially dangerous [3, 9]. In the past, skepticism on the long-term effectiveness of therapeutic efforts suggested a nonaggressive diagnostic conduct limiting the treatment to prebiopsy radiotherapy [10]. In our opinion decompression is the best palliation and it can be better obtained by debulking the mass by appropriate radiation or chemotherapy [4]. To obtain the most effective response to therapy, it is highly desirable to reach a specific tissue diagnosis [11]. Lesser diagnostic procedures may not result in the correct diagnosis. A prescalenic lymph node suitable for excisional biopsy is usually absent. Cytology diagnosis by sputum and fine needle aspiration have often scant sensibility in exactly differentiating the type of tumor [6] and they may be ineffective to start appropriate therapy. Bronchoscopy can be considered the best noninvasive alternative to mediastinoscopy. Nevertheless, it is not risk free [7] and it often requires lavage or brushing cytologic techniques, which are as undiagnostic as sputum and aspiration cytology. In our experience at bronchoscopy we found a low rate (16 of 80 patients [20.0%]) of endoluminal growing lesions available for biopsy and histologic examination. In our opinion mediastinoscopy has a convenient balance of risk and benefit. The examination can provide the highest possible diagnostic accuracy. It can be often performed on as an ambulatory regimen and complication risk is relatively low as shown in Table 2. It is not higher than that in the other patients without caval obstruction [4 – 6, 8, 11–18]. Usually, the lesion causing obstruction is sited in proximity to the thoracic inlet and it is relatively easy to obtain large pieces of tissue for histologic studies. The superior mediastinum is usually not affected by the increased venous pressure as the venous drainage is not directly to the superior vena cava as in the head–neck
226
MINEO ET AL MEDIASTINOSCOPY IN CAVAL OBSTRUCTION
Ann Thorac Surg 1999;68:223– 6
Table 2. Review of the Main Series of Mediastinoscopy in Superior Vena Caval Obstruction Reference Little et al [5] Schraufnagel et al [11] Parish et al [13] Lewis et al [4] Gamondes et al [14] Vanderhoeft [15] Yellin et al [12] Chen et al [16] Callejas et al [17] Bigsby et al [8] Jahangiri and Goldstraw [6] Gamez-Garcia et al [18] Present study
Period (yr)
No. of Patients
No. of Mediastinoscopies
No. of Complications
1960 – 85 1961–79 1961– 81 1966 – 80 1977– 83 1965– 85 1972– 87 1976 – 88 1985– 89 1987–92 1982–93 1974 –96 1974 –97
42 107 86 29 13 93 63 45 8 18 44 29 111
8 14 3 15 12 93 15 1 8 7 34 23 80
0 1 0 0 0 3 1 0 2 2 1 0 5
and the upper extremity. This is one of the reasons why mediastinoscopy is not a difficult procedure in the face of the cava obstruction. The vast majority of the patients with caval obstruction are unlikely to have rapid and life-threatening complications; therefore, we believe that it is worthwhile to try to achieve histologic diagnosis with lesser examinations. Immediate mediastinoscopy must be reserved to those patients who present clinical progression of the caval obstruction in spite of aggressive and prolonged symptomatic therapy. The use of emergency prebiopsy radiotherapy, so-called blind, seems unjustifiable in these conditions. Whenever it was adopted, there were problems in dissection due to fibrosis and in the interpretation of the histology [19]. In 50% of our patients “blind” radiotherapy would have been inappropriate (eg, lymphoma and small cell lung cancer) or harmful as in benign caval obstruction. In conclusion, we consider mediastinoscopy a simple, low risk, and highly effective method in obtaining tissue diagnosis for initiating specific therapy. It should be routinely included after less invasive examinations and in some instances it could be prospected as a first diagnostic urgent procedure.
5.
6. 7. 8. 9. 10. 11. 12. 13. 14.
This study has been supported by a grant from MURST (60%). This study has been carried out within the Research Fellowship Program Dottorato di Ricerca in Oncologia Toracica, appointed by Tor Vergata University of Rome.
15. 16.
References 1. Carlens E. Mediastinoscopy: a method for inspection and tissue biopsy in the superior mediastinum. Dis Chest 1959; 36:343–7. 2. Pearson FG, Nemels JM, Henderson RD, Delarue NC. The role of mediastinoscopy in the selection of treatment for bronchial carcinoma with involvement of superior mediastinal lymph nodes. J Thorac Cardiovasc Surg 1987;64:382–90. 3. Lochridge SK, Knibbe WP, Doty DB. Obstruction of superior vena cava. Surgery 1979;85:14–24. 4. Lewis RJ, Sisler GE, Mackenzie JW. Mediastinoscopy in
17.
18.
19.
advanced superior vena cava obstruction. Ann Thorac Surg 1981;32:458– 62. Little AG, Golomb HM, Ferguson MK, Skosey C, Skinner DB. Malignant superior vena cava obstruction reconsidered: the role of diagnostic surgical intervention. Ann Thorac Surg 1985;40:285– 8. Jahangiri M, Goldstraw P. The role of mediastinoscopy in superior vena caval obstruction. Ann Thorac Surg 1995;59: 453–5. Stanford W, Doty DB. The role of venography and surgery in the management of patients with superior vena cava obstruction. Ann Thorac Surg 1986;41:158– 63. Bigsby R, Greengrass R, Unruh H. Diagnostic algorithm for acute superior vena caval obstruction (SVCO). J Cardiovasc Surg 1993;34:347–50. Nieto AF, Doty DB. Superior vena cava obstruction: clinical syndrome, etiology and treatment. Curr Probl Cancer 1986; 10:441– 81. Davenport D, Ferree C, Blake D, Raben M. Radiation therapy in the treatment of superior vena caval obstruction. Cancer 1978;42:2600–3. Schraufnagel DE, Hill R, Leech JA, Pare JAP. Superior vena cava obstruction: is it a medical emergency? Am J Med 1981; 70:1169–74. Yellin A, Rosen A, Reichert N, Liebermann Y. Superior vena cava syndrome. Am Rev Resp Dis 1990;141:1114– 8. Parish JM, Marshke RF, Dines DE, Lee RE. Etiologic consideration in superior vena cava syndrome. Mayo Clin Proc 1981;56:407–15. Gamondes JP, Vincent M, Weynants P, et al. Mediastinoscopie et syndrome de compression ou d’obstruction de la veine cave superieure. A propos de douze observations. Ann Chir 1985;39:482–5. Vanderhoeft P. La mediastinoscopie dans les syndromes de la veine cave superieure. Ann Chir 1985;39:521. Chen JC, Bongard F, Klein SR. Contemporary perspective on superior vena cava syndrome. Am J Surg 1990;160:207–11. Callejas MA, Rami R, Catalan M, Mainer A, Sanchez-Lloret J. Mediastinoscopy as an emergency diagnostic procedure in superior vena cava syndrome. Scand J Thorac Cardiovasc Surg 1991;25:137–9. Gamez-Garcia AP, Martin de Nicolas Serrahima JL, Marron Fernandez C, Garcia Barajas S, Diaz-Hellin Gude V, Toledo Gonzalez J. Surgical diagnosis procedures in superior vena cava syndrome. Arch Bronconeumol 1997;33:284– 8. Loeffler JS, Leopold KA, Techt A, Weinstein HJ, Tarbell NJ. Emergency prebiopsy radiation for mediastinal masses: impact on subsequent pathologic diagnosis and outcome. J Clin Oncol 1986;4:716–21.
Background. Prejudices against mediastinoscopy in superior vena cava obstruction still remain. Hereby we analyze risk/benefit balance in a large series of patients. Methods. Eighty consecutive patients underwent cervical mediastinoscopy for caval obstruction, 51 after uncertain diagnosis obtained by lesser techniques, 17 after ineffective chemotherapy (n ⴝ 9) or radiotherapy (n ⴝ 8). In 12 patients we immediately performed mediastinoscopy as an urgent procedure. In addition the examination was combined with left anterior mediastinotomy (n ⴝ 7) for staging purposes. Results. No perioperative mortality was recorded. Five patients had significant bleeding, but only one required sternotomy. Definitive diagnosis was obtained in all
patients: 50 lung cancer, 17 lymphoma, 7 invasive thymoma, 3 postradiation fibrosis, 2 metastatic lymph nodes from renal carcinoma, and 1 fibrosing mediastinitis. Specific therapy had excellent effects in 71 patients, negligible in 7, and adverse in 2. Postmediastinoscopy brachial venous pressure had a mean significant decrease (p < 0.0001). Lung cancer was the sole variable significantly associated with unfavorable outcome (p < 0.0004). Conclusions. Mediastinoscopy should be routinely included after less invasive procedures in the diagnostic program because it is simple, low risk, and effective.
I
hospital at Tor Vergata University of Rome. In all patients the examination was conducted by the same surgeon (T.C.M.) so that the two series can be considered homogeneous and available for a unique evaluation. Patient population consisted of 59 men and 21 women ranging in age from 19 to 83 years (mean, 55.5 ⫾ 7.3 standard deviation). The experience was the result of 111 patients with superior vena caval obstruction who underwent preliminary evaluation in the two university hospitals, La Sapienza (n ⫽ 81) and Tor Vergata (n ⫽ 30). Information was retrieved from both operative notes and medical records. Clinical evolution of the disease was retrieved from the oncologist or general practitioner who treated and followed-up the patient. According to our retrospective analysis, caval syndrome was present as the first sign of the underlying disease in 76 of these patients (69%). In 34 patients the obstruction had a rapid onset. Major reason for presentation was always neck swelling. Other symptoms included moderate (n ⫽ 73) or severe dyspnea (n ⫽ 19), cough (n ⫽ 64), hemoptysis (n ⫽ 12), phrenic nerve paresis (n ⫽ 18), weight loss (n ⫽ 34), hoarseness (n ⫽ 15), and dysphagia (n ⫽ 11). No patient presented or developed major neurologic symptoms (eg, seizure, coma). In 85 patients, the clinical diagnosis of obstruction was confirmed by contrast studies, either venography (n ⫽ 31) and angio-computed tomographic scan (n ⫽ 54). Caval outflow was judged as type I according to the classification of Stanford and Doty [7] in 5 patients, type II in 17, type III in 25, and type IV in 38.
n 1959 Carlens [1] first introduced the mediastinoscopy procedure. It filled an important void by reducing the number of patients undergoing thoracotomy for unresectable bronchial carcinoma. Since then, mediastinoscopy has gained an increasing importance in the staging of lung cancer [2]. For many years it has been claimed that mediastinoscopy in patients with superior vena cava obstruction was an inappropriate procedure with a high mortality and morbidity rate [3]. Nevertheless, mediastinoscopy in caval obstruction has been advocated as the most reliable technique in obtaining rapid histologic diagnosis [4 – 6]. Besides, there is no evidence of such high mortality or morbidity rates in literature [6]. In spite of this statement, a common and silent prejudice against the examination still persists. We decided to review our experience with the aim of emphasizing the safety and the advantages of the mediastinoscopy in superior vena caval obstruction.
Patients and Methods Our review embraces 23 years (1974 to 1997) with a total of 80 patients who underwent cervical mediastinoscopy for superior vena caval obstruction in two departments of Thoracic Surgery in consecutive periods. The first 57 patients were observed at La Sapienza University of Rome until 1986 and the other 23 at our present teaching Accepted for publication Feb 2, 1999. Address reprint requests to Dr Mineo, Cattedra di Chirurgia Toracica, Ospedale S. Eugenio, P.le Umanesimo 10, 00144 Rome, Italy; e-mail: [email protected].
© 1999 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 1999;68:223– 6) © 1999 by The Society of Thoracic Surgeons
0003-4975/99/$20.00 PII S0003-4975(99)00455-5
224
MINEO ET AL MEDIASTINOSCOPY IN CAVAL OBSTRUCTION
As far as we know a preliminary bronchoscopic examination was performed in 92 patients, 80 of these in our department. Among this group histologic biopsies were possible in 16 patients, in the other cases diagnosis was attempted by bronchoalveolar lavage cytology. The examination was effective in establishing a diagnosis in 21 patients (26.2%). It was negative in 19 patients and nondiagnostic to start therapy in 40. Fine needle biopsy was done in 24 patients with negative bronchoscopy. Sufficient material to establish a correct histologic diagnosis was obtained only in four instances (16.7%). As a result of the initial screening, 31 patients were not considered for mediastinoscopy as a result of an already known diagnosis by lesser diagnostic examinations (n ⫽ 25) or they refused consent to further examinations (n ⫽ 6). Informed consent for mediastinoscopy was obtained in all other patients. We decided for immediate mediastinoscopy on an urgent basis as first diagnostic procedure in 12 patients. In 51, the mediastinoscopy was performed after less invasive diagnostic techniques (bronchoscopy and fine needle biopsy), which either failed to achieve sufficient sample for a correct diagnosis (n ⫽ 36) or provided negative results (n ⫽ 15). Another 17 patients had been studied elsewhere and referred to us for mediastinoscopy after initiating ineffective therapy (incomplete radiation and chemotherapeutic cycles [n ⫽ 9] or prebiopsy radiotherapy [n ⫽ 8]). Venous pressure values in the brachial district were available in 72 patients immediately before the induction and 6 hours after the procedure. It was calculated by the height (in centimeters) above the right atrium level reached by a column of water in a No. 12 catheter connected to the brachial venous system with the patient lying on supine decubitus. Mediastinoscopy was always conducted with the patient in a supine decubitus position with the head and truncus elevated. The operative field was widely prepared in the eventuality of an emergency thoracotomy. General anesthesia was preferred and the patient was always ventilated through a single lumen endotracheal tube. The operative risk was classified according to Bigsby and colleagues [8] in low (n ⫽ 47) and high risk (n ⫽ 33). Before commencing the procedure, the presence of enlarged scalene nodes was always researched and excluded. Transcervical mediastinoscopy in a superior vena cava obstruction did not substantially differ from the usual procedure, apart from a more careful dissection and accurate hemostasis. As usual, samples were sent for frozen section analysis. The examination ended as soon as the pathologist indicated that the tissue was suitable for histologic diagnosis. In 5 patients, the presence of enlarged prethyroid lymph nodes impeded access to the pretracheal space, but biopsies of these lymph nodes were diagnostic. Statistical evaluation did not include these patients. On the basis of computed tomographic scan findings, in 7 patients further biopsies were provided through a second access during the same operation by a left ante-
Ann Thorac Surg 1999;68:223– 6
rior mediastinotomy. Thus, we were able to reach concomitant left-sided masses, which were the primitive lung cancer (n ⫽ 2) and controlateral neoplastic subaortic lymph nodes (n ⫽ 2). In the last 3 patients anterior mediastinotomy was adopted after a preliminary response of fibrous tissue. This was done to achieve supplementary biopsies from other mediastinal areas. Dependency between survival and subcategories was tested by 2 test. Because of the abnormal distribution of the data, comparison between caval pressure values before and after the procedure was performed by the Wilcoxon test.
Results The procedure, recorded for the last series of 23 patients, had a mean duration of 26.45 minutes including the time for frozen section. No perioperative mortality was recorded. No airway obstruction and cardiovascular collapse at general anesthesia induction were observed. We only noted episodes of mild bronchospasm, which were controlled by steroid injection. Cardiac arrhythmias and temporary arterial hypotension were monitored and successfully treated. We had five cases of major bleeding. Histology in these patients was small cell carcinoma (n ⫽ 3), sclerosis after radiation therapy (n ⫽ 1), and invasive thymoma (n ⫽ 1). Mean venous pressure was 30.8 cm H2O (range, 26 to 38), which is less than the analogous value measured in the entire population at study; types of venous outflow detected in the patients who bled were II (n ⫽ 2), III (n ⫽ 2), and IV (n ⫽ 1). No apparent correlation between risk of bleeding and venous pressure or type of outflow was found. In 4 patients hemorrhage was provoked by tear lesions of the minor venous vessels and it was controlled by local compression and positioning metallic clips or oxidized regenerated cellulose. Only 1 patient required total median sternotomy through which resection of an invasive thymoma and grafting of a superior cava vein by a Gore-Tex (WL Gore & Associates Inc, Flagstaff, AZ) prosthesis were possible. Anesthetic recovery was rapid and all the patients were soon able to breathe spontaneously. The postoperative period was uneventful and wounds healed without major infection. In the last series of 23 patients, 14 procedures were performed on an outpatient basis discharging them after 6 hours of observation. Brachial venous pressure showed a significant decrease between before and after mediastinoscopy mean values by Wilcoxon test (Table 1). Thirty-four patients presented a subjective improvement of symptoms within a few hours after the operation even in the absence of specific therapy. Mean venous pressure diminution was more significant (7.1 versus 3.2 cm H2O; p ⬍ 0.0001) in this group of patients. Biopsies performed by transcervical mediastinoscopy established exact definitive tissue diagnosis in all patients. Fifty-three patients had lung cancer, squamous cell carcinoma in 16, adenocarcinoma in 11, large cell carcinoma in 4, and small cell carcinoma in 22. The
Ann Thorac Surg 1999;68:223– 6
MINEO ET AL MEDIASTINOSCOPY IN CAVAL OBSTRUCTION
225
Table 1. Mediastinoscopic Changes of Venous Pressure Categories Lung cancer Lymphoma Thymoma Type 1–2a Type 3 Type 4 Early improved Total a
No. of Patients
Mean Preoperative Pressure (cm H2O)
Mean Postoperative Pressure (cm H2O)
Significancy (Wilcoxon test)
46 15 6 15 24 33 34 72
32.6 ⫾ 8.29 33.3 ⫾ 8.94 30.0 ⫾ 6.26 24.3 ⫾ 4.83 31.2 ⫾ 6.60 37.1 ⫾ 7.89 33.6 ⫾ 9.34 32.6 ⫾ 8.37
27.4 ⫾ 8.51 30.1 ⫾ 7.49 26.3 ⫾ 6.88 22.1 ⫾ 5.17 26.8 ⫾ 6.22 30.7 ⫾ 9.37 26.5 ⫾ 9.32 27.7 ⫾ 8.29
p ⬍ 0.0001 p ⬍ 0.03 NS NS p ⬍ 0.0006 p ⬍ 0.0001 p ⬍ 0.0001 p ⬍ 0.0001
Type according to the Stanford and Doty [7] classification.
Mean ⫾ standard deviation values for peculiar categories and relative significancy at Wilcoxon test are shown.
majority of the tumors originated in the right upper lobe (n ⫽ 27) and most of these (n ⫽ 21) had contiguity with the cava vein. In the other patients, caval obstruction was provoked by mediastinal lymph nodal involvement. In 3 patients with lung cancer, the syndrome was secondary to sclerosis after radiation therapy and it was completely resolved by steroid administration. In another 17 patients obstruction was secondary to enlarged mediastinal lymph nodes due to lymphoma, 11 nonHodgkin’s and 6 Hodgkin’s. Seven patients had an invasive thymoma, 2 had metastatic lymph nodes from a renal carcinoma, and 1 had fibrosing mediastinitis. In the 12 patients who underwent urgent mediastinoscopy, we were able to start appropriate therapy within 24 hours after the operation. The majority of patients (n ⫽ 71) showed a dramatic improvement of clinical signs of caval obstruction and related symptoms within the first week of treatment. In 7 patients therapy produced only mild or nonsignificant effects and in 2 patients, worsening the general condition. All patients with lung carcinoma for whom we have information (n ⫽ 24) died with a median survival of 6 months. Death was attributable to generalized disease in almost all instances. Only 1 patient died for direct progression of the caval syndrome. At present, 5 of 12 lymphomatous and 4 of the 6 thymomatous patients have remission of disease. In 9 patients with lymphoma we were able to restage the neoplasm after chemotherapy by repeat mediastinoscopy. Once again the examination was easily performed without complications. The 2 test disclosed no significant association between mortality and any of the following factors: Bigsby’s high risk scale, Stanford and Doty III or IV types, caval pressure more than 40 cm H2O, rapid obstruction and postmediastinoscopy brachial venous pressure variation. On the other hand, lung cancer was significantly associated with unfavorable outcome ( p ⬍ 0.0004).
Comment In more than 90% of the patients, caval obstruction is attributable to a malignant disease [8, 9]. Bronchogenic carcinoma is reported in about 80% of these patients,
with lymphoma and metastatic disease comprising 15% and 5%, respectively [8, 9]. In our observations, superior vena cava obstruction arose with the following frequency: 1.73% of lung cancer, 2.77% of the nonHodgkin’s and 3.07% of the Hodgkin’s lymphomas, and 4.25% of the thymomas. Because of the risks related to general anesthetic and local dissection, mediastinoscopy was and still is considered potentially dangerous [3, 9]. In the past, skepticism on the long-term effectiveness of therapeutic efforts suggested a nonaggressive diagnostic conduct limiting the treatment to prebiopsy radiotherapy [10]. In our opinion decompression is the best palliation and it can be better obtained by debulking the mass by appropriate radiation or chemotherapy [4]. To obtain the most effective response to therapy, it is highly desirable to reach a specific tissue diagnosis [11]. Lesser diagnostic procedures may not result in the correct diagnosis. A prescalenic lymph node suitable for excisional biopsy is usually absent. Cytology diagnosis by sputum and fine needle aspiration have often scant sensibility in exactly differentiating the type of tumor [6] and they may be ineffective to start appropriate therapy. Bronchoscopy can be considered the best noninvasive alternative to mediastinoscopy. Nevertheless, it is not risk free [7] and it often requires lavage or brushing cytologic techniques, which are as undiagnostic as sputum and aspiration cytology. In our experience at bronchoscopy we found a low rate (16 of 80 patients [20.0%]) of endoluminal growing lesions available for biopsy and histologic examination. In our opinion mediastinoscopy has a convenient balance of risk and benefit. The examination can provide the highest possible diagnostic accuracy. It can be often performed on as an ambulatory regimen and complication risk is relatively low as shown in Table 2. It is not higher than that in the other patients without caval obstruction [4 – 6, 8, 11–18]. Usually, the lesion causing obstruction is sited in proximity to the thoracic inlet and it is relatively easy to obtain large pieces of tissue for histologic studies. The superior mediastinum is usually not affected by the increased venous pressure as the venous drainage is not directly to the superior vena cava as in the head–neck
226
MINEO ET AL MEDIASTINOSCOPY IN CAVAL OBSTRUCTION
Ann Thorac Surg 1999;68:223– 6
Table 2. Review of the Main Series of Mediastinoscopy in Superior Vena Caval Obstruction Reference Little et al [5] Schraufnagel et al [11] Parish et al [13] Lewis et al [4] Gamondes et al [14] Vanderhoeft [15] Yellin et al [12] Chen et al [16] Callejas et al [17] Bigsby et al [8] Jahangiri and Goldstraw [6] Gamez-Garcia et al [18] Present study
Period (yr)
No. of Patients
No. of Mediastinoscopies
No. of Complications
1960 – 85 1961–79 1961– 81 1966 – 80 1977– 83 1965– 85 1972– 87 1976 – 88 1985– 89 1987–92 1982–93 1974 –96 1974 –97
42 107 86 29 13 93 63 45 8 18 44 29 111
8 14 3 15 12 93 15 1 8 7 34 23 80
0 1 0 0 0 3 1 0 2 2 1 0 5
and the upper extremity. This is one of the reasons why mediastinoscopy is not a difficult procedure in the face of the cava obstruction. The vast majority of the patients with caval obstruction are unlikely to have rapid and life-threatening complications; therefore, we believe that it is worthwhile to try to achieve histologic diagnosis with lesser examinations. Immediate mediastinoscopy must be reserved to those patients who present clinical progression of the caval obstruction in spite of aggressive and prolonged symptomatic therapy. The use of emergency prebiopsy radiotherapy, so-called blind, seems unjustifiable in these conditions. Whenever it was adopted, there were problems in dissection due to fibrosis and in the interpretation of the histology [19]. In 50% of our patients “blind” radiotherapy would have been inappropriate (eg, lymphoma and small cell lung cancer) or harmful as in benign caval obstruction. In conclusion, we consider mediastinoscopy a simple, low risk, and highly effective method in obtaining tissue diagnosis for initiating specific therapy. It should be routinely included after less invasive examinations and in some instances it could be prospected as a first diagnostic urgent procedure.
5.
6. 7. 8. 9. 10. 11. 12. 13. 14.
This study has been supported by a grant from MURST (60%). This study has been carried out within the Research Fellowship Program Dottorato di Ricerca in Oncologia Toracica, appointed by Tor Vergata University of Rome.
15. 16.
References 1. Carlens E. Mediastinoscopy: a method for inspection and tissue biopsy in the superior mediastinum. Dis Chest 1959; 36:343–7. 2. Pearson FG, Nemels JM, Henderson RD, Delarue NC. The role of mediastinoscopy in the selection of treatment for bronchial carcinoma with involvement of superior mediastinal lymph nodes. J Thorac Cardiovasc Surg 1987;64:382–90. 3. Lochridge SK, Knibbe WP, Doty DB. Obstruction of superior vena cava. Surgery 1979;85:14–24. 4. Lewis RJ, Sisler GE, Mackenzie JW. Mediastinoscopy in
17.
18.
19.
advanced superior vena cava obstruction. Ann Thorac Surg 1981;32:458– 62. Little AG, Golomb HM, Ferguson MK, Skosey C, Skinner DB. Malignant superior vena cava obstruction reconsidered: the role of diagnostic surgical intervention. Ann Thorac Surg 1985;40:285– 8. Jahangiri M, Goldstraw P. The role of mediastinoscopy in superior vena caval obstruction. Ann Thorac Surg 1995;59: 453–5. Stanford W, Doty DB. The role of venography and surgery in the management of patients with superior vena cava obstruction. Ann Thorac Surg 1986;41:158– 63. Bigsby R, Greengrass R, Unruh H. Diagnostic algorithm for acute superior vena caval obstruction (SVCO). J Cardiovasc Surg 1993;34:347–50. Nieto AF, Doty DB. Superior vena cava obstruction: clinical syndrome, etiology and treatment. Curr Probl Cancer 1986; 10:441– 81. Davenport D, Ferree C, Blake D, Raben M. Radiation therapy in the treatment of superior vena caval obstruction. Cancer 1978;42:2600–3. Schraufnagel DE, Hill R, Leech JA, Pare JAP. Superior vena cava obstruction: is it a medical emergency? Am J Med 1981; 70:1169–74. Yellin A, Rosen A, Reichert N, Liebermann Y. Superior vena cava syndrome. Am Rev Resp Dis 1990;141:1114– 8. Parish JM, Marshke RF, Dines DE, Lee RE. Etiologic consideration in superior vena cava syndrome. Mayo Clin Proc 1981;56:407–15. Gamondes JP, Vincent M, Weynants P, et al. Mediastinoscopie et syndrome de compression ou d’obstruction de la veine cave superieure. A propos de douze observations. Ann Chir 1985;39:482–5. Vanderhoeft P. La mediastinoscopie dans les syndromes de la veine cave superieure. Ann Chir 1985;39:521. Chen JC, Bongard F, Klein SR. Contemporary perspective on superior vena cava syndrome. Am J Surg 1990;160:207–11. Callejas MA, Rami R, Catalan M, Mainer A, Sanchez-Lloret J. Mediastinoscopy as an emergency diagnostic procedure in superior vena cava syndrome. Scand J Thorac Cardiovasc Surg 1991;25:137–9. Gamez-Garcia AP, Martin de Nicolas Serrahima JL, Marron Fernandez C, Garcia Barajas S, Diaz-Hellin Gude V, Toledo Gonzalez J. Surgical diagnosis procedures in superior vena cava syndrome. Arch Bronconeumol 1997;33:284– 8. Loeffler JS, Leopold KA, Techt A, Weinstein HJ, Tarbell NJ. Emergency prebiopsy radiation for mediastinal masses: impact on subsequent pathologic diagnosis and outcome. J Clin Oncol 1986;4:716–21.