Mediastinal Mass with Superior Vena Cava Syndrome

Primary cardiac tumors do not necessarily require anticoagulation. However, anticoagulation is indicated for any cardiac mass in which associated thrombus is suspected, or the underlying substrate is deemed to be high risk for thrombus formation. For example, a typical infant fibroma does not warrant anticoagulation, but if the tumor is associated with poor ventricular function and/or stasis within a cardiac chamber, anticoagulation should be considered. Likewise, any tumor with mobile extension(s), appearing at risk for embolization, warrants anticoagulation. Magnetic resonance imaging (MRI) is regarded as a superior modality for characterizing cardiac masses, but differentiation from thrombus remains a challenge. If there is doubt—and no significant contraindications to anticoagulation exist—then anticoagulation is often initiated. ¹

Cardiac tumors and mediastinal masses with intracardiac extension raise concern for embolic phenoma, which can cause vascular occlusion as well as tumor seeding. Vascular occlusion can be mitigated by anticoagulation, but systemic anticoagulation is not necessarily indicated for all masses involving the heart. Although there are no published studies to guide this decision, influential factors are the degree of mobility of the intracardiac component, size and shape (pedunculation), debris evident on echocardiography, or evidence of embolization such as the demonstration of pulmonary hypertension. Masses that are particularly large or invasive of myocardium may diminish contractile function; the resulting diminished ejection fraction and/or low-flow currents lead to pooling of blood, which also predisposes to thrombus formation.

The matter of etiology/histology is a critical one, because it makes clear what actual therapy is indicated. In general, approaching treatment of neoplasms begins with staging: determination of tumor type/histology, local extent, regional spread, and distant metastases. With a markedly invasive mass such as this—causing clear physiologic derangement—tissue diagnosis is particularly important in directing therapy. The differential diagnosis includes several neoplasms, each of which respond vastly different to neoadjuvant therapies.

In this patient, one presumes that the mass is a recurrent malignant peripheral nerve sheath tumor (MPNST) based on the history of neurofibromatosis with prior neurofibromas, and MPNST. However, the more common masses of this region would be lymphomas or sarcomas. While a MPNST would require resection, these latter entities are typically more susceptible to neoadjuvant therapies and could therefore be treated with chemotherapy, steroids, or radiotherapy. The upfront treatment in sarcoma and lymphoma patients can provide rapid relief of immediate life-threatening obstruction thus avoiding a technically challenging and high-morbidity operation. Though the likelihood of a sarcoma or lymphoma is small in this patient, the difference in treatment approach would be dramatic.

The subacute symptoms of superior vena caval obstruction required prompt intervention, but since she was in metastable condition without respiratory or cardiac support, we deferred emergent resection and proceeded with establishing histologic diagnosis, using core needle technique. Transeso-phageal echocardiography was also required to better characterize the mass, define its intracardiac extent, and evaluate for evidence of pulmonary embolization. These procedures were performed under the same anesthetic with standard induction techniques to minimize compressive complications of the mediastinal mass (eg, reverse Trendelenberg position, spontaneous ventilation).^2-4

The identification of synchronous or metastatic disease is critical because it helps inform the discussion with the patient and family regarding goals of care, and provides a context for interpreting the risks and benefits of the operation. The presence of synchronous or metastatic tumors may severely affect expected longevity and quality of life. If there is advanced synchronous or metastatic disease, which is anticipated to lead to short-term death, it is critical to discuss the risks and morbidity of the operation and whether it is warranted. If the distant disease is not immediately life threatening, it would seem more reasonable to proceed with aggressive treatment of the mediastinal tumor. In this patient’s case, she did have evidence of distant disease, but not immediately life threatening. In light of these findings, she elected to proceed with aggressive resection of the mediastinal mass to help restore her physical endurance and improve her quality of life.

There are several important anesthetic considerations for this patient that relate both to the anterior mediastinal mass and the SVC obstruction/syndrome. ⁵ Induction of anesthesia can compromise both the cardiovascular status of the patient, but importantly, can also compromise the integrity of the airway already marginalized due to tissue edema (secondary to the SVC syndrome) as well as possible direct compression by the tumor. ⁶ As such, a detailed preoperative assessment of the airway and cardiovascular system is required. This should include both visual inspection of the upper airway, as well by review of the chest computed tomography (CT) scan to examine the relationship between the tumor and the major intrathoracic airways. In addition, echocardiography is critical to understanding the dynamic relationship of possible tumor compressive effects. Importantly, both CT and echo can be performed in various patient positions, further evaluating the variable compression by the mass of the major vessels (SVC, pulmonary artery, aorta) and the heart that can vary with patient position. These imaging modalities can also aid in assessing for the presence of concomitant pleural and pericardial effusions. In theory, respiratory flow-volume loops may help define the dynamic extrinsic compression of the tumor on the major airways. However, in reality there is general consensus that flow-volume loops offer very little practical information. ⁷

The presence and severity of cardiorespiratory distress is of primary concern. Orthopnea and the refusal to lie flat are key signals of severe respiratory distress. The rapid development of facial, neck, and upper extremity swelling suggests significant venous obstruction. Chest radiographs provided evidence of a widened mediastinum, tracheobronchial compression, pleural effusion, and elevation of the right hemidiaphragm. CT scan more accurately defines the extent of tracheal compression. CT angiography and/or MRI also inform the clinician of vascular anatomy. Echocardiography and cardiac MRI can provide details of cardiac function as well as the location of the mass in relation to cardiovascular structures. Unfortunately, dyspnea and discomfort associated with SVC syndrome may preclude MRI studies. Although pulmonary function tests (PFTs) including flow volume loops have historically been recommended prior to surgery, Hnatiuk et al ⁸ found that these tests are ordered sporadically, may not be reviewed by the surgeon and/or anesthesiologist prior to surgery, and do not reliably predict complications. Chest radiography (CXR), CT, and cardiac echocardiography should provide adequate preoperative data for optimal anesthetic management.

Patients with rapidly progressive disease processes often have significant anxiety related to dyspnea and, in this case, disfigurement related to SVC syndrome, and pain. Multiple other neurofibromas have contributed to severe, long-standing chronic pain and narcotic tolerance. Successful management of acute postsurgical pain will be challenging. This patient’s baseline narcotic therapy—including fentanyl patch, oxycodone IR (immediate release) and SR (sustained release), and hydromorphone—provided the patient with a morphine equivalent dose of nearly 800 mg (12 mg/kg/d) orally, or 3.5 mg/kg/d parenterally.

With respect to monitoring, the pre-anesthetic induction placement of invasive arterial pressure monitoring is critical. Central venous catheterization, although ideal, is optional before induction of anesthesia and should be guided by the urgency of the need for treatment and other issues related to patient comfort during insertion. However, securing vascular access, as well as invasive monitoring is ultimately critical, as the operative procedure itself can be fraught with hemodynamic compromise, due to tumor compression as well as the possibility of life-threatening blood loss. Accordingly, large bore intravenous (IV) access, both above and below the level of the tumor is critical. In particular, because of the SVC syndrome, IV access in the inferior vena cava system is important as the SVC distribution will be compromised by the tumor obstruction and SVC resection.

The induction of anesthesia can be carried out in a number of different ways. ⁹ Each of these depends on the size of the tumor and severity of the patient’s symptoms. In the cooperative patient, it may be advantageous to perform an awake intubation after topicalization of the airway in order to maintain an airway that may be compromised by the tumor, as well as to maintain spontaneous ventilation, which may be important in minimizing both the extrinsic compression of the airway and the hemodynamic compromise of positive pressure ventilation. However, in the uncooperative patient, an inhalational induction may also be considered. An IV induction might have to be pursued, however, but should only be done with the immediate availability of resuscitation drugs, as well as the possibility for sternotomy and/or cannulation for cardiopulmonary bypass (CPB) if cardiac arrest should be imminent. CPB standby, however, should not be relied on as it can take a prolonged period (>10 minutes) to institute should arrest occur. Pre-induction femoral cannulation for CPB is also an option but is difficult in the patient unable to be positioned supine and also requires heparinization.

For diagnostic biopsy of mediastinal masses, the literature consistently recommends local anesthesia with sedation and spontaneous ventilation. However, when surgical resection is planned, general anesthesia is required and evidence-based recommendations are limited. Induction of anesthesia should be conducted in the position that optimizes ventilation and hemodynamics. This may require the seated or semiseated (Fowler’s) position. Adjustments to position should occur gradually with continuous reassessment of the effect of change on cardiorespiratory stability. If the supine position is not tolerated, the lateral decubitus or prone positions may relieve the external compression contributing to hemodynamic or respiratory compromise. Spontaneous ventilation and a gradual progression to deeper levels of anesthesia are advisable. The use of either intravenous or inhalational agents is acceptable. Airway compromise has been reported in association with change of position, loss of spontaneous ventilation, endotracheal intubation, and muscle relaxation. There are rare reports of acute cardiovascular collapse associated with anesthesia and mediastinal masses. If preoperative assessment suggests that compromise of venous return or cardiac output might result in cardiovascular collapse, the immediate availability of extracorporeal membrane oxygenation or CPB is indicated.

In addition to the standard ASA (American Society of Anesthesiologists) monitors, this patient will require invasive arterial pressure monitoring, which was placed after sedation during assisted positive pressure ventilation in the semi-Fowler’s position. Central venous access is indicated for both monitoring and the administration of inotropic agents. One may consider the relative merits of measuring central pressures in the inferior vena cava against pressures in the superior circulation proximal to the SVC mass. In this case, we placed a left internal jugular line whose distal tip lay in the left brachiocephalic vein.

Unilateral paralysis of the diaphragm is well compensated in older patients unless an underlying respiratory or neuromuscular disease further compromises respiratory function. The diaphragm is relatively resistant to the depressant effects of inhalation anesthetics and muscle relaxants. The supine position is associated with elevation of the diaphragm, diminished functional residual capacity (FRC) and atelectasis when the closing capacity is greater than FRC and these effects are exaggerated under inhalation anesthesia. However, these phenomena are unlikely to result in acute respiratory collapse. Negative intrathoracic pressures associated with spontaneous ventilation increase venous return and cardiac output and may preserve the patency of the tracheobronchial tree in the setting of external compression by tumor or other mass. Rigid bronchoscopy is useful when complete airway obstruction occurs. Advancing the bronchoscope beyond the point of obstruction has permitted adequate ventilation in the setting of near total tracheal or bronchial compression. ¹⁰ A slow transition from spontaneous to assisted positive pressure ventilation is usually well tolerated. If not, one may return to spontaneous ventilation as long as muscle relaxants have not been administered. Spontaneous respiration is not always necessary, but the opportunity to return to spontaneous ventilation should remain prior to sternotomy.

Rigid bronchoscopy is often discussed as an option for obtaining an airway in patients with mediastinal masses. Because of direct compression of the major airways, a rigid bronchoscope can be used to bypass the compromised airway. However, it is largely used as rescue therapy after the demonstrated inability to pass a conventional endotracheal tube beyond an obstruction. However, because of the fact that this patient has an SVC syndrome with edema of the airway, there is considerable bleeding risk from the trauma that can occur in airway acquired using rigid bronchoscopy. Importantly, rigid bronchoscopy should only be performed by a skilled bronchoscopist. As positive pressure ventilation may enhance the extrinsic collapse of the airway due to increases in intrathoracic pressure, maintenance of spontaneous respiration is particularly important.

(In this patient) The preoperative echocardiogram demonstrated tumor and/or thrombus extending in to the right atrium. I believe this is a strong indication for the use of CPB. Preoperative diagnostic studies in this patient revealed a near total obstruction of the SVC and numerous intercostal collaterals. Further occlusion of the SVC during tumor resection would be unlikely to produce acute adverse sequelae. Acute clamping of a patent SVC results in increased pressures proximal to the clamp, which will result in decreased cerebral perfusion pressure if the arterial pressure remains constant. However, within the limits of autoregulation, cerebral perfusion should be maintained. A primate model ¹¹ demonstrated increased intracranial pressure and decreased regional cerebral blood flow but adequate CPP and preservation of electroencephalogram activity during a one-hour clamp of SVC.

CPB is often considered in the emergency management of mediastinal masses. If the need for it is highly suspected, it is far better to institute under more controlled conditions than as rescue therapy in the patient who is near hemodynamic collapse. In the present case, the need for it, assuming that the induction of anesthesia is uneventful, will be dictated by the surgical plan. Whereas it may be possible to resect this tumor without its use, one cannot guarantee that the tumor may not compromise the SVC significantly enough that bypass may be needed. The clamping of the SVC can be expected to cause a life-threatening reduction in venous return, necessitating cannulation of the vena cava above the tumor (distal to the heart) to ensure the maintenance of venous return during the onset of CPB. One of the difficulties with CPB in these patients is the necessity for systemic heparinization with an overall increase in bleeding. Depending on surgical experience, full CPB may be avoided by using a veno-venous shunt around the tumor to provide the uninterrupted flow of venous return, similar to what has been reported with inferior vena cava clamping during hepatic transplantation. ¹²

Although a mass with intracaval extension into a cardiac chamber can at times be removed without CPB support, it is always prudent to have CPB available. For cardiac tumors or tumors invading through the wall of the cava into the heart, CPB will be necessary for complete resection. Preoperative imaging is of paramount importance in planning the surgical strategy. For right-sided tumors, as in this case, it is necessary to know the degree of involvement of both cavoatrial junctions as well as access to the ascending aorta, as aorto-bicavalcannulation, or the equivalent thereof, will be necessary.

The nature of the tumor and risk of embolization with cardiac manipulation further modifies the decision to cannulate centrally versus peripherally. In extremely high-risk cases, it may be prudent to cannulate through the neck (for vein) and the groin for artery and vein (although the carotid artery can be used in an otherwise healthy child) prior to opening the chest. This carries the acceptable attendant drawback of being fully heparinized for the duration of the tumor dissection. In this case, the inferior cavoatrial junction was relatively free, and the tumor was regarded as carrying high but not prohibitive risk for a primary approach in the chest. However, given the invasion of the innominate/SVC junction, drainage for the brain was obtained via the jugular vein in the neck. In principle, one could cannulate both jugular veins; however, based on the premise of assistance by some degree of collateralization, only the right jugular vein was cannulated.

Studies evaluating the impact on the brain of brief occlusion of the SVC (<40-60 minutes) have produced mixed results, ranging from no obvious impact to frank stroke. Impact depends on the duration of prior severe stenosis or occlusion and the degree of collateralization. For a chronically occluded SVC with a high degree of collateralization, one can likely safely clamp around the area of obstruction with little impact. Test occlusion with measurement of the central pressure proximally is one approach, with pressures exceeding approximately 20 to 25 mm Hg considered borderline. This of course requires concomitant consideration of the mean arterial pressure to better estimate the cerebral perfusion pressure. For acute cases, the jugular pressure is already excessively high. With the availability of near-infrared spectroscopy (NIRS) monitoring, brain saturations can be monitored to help ensure tolerance. If heparinization and CPB are going to be used, there is little drawback to cannulation proximal to anticipated clamp sites, and in general, this would be the preferred approach. In this case, involvement of the distal or low jugular/innominate/SVC junction was suspected. Given prior irradiation, adhesions, and engorged vasculature, it was my judgment that decompressing the venous vasculature from above would not only simplify the dissection but would also help minimize blood loss. ¹

The possibility of significant compromise of venous outflow from the brain due to obstruction of the SVC should always be considered with any mediastinal mass but is of particular importance in this patient with a known SVC syndrome. The cerebral perfusion pressure (mean arterial pressure minus intracranial or central venous pressure, whichever is greater), may be compromised with clamping the SVC. As a result, maintenance of a greater than normal mean arterial blood pressure may be needed. Adequate drainage of the SVC will be required either directly after sternotomy or percutaneously via the internal jugular vein. Specific cerebral monitoring may aid in the determination of when the brain has been compromised and for the guidance of the optimal hemodynamic targets. Specifically, the placement of a cerebral oximeter may allow for the determination as to whether the SVC obstruction is causing significant impairment of cerebral oxygenation. Transcranial Doppler and/or multichannel electroencephalography (EEG) monitoring can also be used to determine the adequacy of cerebral blood flow; however, these are not widely available, and require significant training and expertise.

Preservation of neurologic function following surgery requiring CPB is of significant concern. Unfortunately, there is little in the literature to support an evidence-based approach to monitoring. Current technologies available include EEG, transcranial Doppler ultrasound, and NIRS to monitor cerebral oxygenation. EEG and processed EEG, including the Bispectral Index (BIS) and Patient State Index (PSI) have been used primarily to assess depth of anesthesia, and to date have not been useful in monitoring cerebral function. ¹³ The transcranial Doppler measures blood flow velocity, typically in the middle cerebral artery. Despite its utility in detecting emboli and as a research device to understand cerebral perfusion, it has not demonstrated significant utility in the clinical setting. Trend monitoring of cerebral oxygen saturation using NIRS is an accepted practice in our institution for all patients undergoing cardiac surgery and other high-risk procedures. Clinical studies to date have not proven the efficacy of this monitor in improving neurological outcomes in patients undergoing cardiopulmonary bypass. However, piglet models demonstrate that sustained decreases in regional oxygen saturation measured by NIRS are associated with adverse neurological outcomes. A decrease in cerebral oxygen saturation is also seen in association with SVC occlusion.

In the 1980s, there were 5 pediatric cases in the ASA Closed Claims Project related to anesthesia and surgery for mediastinal masses. The most recent database (2000-2007) had no similar claims. ¹⁴ Heightened awareness of the risks associated with these lesions, prudent sedation and anesthetic guidelines, including local anesthesia when possible, spontaneous ventilation and avoidance of muscle relaxants, and the immediate availability of specialists skilled in rigid bronchoscopy and extracorporeal circulation have reduced the incidence of severe adverse outcomes.

Given the involvement of both innominate veins and SVC, extensive complex reconstruction was anticipated. Plans were also made for superior cavoatrial and right atrial resection, potentially with placement of a pacemaker should the underlying atrial rate be insufficient for adequate cardiac output. Replacement options require consideration of ease or simplicity, function, durability, vulnerability to infection, and contraindications to intermediate-term anticoagulation. One important consideration is the chronicity of the obstruction and degree of collateralization. In the setting of extensive collateralization, flow through the reconstructed cava may be inadequate and pose high risk for thrombosis. A variety of materials have been used successfully for central venous reconstruction, including spiral autologous vein (which requires a lot of suture lines, and in this case would require a lot of vein to provide the extent of reconstruction needed), internal iliac vein (in this case, not adequate amount of tissue), Contegra conduit (bovine jugular vein—more expensive, but very functional and probably a reasonable choice for short- to intermediate-term duration), and Goretex. Dacron has been used but very infrequently. In this case, Goretex was chosen because of simplicity, extent of reconstruction needed, durability, and proven track record for central venous reconstruction. Goretex is also considered relatively resistant to infection.^15,16