Imaging studies are crucial to rule out the presence of metastatic disease and to define the level of the thrombus (Fig. 5.1). The accurate delineation of the proximal extent of the thrombus is of paramount importance because it determines the approach, the position of the patient on the operating table, and the need for bypass procedures.

Ultrasonography and computed tomography (CT) have demonstrated good specificity in detecting the presence of tumor thrombus, with a reported sensitivity of 65–90 %, reaching 87 % when used in combination. Conversely, Doppler examination is suboptimal in visualizing the distal renal vein and the infrahepatic vena cava. Ultrasound accuracy is also strongly operator dependent and may be influenced by the patient’s corporal habitus.

The improved multidetector CT would represent an option in candidates unsuitable to magnetic resonance imaging (MRI). The accuracy of multidetector CT for assessing venous tumor thrombus has been evaluated recently. These scans provide increased anatomical detail compared to conventional CT through reconstructed images [22, 23]. Guzzo et al. studied 41 patients and noted an 84 % concordance rate between multidetector CT and surgical pathology, with the thrombus level accurately predicted in 96 % [24]. However, no direct comparison to MRI was made in this study, and additional larger series will be needed to confirm the potential usefulness of this imaging modality.

MRI is currently considered the gold standard in the evaluation of these patients because T1-weighted images provide precise and clear anatomic depiction of the cephalad extension of the thrombus and the relationship of the thrombus to the liver, diaphragm, and right atrium. MRI also provides an accurate evaluation of the degree of IVC occlusion and the presence of associated bland (platelet) thrombus, which is often present in the infrarenal IVC in these patients [25, 26]. In addition, Zini et al. noted that invasion of the renal vein ostium wall may be predicted on MRI, specifically by measuring the renal vein and IVC diameter [25]. They reported a 90 % sensitivity for wall invasion with a renal vein or IVC diameter greater than 1.4 cm and 1.8 cm, respectively.

Transesophageal echocardiography (TEE) could be used preoperatively or intraoperatively and is especially indicated in cases with conflicting findings on MRI or in determining the presence of thrombus invading the major hepatic veins [26]. In addition, when thrombus is found in the right atrium or if pulmonary emboli are suspected, transesophageal echocardiography and CT angiography of the pulmonary vasculature should be performed.

As for all patients with suspected RCC, evaluation for potential metastatic disease necessitates chest imaging and liver function tests. When considering a possible need for cardiopulmonary bypass intraoperatively, cardiology consultation should be obtained for medical optimization and possible cardiac angiography. Moreover, cross-sectional imaging of the brain for potential metastatic disease that would be at risk for hemorrhage during anticoagulation with cardiopulmonary bypass (CPB) has also been suggested for patients with an atrial thrombus [4].

It is highly recommended for the last imaging study previous to a planned intervention to be done within 7–14 days of surgery for level II, III, and IV tumor thrombi, as propagation of thrombus may occur and accordingly that variation in cranial extension can significantly alter the surgical approach [27, 28].

Multiple staging systems have been proposed to classify RCC with caval involvement. The recently revisited Union for International Cancer Control/American Joint Cancer Committee (UICC/AJCC) TNM staging system for renal carcinoma has supplanted Robson RCC staging classification for prognostic and, thus, therapeutic purposes. In previous TNM classifications, the pT3b group included both renal vein and IVC invasions. As the result of many studies into the independent prognostic value of vena cava compared to renal vein invasion alone [29–31], these two groups have now been separated in the latest version of the TNM classification [32]. Accordingly, this staging system stratifies RCC cases with tumor thrombus into the renal vein or segmental renal vein branches as pT3a, thrombus in the IVC below the diaphragm as pT3b, and supradiaphragmatic thrombus and/or thrombus that invades the IVC wall as pT3c [33].

Since the anatomic level of the tumor thrombus within the IVC often impacts surgical planning, for surgical considerations, the most widely used classification remains that proposed by Neves and Zincke (i.e., Mayo Classification System) [34]. According to this system, tumor thrombus in the renal vein only is classified as level 0, thrombus extending into the IVC <2 cm from the renal vein ostium is level I, IVC extension >2 cm from the renal vein ostium and below the hepatic veins is level II, thrombus at the hepatic veins and below the diaphragm is level III, and supradiaphragmatic or atrial tumor thrombus is level IV (Fig. 5.2).

However, the Mayo Classification System may not provide complete information regarding some retrohepatic and suprahepatic/infradiaphragmatic (i.e., level III) tumor thrombi. Conversely, the Miami Classification System may be used to improve decision-making in this particular setting [35]. In this modified classification system, level IIIa tumors are defined as those with thrombus extending into the retrohepatic IVC but ending below the origins of the major hepatic veins, level IIIb as extending to the ostia of the major hepatic veins, level IIIc as extending above the major hepatic veins but below the diaphragm, and level IIId as extending above the diaphragm but not into the right heart (Fig. 5.3).

Radical nephrectomy and tumor thrombectomy were documented as early as 1913 [36]. In the past, however, patients with IVC involvement were not operated routinely due to high morbidity and poor survival rates. The widened scope of surgical alternatives with the advent of bypass procedures in the 1970s [37] along with overall improvement in perioperative care has significantly extended long-term free survival rates [38–43]. The virtually unanimous acceptance of the lack of tumor response to standard adjuvant therapy protocols, the acceptable survival rates reported in advanced stages of the disease, and the significant improvement in quality of life provided by thrombus removal in symptomatic patients also support a surgical strategy [44]. Likewise, patients with all forms of renal tumor with venous extension and nonmetastatic disease are suitable candidates for surgery. Surgery should be strongly considered after counseling in those patients with metastatic disease, given that the natural history of the disease in its free evolution ends inevitably with the death of a patient in extremely poor conditions.

The choice of surgical technique should be individualized for each specific case based on the features of the disease process, which include (i) the comorbid status of the patient at clinical presentation, (ii) the malignancy burden, (iii) the tumor laterality, (iv) the extent of tumor thrombus inside the IVC lumen, and (v) the presence of accompanying embolic events.

In this way, each of the small technical “blocks,” or surgical steps, will be integrated to form a unique surgical technique, resulting in a procedure that is tailored to the precise requirements of the individual patient.

Preoperative comorbid status. Worse outcomes after surgery have been found in patients with overall poorer health and functional status preoperatively [7, 10].

Malignancy burden. Postoperative prognostic factors for RCC invading the IVC include a set of features associated with more aggressive tumor behavior and contributing to higher levels of local and distant invasion [10, 45]. All of these features should be considered extensively during surgical planning. However, it is possible that preoperative imaging does not provide an accurate depiction of certain important details [46], and therefore, a shift in strategy may occasionally be needed during the surgical procedure to address these contingencies.

Extent of tumor thrombus. The extent of tumor thrombus inside the IVC is the most important consideration in planning the operative strategy [47]. To adequately assess the extent of venous involvement, two parameters must be addressed: (i) the anatomic level of the tumor thrombus in the IVC and (ii) the degree of IVC occlusion generated by the thrombus intraluminal growth.

The tumor thrombus level is a critical consideration in the selection of surgical technique, given that the extent of dissection is generally predicated on the cephalad level of tumor thrombus, thus dictating the number and type of surgical maneuvers for its successful removal. There is general agreement among authors with regard to the combination of surgical steps required to remove lower thrombus cases (levels I and II) and most cases involving the right atrium (level IV). However, there is no consensus on the appropriate strategy in cases affecting the retrohepatic or suprahepatic IVC segments (level III).

The degree of IVC occlusion represents another important feature with regard to the extent of thrombus, and totally or partially occluding caval thrombus may be encountered. An extreme degree of occlusion may lead to the invasion of the venous wall containing the thrombus.

Bland thrombus may be detected on MRI, most frequently in the infrarenal IVC. As opposed to tumor thrombus, it is characterized by a lack of contrast enhancement on MRI and may be found contiguous with tumor thrombus or as a separate distant clot, frequently within the common iliac veins.

Although these embolic events occur at a relatively low frequency (10–15 % and 1.5–3.4 % of cases for bland thrombus and PE, respectively), they cannot be overlooked during surgical planning, given that their presence may demand a radical shift in surgical strategy [48–50]. The proper management of a coexistent bland thrombus commonly requires blood flow interruption through the IVC, while evidence of a preexisting PE (or its sudden intraoperative onset) usually entails installation of extracorporeal circulation (e.g., removal of tumor thrombus from the pulmonary arteries).

Additional important considerations during the preoperative assessment of RCC and venous tumor thrombus are the potential roles for systemic anticoagulation, renal angioembolization, IVC filters, and tyrosine-kinase inhibitors (TKIs).

In cases of associated bland thrombus, systemic anticoagulation is recommended to prevent thrombus propagation. Propagation risks contralateral renal vein thrombosis and debilitating lower extremity edema and may serve as a nidus for distal or proximal embolization. Efforts should be made for expeditious surgery, and, as such, these patients are placed on low molecular weight heparin in an outpatient or on intravenous heparin in the inpatient setting to allow for rapid discontinuation before surgery.

Embolization of the renal artery. This maneuver aims to reduce blood supply, mass size, and collateral blood flow around the tumor. The purported benefits of this management strategy also include a reduction in tumor thrombus extent, a decrease of intraoperative blood loss, and facilitation of renal hilar dissection [51].

Although isolated single institutional experiences support this procedure as a method to decrease the overall complexity of the intervention [52], this technique presents certain disadvantages, which may advise against its use. Subramanian et al. [53] showed that there is no significant advantage in preoperative embolization for the treatment of RCC with an IVC thrombus, and in fact, this procedure may increase the risk of complications and mortality probably by inducing a significant reaction around the kidney and surgical field. Hence, both the high frequency of postembolization syndrome and in many cases its severe clinical presentation led experts to no longer recommend this maneuver [54].

Preoperative IVC filter deployment. Possible migration of dislodged thrombus fragments into the pulmonary circulation favored the presurgical use of IVC filters as PE preventive strategy [55]. Currently, this recommendation remains controversial, due in part to different reports on the rupture of the caval wall during the device deployment [56], not to mention the infrequent proximal migration of the filter into the right heart chambers causing a lethal cardiac tamponade [57].

In our opinion, if the patient presents an established PE at the time of diagnosis, there is no indication for IVC filter use. In most of these cases, if not all, PE is produced by a mix of tumor and bland thrombus fragments. Tumor thrombus fragments are completely insensitive to anticoagulant therapy. Under these circumstances, CPB is advisable for a complete tumor removal from the pulmonary arteries. IVC filters are not capable of preventing tumor thrombus enlargement. Therefore, the device can be progressively entrapped within the neoplastic tissue after placement. If this occurs, the complexity of the procedure is multiplied exponentially, and what apparently would be a resectable case may become almost unresectable (Fig. 5.4).

The anatomical location of the proximal thrombus limit and the degree of IVC occlusion may also contraindicate the deployment of a filter. Filter placement may not be warranted in higher thrombus level cases (i.e., levels III–IV) due to a space conflict above the major hepatic veins (MHVs). Obviously, the filter cannot be placed in the right atrium. In addition, the use of a distal (i.e., femoral) instead of a proximal (i.e., transjugular) percutaneous approach for filter deployment may potentially induce a partial dislodgement of the thrombus with devastating consequences.

In cases of complete IVC flow interruption, filter deployment may be unnecessary since the thrombus would act as a filter itself (i.e., completely occluding the lumen of the IVC). In addition, as a result of complete IVC occlusion, venous flow redistributes through a secondary network of variable size collaterals. The diameter of these vessels is occasionally wide enough to permit the passage of thrombus fragments. Placing a filter in the IVC would not prevent an eventual PE under these circumstances. Nevertheless, if the filter is thought strongly indicated (i.e., level II, not completely obstructing, tumor thrombus cases, with or without associated bland thrombus), it should be deployed <48 h before surgery to reduce the incidence of thrombus entrapment [27].

Tyrosine-kinase inhibitors. Although the concept of neoadjuvant therapy to down stage locally advanced tumors and improve survival has been incorporated into the treatment approach for a variety of malignancies, to date its use in kidney cancer has been limited. Nevertheless, tyrosine-kinase inhibitors (TKIs) have an interesting ability to reduce tumor size in RCC that would have also a role in decreasing the thrombus anatomic level before nephrectomy [58]. In fact, some cases had been reported in the literature with the use of sorafenib and sunitinib [59, 60]. However, existing information on the response of primary tumors and thrombus to these agents remains limited.

Initial Requirements. Most authors agree that there are a number of requirements that must be met in order to carry out this type of intervention. These include (i) an adequate environment, (ii) close collaboration among the members of an experienced multidisciplinary team, and (iii) an accurate preoperative imaging assessment [61].

Minimally invasive approaches. The application of advanced techniques has allowed renal tumors invading the IVC to be managed in purely minimally invasive fashion, but the standard approach for tumors with intracaval extension remains open surgery because the benefits of minimally invasive techniques are temporary or short term. Further experience will be necessary to decide if the benefits obtained justify the approach [62].

Patient positioning, incision, and self-retaining retractor. The positioning of the patient should be based on the choice of incision [63]. Supine position offers advantages for anesthetic and surgical control, since it allows better access to the patient’s head/chest when TEE or CPB is required.

Surgical incision should be chosen under the basis of an optimal approach to the tumor and vascular control. A large number of different incisions have been used in the treatment of RCC with caval involvement. Generally, the use of a particular type of incision is in relation to the volume of the renal mass, its relationship to the surrounding structures, and the anticipated level of tumor thrombus. Thoracoabdominal and flank approaches were frequently used in the past [64] but are less common today, since thoracoabdominal access involves the use of a postsurgical chest tube and the approach through the flank does not allow proper control of the great retroperitoneal vascular structures that commonly lie hidden by the renal mass.

Midline xipho-pubic and transverse subcostal incisions are now preferred, as they avoid the postoperative need for a thoracic tube. In addition, both can be combined with a midline sternotomy if CPB is required. Although the approach through the midline is easier to learn and quicker to perform and can provide excellent exposure, this incision entails a typical telescopic effect, which increases with depth of surgical field, putting at risk the adequate control of “deep” areas, including vascular structures, and thus the overall safety of the procedure [65]. Conversely, transverse incisions are based on better physiological principles and should be recommended, as there are fewer complications in the early postoperative period and a lower incidence of late incisional hernia [66]. However, this does not appear to be clinically significant, as complication rates and recovery times are comparable to those obtained with midline incisions [64].

The triradiate Chevron incision combines the advantages of midline and subcostal approaches without increasing the rate of incision-related complications. This incision may represent an excellent alternative for the more complex cases [67]. Nevertheless, the optimal incision for abdominal surgery still remains the preference of the surgeon [64].

The choice of self-retaining retractor should be based on the type of incision used. While the Omni-Tract FastSystem® (Omni-Tract Surgical/Minnesota Scientific, Inc., St. Paul, MN, USA) and Bookwalter^TM abdominal retractor (Codman & Shurtleff, Inc., Raynham, MA, USA) are excellent options for midline incisions, in the event that a triradiate Chevron incision is used, a retractor designed for liver surgery is preferable (e.g., Rochard and Thompson retractors). Liver retractors have the advantage of moving the costal margins toward the axillae, which flattens the diaphragmatic domes, thereby increasing exposure in areas that are difficult to access such as the upper abdominal quadrants [68] (Fig. 5.5).

Control of the renal artery. Access to the main renal artery can be achieved through either an anterior or posterior approach. The anterior approach requires the full mobilization of the peritoneal structures to enter the retroperitoneum, while the posterior approach uses en bloc mobilization of the kidney with the peritoneal structures lying above it (Cattell–Braasch and Mattox maneuvers) [69], thus creating a plane of cleavage anterior to the posterior abdominal wall. Each plane of dissection has specific advantages that may facilitate arterial control in particular situations. The posterior approach requires the division of all adhesions between Gerota’s fascia and the posterior abdominal wall [65]. Although it may seem more tedious to perform, this technique avoids potential engorged vessels in the anterior renal surface, providing quick and safe access to the main renal artery near its takeoff in the aorta. This type of access may represent the best option in cases of marked venous collateral circulation or when access to the anterior aspect of the renal hilum is difficult (e.g., hampered by bulky lymphadenopathy) (Fig. 5.6).

Conversely, the anterior approach [70] may be the best option when the renal mass cannot be mobilized (e.g., extremely large size). This approach is perhaps faster and easier for the surgeon, providing vascular control of both renal hilar structures simultaneously. However, it requires a relatively free anterior plane and extensive dissection in the proximity of the great retroperitoneal vessels, which may more readily result in injury.

Exposure. Anterior access to the kidney is achieved through mobilization of the ipsilateral colon segment. A peritoneal incision on the avascular line of Toldt is mandatory to gain access. Medial colon rotation progressively exposes the anterior surface of Gerota’s fascia. The Kocher maneuver and liver mobilization [69] complete the right renal exposure, while the dissection of the visceral complex formed by the stomach, spleen, and pancreas allows full inspection of the left anterior renal plane [71] (Fig. 5.7).

Handling of IVC. Management of the obstructed IVC involves surgical maneuvers to gain vascular control at three different levels: (i) infrahepatic/renal, (ii) retrohepatic/suprahepatic/infradiaphragmatic, and (iii) supradiaphragmatic [67, 72].

Infrahepatic IVC segment and renal veins: Every vestige of lymphatic tissue must be cleared from the anterior aspect of the infrahepatic IVC segment. Both renal veins should be encircled and clamped before opening the IVC. The posterior surface of the IVC needs to be detached from the posterior abdominal wall by ligating and dividing all of the lumbar veins found at this level, thus gaining complete circumferential control. This step should be performed with great care, given that lumbar venous vessels may be engorged in response to IVC occlusion and uncontrolled bleeding at this level can be extremely dangerous.

Retrohepatic/suprahepatic infradiaphragmatic IVC segment: Exposure of this segment requires liver mobilization. The level reached by tumor thrombus inside the IVC determines the extent of liver dissection. For example, level II tumor thrombi may require only a classical mobilization of the right hepatic lobe (Langenbuch maneuver) (Fig. 5.8), whereas more proximal tumor thrombus levels (levels IIIa–IIIc) require full liver mobilization, enabling total vascular control on the abdominal segments of IVC behind and above the liver (“piggyback” liver dissection) (Fig. 5.9) [69]. Classical mobilization of the right hepatic lobe involves division of the right triangular and coronary ligaments [73]. This allows a gradual rotation of the right hepatic lobe toward the midline and allows access to the right lateral surface of the IVC, thus completely exposing the right renal upper pole and the ipsilateral adrenal gland and vein. However, this maneuver may be insufficient if there is a need for circumferential control on the IVC and CPB is not planned.

In 1989, Tzakis et al. [74] described the so-called piggyback liver transplantation technique, which is based on a tangential clamping of IVC at the level of the MHVs, avoiding complete caval occlusion during the anhepatic phase of the procedure. With this technique, the liver is fully mobilized by dividing all of its attachments (i.e., complete posterior ligament detachment and division of the short hepatic veins between the right and caudate lobes of the liver and the anterior caval aspect) until it lies fixed to the IVC only by the MHVs. It then becomes possible to freely move the complete liver, thus facilitating full circumferential control of the entire retrohepatic and suprahepatic IVC segments [69].

Supradiaphragmatic IVC segment and right cardiac chambers: Adequate vascular control of the right heart chambers and supradiaphragmatic segment of the IVC is perhaps more controversial, since access can be gained from the abdomen or the thorax. Although it is clear that a large mass inside the right atrium almost invariably requires a thoracic approach and extracorporeal circulation, vascular control required in levels IIIb–d may vary according to the preference of the surgical team.

Diaphragmatic caval release, by opening the central tendon of the diaphragm and encircling the IVC, allows access to the intrathoracic IVC segment from the abdominal field [69, 72]. Thereafter, the pericardium can be opened, and the right atrium can be gently pulled through the diaphragm, to be controlled inside the abdomen. This maneuver permits the resection of level III–IV tumor thrombi without the need for extracorporeal circulation [69, 72] (Fig. 5.10).

Cardiac control can also be achieved, with good results, through minimally invasive approaches if the intent is only to place the patient on CPB [75]. However, this access may be severely limited if a clear view of the right chambers of the heart is necessary. In these cases, it may be preferable to use a wider thoracic approach by means of a midline sternotomy or thoracoabdominal access [5] (Fig. 5.11).

Thrombectomy. The extension of caval incision for tumor thrombus withdrawal should be established according to the thrombus level inside the IVC [76, 77].

Level I tumor thrombi are usually “floating” in the lumen of the IVC, which is commonly partially obstructed. They can be easily “milked” back into the renal vein. Thereafter, a side bite of the IVC with an appropriate vascular clamp permits complete thrombus control, preserving blood flow though the IVC and preventing the embolization of eventual dislodged thrombus fragments to the pulmonary circulation. The IVC is then incised, permitting the complete removal of the thrombus under direct vision. After radical nephrectomy and thrombectomy are complete, the venotomy can be sutured closed in the usual manner (commonly with a double row of continuous 4-0 polypropylene suture) (Fig. 5.12).

In level II tumor thrombi, vascular control is ensured by placing clamps or Rummel tourniquets sequentially on the infrarenal vena cava, contralateral renal vein, and suprarenal vena cava above the upper thrombus limit. Likewise, the thrombus is dissected and removed with the entire renal ostium. Once the lumen of the cava is flushed with heparin solution and the absence of residual tumor is ascertained, the cavotomy is sutured closed and the clamps are removed in a cephalad direction. To preserve renal function, it is advisable to permit venous flow return from the contralateral renal vein and lower IVC. Many times, the shape and position of the thrombus inside the IVC allow partial occlusion with a long curved vascular clamp. In some cases, the tumor configuration permits the placement of the vascular clamp in an oblique fashion so that the contralateral vein lies above the clamp and venous drainage into the proximal IVC can be maintained. However, one clamp may be insufficient to encircle the thrombus in the case of a large irregular configuration, and an additional clamp will be needed for complete IVC occlusion.