CURRENT REVIEW
Superior
Vena Cava Obstruction
A Review of the Literature and Report of 2 Cases due to Benign Intrathoracic Tumors Mario N. Gomes, M.D., and Charles A. Hufnagel, M.D. ABSTRACT A review of the literature shows an increasing number of cases of superior vena cava obstruction associated with malignancy and a marked decrease in the number of patients with caval obstruction of benign origin. In contrast to granulomatous diseases and aneurysms of the ascending thoracic aorta, which have decreased, the incidence of benign tumors is essentially unchanged. Clinical featuresof superior vena cava obstruction in relation to the acatomical site of obstruction and collateral pathways are correlated. Diagnostic approaches, including angiography and technetium scanning, are usually definitive in outlining the site of obstruction. Experimental data and the numerous avaiIable techniques for surgical correction indicate that an entirely satisfactory procedure is not available for all patients. Methods includethe use of venous bypass or Teflon prostheses and the addition of a small arteriovenous fistula proximally. Two new cases of superior caval obstruction due to benign tumor are reported. In 1patient, who had intrapericardialbronchogenic cyst with fibrotic caval obstruction and thrombosis, a method for caval reconstruction while maintaining venous return to the right atrium is described. The second patient had an intrathoracic thyroid adenoma and caval obstruction without thrombosis.
S
uperior vena cava obstruction secondary to benign intrathoracic tumors is very unusual. Since the first authentic description in 1757 by William Hunter [223 of superior vena caval obstruction due to a luetic saGular aortic aneurysm, multiple other etiologies have been reported. In recent years the presence of superior vena cava syndrome has usually been synonymous with obstruction secondary to malignancy. In a review in 1967, Kamiya and associates 1231, studying 734 patients reported since 1949, verified From the Division of Cardiovascular Surgery, Georgetown University Medical Center, Washington, D.C. Address reprint requests to Dr. Gomes, Division of Cardiovascular Surgery, Georgetown University Medical Center, 3800 Reservoir Road, N.W., Washington, D.C. 20007.
344
T H E ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
Superior Vena Cava Obstruction
this conclusion and found that 540 of them had thoracic malignancy as a cause. Tumors were primary in 521 patients and metastatic in the remainder. Other etiologies were, in order of frequency, mediastinitis (74 patients), thoracic aortic aneurysm (34 patients), and idiopathic thrombophlebitis (24 patients). The cause was unknown in 25 cases, and 37 were due to benign or unclassified tumors. Banker and Maddison [2] compiled a series of patients reported by other authors from 195 1 to 1966 and drew similar conclusions. Their figures include 438 cases, of which 80% were due to malignant tumors. Anaplastic carcinoma of the lung was the cause of the syndrome in 65% of the patients, and other mediastinal malignancies, including lymphoma and metastatic carcinoma, were responsible for the remaining 15%. Similar findings have been reported by various other authors in smaller series [7, 8, 34, 39,461. Szur and Bromley [46] studied 2,000 patients with bronchogenic carcinoma and found 107 with obstruction of the superior vena cava. It has been estimated by others [441 to occur in about 10 to 15% of all bronchogenic carcinomas. A very significant difference is apparent when these figures are compared with older reviews. Before 1949 only one-third of the obstructions were caused by primary intrathoracic malignancies, compared with the last figure compiled by Banker and Maddison [2] of 76%.Of the 502 combined cases reported in 1949 by McIntire and Sykes [30], 30% were secondary to aortic aneurysms in contrast to 4%of the total cases reported since. Luetic aneurysms, the main cause of this syndrome in the past, have rapidly decreased. The incidence of dissecting aneurysm, however, appears to be increasing [45] and may be the cause of a large number of superior vena cava obstructions. We have observed several. Of McIntire’s series [30] 15.4% were due to mediastinal fibrosis, which also has decreased in incidence due to better control of granulomatous diseases. Obstruction of the superior vena cava is essentially due to extrinsic pressure, invasion of vein wall by neoplasm, or thrombosis. Thrombosis of the superior vena cava has been responsible for about 13%of the cases in Calkins’series [8]. Ofthese, 36%were due to phlebitis listed as idiopathic, syphilitic, tuberculous, pyogenic, or traumatic; 29% were secondary to external compression; 23% were due to mediastinitis; and 12% were of unknown cause. With the decrease in granulomatous and infectious lesions of the mediash u m , benign causes of superior vena cava obstruction have become increasingly uncommon. Benign tumors, even though rare, are now the major etiological factor in curable superior vena cava obstructions. In contrast to infectious lesions and aneurysms of the thoracic aorta, the incidence of benign mediastinal tumors causing caval obstruction appears to have remained essentially the same. In the series of McIntire and Sykes, in which 250 patients of their own were reviewed, only 3 cases of benign thoracic tumor were reported. These included 1 dermoid cyst, 1 probable goiter, and 1 benign thymic tumor. They compared this incidence of 1.2%with the 4% in Fischer’s earlier series of 252 patients ( 1 904). In these two combined reports the total percentage of benign thoracic tumors is only 2.6 and is similar to the 2% figure of Banker and Maddison, whose 12 patients represented 9 VOL. 20, NO. 3, SEPTEMBER, 1975
345
GOMES AND HUFNAGEL cases of substernal thyroid, 2 of teratoma, and 1 of atrial myxoma. Kamiya and his co-workers found only 9 cases among the 733 patients they studied. Included among them were 4 intrathoracic goiters [8,11], 2 teratomas [14,24], and isolated cases of mediastinal herniation of a cystic lung [29], constriction by pericardial ring, and neuroblastoma [241.
Clinical Features The signs and symptoms that appear in a well-established case of superior vena cava obstruction are usually easily recognized. Early in its course, however, caval compression may be asymptomatic and minimal signs overlooked or disregarded. Very exceptionally, complete obstruction may take place without noticeable signs or symptoms developing. The clinical picture is due to venous hypertension in the areas normally drained by the superior vena cava or its tributaries, delayed circulation time, development of venous collaterals, and the associated manifestations of the primary pathological process causing the obstruction. Symptoms vary in severity depending upon the degree, location, and rapidity of the obstruction as well as on the development and adequacy of collateral circulation and lack of obstruction in other venous pathways. The main features of this syndrome include dilated venous channels on the trunk, upper extremities, and neck. Edema may also be present in these areas but is not easily apparent. A change in collar size or the size of clothing around the neck, noted by the patient, may be an early sign. The skin may be somewhat cyanotic, flushed, or of a peculiar violet hue. Headache, nausea, dizziness, distortion of vision, hoarseness, stridor, stupor, respiratory distress, and even convulsions may result because of the associated cerebral edema. Periorbital edema with proptosis is seen frequently. Syncope also may occur. All symptoms are increased by bending forward or by the horizontal position, and patients who become dyspneic when recumbent may sleep sitting up. Coughing, dizziness, and increased cyanosis occur when the head is not elevated. The venous return to the heart from the upper half of the body is through four principal collateral pathways and the multiple anastomoses that develop between them. The site of obstruction, especially in relation to the entrance of the azygos vein and availability of alternative venous channels, determines the most prominent pathway. The azygos system, including the azygos and hemiazygos veins with their connecting lumbar veins, is the principal pathway for collateralization when caval obstruction is distal to the entrance of the azygos vein. The internal mammary system, with its tributaries and connections with the superior and inferior epigastric veins, allows drainage into the inferior vena cava by way of the external iliac veins. This pathway is primarily involved when the azygos vein is occluded also. The long thoracic venous system is usually enlarged whenever superior caval obstruction is present. It drains into the femoral veins through connections with the long saphenous system. The vertebral veins and their tributaries are frequently enlarged.
346
T H E ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
Superior Vena Cava Obstruction
These are the most important collateral routes, but others are also available. They are all interconnected and vary in their development in relation to the site and extension of the obstruction. A comprehensive review by Fisher [ 151 showed that among 166 patients, the obstruction occurred above the azygosjunction in 14, included the azygosjunction in 121, and occurred below the azygosjunction in 3 1 . Localizing the obstruction in relation to the azygos vein is important. When both the superior vena cava and the azygos vein are obstructed, an extensive superficial plexus of collateral veins soon becomes prominent and the venous pressure is usually higher. If the obstruction is above the entrance of the azygos vein, the collateral venous pattern is limited to the upper chest or may even be absent. In general, when the collateral channels are given enough time, sufficient blood is shunted around the obstruction to avoid major symptoms under basal conditions. The anatomy of the superior vena cava adequately explains the clinical picture that results from obstruction or compression of this vessel. A major portion of the superior vena cava is enclosed within the pericardial reflection. Klassen and colleagues [24] reported a unique case of obstruction due to constriction at this site. The fixation of the cava within the pericardial sac limits its displacement in the presence of a space-occupying lesion. It is therefore particularly vulnerable to obstruction because it is a thin-walled, low-pressure vessel within this tight compartment. A slowly developing obstruction such as that seen with a benign lesion is tolerated well; a rapidly developing obstruction, as from a malignant lesion, is tolerated poorly. Glushien and Mansuy [ 171 reported a patient with this syndrome who survived 36 years. Among the 23 patients reported by Briickner [71, whenever superior vena cava obstruction had been present for longer than 6 months, the primary lesion was benign. Such a finding may be a clue leading to surgical therapy and cure. One patient with complete obstruction remained asymptomatic for 50 years and died from another cause [291.
Diagnosis Based on the symptoms and clinical findings alone, the diagnosis of superior vena cava obstruction can be made easily when the syndrome is well established. In order to determine the primary cause of the obstruction, its location, and the extension of collateral venous development, however, further investigation is necessary. Accessory diagnostic procedures include determination of venous pressure, roentgenograms, infrared thermography, and superior vena cavography. Venous pressure is measured in the upper and lower extremities to substantiate the diagnosis; it is markedly increased in the upper extremities while normal or even decreased in the lower extremities. This is further accentuated with exercise. Although plain roentgenograms of the chest do not permit the etiology of superior vena cava syndrome to be determined, they do furnish pertinent information on the location of the obvious intrathoracic abnormalities.
VOL. 20, NO. 3, SEPTEMBER, 1975
347
GOMES AND HUFNAGEL Phlebography is a standard procedure for detecting the site of obstruction and delineating collateral flow. It provides accurate information concerning the exact location, the degree of obstruction, and alterations in collateral circulation. However, passage of the contrast medium in high concentration through multiple dilated and tortuous collaterals may produce a greater degree of discomfort than is usually encountered in angiographic studies. Infrared thermography demonstrates the superficial collateral pattern, which may indicate the general site of obstruction. An alternative method utilizes technetium 99m pertechnetate (99mTc). The short-half-life tracers used for these studies have not been reported to produce significant discomfort, morbidity, or mortality. While venous angiography provides excellent anatomical detail, the g 9 m Tstudy ~ gives an estimation of flow. In a series of 20 patients with obstruction of the superior vena cava studied by this technique [28], the correlation was excellent in the patients in whom venous angiography was also performed. The main advantage of the 99mTc method is that it can be used for serial evaluation of the response to therapy without additional morbidity. Its high degree of accuracy in detecting vascular blockage and its lack of morbidity seem to make superior vena cava cavography with 9 g m T a~ useful procedure for initial screening and follow-up of patients with superior vena cava obstruction.
Treatment No specific surgical method has yet been described that is fully satisfactory, and the large number of approaches reported reflects the lack of uniform therapeutic success. Direct vascular operation has usually consisted of resection of the thrombosed vena cava with reconstruction or bypass of the caval obstruction by a shunt. Since the initial description of endvenectomy by ONeil [36], only a few similar reports have followed [4, 13, 20,471. They demonstrate the feasibility of removing an obstructing lesion from the lumen of the superior vena cava to restore normal blood flow and afford relief of superior vena cava syndrome. The results seem to indicate that endvenectomy and thrombectomy should be attempted before resorting to graft bypass or replacement. Arterial allografts, aortic allografts or xenografts, fresh aortic isografts, caval allografts, venous autografts, and Teflon, Dacron, and Ivalon prostheses have all been extensively studied in the laboratory with varying degrees of success. Early stenosis and thrombosis with collapse of the graft have been the main causes of dissatisfaction with many of these prostheses. These experiments have not clearly indicated the best material to be used as a superior vena cava substitute, but they do suggest that it should be semirigid, of considerable size, and not readily compressible or subject to fibrosis. In some reports [ 10,2?, 351 aortic allografts seem to have produced the most consistently good results. In contrast to arterial replacements, late aneurysm formation is not likely to occur in the low-pressure-venoussystem, and narrowing 348
T H E ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
Superior Vena Cava Obstruction
of the allograft by atheromatous deposits should not commonly occur. The mediastinum also provides quite a suitable bed for the allograft, permitting vascular and tissue ingrowth. Other authors [41] consider vein graft replacement preferable. Kay 1361 in 1948 experimented with vena cava allografts. The initial patency rate was excellent, but at 6 months all the grafts showed extensive fibrosis and none were adequate when the animals were sacrificed later. More recently Moore and associates [33] carried out a study to evaluate the suitability of fresh caval allografts for superior vena cava replacement in dogs and their histological fate in this low-pressure system. Patency was 71% from 2 to 467 days, and no evidence of graft dilatation or aneurysm formation in any of the patent allografts was found at postmortem examination. Careful histological examination suggested that there was no progressive alteration or degeneration of the grafts after a 51-day postoperative interval. These findings give added support to the cavogram observations that fresh allografts which remain patent in the superior vena cava after an early critical period may be expected to remain patent. Vein isografts preferred initially (1947) by Gerbode and his associates [16] have been used clinically with similar success. A composite saphenous vein graft constructed by splitting the veins longitudinally, resuturing, and thus doubling the lumen size has been used [18, 43, 481. Autogenous fascia grafts were studied in dogs and found to be uniformly occluded within 3 weeks to 3 months after operation when used as vena cava replacements [6]. The technical aspects are much more important in venous grafting than in arterial replacement, requiring precise fitting and anastomosing as well as delicate handling. Experimental work by Lloyd-DSilva and associates [25], testing different types of anastomoses, suggests that an end-to-end type of squarely cut prosthesis offers a better chance for long-term patency. They also indicate that the best location for a graft is in the vena cava, placing it as an end-to-end anastomosis at each end, whether square or oblique. The second choice seems to be between the superior vena cava and atrial appendage, and the third, between the vena cava and atrial wall. Early experimental use of synthetic tubes for superior vena cava replacement was reported in 1955 by Deterling and Bhonslay [lo]. They employed Ivalon tubes, cylinders made from sheets of woven Dacron or nylon, and woven seamless tubes of nylon, all with disappointing results. Encouraging short-term experience with the use of Teflon prostheses for superior vena cava replacement has been reported by Moore and co-workers [32] as well as by Botham and his associates 151. They showed that in dogs the cava may be replaced with a relatively rigid crimped Teflon prosthesis, slightly to moderately larger than the replaced caval segment, with an expected success rate in the neighborhood of 90% (follow-up to 14 months in some). Using smaller grafts the patency rate was lower (around 65%). Further reports by Hache [19], Hill [213, and Lloyd-D’Silva [251 and their associates, among others, attest to the efficacy of the material used. The relatively rigid nature of the graft, along with crimping, VOL. 20, NO. 3, SEPTEMBER, 1975
349
GOMES AND HUFNAGEL serves to maintain the lumen of this low-pressure venous substitute, rendering it less vulnerable to compression by extraneous forces. The percentage of graft patency seems to increase when a temporary carotid artery-to-external jugular vein fistula is added. This was particularly studied by Scheinin and Jude [40] after previous experiments had shown that the best results with superior venacava grafts occurred when the azygos vein had been ligated and the volume and rate of venous flow through the graft thereby increased. They determined in dogs the effect on graft patency of increased blood flow induced by a temporary peripheral arteriovenous fistula. The fistula was maintained only during the crucial period immediately after replacement - 1 month - but seemed quite effective in reducing the incidence of thrombosis. With Teflon or autogenous vein there were no failures in the fistula series, and even with such a relatively poor caval substitute as Dacron the increased blood flow was successful in maintaining patency in 60% of the experiments. Grafts of all types that were patent at 1 month remained patent, with no instances of subsequent narrowing or late thrombosis. Thrombosis occurred in 67% of the control group without the peripheral arteriovenous fistula and in only 12%of the animals with a temporary fistula. The perfect venacava substitute and grafting procedure have yet to be found. When faced with the need for rerouting blood from the superior vena cava to the right atrium, one must be guided by the anatomical situation. The following types of shunts have been reported to reverse the syndrome successfully. 1. A graft from the superior vena cava or innominate vein to the right atrial
appendage [161. 2. A graft from the superior vena cava or innominate vein to the intrapericardial superior vena cava [I]. 3. Anastomosis of the azygos vein to the right atrial appendage [16] or intrapericardial superior vena cava by direct suture or interposed venous graft 1243. 4. Anastomosis of the azygos vein directly to the inferior vena cava [9]. 5. A graft from the jugular vein to the right atrial appendage [381 or intrapericardial superior vena cava. A simpler rerouting from the upper to the lower caval system is provided by Schramel and Olinde’s [41] subcutaneous bridge from the external jugular to the femoral vein using a reversed long saphenous vein. This technique is much simpler than intrathoracic procedures and may be preferable in patients with nonresectable malignancy. The current availability of large-caliber bovine grafts makes the possibility of this type of bypass quite attractive. The internal jugular, subclavian, or axillary veins are alternative sites for the proximal anastomosis in specific situations. The temporary addition of an arteriovenous fistula from the external carotid artery would be helpful in assuring long-term patency. Berman and co-workers [3] described a simple method of extracorporeal shunting for temporary relief of symptomatic superior vena cava obstruction. A 350
THE ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
Superwr Vena Caua Obstruction
long Silastic tube was inserted into the distal superior vena cava through the saphenous vein and connected with another cannula inserted through the brachial vein into the proximal superior vena cava. This extracorporeal shunt was continued intermittently for 72 hours. Three weeks after completion of the shunting procedure symptoms of caval obstruction recurred in this patient. Such a technique could be of special benefit to patients with advanced malignant disease in whom other procedures are of limited value or rarely justified. Considering the anatomical features and especially the relation of the block to the azygos vein, any one of the above procedures can be used in patients with benign obstruction of the superior vena cava or in idiopathic mediastinal fibrosis for relief of the obstructed channel. Selected patients with low-grade radioresistant malignancy may also profit from decompression.
Case Reports Two new cases of benign caval obstruction are reported here. As far as we can ascertain from reviewing the literature, our Patient 1 is the first to be reported in whom obstruction of the superior vena cava was due to an intrapericardial bronchogenic cyst. Patient 2 had a substernal goiter causing compression of the superior vena cava. PATIENT 1
A 45-year-old woman was admitted with the diagnosis of superior vena cava obstruction. Slight distention of the neck veins, moderate edema of the head and neck, and cyanosis on recumbency at the same areas were noted. The chest roentgenogram showed a rounded density at the right mediastinal border over the junction of the superior vena cava and right atrium. Laminagraphy of this region (Fig. 1) confirmed the finding. Angiography showed gradual tapering of the superior vena cava with complete occlusion at its entrance into the right atrium. Lateral caval displacement to the right was also present. A very large azygos vein was demonstrated draining the superior into the inferior vena cava system (Fig. 2). Another catheter was passed into the right atrium from the femoral vein, and upon injection of the contrast medium, a smooth, crescentshaped filling defect was seen displacing approximately 30% of the right atrium on its superior and medial aspect. All intracardiac right heart pressures were within normal range, but a 14 mm Hg pressure differential was present between the superior vena cava and right atrium. Exploration of the heart and mediastinum was performed through a midline sternotomy incision. The pericardium was opened. A cyst was found that was compressing the medial aspect of the superior vena cava and superior aspect of the right atrium and that was markedly adherent to the anterior pericardium. It was rounded and approximately 10 to 12 cm wide, and it extended posteriorly to the right atrial appendage, was adherent to the left atrium, and spread over the anterior and inferior surfaces of the right pulmonary artery (Fig. 3). The cyst was dissected from the cava, both atria, and the right pulmonary artery. After its
VOL. 20, NO. 3, SEPTEMBER, 1975
351
GOMES AND HUFNAGEL
FIG. 1 . (Patient 1.) Laminagram showing mass along the right superior medktinal border.
removal the superior vena cava remained completely obstructed at the point of previous compression. Reconstruction of the vena cava was then undertaken. Umbilical tapes were looped around the superior vena cava, proximally near the right atrium and distally at the junction of the innominate veins as well as around the azygos vein. The proximal and distal cava and azygos veins were occluded. A longitudinal incision was then made in the cava, and the involved portion was found to be thickened, narrowed, and completely obstructed by compression and thrombus (Fig. 4). Following thromboendvenectomy, a 40F catheter was passed into the right atrial appendage and upward into the patent cava through the area of narrowing. The tapes were tightened around the catheter, which had been previously prepared by cutting an additional hole in that portion resting in the right atrium. The superior vena cava was bypassed because to reconstruct the cava it is necessary to occlude the azygos vein, the collateral route for all of the superior venous bed. The superior vena cava was reconstructed over the bypassing catheter using an ovoid patch of pericardium extending from the base of the right atrium to a point just inferior to the junction of the innominate veins (Fig. 5). Following operation the venous pressures in the upper extremities returned to normal and have remained that way over the last 3 years. The postoperative period was uncomplicated, and the patient has remained asymptomatic. A pathological diagnosis of bronchogenic cyst was made. PATIENT 2
A 61-year-old woman was admitted with the diagnosis of superior vena cava obstruction. She had noted an enlarged thyroid gland for many years, and 3 years prior to this admission a mediastinal mass had been seen in a preemployment
352
T H E ANNALS OF THORACIC SURGERY
CURRENT REVIEW: Superior
Vena Cava Obstruction
FIG. 2. (Patient 1 .) Venogram of the superior vena cava showing smooth conical caval constriction and dilated azygos vein. (SVC = superior vena cava; RA = right atrium.)
chest roentgenogram. On a recent physical examination prompted by increasing fatigue, many enlarged veins were present over the upper thorax. Further enlargement of the upper mediastinal mass was seen in consecutive roentgenograms over the 3-year period. N o symptoms, including dysphagia or respiratory difficulty, were present. There was no evidence of hyperthyroidism or hypot hyroidism. Relevant findings were bilateral periorbital edema and less pronounced conjunctival injection. The thyroid gland was firm, multinodular, and enlarged to approximately three times normal size; it was palpable at the suprasternal area, FIG. 3. (Patient 1 . ) Thrombosed superior vena cava in relation to the intrapericardial cyst.
vena cava
atrial appendage
VOL. 20, NO. 3, S E n E M B E R , 1975
3 53
GOMES AND HUFNAGEL
FIG. 4 . (Patient I .) Extraction of thrombus by endvewctomy prior to cavoatrial intubation.
but no clear delineation of the lower poles could be ascertained. The external jugular veins were markedly distended, and very prominent, tortuous, distended veins were also noted on the upper chest. All thyroid studies confirmed normal thyroid function. The thyroid scan demonstrated a massively enlarged gland with a generalized irregular uptake in which almost half of each lobe was located below the level of the suprasternal notch. On the chest roentgenogram a large mediastinal mass (Fig. 6), smooth in contour, was present within the superior mediastinum. The trachea was slightly deviated to the left. Laminagrams confirmed these findings and showed that the mass, although located mostly anteriorly, also extended posteriorly. A retrograde arch aortogram (Fig. 7) demonstrated a superior mediastinal mass between the innominate and left carotid arteries that markedly displaced these vessels laterally but did not compress them. A superior vena cavogram showed definite compression of the right innominate vein and superior vena cava (Fig. 8); later studies demonstrated an increased collateral venous pattern. With the diagnosis of substernal goiter the patient was operated upon. The tumor was satisfactorily removed through a midsternotomy incision. Relief of compression allowed the cava to resume its normal caliber. There was no evidence of thrombosis or scarring. The previous signs of caval compression disappeared, and the patient has remained asymptomatic over the last 2% years. Pathological examination showed 240 gm of thyroid tissue with features of nodular adenomatous goiter and old hemorrhage. 354
THE ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
FIG. 5 . (Patient 1.) Caua intubated for patch angioplasty. Catheter has holes aboue and below the sites of the caual tourniquets, permitting blood to flow from the innominate veins to the right atrium.
Superior Vena Cava Obstruction
atiial appendage
Comment Effler and Groves [l 11 have pointed out that most patients with chronic benign obstruction of the superior vena cava adapt satisfactorily to their symptoms and usually do not require surgical therapy, though some have persistent disability. Symptomatic treatment until sufficient collateral circulation devel-
FIG. 6. (Patient 2.) Roentgenogram of the chest showing superb- mediastinal mass extending into the neck.
VOL. 20, NO. 3, SEPTEMBER, 1975
355
GOMES AND HUFNAGEL
FIG. 7. (Patient 2.) Arch aortogram shows displacement of the innominate and carotid arteries by the mass.
ops therefore is usually sufficient. The collateral circulation has a remarkable capacity to circumvent block when it is located above the entry of the azygos vein. When occlusion is below the azygos vein entry, and especially when it obliterates the entire superior vena cava including the azygos entry, there is a greater possibility that collateral circulation will be inadequate and noticeable symptoms may persist. In these circumstances,operative intervention may be indicated if the procedure is considered safe and offers a good possibility of sustained relief. When removing the cause of the obstruction will not in itself restore flow through the cava, relief of symptoms may require caval reconstruction or bypass. Superior vena cava syndrome secondary to malignant disease is usually not suitable for curative resection, and operations to relieve obstruction in these circumstances rarely achieve either early or lasting success; approximately 85%of these patients are terminal. Indications for operation in such circumstances would include the need for a definite diagnosis to be established by biopsy or a good possibility that progressive and severe symptoms could be relieved by vascular reconstruction or bypass. Because of the dubious long-term results, some workers think decompression is achieved more effectively by the patient’s own collateral circulation than by any operation. The first patient reported here is an example of a rare case of longstanding caval obstruction, with recurrent symptoms due to progressive changes occurring in the cyst associated with fibrotic constriction and secondary caval thrombosis. The technique of removing the thrombus and widening the cava was a simple and very satisfactory method. The cavoatrial bypass permitted the azygos collateral system to be occluded without the danger of markedly elevating intracranial venous pressures or decreasing cardiac venous return during the repair. The second patient represents an unusual manifestation of the relatively common problem of substernal thyroid. The caval compression in these instances
356
THE ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
Superior Vena Cava Obstruction
FIG. 8. (Patient 2.) Venogram illustrating compression of the right innominate vein and superior vena cava.
follows the m o r e usual pattern of simple extrinsic obstruction without permanent constriction of the cava. The recent onset o f caval obstruction was probably associated with hemorrhage into t h e goiter, which was evident on gross a n d microscopical examination. T h e s e patients illustrate t h e two extremes of caval compression by benign lesions. T h e simple extrinsic compression produced by a large intrathoracic goiter typifies relief of obstruction by removal of t h e compressing mass. The intrapericardial cyst producing not only caval thrombosis but severe constriction was not corrected by removal of the primary lesion alone. The method used provides a simple way o f treating caval obstruction o f benign origin and can be adapted t o the specific anatomical variations associated with this problem.
References 1 . Allansmith, R., and Richards, V.
2. 3. 4. 5. 6. 7. 8.
Superior vena caval obstruction. AmJ Surg 96:353, 1958. Banker, V. P., and Maddison, F. E. Superior vena cava syndrome secondary to aortic disease: Report of two cases and review of the literature. Dis Chest 51:656, 1967. Berman, I. R., Mergenthaler, F. W., and Clauss, R. H. An extracorporeal venous shunting procedure for the symptomatic relief of superior vena caval obstruction and report of a case. Ann Surg 167:269, 1968. Blondeau, P., Wapler, C., Piwnica, A., and DuBost, C. Deux cas de syndrome de la veine cava supkrieure traitks chirurgicalement avec succes, I'un par dksobstruction, l'autre par greffe. Arch Ma1 Coeur 52:504, 1959. Botham, R. J., Dracopoulus, T. T., and Gale, J. W. Superior vena caval replacement with sponge Teflon prosthesis. J Thorac Cardiouasc Surg 39:202, 1960. Brea, C. A., Allen, M. S., and Muller, W. H. Superior vena cava replacement with autogenous fascia: An experimental study. Am Surg 31:611, 1965. Bruckner, W. J. Significance of the superior vena caval syndrome. Arch Intern Med 102:88, 1958. Calkins, E. A, The superior vena caval syndrome: Report of twenty-one cases. Dis Chest 30:404,1956.
VOL. 20, NO. 3, SEPTEMBER, 1975
357
GOMES AND HUFNAGEL 9. Cooley, D. A., and Hallman, G. L. Superior vena caval syndrome treated by azygos vein-inferior vena cava anastomosis. J Thorac Cardiovasc Surg 43:574, 1962. 10. Deterling, R. A., Jr., and Bhonslay, S. B. Use of vessel grafts and plastic prostheses for relief of superior vena cava obstruction. Surgery 38:1008, 1955. 11. Effler, D. B., and Groves, L. K. Superior vena caval obstruction.J Thorac Cardiovasc Surg 43:574, 1962. 12. Emery, W. F. Dissecting aneurysm with an account of 24 cases. Med J Aust 2:690, 1946. 13. Fadhli, H. A. Endvenectomy and decompression in fibrosing mediastinitis causing obstruction of the superior vena cava.J Thorac Cardiovasc Surg 53:881, 1967. 14. Failor, H. J., Edwards, J. E., and Hodgson, C. H. Etiologic factors in obstruction of the superior vena cava: A pathologic study. Proc Mayo Clin 33:671, 1958. 15. Fisher, J. (Quoted by McIntire [30].) U ber Verengerung und Verschliesung der Vena Cava Superior. Inaugural dissertation, Halle, 1904. 16. Gerbode, F., Yee, J., and Rundle, F. Experimental anastomosis of vessels to the heart. Surgery 25:556, 1949. 17. Glushien, A. S., and Mansuy, M. M. Superior vena cava obstruction with survival after 36 years. Angiology 2:210, 1951. 18. Gundersen, E. (Quoted by Skinner [43]. Personal communication. 19. Hache, L., Woolner, L. B., and Bernatz, P. E. Idiopathic fibrous mediastinitis. Db Chest 41:9, 1962. 20. Hanlon, C. R., and Davis, R. K. Superior vena cava obstruction: Indications for diagnostic thoracotomy. Am Surg 161:771, 1965. 21. Hill, L. D., Lawrence, G. H., and Herron, P. Surgical management of obstruction of the superior vena cava. Db Chest 42:198, 1962. 22. Hunter, W. The history of an aneurysm of the aorta with some remarks on aneurysms in general. Med Observ Znq (Lond) 1:323, 1957. 23. Kamiya, K., Nakata, Y., Naiki, K., and Hayashi, H. Superior vena cava syndrome. Vasc Dis 4:59, 1967. 24. Klassen, K. P., Andrews, N. C., and Curtis, G. M. Diagnosis and treatment of superior vena cava obstruction. Arch Surg 63:311, 1951. 25. Lloyd-D'Silva,J., Crampton, A. R., and Hohf, R. P. Superior vena caval prosthesis: A comparison of types and site of anastomosis.J Thorac Cardiovasc Surg 48:276, 1964. 26. Lowenburg, E. L., and Hotchkiss, W. S. The superior vena cava syndrome. J Cardiovasc Surg (Torino) 4:536, 1963. 27. Lowenburg, E. L., Hotchkiss, W., and McDaniel, S. The superior vena cava syndrome: Diagnostic and surgical considerations (case reports). Dis Chest 47:323, 1965. 28. Maxfield, W. S., and Meckstroth, G. R. Technetium-99m superior vena cavography. Radiology 92:913, 1969. 29. McCord, M. C., Edlin, P., and Block, M. Superior vena caval system obstruction. D b Chest 19:19, 1951. 30. McIntire, F. T., and Sykes, E. M., Jr. Obstruction of the superior vena cava: A review of the literature and report of two personal cases. Ann Intern Med 30:925, 1949. 31. Moore, T. C., and Riberi, A. Superior vena caval replacement: Successful use of fresh autogenous aorta. Surgery 44:898, 1958. 32. Moore, T. C., Teramoto, S., and Heimburger, I. L. Successful use of Teflon grafts for superior vena cava replacement. Surg Gynecol Obstet 111:475, 1960. 33. Moore, T. C., Teramoto, S., Heimburger, I. L., and Harshman, J. A. Superior vena cava replacement: Use of fresh homografts of superior vena cava.J Thorac Cardiovasc Surg 42:379, 1961. 34. Roswit, B., Kaplan, G., and Jacobson, H. G. The superior vena caval syndrome in bronchogenic carcinoma: Pathologic, physiologic and therapeutic management. Radiolorn 61:722. 1953. 35. Salyer,?. C., and Riberi, A. Superior vena caval replacement: Successful use of fresh autogenous aorta. Surgery 44:898, 1958. 36. Samson, P., ONeill, T., Kay, E., Hanlon, R., Klassen, K., Deterling, R., and Gerbode, F. Discussion of Scannell and Shaw [38]. 37. Scannell, J. G. Etiology and surgical approaches in superior vena caval obstruction. Radiology 81:378, 1963. 358
THE ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
Superior Vena Cava Obstruction
38. Scannell, J. G., and Shaw, R. S. Surgical reconstruction of superior vena cava. J Thorac Surg 28:163, 1954. 39. Schechter, M. M. The superior vena caval syndrome. Am J Med Sci 227:46, 1954. 40. Scheinin, T. M., and Jude, J. R. Experimental replacement of the superior venacava: Effect of temporary increase in blood flow.J Thoruc Cardiouasc Surg 48:781, 1964. 41. Schramel, R., and Olinde, H. D. A new method of bypassing the obstructed vena cava. J Thorac Cardiouasc Surg 41:375, 1961. 42. Silverstein, G. E., Burke, G., Goldberg, D., and Halko, A. Superior vena caval system obstruction caused by benign endothoracic goiter. Dis Chest 56:519, 1969. 43. Skinner, D. B., Salzman, E. W., and Scannell, J. G. The challenge of superior vena caval obstruction. J Thorac Cardiouasc Surg 49:824, 1965. 44. Steinberg, I., and Dotter, C. T. Lung cancer: Angiocardiographic findings in 100 consecutive proved cases. Arch Surg 64: 10, 1952. 45. Strax, T. E., Ryvicker, M. J., and Elguezabal, A. Superior vena caval syndrome due to a mediastinal hematoma secondary to a dissecting aortic aneurysm. Dis Chest 55:338, 1969. 46. Szur, L., and Bromley, L. L. Superior vena cava in carcinoma of the bronchus. Br Med J 2:1273, 1956. 47. Templeton, J. Y. Endvenectomy for the relief of obstruction of the superior vena cava. Am J Surg 104:70, 1962. 48. Urschel, H. C., and Paulson, D. L. Superior vena caval obstruction. Dis Chest 49:155, 1966. 49. Yong, N. K., Klein, N., and Moore, T. C. Superior vena cava replacement: Experimental use of fresh autologous vein grafts. Br J Surg 51:374, 1964.
VOL. 20, NO. 3, SEPTEMBER, 1975
3 59
Superior
Vena Cava Obstruction
A Review of the Literature and Report of 2 Cases due to Benign Intrathoracic Tumors Mario N. Gomes, M.D., and Charles A. Hufnagel, M.D. ABSTRACT A review of the literature shows an increasing number of cases of superior vena cava obstruction associated with malignancy and a marked decrease in the number of patients with caval obstruction of benign origin. In contrast to granulomatous diseases and aneurysms of the ascending thoracic aorta, which have decreased, the incidence of benign tumors is essentially unchanged. Clinical featuresof superior vena cava obstruction in relation to the acatomical site of obstruction and collateral pathways are correlated. Diagnostic approaches, including angiography and technetium scanning, are usually definitive in outlining the site of obstruction. Experimental data and the numerous avaiIable techniques for surgical correction indicate that an entirely satisfactory procedure is not available for all patients. Methods includethe use of venous bypass or Teflon prostheses and the addition of a small arteriovenous fistula proximally. Two new cases of superior caval obstruction due to benign tumor are reported. In 1patient, who had intrapericardialbronchogenic cyst with fibrotic caval obstruction and thrombosis, a method for caval reconstruction while maintaining venous return to the right atrium is described. The second patient had an intrathoracic thyroid adenoma and caval obstruction without thrombosis.
S
uperior vena cava obstruction secondary to benign intrathoracic tumors is very unusual. Since the first authentic description in 1757 by William Hunter [223 of superior vena caval obstruction due to a luetic saGular aortic aneurysm, multiple other etiologies have been reported. In recent years the presence of superior vena cava syndrome has usually been synonymous with obstruction secondary to malignancy. In a review in 1967, Kamiya and associates 1231, studying 734 patients reported since 1949, verified From the Division of Cardiovascular Surgery, Georgetown University Medical Center, Washington, D.C. Address reprint requests to Dr. Gomes, Division of Cardiovascular Surgery, Georgetown University Medical Center, 3800 Reservoir Road, N.W., Washington, D.C. 20007.
344
T H E ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
Superior Vena Cava Obstruction
this conclusion and found that 540 of them had thoracic malignancy as a cause. Tumors were primary in 521 patients and metastatic in the remainder. Other etiologies were, in order of frequency, mediastinitis (74 patients), thoracic aortic aneurysm (34 patients), and idiopathic thrombophlebitis (24 patients). The cause was unknown in 25 cases, and 37 were due to benign or unclassified tumors. Banker and Maddison [2] compiled a series of patients reported by other authors from 195 1 to 1966 and drew similar conclusions. Their figures include 438 cases, of which 80% were due to malignant tumors. Anaplastic carcinoma of the lung was the cause of the syndrome in 65% of the patients, and other mediastinal malignancies, including lymphoma and metastatic carcinoma, were responsible for the remaining 15%. Similar findings have been reported by various other authors in smaller series [7, 8, 34, 39,461. Szur and Bromley [46] studied 2,000 patients with bronchogenic carcinoma and found 107 with obstruction of the superior vena cava. It has been estimated by others [441 to occur in about 10 to 15% of all bronchogenic carcinomas. A very significant difference is apparent when these figures are compared with older reviews. Before 1949 only one-third of the obstructions were caused by primary intrathoracic malignancies, compared with the last figure compiled by Banker and Maddison [2] of 76%.Of the 502 combined cases reported in 1949 by McIntire and Sykes [30], 30% were secondary to aortic aneurysms in contrast to 4%of the total cases reported since. Luetic aneurysms, the main cause of this syndrome in the past, have rapidly decreased. The incidence of dissecting aneurysm, however, appears to be increasing [45] and may be the cause of a large number of superior vena cava obstructions. We have observed several. Of McIntire’s series [30] 15.4% were due to mediastinal fibrosis, which also has decreased in incidence due to better control of granulomatous diseases. Obstruction of the superior vena cava is essentially due to extrinsic pressure, invasion of vein wall by neoplasm, or thrombosis. Thrombosis of the superior vena cava has been responsible for about 13%of the cases in Calkins’series [8]. Ofthese, 36%were due to phlebitis listed as idiopathic, syphilitic, tuberculous, pyogenic, or traumatic; 29% were secondary to external compression; 23% were due to mediastinitis; and 12% were of unknown cause. With the decrease in granulomatous and infectious lesions of the mediash u m , benign causes of superior vena cava obstruction have become increasingly uncommon. Benign tumors, even though rare, are now the major etiological factor in curable superior vena cava obstructions. In contrast to infectious lesions and aneurysms of the thoracic aorta, the incidence of benign mediastinal tumors causing caval obstruction appears to have remained essentially the same. In the series of McIntire and Sykes, in which 250 patients of their own were reviewed, only 3 cases of benign thoracic tumor were reported. These included 1 dermoid cyst, 1 probable goiter, and 1 benign thymic tumor. They compared this incidence of 1.2%with the 4% in Fischer’s earlier series of 252 patients ( 1 904). In these two combined reports the total percentage of benign thoracic tumors is only 2.6 and is similar to the 2% figure of Banker and Maddison, whose 12 patients represented 9 VOL. 20, NO. 3, SEPTEMBER, 1975
345
GOMES AND HUFNAGEL cases of substernal thyroid, 2 of teratoma, and 1 of atrial myxoma. Kamiya and his co-workers found only 9 cases among the 733 patients they studied. Included among them were 4 intrathoracic goiters [8,11], 2 teratomas [14,24], and isolated cases of mediastinal herniation of a cystic lung [29], constriction by pericardial ring, and neuroblastoma [241.
Clinical Features The signs and symptoms that appear in a well-established case of superior vena cava obstruction are usually easily recognized. Early in its course, however, caval compression may be asymptomatic and minimal signs overlooked or disregarded. Very exceptionally, complete obstruction may take place without noticeable signs or symptoms developing. The clinical picture is due to venous hypertension in the areas normally drained by the superior vena cava or its tributaries, delayed circulation time, development of venous collaterals, and the associated manifestations of the primary pathological process causing the obstruction. Symptoms vary in severity depending upon the degree, location, and rapidity of the obstruction as well as on the development and adequacy of collateral circulation and lack of obstruction in other venous pathways. The main features of this syndrome include dilated venous channels on the trunk, upper extremities, and neck. Edema may also be present in these areas but is not easily apparent. A change in collar size or the size of clothing around the neck, noted by the patient, may be an early sign. The skin may be somewhat cyanotic, flushed, or of a peculiar violet hue. Headache, nausea, dizziness, distortion of vision, hoarseness, stridor, stupor, respiratory distress, and even convulsions may result because of the associated cerebral edema. Periorbital edema with proptosis is seen frequently. Syncope also may occur. All symptoms are increased by bending forward or by the horizontal position, and patients who become dyspneic when recumbent may sleep sitting up. Coughing, dizziness, and increased cyanosis occur when the head is not elevated. The venous return to the heart from the upper half of the body is through four principal collateral pathways and the multiple anastomoses that develop between them. The site of obstruction, especially in relation to the entrance of the azygos vein and availability of alternative venous channels, determines the most prominent pathway. The azygos system, including the azygos and hemiazygos veins with their connecting lumbar veins, is the principal pathway for collateralization when caval obstruction is distal to the entrance of the azygos vein. The internal mammary system, with its tributaries and connections with the superior and inferior epigastric veins, allows drainage into the inferior vena cava by way of the external iliac veins. This pathway is primarily involved when the azygos vein is occluded also. The long thoracic venous system is usually enlarged whenever superior caval obstruction is present. It drains into the femoral veins through connections with the long saphenous system. The vertebral veins and their tributaries are frequently enlarged.
346
T H E ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
Superior Vena Cava Obstruction
These are the most important collateral routes, but others are also available. They are all interconnected and vary in their development in relation to the site and extension of the obstruction. A comprehensive review by Fisher [ 151 showed that among 166 patients, the obstruction occurred above the azygosjunction in 14, included the azygosjunction in 121, and occurred below the azygosjunction in 3 1 . Localizing the obstruction in relation to the azygos vein is important. When both the superior vena cava and the azygos vein are obstructed, an extensive superficial plexus of collateral veins soon becomes prominent and the venous pressure is usually higher. If the obstruction is above the entrance of the azygos vein, the collateral venous pattern is limited to the upper chest or may even be absent. In general, when the collateral channels are given enough time, sufficient blood is shunted around the obstruction to avoid major symptoms under basal conditions. The anatomy of the superior vena cava adequately explains the clinical picture that results from obstruction or compression of this vessel. A major portion of the superior vena cava is enclosed within the pericardial reflection. Klassen and colleagues [24] reported a unique case of obstruction due to constriction at this site. The fixation of the cava within the pericardial sac limits its displacement in the presence of a space-occupying lesion. It is therefore particularly vulnerable to obstruction because it is a thin-walled, low-pressure vessel within this tight compartment. A slowly developing obstruction such as that seen with a benign lesion is tolerated well; a rapidly developing obstruction, as from a malignant lesion, is tolerated poorly. Glushien and Mansuy [ 171 reported a patient with this syndrome who survived 36 years. Among the 23 patients reported by Briickner [71, whenever superior vena cava obstruction had been present for longer than 6 months, the primary lesion was benign. Such a finding may be a clue leading to surgical therapy and cure. One patient with complete obstruction remained asymptomatic for 50 years and died from another cause [291.
Diagnosis Based on the symptoms and clinical findings alone, the diagnosis of superior vena cava obstruction can be made easily when the syndrome is well established. In order to determine the primary cause of the obstruction, its location, and the extension of collateral venous development, however, further investigation is necessary. Accessory diagnostic procedures include determination of venous pressure, roentgenograms, infrared thermography, and superior vena cavography. Venous pressure is measured in the upper and lower extremities to substantiate the diagnosis; it is markedly increased in the upper extremities while normal or even decreased in the lower extremities. This is further accentuated with exercise. Although plain roentgenograms of the chest do not permit the etiology of superior vena cava syndrome to be determined, they do furnish pertinent information on the location of the obvious intrathoracic abnormalities.
VOL. 20, NO. 3, SEPTEMBER, 1975
347
GOMES AND HUFNAGEL Phlebography is a standard procedure for detecting the site of obstruction and delineating collateral flow. It provides accurate information concerning the exact location, the degree of obstruction, and alterations in collateral circulation. However, passage of the contrast medium in high concentration through multiple dilated and tortuous collaterals may produce a greater degree of discomfort than is usually encountered in angiographic studies. Infrared thermography demonstrates the superficial collateral pattern, which may indicate the general site of obstruction. An alternative method utilizes technetium 99m pertechnetate (99mTc). The short-half-life tracers used for these studies have not been reported to produce significant discomfort, morbidity, or mortality. While venous angiography provides excellent anatomical detail, the g 9 m Tstudy ~ gives an estimation of flow. In a series of 20 patients with obstruction of the superior vena cava studied by this technique [28], the correlation was excellent in the patients in whom venous angiography was also performed. The main advantage of the 99mTc method is that it can be used for serial evaluation of the response to therapy without additional morbidity. Its high degree of accuracy in detecting vascular blockage and its lack of morbidity seem to make superior vena cava cavography with 9 g m T a~ useful procedure for initial screening and follow-up of patients with superior vena cava obstruction.
Treatment No specific surgical method has yet been described that is fully satisfactory, and the large number of approaches reported reflects the lack of uniform therapeutic success. Direct vascular operation has usually consisted of resection of the thrombosed vena cava with reconstruction or bypass of the caval obstruction by a shunt. Since the initial description of endvenectomy by ONeil [36], only a few similar reports have followed [4, 13, 20,471. They demonstrate the feasibility of removing an obstructing lesion from the lumen of the superior vena cava to restore normal blood flow and afford relief of superior vena cava syndrome. The results seem to indicate that endvenectomy and thrombectomy should be attempted before resorting to graft bypass or replacement. Arterial allografts, aortic allografts or xenografts, fresh aortic isografts, caval allografts, venous autografts, and Teflon, Dacron, and Ivalon prostheses have all been extensively studied in the laboratory with varying degrees of success. Early stenosis and thrombosis with collapse of the graft have been the main causes of dissatisfaction with many of these prostheses. These experiments have not clearly indicated the best material to be used as a superior vena cava substitute, but they do suggest that it should be semirigid, of considerable size, and not readily compressible or subject to fibrosis. In some reports [ 10,2?, 351 aortic allografts seem to have produced the most consistently good results. In contrast to arterial replacements, late aneurysm formation is not likely to occur in the low-pressure-venoussystem, and narrowing 348
T H E ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
Superior Vena Cava Obstruction
of the allograft by atheromatous deposits should not commonly occur. The mediastinum also provides quite a suitable bed for the allograft, permitting vascular and tissue ingrowth. Other authors [41] consider vein graft replacement preferable. Kay 1361 in 1948 experimented with vena cava allografts. The initial patency rate was excellent, but at 6 months all the grafts showed extensive fibrosis and none were adequate when the animals were sacrificed later. More recently Moore and associates [33] carried out a study to evaluate the suitability of fresh caval allografts for superior vena cava replacement in dogs and their histological fate in this low-pressure system. Patency was 71% from 2 to 467 days, and no evidence of graft dilatation or aneurysm formation in any of the patent allografts was found at postmortem examination. Careful histological examination suggested that there was no progressive alteration or degeneration of the grafts after a 51-day postoperative interval. These findings give added support to the cavogram observations that fresh allografts which remain patent in the superior vena cava after an early critical period may be expected to remain patent. Vein isografts preferred initially (1947) by Gerbode and his associates [16] have been used clinically with similar success. A composite saphenous vein graft constructed by splitting the veins longitudinally, resuturing, and thus doubling the lumen size has been used [18, 43, 481. Autogenous fascia grafts were studied in dogs and found to be uniformly occluded within 3 weeks to 3 months after operation when used as vena cava replacements [6]. The technical aspects are much more important in venous grafting than in arterial replacement, requiring precise fitting and anastomosing as well as delicate handling. Experimental work by Lloyd-DSilva and associates [25], testing different types of anastomoses, suggests that an end-to-end type of squarely cut prosthesis offers a better chance for long-term patency. They also indicate that the best location for a graft is in the vena cava, placing it as an end-to-end anastomosis at each end, whether square or oblique. The second choice seems to be between the superior vena cava and atrial appendage, and the third, between the vena cava and atrial wall. Early experimental use of synthetic tubes for superior vena cava replacement was reported in 1955 by Deterling and Bhonslay [lo]. They employed Ivalon tubes, cylinders made from sheets of woven Dacron or nylon, and woven seamless tubes of nylon, all with disappointing results. Encouraging short-term experience with the use of Teflon prostheses for superior vena cava replacement has been reported by Moore and co-workers [32] as well as by Botham and his associates 151. They showed that in dogs the cava may be replaced with a relatively rigid crimped Teflon prosthesis, slightly to moderately larger than the replaced caval segment, with an expected success rate in the neighborhood of 90% (follow-up to 14 months in some). Using smaller grafts the patency rate was lower (around 65%). Further reports by Hache [19], Hill [213, and Lloyd-D’Silva [251 and their associates, among others, attest to the efficacy of the material used. The relatively rigid nature of the graft, along with crimping, VOL. 20, NO. 3, SEPTEMBER, 1975
349
GOMES AND HUFNAGEL serves to maintain the lumen of this low-pressure venous substitute, rendering it less vulnerable to compression by extraneous forces. The percentage of graft patency seems to increase when a temporary carotid artery-to-external jugular vein fistula is added. This was particularly studied by Scheinin and Jude [40] after previous experiments had shown that the best results with superior venacava grafts occurred when the azygos vein had been ligated and the volume and rate of venous flow through the graft thereby increased. They determined in dogs the effect on graft patency of increased blood flow induced by a temporary peripheral arteriovenous fistula. The fistula was maintained only during the crucial period immediately after replacement - 1 month - but seemed quite effective in reducing the incidence of thrombosis. With Teflon or autogenous vein there were no failures in the fistula series, and even with such a relatively poor caval substitute as Dacron the increased blood flow was successful in maintaining patency in 60% of the experiments. Grafts of all types that were patent at 1 month remained patent, with no instances of subsequent narrowing or late thrombosis. Thrombosis occurred in 67% of the control group without the peripheral arteriovenous fistula and in only 12%of the animals with a temporary fistula. The perfect venacava substitute and grafting procedure have yet to be found. When faced with the need for rerouting blood from the superior vena cava to the right atrium, one must be guided by the anatomical situation. The following types of shunts have been reported to reverse the syndrome successfully. 1. A graft from the superior vena cava or innominate vein to the right atrial
appendage [161. 2. A graft from the superior vena cava or innominate vein to the intrapericardial superior vena cava [I]. 3. Anastomosis of the azygos vein to the right atrial appendage [16] or intrapericardial superior vena cava by direct suture or interposed venous graft 1243. 4. Anastomosis of the azygos vein directly to the inferior vena cava [9]. 5. A graft from the jugular vein to the right atrial appendage [381 or intrapericardial superior vena cava. A simpler rerouting from the upper to the lower caval system is provided by Schramel and Olinde’s [41] subcutaneous bridge from the external jugular to the femoral vein using a reversed long saphenous vein. This technique is much simpler than intrathoracic procedures and may be preferable in patients with nonresectable malignancy. The current availability of large-caliber bovine grafts makes the possibility of this type of bypass quite attractive. The internal jugular, subclavian, or axillary veins are alternative sites for the proximal anastomosis in specific situations. The temporary addition of an arteriovenous fistula from the external carotid artery would be helpful in assuring long-term patency. Berman and co-workers [3] described a simple method of extracorporeal shunting for temporary relief of symptomatic superior vena cava obstruction. A 350
THE ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
Superwr Vena Caua Obstruction
long Silastic tube was inserted into the distal superior vena cava through the saphenous vein and connected with another cannula inserted through the brachial vein into the proximal superior vena cava. This extracorporeal shunt was continued intermittently for 72 hours. Three weeks after completion of the shunting procedure symptoms of caval obstruction recurred in this patient. Such a technique could be of special benefit to patients with advanced malignant disease in whom other procedures are of limited value or rarely justified. Considering the anatomical features and especially the relation of the block to the azygos vein, any one of the above procedures can be used in patients with benign obstruction of the superior vena cava or in idiopathic mediastinal fibrosis for relief of the obstructed channel. Selected patients with low-grade radioresistant malignancy may also profit from decompression.
Case Reports Two new cases of benign caval obstruction are reported here. As far as we can ascertain from reviewing the literature, our Patient 1 is the first to be reported in whom obstruction of the superior vena cava was due to an intrapericardial bronchogenic cyst. Patient 2 had a substernal goiter causing compression of the superior vena cava. PATIENT 1
A 45-year-old woman was admitted with the diagnosis of superior vena cava obstruction. Slight distention of the neck veins, moderate edema of the head and neck, and cyanosis on recumbency at the same areas were noted. The chest roentgenogram showed a rounded density at the right mediastinal border over the junction of the superior vena cava and right atrium. Laminagraphy of this region (Fig. 1) confirmed the finding. Angiography showed gradual tapering of the superior vena cava with complete occlusion at its entrance into the right atrium. Lateral caval displacement to the right was also present. A very large azygos vein was demonstrated draining the superior into the inferior vena cava system (Fig. 2). Another catheter was passed into the right atrium from the femoral vein, and upon injection of the contrast medium, a smooth, crescentshaped filling defect was seen displacing approximately 30% of the right atrium on its superior and medial aspect. All intracardiac right heart pressures were within normal range, but a 14 mm Hg pressure differential was present between the superior vena cava and right atrium. Exploration of the heart and mediastinum was performed through a midline sternotomy incision. The pericardium was opened. A cyst was found that was compressing the medial aspect of the superior vena cava and superior aspect of the right atrium and that was markedly adherent to the anterior pericardium. It was rounded and approximately 10 to 12 cm wide, and it extended posteriorly to the right atrial appendage, was adherent to the left atrium, and spread over the anterior and inferior surfaces of the right pulmonary artery (Fig. 3). The cyst was dissected from the cava, both atria, and the right pulmonary artery. After its
VOL. 20, NO. 3, SEPTEMBER, 1975
351
GOMES AND HUFNAGEL
FIG. 1 . (Patient 1.) Laminagram showing mass along the right superior medktinal border.
removal the superior vena cava remained completely obstructed at the point of previous compression. Reconstruction of the vena cava was then undertaken. Umbilical tapes were looped around the superior vena cava, proximally near the right atrium and distally at the junction of the innominate veins as well as around the azygos vein. The proximal and distal cava and azygos veins were occluded. A longitudinal incision was then made in the cava, and the involved portion was found to be thickened, narrowed, and completely obstructed by compression and thrombus (Fig. 4). Following thromboendvenectomy, a 40F catheter was passed into the right atrial appendage and upward into the patent cava through the area of narrowing. The tapes were tightened around the catheter, which had been previously prepared by cutting an additional hole in that portion resting in the right atrium. The superior vena cava was bypassed because to reconstruct the cava it is necessary to occlude the azygos vein, the collateral route for all of the superior venous bed. The superior vena cava was reconstructed over the bypassing catheter using an ovoid patch of pericardium extending from the base of the right atrium to a point just inferior to the junction of the innominate veins (Fig. 5). Following operation the venous pressures in the upper extremities returned to normal and have remained that way over the last 3 years. The postoperative period was uncomplicated, and the patient has remained asymptomatic. A pathological diagnosis of bronchogenic cyst was made. PATIENT 2
A 61-year-old woman was admitted with the diagnosis of superior vena cava obstruction. She had noted an enlarged thyroid gland for many years, and 3 years prior to this admission a mediastinal mass had been seen in a preemployment
352
T H E ANNALS OF THORACIC SURGERY
CURRENT REVIEW: Superior
Vena Cava Obstruction
FIG. 2. (Patient 1 .) Venogram of the superior vena cava showing smooth conical caval constriction and dilated azygos vein. (SVC = superior vena cava; RA = right atrium.)
chest roentgenogram. On a recent physical examination prompted by increasing fatigue, many enlarged veins were present over the upper thorax. Further enlargement of the upper mediastinal mass was seen in consecutive roentgenograms over the 3-year period. N o symptoms, including dysphagia or respiratory difficulty, were present. There was no evidence of hyperthyroidism or hypot hyroidism. Relevant findings were bilateral periorbital edema and less pronounced conjunctival injection. The thyroid gland was firm, multinodular, and enlarged to approximately three times normal size; it was palpable at the suprasternal area, FIG. 3. (Patient 1 . ) Thrombosed superior vena cava in relation to the intrapericardial cyst.
vena cava
atrial appendage
VOL. 20, NO. 3, S E n E M B E R , 1975
3 53
GOMES AND HUFNAGEL
FIG. 4 . (Patient I .) Extraction of thrombus by endvewctomy prior to cavoatrial intubation.
but no clear delineation of the lower poles could be ascertained. The external jugular veins were markedly distended, and very prominent, tortuous, distended veins were also noted on the upper chest. All thyroid studies confirmed normal thyroid function. The thyroid scan demonstrated a massively enlarged gland with a generalized irregular uptake in which almost half of each lobe was located below the level of the suprasternal notch. On the chest roentgenogram a large mediastinal mass (Fig. 6), smooth in contour, was present within the superior mediastinum. The trachea was slightly deviated to the left. Laminagrams confirmed these findings and showed that the mass, although located mostly anteriorly, also extended posteriorly. A retrograde arch aortogram (Fig. 7) demonstrated a superior mediastinal mass between the innominate and left carotid arteries that markedly displaced these vessels laterally but did not compress them. A superior vena cavogram showed definite compression of the right innominate vein and superior vena cava (Fig. 8); later studies demonstrated an increased collateral venous pattern. With the diagnosis of substernal goiter the patient was operated upon. The tumor was satisfactorily removed through a midsternotomy incision. Relief of compression allowed the cava to resume its normal caliber. There was no evidence of thrombosis or scarring. The previous signs of caval compression disappeared, and the patient has remained asymptomatic over the last 2% years. Pathological examination showed 240 gm of thyroid tissue with features of nodular adenomatous goiter and old hemorrhage. 354
THE ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
FIG. 5 . (Patient 1.) Caua intubated for patch angioplasty. Catheter has holes aboue and below the sites of the caual tourniquets, permitting blood to flow from the innominate veins to the right atrium.
Superior Vena Cava Obstruction
atiial appendage
Comment Effler and Groves [l 11 have pointed out that most patients with chronic benign obstruction of the superior vena cava adapt satisfactorily to their symptoms and usually do not require surgical therapy, though some have persistent disability. Symptomatic treatment until sufficient collateral circulation devel-
FIG. 6. (Patient 2.) Roentgenogram of the chest showing superb- mediastinal mass extending into the neck.
VOL. 20, NO. 3, SEPTEMBER, 1975
355
GOMES AND HUFNAGEL
FIG. 7. (Patient 2.) Arch aortogram shows displacement of the innominate and carotid arteries by the mass.
ops therefore is usually sufficient. The collateral circulation has a remarkable capacity to circumvent block when it is located above the entry of the azygos vein. When occlusion is below the azygos vein entry, and especially when it obliterates the entire superior vena cava including the azygos entry, there is a greater possibility that collateral circulation will be inadequate and noticeable symptoms may persist. In these circumstances,operative intervention may be indicated if the procedure is considered safe and offers a good possibility of sustained relief. When removing the cause of the obstruction will not in itself restore flow through the cava, relief of symptoms may require caval reconstruction or bypass. Superior vena cava syndrome secondary to malignant disease is usually not suitable for curative resection, and operations to relieve obstruction in these circumstances rarely achieve either early or lasting success; approximately 85%of these patients are terminal. Indications for operation in such circumstances would include the need for a definite diagnosis to be established by biopsy or a good possibility that progressive and severe symptoms could be relieved by vascular reconstruction or bypass. Because of the dubious long-term results, some workers think decompression is achieved more effectively by the patient’s own collateral circulation than by any operation. The first patient reported here is an example of a rare case of longstanding caval obstruction, with recurrent symptoms due to progressive changes occurring in the cyst associated with fibrotic constriction and secondary caval thrombosis. The technique of removing the thrombus and widening the cava was a simple and very satisfactory method. The cavoatrial bypass permitted the azygos collateral system to be occluded without the danger of markedly elevating intracranial venous pressures or decreasing cardiac venous return during the repair. The second patient represents an unusual manifestation of the relatively common problem of substernal thyroid. The caval compression in these instances
356
THE ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
Superior Vena Cava Obstruction
FIG. 8. (Patient 2.) Venogram illustrating compression of the right innominate vein and superior vena cava.
follows the m o r e usual pattern of simple extrinsic obstruction without permanent constriction of the cava. The recent onset o f caval obstruction was probably associated with hemorrhage into t h e goiter, which was evident on gross a n d microscopical examination. T h e s e patients illustrate t h e two extremes of caval compression by benign lesions. T h e simple extrinsic compression produced by a large intrathoracic goiter typifies relief of obstruction by removal of t h e compressing mass. The intrapericardial cyst producing not only caval thrombosis but severe constriction was not corrected by removal of the primary lesion alone. The method used provides a simple way o f treating caval obstruction o f benign origin and can be adapted t o the specific anatomical variations associated with this problem.
References 1 . Allansmith, R., and Richards, V.
2. 3. 4. 5. 6. 7. 8.
Superior vena caval obstruction. AmJ Surg 96:353, 1958. Banker, V. P., and Maddison, F. E. Superior vena cava syndrome secondary to aortic disease: Report of two cases and review of the literature. Dis Chest 51:656, 1967. Berman, I. R., Mergenthaler, F. W., and Clauss, R. H. An extracorporeal venous shunting procedure for the symptomatic relief of superior vena caval obstruction and report of a case. Ann Surg 167:269, 1968. Blondeau, P., Wapler, C., Piwnica, A., and DuBost, C. Deux cas de syndrome de la veine cava supkrieure traitks chirurgicalement avec succes, I'un par dksobstruction, l'autre par greffe. Arch Ma1 Coeur 52:504, 1959. Botham, R. J., Dracopoulus, T. T., and Gale, J. W. Superior vena caval replacement with sponge Teflon prosthesis. J Thorac Cardiouasc Surg 39:202, 1960. Brea, C. A., Allen, M. S., and Muller, W. H. Superior vena cava replacement with autogenous fascia: An experimental study. Am Surg 31:611, 1965. Bruckner, W. J. Significance of the superior vena caval syndrome. Arch Intern Med 102:88, 1958. Calkins, E. A, The superior vena caval syndrome: Report of twenty-one cases. Dis Chest 30:404,1956.
VOL. 20, NO. 3, SEPTEMBER, 1975
357
GOMES AND HUFNAGEL 9. Cooley, D. A., and Hallman, G. L. Superior vena caval syndrome treated by azygos vein-inferior vena cava anastomosis. J Thorac Cardiovasc Surg 43:574, 1962. 10. Deterling, R. A., Jr., and Bhonslay, S. B. Use of vessel grafts and plastic prostheses for relief of superior vena cava obstruction. Surgery 38:1008, 1955. 11. Effler, D. B., and Groves, L. K. Superior vena caval obstruction.J Thorac Cardiovasc Surg 43:574, 1962. 12. Emery, W. F. Dissecting aneurysm with an account of 24 cases. Med J Aust 2:690, 1946. 13. Fadhli, H. A. Endvenectomy and decompression in fibrosing mediastinitis causing obstruction of the superior vena cava.J Thorac Cardiovasc Surg 53:881, 1967. 14. Failor, H. J., Edwards, J. E., and Hodgson, C. H. Etiologic factors in obstruction of the superior vena cava: A pathologic study. Proc Mayo Clin 33:671, 1958. 15. Fisher, J. (Quoted by McIntire [30].) U ber Verengerung und Verschliesung der Vena Cava Superior. Inaugural dissertation, Halle, 1904. 16. Gerbode, F., Yee, J., and Rundle, F. Experimental anastomosis of vessels to the heart. Surgery 25:556, 1949. 17. Glushien, A. S., and Mansuy, M. M. Superior vena cava obstruction with survival after 36 years. Angiology 2:210, 1951. 18. Gundersen, E. (Quoted by Skinner [43]. Personal communication. 19. Hache, L., Woolner, L. B., and Bernatz, P. E. Idiopathic fibrous mediastinitis. Db Chest 41:9, 1962. 20. Hanlon, C. R., and Davis, R. K. Superior vena cava obstruction: Indications for diagnostic thoracotomy. Am Surg 161:771, 1965. 21. Hill, L. D., Lawrence, G. H., and Herron, P. Surgical management of obstruction of the superior vena cava. Db Chest 42:198, 1962. 22. Hunter, W. The history of an aneurysm of the aorta with some remarks on aneurysms in general. Med Observ Znq (Lond) 1:323, 1957. 23. Kamiya, K., Nakata, Y., Naiki, K., and Hayashi, H. Superior vena cava syndrome. Vasc Dis 4:59, 1967. 24. Klassen, K. P., Andrews, N. C., and Curtis, G. M. Diagnosis and treatment of superior vena cava obstruction. Arch Surg 63:311, 1951. 25. Lloyd-D'Silva,J., Crampton, A. R., and Hohf, R. P. Superior vena caval prosthesis: A comparison of types and site of anastomosis.J Thorac Cardiovasc Surg 48:276, 1964. 26. Lowenburg, E. L., and Hotchkiss, W. S. The superior vena cava syndrome. J Cardiovasc Surg (Torino) 4:536, 1963. 27. Lowenburg, E. L., Hotchkiss, W., and McDaniel, S. The superior vena cava syndrome: Diagnostic and surgical considerations (case reports). Dis Chest 47:323, 1965. 28. Maxfield, W. S., and Meckstroth, G. R. Technetium-99m superior vena cavography. Radiology 92:913, 1969. 29. McCord, M. C., Edlin, P., and Block, M. Superior vena caval system obstruction. D b Chest 19:19, 1951. 30. McIntire, F. T., and Sykes, E. M., Jr. Obstruction of the superior vena cava: A review of the literature and report of two personal cases. Ann Intern Med 30:925, 1949. 31. Moore, T. C., and Riberi, A. Superior vena caval replacement: Successful use of fresh autogenous aorta. Surgery 44:898, 1958. 32. Moore, T. C., Teramoto, S., and Heimburger, I. L. Successful use of Teflon grafts for superior vena cava replacement. Surg Gynecol Obstet 111:475, 1960. 33. Moore, T. C., Teramoto, S., Heimburger, I. L., and Harshman, J. A. Superior vena cava replacement: Use of fresh homografts of superior vena cava.J Thorac Cardiovasc Surg 42:379, 1961. 34. Roswit, B., Kaplan, G., and Jacobson, H. G. The superior vena caval syndrome in bronchogenic carcinoma: Pathologic, physiologic and therapeutic management. Radiolorn 61:722. 1953. 35. Salyer,?. C., and Riberi, A. Superior vena caval replacement: Successful use of fresh autogenous aorta. Surgery 44:898, 1958. 36. Samson, P., ONeill, T., Kay, E., Hanlon, R., Klassen, K., Deterling, R., and Gerbode, F. Discussion of Scannell and Shaw [38]. 37. Scannell, J. G. Etiology and surgical approaches in superior vena caval obstruction. Radiology 81:378, 1963. 358
THE ANNALS OF THORACIC SURGERY
CURRENT REVIEW:
Superior Vena Cava Obstruction
38. Scannell, J. G., and Shaw, R. S. Surgical reconstruction of superior vena cava. J Thorac Surg 28:163, 1954. 39. Schechter, M. M. The superior vena caval syndrome. Am J Med Sci 227:46, 1954. 40. Scheinin, T. M., and Jude, J. R. Experimental replacement of the superior venacava: Effect of temporary increase in blood flow.J Thoruc Cardiouasc Surg 48:781, 1964. 41. Schramel, R., and Olinde, H. D. A new method of bypassing the obstructed vena cava. J Thorac Cardiouasc Surg 41:375, 1961. 42. Silverstein, G. E., Burke, G., Goldberg, D., and Halko, A. Superior vena caval system obstruction caused by benign endothoracic goiter. Dis Chest 56:519, 1969. 43. Skinner, D. B., Salzman, E. W., and Scannell, J. G. The challenge of superior vena caval obstruction. J Thorac Cardiouasc Surg 49:824, 1965. 44. Steinberg, I., and Dotter, C. T. Lung cancer: Angiocardiographic findings in 100 consecutive proved cases. Arch Surg 64: 10, 1952. 45. Strax, T. E., Ryvicker, M. J., and Elguezabal, A. Superior vena caval syndrome due to a mediastinal hematoma secondary to a dissecting aortic aneurysm. Dis Chest 55:338, 1969. 46. Szur, L., and Bromley, L. L. Superior vena cava in carcinoma of the bronchus. Br Med J 2:1273, 1956. 47. Templeton, J. Y. Endvenectomy for the relief of obstruction of the superior vena cava. Am J Surg 104:70, 1962. 48. Urschel, H. C., and Paulson, D. L. Superior vena caval obstruction. Dis Chest 49:155, 1966. 49. Yong, N. K., Klein, N., and Moore, T. C. Superior vena cava replacement: Experimental use of fresh autologous vein grafts. Br J Surg 51:374, 1964.
VOL. 20, NO. 3, SEPTEMBER, 1975
3 59
