Introduction

Over 200,000 new cases of kidney cancer are diagnosed annually and more than 100,000 deaths occur from this disease each year worldwide, with the highest incidence in North America, Europe and Australia [89]. Renal cell carcinoma (RCC) accounts for 3% of all adult malignancies and is steadily increasing at a rate of about 2.5% per year across population groups [14]. Kidney cancer is the most lethal of the urologic malignancies with more than 40% of patients dying from their cancer [51]. Approximately, 20–30% of patients present with metastatic disease and 20–40% of patients undergoing nephrectomy for clinically localized RCC will develop metastases [61]. Significant advances in the diagnosis, staging, and treatment of patients with RCC during the last 20 years have resulted in improved survival of a select group of patients and an overall change in the natural history of the disease [86]. Recent clinical surveys have revealed that the majority of kidney cancers are now diagnosed incidentally during investigation of unrelated complaints, in part because of increased use of imaging procedures, such as ultrasonography, computed tomography (CT), and magnetic resonance imaging (MRI) [14, 86]. The classic clinical triad of pain, hematuria, and a flank mass is less frequent than previously seen, but RCC remains a great mimic by virtue of its various paraneoplastic manifestations [54]. The evolution of the surgical management of RCC has been predicated by these changes.

Staging and cancer-specific survival

The tumor, nodes, metastasis (TNM) staging system is currently the most commonly used staging system for RCC. It has undergone several systematic revisions to mirror the improved results attained in the management of RCC. In 1997, the cutoff size for T1 stage was expanded from 2.5 to 7 cm [41]. The increase in cutoff size to 7 cm. for T1 tumors has led to a great deal of controversy. Several studies have evaluated the optimal T1 size criteria for patients undergoing surgical resection suggesting alternative cutoffs and while these studies disagree about the optimal cutoff, they all demonstrate that primary tumor size is an important determinant of prognosis [61]. In addition, inferior vena cava (IVC) tumor thrombus extension located above the diaphragm, previously stage T4, was changed to T3c, and thrombus extension below the diaphragm, previously staged T3c, was changed to T3b with renal vein (RV) involvement [41]. The 5-year cancer-specific survival for RCC using the 1997 TNM staging system from several contemporary cohorts has been reported to be 75–96, 63–95, 38–70, and 11–32% for stages I to IV [60].

The impact of various anatomic, histologic and symptomatic factors has led to improved prognostication in patients with RCC and the establishment of more comprehensive staging systems [61]. At the University of California-Los Angeles (UCLA), an extensive prognostic system has been created for both localized and metastatic RCC designated the UCLA Integrated Staging System (UISS). The initial UISS contained five groups based on TNM stage, Fuhrman nuclear grade, and Eastern Cooperative Oncology Group (ECOG) performance status (PS). Projected 2- and 5-year survival for patients in UISS groups I to V are: 96 and 94%, 89 and 67%, 66 and 39%, 42 and 23%, and 9 and 0%, respectively [137]. The UISS has been subsequently modified into a simplified system, based on separate stratification of metastatic and non-metastatic patients into low-, intermediate-, and high-risk groups [138]. This provides a clinically useful system for predicting postoperative outcome, and provides a unique tool for risk assignment and outcome analysis to help determine follow-up regimens and eligibility for clinical trials.

Radical nephrectomy: the gold standard in estimating long-term outcomes of RCC surgery

In 1969, Robson et al. [99] published what has been considered the most influential as well as the most frequently cited article on RCC published in the urologic literature. This was a retrospective review of 88 RCC cases between 1949 and 1964 that demonstrated improved surgical results compared to other series of renal tumors managed by simple nephrectomy reporting an overall 52% 5-year survival and a 66% 5-year survival in patients with localized disease [99]. This study drew attention to the basic surgical oncological principles for the successful management of renal tumors and the importance of stratifying the anatomic spread of the tumor for the purpose of patient prognostication. The basic principles established included early ligation of the renal vessels to minimize risk of vascular tumor emboli, excision of Gerota’s fascia including the kidney and adrenal gland, and extensive lymphadenectomy including the paraaortic and paracaval nodes from the crus of the diaphragm to the bifurcation of the aorta. This surgical approach was designed to achieve several objectives that included excision of the ipsilateral adrenal gland, which historically was involved in 10% of cases; removal of lymphatic metastases, which may diffuse through the perirenal fat; and a greater margin away from the tumor was achieved, especially when the tumor invaded the perirenal fat, which occurs in 25% of patients. Subsequently, multiple studies have reported the outcome of patients treated with radical nephrectomy. Five-year survival after radical nephrectomy from contemporary cohorts ranges from 75 to 95% for patients with organ-confined disease, 65–80% for perinephric fat or adrenal involvement, 40–60% for patients with vena cava thrombus, 10–20% for lymph node involvement, and 0–5% for metastatic disease [29, 41, 45, 50, 90, 126].

As we have previously indicated, Robson’s approach was developed in the era when the diagnosis of a renal mass was based primarily on intravenous pyelography and angiography, and the difficulty in accurately defining the status of the primary tumor and the ipsilateral adrenal gland justified wide excision of normal tissue along with the tumor. However, with the development of high resolution CT scanning, more accurate preoperative clinical staging of renal tumors has become possible, and the necessity of all of the components of the radical nephrectomy (adrenalectomy, extensive retroperitoneal lymph node dissection, necessity for removing the entire kidney, open incisional approach) for all renal masses has come under scrutiny. Outcomes related to several of these controversial areas are reviewed in the following sections.

Necessity for routine adrenalectomy

Several studies have suggested that removal of the ipsilateral adrenal gland is not routinely necessary during radical nephrectomy [56, 92, 108, 110, 125]. The low local recurrence rate following NSS for RCC, in which by definition adrenalectomy is not performed, further questions the value of routine adrenalectomy [22]. Paul et al. [92] examined 866 patients who underwent radical nephrectomy plus ipsilateral adrenalectomy for RCC. Twenty-seven (3.1%) patients had adrenal metastasis, of whom four had bilateral or contralateral adrenal involvement. In this study, tumor size greater than 8 cm. and M stage were independent predictors of adrenal involvement. Current evidence suggests that the rate of adrenal metastasis is low and that modern day imaging modalities are sensitive enough to pick up adrenal lesions. These findings were confirmed in a contemporary prospective series of 511 nephrectomies from UCLA, in which CT was greater than 99% specific and nearly 90% sensitive to detect adrenal involvement preoperatively [125]. Given the low percentage of patients with adrenal involvement and use of detailed preoperative imaging, the vast majority of those with RCC can be spared the potential morbidity associated with ipsilateral adrenalectomy. Siemer et al. [110] reported on a retrospective review of 1,635 patients that underwent radical nephrectomy for RCC in which 1,010 patients underwent simultaneous ipsilateral adrenalectomy were compared to 625 patients that did not have simultaneous adrenalectomy. Adrenal gland metastases were found in 56 of 1,010 patients (5.5%) undergoing nephrectomy plus adrenalectomy. Of 30 patients with adrenal metastasis and a preoperative CT/MRI, 23 were found to have histological evidence of cancer, approaching a false-negative rate of 23.3%. All patients with false-negative CT/MRI had a primary tumor of greater than 4 cm. and patients with adrenal metastases predominately had pT3 or greater tumor stage (82%). However, there was no significant difference in cancer-specific survival rates (75 vs. 73%) and postoperative complications rates (7 vs. 8%) between patients that underwent adrenalectomy compared to those that did not. One final caveat to consider regarding routine adrenalectomy for all patients with RCC is the possibility of developing a future, contralateral adrenal metastasis. If these patients become functionally or surgically adrenal insufficient and require steroid replacement, this could negatively impact on their eligibility to receive systemic immunotherapy.

The current TNM staging system categorizes patients with adrenal involvement into the T3a group, which includes tumors that extend into the perirenal fat, but not beyond Gerota’s fascia. Tsui et al. [126] reported a 42% 5-year cancer-specific survival rate in patients with stage T3 disease based on a retrospective analysis of 643 patients, including 270 with stage T3 disease. Ficarra et al. [24] reported similar 5-year cancer-specific survival rates of 57% for T3a and 47% for T3b tumors in a retrospective analysis of 675 cases. However, neither of these, or other studies investigated whether survival differences exist within the T3a category when comparing tumors with direct extension into the adrenal gland versus those with perinephric/renal sinus fat involvement. Several recent studies have investigated whether survival differences exist within the T3a category when comparing tumors with direct extension into the adrenal gland versus those with perinephric/renal sinus fat involvement [43, 111, 121]. Han et al. [43] from UCLA reported survival following radical nephrectomy of 27 patients with RCC invading the adrenal gland was significantly worse than the 187 patients with perinephric fat invasion and no adrenal involvement. Among patients with adrenal involvement, the median survival was 12.5 months and the 5-year cancer-specific survival rate was 0%, whereas, among patients with perinephric fat invasion and no adrenal involvement, the median survival was 36 months and the 5-year cancer-specific survival rate was 36%. Furthermore, the survival of patients with T3a disease and adrenal involvement was not better than patients with T4 tumors. Although a correlation existed between adrenal invasion and higher tumor grade, lymph node involvement, and metastatic disease, multivariate analysis demonstrated that adrenal invasion was an independent predictor of poor prognosis.

Benefits of lymphadenectomy

The overall risk of lymph node metastasis is approximately 20% and 5-year survival rates of patients with lymph node metastases ranges from 11 to 35% [61]. Although there remains debate as to the value of a retroperitoneal lymphadenectomy in treating RCC [32, 46, 77, 88, 103, 112], its prognostic role is undoubted [88, 99, 131]. Life expectancy decreases considerably when lymph node metastases are present, with overall 5-year survival rates of 11–35% [88, 131]. Potential benefits to performing a lymphadenectomy at the time of nephrectomy include more accurate staging and potentially a lower risk of positive margins due to more extensive dissection of perinephric tissue, and thus in theory leading to a lower risk of local recurrence. In addition, lymphadenectomy may potentially cure a select group of patients with metastatic disease limited to the resected nodes and lead to improved survival.

The risk of lymph node involvement varies greatly depending upon primary tumor stage and size; RV involvement; presence of metastases; and extent of lymphadenectomy performed [88, 120]. Patients with clinically localized disease have a relatively low incidence (2–9%) of nodal involvement [7, 77, 112, 120], whereas the incidence in patients with metastatic disease or RV involvement is as high as 45% [44, 88]. Local recurrences are rare even when lymphadenectomy is not performed. Rassweiler et al. [98] reported a local recurrence rate of 2.2% following laparoscopic radical nephrectomy (LRN) for localized RCC. At UCLA, the local recurrence rate is approximately 2.8% following radical nephrectomy, and there is no significant difference in local recurrence rate between patients with and without lymphadenectomy performed [88].

Several retrospective studies have suggested that there may be a survival advantage to performing lymphadenectomy at the time of nephrectomy [32, 46, 94]. However, other retrospective studies have not demonstrated a survival advantage in patients treated with a lymphadenectomy [77, 103, 112]. The European Organization for Research and Treatment of Cancer is conducting the only prospective, randomized controlled study (EORTC 30881) comprised of 772 patients with clinically localized disease randomized to nephrectomy with or without a standardized lymphadenectomy. Although the data are still immature, there were no differences in progression or survival between patients treated with or without lymphadenectomy at 5-year median follow-up [7]. These authors noted that the overall 5-year survival (82%) is high and longer follow-up is needed for more events to occur before a difference is survival may become apparent.

Several studies examined the value of lymphadenectomy for patients with both nodal and disseminated metastases. Vasselli et al. [131] examined the impact of regional lymph node involvement on survival of patients undergoing radical nephrectomy for metastatic RCC in preparation for IL-2 based immunotherapy. Patients with no preoperative evidence of lymph node involvement had a significantly longer median survival than those with lymph node involvement (14.7 vs. 8.5 months, respectively). Moreover, the degree of lymphadenopathy was predictive of survival, with less than 50 cm3 retroperitoneal lymphadenopathy having improved survival compared to those with greater than 50 cm3 (8.5 vs. 5.3 months, respectively). Multivariate analysis identified performance status and lack of lymphadenopathy as predictors of improved survival. In addition, patients with minimal or extensive lymphadenectomy had better survival than patients considered to have unresectable disease. For patients with grossly positive lymph nodes and metastases treated at UCLA, the median survival after IL-2 immunotherapy was approximately 5 months longer for patients who underwent lymphadenectomy compared to those who did not [87]. This difference may be partly explained by the observation that positive lymph nodes rarely respond to immunotherapy [87]. Therefore, surgical resection of involved lymph nodes may offer the best chance for a complete response.

There is currently no accepted standard for managing lymph nodes at the time of nephrectomy. With adequate preoperative imaging, the occurrence of unsuspected lymph node involvement will be rare. Therefore, staging accuracy is unlikely to be significantly improved by a formal lymphadenectomy when preoperative imaging is negative. However, given the high false positive rate associated with CT and MRI, a lymphadenectomy may be useful for staging patients more accurately when the lymph nodes are enlarged preoperatively. For patients with localized RCC, a lymphadenectomy at the time of radical nephrectomy is unlikely to lead to a lower risk of local recurrence or improved survival. For patients with lymphatic involvement, several retrospective studies suggest that lymphadenectomy may improve survival. However, in the absence of a randomized, prospective study, strong recommendations cannot be made regarding the role or optimal surgical boundaries of lymphadenectomy in this population.

Open partial nephrectomy

Advances in imaging, the increased incidence of smaller incidentally discovered renal masses, and advances in advanced renal surgical techniques made possible the routine use of partial as opposed to radical nephrectomy. The long-term outcomes of partial nephrectomy which have subsequently become available, though not based on randomized studies, convincingly support the concept that nephron-sparing approaches can achieve equivalent cancer-specific survival to the radical approach in properly selected patients. The contemporary indications for partial nephrectomy or NSS can be categorized as absolute, relative, or elective. The goal of NSS is twofold: (1) oncologically-complete local excision of tumor with minimal technical complications, and (2) optimal functional preservation of the renal remnant. The absolute and relative indications for NSS are widely accepted. In current practice, elective indications are perhaps the most controversial, however, its indications continue to expand. The major argument against elective NSS is the risk of local recurrence in the spared kidney due to incomplete excision or the risk of multifocal disease. Multifocal tumor development can occur in 3–25% of patients [13, 81, 134]. Despite these concerns, the role of NSS in the treatment of RCC continued to be investigated. With almost 60% of RCC being diagnosed incidentally using modern imaging modalities, there is a large percentage of tumors that are small (less than 4 cm) and organ-confined. The fact that contralateral tumors of the kidney may develop metachronously must also be taken into consideration when deciding between total and partial nephrectomy. A small renal mass, usually less than 4 cm, located peripherally and easily amenable to resection is the generally suggested indication based on existing data to date.

Hafez et al. [42] attempted to patients delineate the optimal cutoff size for tumors amenable to NSS and, in so doing, demonstrated that patients with T1 tumors ≤4 cm. who underwent NSS had a significantly better survival than patients with larger tumors. The oncologic efficacy of NSS has been demonstrated in major series by its equivalence to radical nephrectomy with 5-year disease-free survival ranging from 86 to 100% for tumors less than 4 cm [4, 5, 42, 63, 65, 129]. As a result, the 2002 TNM T1 category was amended to T1a and T1b, based on a 4 cm cutoff [115]. Patients with an absolute indication for NSS have lower cancer-specific survival rates compared to those with an elective indication secondary to increased age, tumor size, stage, and overall decreased health [22, 31]. However, the outcomes are relatively consistent with the size- and stage-stratified, disease-free survival reported in historical radical nephrectomy series. NSS is currently being performed most commonly for small tumors less than 4 cm in size in the setting of a normal contralateral kidney, as the 10-year survival rates for patients with these tumors is comparable to patients undergoing radical nephrectomy for tumors of similar size [4, 22]. Ficarra et al. [25] recently reported that the 5 and 10-year cancer-specific survival rates of open radical and NSS were 97.4 and 95.6%, respectively, in patients with T1a disease, 92.5 and 89.8% in T1b, and 89.3 and 78.5% in patients with T2 disease.

The concept of including tumors greater than 4 cm in size is currently being assessed in this elective setting. Emerging data now suggests that NSS can be performed on patients with larger tumors that are anatomically amenable, provided an adequate surgical margin can be safely obtained [64, 91]. The rationale for expanding the indications for elective NSS for larger tumors include an increase in the life expectancy of the general population, along with an increase in incidental diagnosis at an earlier age have led to increased concerns regarding the long-term risk of renal insufficiency or metachronous tumor recurrence in patients that undergo radical nephrectomy [86]. NSS is not recommended for a kidney that cannot be rendered tumor-free with negative surgical margins. Patients with extensive local tumors (minimal remaining normal renal parenchyma), regional adenopathy, or RV or IVC extension should also have a radical nephrectomy.

Aside from the equivalent cancer control, the main perceived benefit of partial nephrectomy is renal parenchyma preservation. Although the value of this preservation is defined in the case of absolute and relative indications, it remains unclear in cases of elective indications. Although several studies have demonstrated a statistically significant decreased risk of chronic renal insufficiency among patients undergoing NSS compared to those undergoing radical nephrectomy, the clinical significance of these findings remains to be defined [62, 74]. NSS has now been shown to be safe and technically feasible [9, 66, 95]. NSS compared to radical nephrectomy has been shown to have equivalent hospital stay, blood loss, and rate of complications [9]. Over the years, complications from NSS have been well characterized and include acute renal failure, permanent dialysis, infection, urinary fistula, hemorrhage, infection, kidney loss, and death [127]. Furthermore, as surgical techniques have evolved, there appears be a decreased rate of complications. Thompson et al. [122] recently reported that patients treated with open NSS in the contemporary era had fewer complications compared to those in the earlier series.

Several investigators have reported on the outcome of partial nephrectomy in individuals with locally advanced and metastatic RCC. In such cases, in addition to the conventional rationale for partial nephrectomy, preservation of renal mass may be desirable to allow adequate renal function to receive systemic therapy. Angermeier et al. [3] reported on a series of nine patients with RCC and IVC involvement in solitary kidneys who underwent NSS. Four of the nine patients died of metastatic disease, two of whom had local recurrences in the renal remnant. Similarly, Sengupta et al. [107] described the NSS approach for 13 patients with RCC in a solitary kidney with extension into the renal vein or IVC (T3b). At a median follow-up of 24 months (range 0–204), eight patients had died, four from RCC (all having had NSS) at a median interval of 9.5 months (range 7–16). Of the five patients alive at a median follow-up of 25 months, four had no identifiable disease, while one had systemic recurrence. Krishnamurthi et al. [59] reported on 15 patients with a solitary kidney and metastatic RCC who underwent NSS. Nine patients had previously undergone contralateral radical nephrectomy and resection of metastatic disease and presented with recurrence in the remaining kidney. Of these patients, 67% were disease-free after partial nephrectomy at a mean follow-up of 31 months.

Laparoscopic radical and partial nephrectomy

In the last 10 years, with improvements in instrumentation and surgeon experience, laparoscopic renal surgery has emerged as an equally efficacious and minimally morbid surgical alternative to open radical nephrectomy for selected kidney tumors. There are a wealth of studies with intermediate to long-term follow-up results that indicate that laparascopic surgery for RCC offers equivalent safety and oncologic efficacy to open surgical approaches. Clayman et al. [15] were the first to perform a LRN for a tumor-bearing kidney. Compared with open surgery, patients benefit from shorter hospitalization, more rapid convalescence, decreased pain, and improved cosmesis [1, 21, 35, 84]. In addition, it has been shown that LRN in elderly and obese patients as well as select individuals with advanced tumors is a feasible and safe procedure with minor technical modifications [4, 5].

LRN has emerged as an alternative to open surgery in the management of low-volume (8 cm or smaller), localized RCC with no local invasion, renal vein involvement, or lymphadenopathy. The three laparoscopic approaches to radical nephrectomy are (1) transperitoneal, (2) retroperitoneal, and (3) hand-assisted. The original description of the LRN [15] was a transperitoneal approach. The retroperitoneal approach was developed to mimic the open flank nephrectomy and allow the hilar vessels to be accessed and controlled without violating the peritoneal cavity. Hand-assisted nephrectomy was developed to facilitate the procedure for the novice surgeon, to provide an alternative to open conversion, and to help expedite the procedure for large renal tumors. Each approach has its own advantages and disadvantages, and surgeons have their own preferences, but the ultimate goal of all three is to perform an adequate oncologic operation while minimizing morbidity. The main problem with LRN relates to the long learning curve, which is often associated with higher complication rates. Nevertheless, multiple series have now demonstrated that with appropriate patient selection and adequate training, LRN can be carried out with complication rates and intermediate-term oncologic outcomes that are equivalent to open surgery [1, 10, 12, 21, 34, 35, 84, 96, 118]. Permpongkosol et al. [93] recently evaluated the long-term oncological efficacy of LRN compared with open radical nephrectomy in 121 patients with clinically localized RCC. The median follow-up was 73 months for the LRN group and 80 months for the open group. There were no significant differences in the 5 and 10-year disease-free survival rates between the laparoscopic and open groups. Furthermore, there were no significant differences in the 5 and 10-year cancer-specific and actuarial survival rates between the two groups.

LRN has traditionally been reserved for patients with organ-confined clinical stage T1 or T2 renal tumors, but the laparoscopic removal of pathologic stage T3 tumors have been reported [12, 135]. The size limitation is dependent on the experience and comfort of the surgeon rather than a limitation of the technique. Although tumors as large as 20 cm have been removed laparoscopically, the need for open specimen removal or conversion to open nephrectomy has been reported to be higher for larger tumors [116, 133]. LRN has been applied to locally advanced or metastatic RCC. Patients with tumor thrombus within the renal vein or IVC are currently not considered to be candidates for LRN due to the technical challenges of the procedure. However, some investigators are attempting to reset the limits of LRN by demonstrating its feasibility [20]. Tumors extending beyond Gerota’s fascia have been approached laparoscopically; however, a greater level of technical expertise may be needed. In addition to being associated with shorter recovery times, patients with metastatic RCC that underwent laparoscopic cytoreductive nephrectomy compared to those that underwent the open approach were shown to have a shorter interval between surgery and initiation of IL-2 therapy [133]. However, there was a significantly higher rate of conversion involving 5 of 11 cases. Furthermore, these patients were not afforded the potential benefit of a lymphadenectomy.

The role of laparoscopy in cancer treatment has been questioned because of the risk of tumor spillage and port site recurrences, which depends on multiple factors including the surgical technique, such as tumor removal without the use of an entrapment sac, and the biological properties of the tumor [75, 124]. In 2003, Tsivian and Sidi comprehensively reviewed port site metastases in urological laparoscopic surgery and found nine published cases [124]. The incidence of port site metastasis following LRN was 0–6.25%. In a study of over 1,098 urologic laparoscopy cases for malignancy, there were eight local recurrences (0.73%) and two port site metastases (0.18%) at a median follow-up of 58 months [98]. Specifically for RCC, the incidence of local recurrence and port site metastasis was 2.2 and 0%, respectively.

The decision to offer a patient a LRN versus an open partial nephrectomy for a small unilateral RCC with a normal contralateral kidney has been debated in the urologic literature since tumors most amendable to LRN (stage T1) are also equally amenable to open partial nephrectomy [5, 64, 71, 91, 104]. LRN leads to short-term benefits in terms of blood loss, hospital stay, and recovery, but at a cost of poorer long-term renal function. Although LRN has become an established procedure, laparoscopic partial nephrectomy (LPN), a technically challenging procedure, has lagged behind. Nevertheless, advances in laparoscopic techniques and surgical skills have allowed the development of a reliable technique for LPN [36] that mimics the open approach. However, at the present LPN remains a technically demanding operation with challenges encountered with securing renal hypothermia, parenchymal hemostasis, pelvicalyceal reconstruction, and parenchymal renorraphy. Even in the best hands, small renal tumors that involve the pelvicalyceal system or are more centrally located pose a particular challenge laparoscopically. Results from pure and hand-assisted LPN have now been reported by multiple centers, which are similar to open surgery [37, 67, 73, 97, 117]. Allaf et al. [2] recently reported intermediate-term results for patients that underwent LPN for clinically localized RCC. The final surgical margin was positive in one patient (2.1%). At 3 years mean follow-up, no recurrences had occurred in 46 of 48 patients (95.8%). One patient with von Hippel-Lindau (VHL) disease was found to have local recurrence 18 months after LPN and observation was elected. The second patient had recurrence in the same kidney approximately 4 years later away from the original tumor site, which was treated by percutaneous ablation.

Other specialized areas of renal surgery

Outcomes of radical nephrectomy for tumors with major venous thrombus

Extension of RCC into the IVC presents a challenge for the urologist from both an oncologic and surgical standpoint. RCC involves the renal vein (RV) and IVC in 16–33% and 4–13% of cases, respectively [68]. Accurate definition of the cranial extent of the thrombus is critical in determining the surgical approach. Level I thrombi extend to within 2 cm of the RV ostium; level II thrombi extend >2 cm from the ostium, but are below the level of the hepatic veins; level III thrombi are above the hepatic veins but below the atrium; finally, level IV thrombi extend into the right atrium. Open surgical management with an aggressive approach to resection of these tumors may provide survival benefit, especially in patients with low grade, organ-confined tumors and the absence of distant metastases. Accurate resection of these tumors requires optimal exposure and technical expertise, which, for the time being, requires an open surgical approach in the vast majority of cases.

Moinzadeh and Libertino [78] recently reviewed 153 patients who underwent nephrectomy and tumor thrombectomy and concluded that long-term survival of patients with RV involvement was significantly greater than those with IVC involvement. Kim et al. [55] compared 226 patients who underwent nephrectomy and RV or IVC tumor thrombectomy with 654 patients that underwent nephrectomy without venous involvement. In patients with localized RCC (N0M0), cancer-specific survival was similar in patients with RV (T3b) and IVC involvement below the diaphragm (T3b). However, patients with IVC involvement above the diaphragm (T3c) had significantly worse survival even after controlling for grade and ECOG-PS in a multivariate analysis. The 3-year cancer-specific survival rates associated with RCC without thrombus, RV involvement (T3b), IVC involvement below the diaphragm (T3b), and IVC involvement above the diaphragm (T3c) were 89, 76, 63 and 23%, respectively. Patients treated for metastatic RCC had a similar prognosis regardless of the level of venous involvement. They concluded that local tumor stage and grade were better predictors of prognosis than extent of venous involvement, supporting the current TNM classification of venous involvement. Recent studies have demonstrated 5-year survival rates of 47–69% for RCC patients with venous involvement and tumor limited to the kidney [39, 139], and with modern advances in surgical technique, surgical resection can be performed with acceptable morbidity. Invasion and infiltration of tumor into the wall of the vena cava has been reported in approximately 20% of patients with IVC thrombi, with the majority of tumors limited to the area of the RV ostium [114]. Surgical management of these tumors generally involves resection of the IVC and repair with a synthetic or autologous pericardium patch graft. This phenomenon may suggest a poor prognosis, but studies are limited.

Role of primary tumor resection in the setting of metastatic disease

Approximately 20–30% of patients present with metastatic disease, and 20–40% undergoing nephrectomy for clinically localized RCC will develop clinically detectable metastases [61]. Patients with metastatic RCC face a poor prognosis, with a median survival of 6–10 months and a 2-year survival of 10–20% [48], however, subsets of patients with advanced disease have experienced improvements in survival. This improved outlook for some patients with advanced and metastatic RCC is related in part to the introduction of immunotherapeutic approaches and a better understanding of the role and timing of cytoreductive nephrectomy [27, 28, 76]. With the emergence of modern immunotherapy, the role of nephrectomy and the relative efficacy of initial biologic response modifier treatment versus nephrectomy has reemerged as a source of controversy.

Historically, the principle behind cytoreductive nephrectomy as a treatment for metastatic RCC was based on the immunological phenomenon of “spontaneous” regression of metastasis following nephrectomy. In a review of 474 patients with metastatic disease who underwent nephrectomy alone, only four (0.8%) experienced spontaneous regression of their metastatic disease [79]. More recently, Marcus et al. reported that of 91 patients, four (4.4%) had complete regression of all metastatic disease following nephrectomy, of which these patients had lung metastases only [70]. Although nephrectomy alone offers no curative benefit in the setting of metastatic disease [19], cytoreductive surgery may have a role when performed in conjunction with cytokine therapy. Accordingly, since the modern era of immunotherapy, a number of studies on prognostic factors in RCC suggested that undergoing nephrectomy was associated with improved survival [80, 82]. Another historical indication for nephrectomy for metastatic RCC has been to improve quality of life. Removal of the malignant kidney may be of palliative benefit in some settings of metastatic RCC and is appropriate when the patient is having pain related to the kidney mass, intractable hematuria, erythrocytosis, uncontrolled hypertension, or persistent hypercalcemia that does not respond to pharmacologic agents [132]. Surgery may also be directed at metastases, in order to control local symptoms, which include the relief of spinal cord compression and fixation of fractures.

The best evidence for performing cytoreductve nephrectomy comes from two prospective randomized clinical trials, SWOG 8949 and EORTC 30947 [27, 76], which revealed a survival benefit for nephrectomy followed by interferon-alfa (IFN-α) compared with IFN-α alone (median survival of 11.1 and 8.1 months, respectively, in the SWOG trial; and 17 and 7 months, respectively, in the EORTC trial). Flanigan et al. [28] performed a combined analysis of these two trials that yielded a median survival of 13.6 months for nephrectomy plus IFN-α versus 7.8 months for IFN-α alone. Cytoreductive nephrectomy appeared to improve overall survival in patients with metastatic RCC treated with IFN-α independent of patient PS, site of metastases, and presence of measurable disease. A retrospective study recently reported by Han et al. [44] demonstrated that patients with more than one metastatic site had a lower response rate to adjuvant immunotherapy following nephrectomy of the primary tumor and a significantly shorter survival than patients with a single metastatic site. Although the observed survival benefit in the SWOG and EORTC-GU studies was independent of the location of metastatic sites, neither trial stratified the number of metastatic sites or overall tumor burden, and therefore this question remains unresolved.

Investigators at UCLA supplemented the SWOG findings with retrospective data to address the relative efficacy of IFN-α versus IL-2 following cytoreductive nephrectomy using a comparable population treated with IL-2 from the UCLA Kidney Cancer Database, containing the records of over 450 metastatic RCC patients treated with immunotherapy [85]. Using the eligibility criteria for the SWOG 8949 study, 89 patients treated with IL-2 based regimens following nephrectomy were identified. Survival of these patients was analyzed and compared to the survival of 120 patients in the SWOG surgery arm. Median survival of the patients treated with nephrectomy plus IL-2 was 16.7 months, which was twice that of the IFN-α only SWOG arm, and 4 months (30%) greater than the nephrectomy plus IFN-α SWOG arm [44]. At 5 years, survival was 19.6% for patients that received IL-2 compared to 10% for those that received IFN-α. These data emphasize the need to determine whether this survival advantage can be further improved with more aggressive immunotherapy or other novel agents in the setting of cytoreductive nephrectomy [61]. However, a randomized trial would be needed to evaluate the precise timing of cytoreductive therapy in combination with these agents with regard to morbidity, overall survival, and quality of life.

Role of metastasectomy

Systemic immunotherapy can be expected to produce objective responses in 10–20% of metastatic RCC patients, of which only 5–10% will be complete responders [101]. Since many patients have only a partial regression of metastases or prolonged stabilization, surgery must often be considered. Clinical situations in which surgery is potentially appropriate include: excision of locally recurrent disease, excision of solitary or multiple metastases, resection of a residual mass after systemic therapy, and palliation. Several series have demonstrated a potential benefit for complete surgical resection of all tumor burden, including removal of both the primary renal mass as well as metastatic deposits in carefully selected patients with minimal volume metastatic disease. Surgical removal of solitary metastasis is widely accepted as potentially effective with 5-year overall survival rates of 29–35% [53, 58, 83, 113, 123, 128]. Best results are obtained if the metastases are pulmonary, metachronous with a long disease-free interval and completely resected [47, 128]. Kavolius et al. [53] have reported on the largest single series of patients undergoing resection of a solitary metastasis. Patients with completely resected solitary metastasis fared better than those who underwent complete resection of multiple metastases, with a 5-year overall survival of 52% compared with 29%. Disease-free survival rates by site were 44% for patients with a solitary metastasis in the lung, 42% gland (pancreas, adrenal, thyroid, and salivary), 20% brain, 40% bone, and 50% for soft tissue metastases. Patients who present with synchronous metastases have an average survival time of approximately 4 months with only 10% surviving 1 year, whereas, patients with metachronous metastases have a median survival time of 11 months regardless of the site of metastasis [26, 69].

Local recurrence in or near the renal fossa is a unique variant of advanced RCC that may not truly represent metastatic disease. It may result from incomplete resection of the primary tumor or persistence of tumor in the regional lymph nodes, but it is generally associated with a poor prognosis, and surgical resection may be considered. Itano et al. [49] reported in the largest series so far that an isolated local recurrence occurred in only 1.8% of 1,737 patients who underwent radical nephrectomy for T1-3N0M0 RCC. Surgical therapy significantly improved the 5-year survival rates when compared with systemic therapies, at 51 and 18% for surgery and conservative treatment, respectively. Subsequent studies reported long-term survival rates of 40–45% after surgical treatment for local recurrences [40, 105]. Tanguay et al. [119] reported on 16 patients with a renal fossa recurrence, in which complete resection was possible in 15 patients and 12 of whom also had negative surgical margins. Twelve patients were alive after a median follow-up of 23.5 months. These reports suggest that aggressive resection of a solitary local recurrence in patients with no other sites of metastases can be associated with long-term survival.

Conclusions

Formidable long-term data currently underlie our understanding of the relative merits of open versus laparascopic, and radical versus partial nephrectomy for RCC. The approach to treating RCC continues to evolve, and further challenges exist as advances are made in technology and understanding the biology and genetics of kidney cancer. The overall strategy for early detection and complete eradication is unchanged, but the tools available for evaluation, diagnosis, and surgical treatments are less invasive and increasingly effective. In addition, as more experience is gained in the surgical management of RCC, definitive answers in areas of controversy will be made available. With an improved understanding of the behavior of RCC and the clinical factors determining prognosis and survival in the management of the disease, minimally invasive treatments (cryotherapy, HIFU, radiofrequency tumor ablation) will be increasingly being offered for a select group of individuals, though long-term outcomes in these areas are still lacking. Despite the trend towards less invasive and less extirpative approaches to RCC, however, there remain a number of indications for the continued practice of open surgical resection in the management of RCC. In 2006, the challenge for the surgical urologist will be not only the mastery of a diverse armamentarium of surgical techniques, but also the wisdom and knowledge to apply the most appropriate surgical technique that maximizes the oncological efficacy while minimizing the surgical morbidity for an individual patient.