Kidney Cancer MD Anderson (Oncology)

From Kidney Cancer Resource

Contents 1 Overview 2 Details 2.1 Kidney Cancer: 2.2 Introduction 2.3 Incidence and Diagnosis 2.4 Epidemiology and Risk Factors 2.5 Genetics and Familial Syndromes 2.6 Von Hippel–Lindau Syndrome 2.7 Types 1 and 2 Hereditary Papillary Renal Cell Carcinoma 2.8 The Birt-Hogg-Dubé Syndrome 2.9 Prognostic Factors 2.10 Clinical Features 2.11 Molecular Markers 2.12 Control of the Primary Tumor 2.13 Radical Nephrectomy 2.14 The Role of Adrenalectomy 2.15 The Role of Lymphadenectomy 2.16 Nephron-Sparing Surgery (Partial Nephrectomy) 2.17 Laparoscopic Radical Nephrectomy 2.18 Minimally Invasive Nephron-Sparing Approaches 2.19 Cytoreductive Nephrectomy 2.20 Adjuvant Therapy 2.21 IL-2 = interleukin-2. 2.22 Management of Metastatic Renal Cell Carcinoma 2.23 Immunotherapy 2.24 Cytotoxic Chemotherapy 2.25 Targeted Therapy and New Agents 3 Disclaimer 4 E&OE - Errors & Omissions Excepted

Overview

MD Anderson Manual of Medical Oncology

Section IX: Genitourinary Carcinomas

Chapter 29: Renal Cell Carcinoma

Details

Kidney Cancer:

MD Anderson Manual of Medical Oncology

Section IX: Genitourinary Carcinomas

Chapter 29: Renal Cell Carcinoma

Introduction

Incidence and Diagnosis

The American Cancer Society predicts that there will be over 35,000 new cases of renal neoplasms in the coming year in the United States (1) and that 12,500 patients will die as a consequence of disease progression. Renal cell carcinoma (RCC) is the most common histology found in kidney tumors, with clear cell (conventional) RCC being the most common histologic subtype (Fig. 29-1). Non–clear cell subtypes include chromophobe, papillary, oncocytoma, collecting duct carcinoma (CDC), and unclassified RCC (2).

Figure 29-1

Photomicrographs of clear cell (conventional) RCC with low-grade A. and high-grade B. nuclear features. Photomicrographs of a type 1 papillary RCC C. showing papillae lined by short cuboidal cells, and type 2 papillary RCC D. showing papillae lined by tall columnar cells, with eosinophilic cytoplasm and high grade nuclear features. (Reprinted by permission from Pheroze Tamboli, MD.)

The incidence of RCC has been increasing over the last several decades (3). Most oncologists attribute this increase in incidence partly to the increased use of better-quality abdominal imaging in patients who present with either unrelated or nonspecific complaints. As a consequence, one would predict an increase in the earlier detection of RCC, in earlier stages, and better outcomes associated with definitive therapy. Several reports have examined large epidemiologic databases to determine whether other causative factors may explain the increased incidence of RCC (3–7). Chow and colleagues (3) examined the database of the National Cancer Institute's Surveillance, Epidemiology, and End Results (SEER) Program looking at patients who were diagnosed with kidney cancer from 1975 through 1995. This database reflects approximately 10% of the U.S. population. They noted an increasing incidence of RCC in men and women regardless of race and also that the increases were greatest for localized tumors, but there was an increased incidence of more advanced and unstaged tumors as well. These data are corroborated in other studies and suggest that there has been an increase in the prevalence of RCC across both gender and racial lines and that, despite the earlier detection, there has not been an improvement in disease-specific mortality rates.

Epidemiology and Risk Factors

It has been well documented that there are racial differences in the incidence of RCC. Both the incidence and mortality of RCC are increasing in African Americans faster than in whites (3). The biology of RCC in African Americans appears more virulent and less responsive to conventional therapies. This may reflect differences in access to health care or dietary influences. It may also reflect a difference in the incidence of comorbidities, such as hypertension or diabetes, that are associated with an increased risk of RCC (see below). In addition to these epigenetic factors, some hypothesize that there are true genetic factors leading to development of a more aggressive form of RCC in African Americans (8).

The link between germline genetic mutations and the development of RCC is well established and applies to a small but biologically important subset of RCC cases (9). While these genetic alterations certainly play an important role in the biology of RCC in both familial and sporadic cases, there are also some environmental factors that contribute to the risk of developing a renal neoplasm. Smoking has long been linked to an increase in the risk of developing RCC, in addition to its association with multiple other malignancies (10–12). As is the case with other malignancies, cessation of smoking can be associated with a diminution in the risk of developing RCC. Interestingly, this diminution in risk appears to be slower and requires a significantly longer time frame to reach baseline than that seen with other malignancies, such as lung cancer.

Obesity or increased body mass index (BMI) has also been linked to increased risk of developing RCC. Several studies have demonstrated a higher incidence of obesity or increased BMI in patients with RCC, suggesting an epidemiologic linkage (13–16). As is the case in most retrospective epidemiologic studies, these observations are confounded by other associated variables, including diet, occupational history, and smoking. In reports by both Kamat et al. and Schips et al., obese patients were not more likely to present with advanced-stage tumors; overall, they had a more favorable prognosis following surgery than did patients with a lower BMI (13,14).

Several studies have pointed to an increase in the risk of developing RCC in association with the diagnosis of diabetes mellitus or hypertension (16–20). Some of the therapies employed for the treatment of hypertension, such as diuretic therapy (20), have been associated with an increased risk of developing RCC. Given the retrospective nature of these studies, it remains unclear whether it is the underlying medical condition or its treatment that increases the risk of developing RCC. Some insight may be forthcoming from a study published in the New England Journal of Medicine (16), which demonstrated that successful treatment of hypertension lowered the risk of developing RCC.

Dietary factors have been linked to the development of RCC. These include a high-fat or high-protein diet or the ingestion of significant quantities of fried foods (21,22). Increased alcohol consumption and increased consumption of red meat have also been identified as risk factors (23,24). A study published by Yuan et al. suggested that increased intake of cruciferous vegetables may decrease the risk of developing RCC (25).

Exposure to specific drugs and environmental carcinogens has also been linked to the development of RCC. It has been reported that increased use of analgesics as well as amphetamines may increase the risk of RCC, although this has been disputed by others (26–29). Occupational exposures to asbestos and cadmium have also been implicated; however, all of these studies suffer from multiple confounding variables that make causality difficult to establish (30,31).

In conclusion, there are multiple factors that may play a significant role in RCC carcinogenesis and progression, through both genetic and epigenetic pathways. With a more thorough understanding of the biology of RCC, we may be able to identify the causal mechanisms behind these epidemiologic factors and their relationship to the development of renal cancer.

Genetics and Familial Syndromes

The large majority of RCC cases are considered to be sporadic. Nonetheless, important lessons about the pathogenesis of RCC can be learned from familial RCC syndromes that have been characterized over the past few decades.

Von Hippel–Lindau Syndrome

Between 1:32,000 and 1:40,000 live births are affected by von Hippel–Lindau (VHL) syndrome every year in the United States (32). People living with this disease must contend with retinal angiomas, cerebellar and spinal hemangioblastomas, pheochromocytomas, clear cell RCC, pancreatic neuroendocrine tumors, pancreatic cysts, endolymphatic sinus tumors, and epididymal cysts. The genetic lesion is found on 3p25.26 and involves truncation, missense, nonsense, or small deletional mutations in the VHL gene. The VHL protein is involved in a number of cellular functions, including regulation of hypoxia-inducible factors (HIFs) and extracellular matrix regulation; it is physically associated with a number of other important regulatory proteins. Up to 80% of sporadic clear cell RCC tumors also possess a mutation in the VHL gene, presenting a key example of how germline mutations can aid in our understanding of somatic defects in sporadic cancers (9).

Types 1 and 2 Hereditary Papillary Renal Cell Carcinoma

The type 1 hereditary papillary RCC syndrome (HPRC) is associated with a mutation in the c-met gene, found on chromosome 7. The clinical syndrome was first described by Zbar and associates in 1994 (33) and the genetic mutation was delineated in 1997, at chromosome 7q31.1-34 (34). Missense mutations in the met gene resulted in constitutive activation of met. This mutation is inherited in an autosomal dominant pattern. Affected patients demonstrate multiple bilateral renal lesions with up to 3400 microscopic tumors per kidney (9). Kidney cancers in HPRC occur in the fourth, fifth, and sixth decades of life. Because the risk of metastases and death increases with size above 3 cm, lesions are generally removed when they reach that size, with nephron-sparing approaches used to maintain renal function as long as possible (see Fig. 29-1) (9).

Hereditary leiomyomatosis RCC (HLRCC, or type 2 papillary RCC) arises in patients with multiple cutaneous leiomyomas (MCLs), uterine leiomyomata, and an increased risk of leiomyosarcoma. The Multiple Leiomyoma Consortium mapped the affected gene to 1q42.3-q43 and ultimately identified it to be fumarate hydratase, an enzyme of the carboxylic acid cycle (35). Type 2 renal tumors tend to be solitary RCCs, are aggressive, and tend to metastasize early. Early surgical intervention is recommended in these cases, although no precise guidelines regarding size and metastatic potential have been proposed.

The Birt-Hogg-Dubé Syndrome

Patients with Birt-Hogg-Dubé (BHD) syndrome have chromophobe RCC, fibrofolliculomas trichodiscomas, and pulmonary cysts. The eponymous cutaneous syndrome was first described in 1977 by Birt et al. (36), and in 1993 Roth et al. reported on a patient with bilateral multifocal chromophobe RCCs. The BHD gene was localized to the short arm of chromosome 17, and the gene folliculin was described in 2002 by Nickerson et al. (37). A study sponsored by the National Cancer Institute (NCI) demonstrated an increased prevalence of renal neoplasia in 152 individuals with BHD, with oncocytomas figuring prominently, followed by papillary RCC. Pavlovich et al. reviewed 130 solid renal tumors resected from 30 patients with BHD in 19 different families. Preoperative computed tomography (CT) scans demonstrated a mean of 5.3 tumors per patient. It was found that 34% of these were chromophobe RCC, 9% were clear cell RCC, 5% were oncocytomas, and 50% were hybrid oncocytic neoplasms (38).

Prognostic Factors

The prognostic classification of patients with RCC presents an important challenge for clinicians. The clinical course is variable, depending on tumor characteristics, the age and general medical condition of the patient, and host factors such as angiogenesis and the immune response. Patients undergoing nephrectomy for localized disease remain at risk of recurrence for many years and require appropriate counseling and surveillance. The traditional measures of performance status (PS), tumor stage, and tumor grade each demonstrate good correlation with clinical outcome after nephrectomy. The integration of these parameters with laboratory markers—including anemia, hypercalcemia, and elevated lactate dehydrogenase (LDH)—allows one to further classify patients with metastatic disease into groups with clinically relevant differences in outcome. A better understanding of RCC tumor biology and its underlying mo-lecular diversity is needed in order to refine our classification systems and ultimately to develop more effective treatment for this heterogeneous disease.

Clinical Features

The UCLA Integrated Staging System (UISS) for patients undergoing nephrectomy for RCC was reported in 2001 and integrated the 1997 IUAC staging system, Fuhrman nuclear grade, and PS (Tables 29-1 and 29-2) (39,40).

Table 29-1 Stage Definitions for Renal Cell Carcinoma

Stage	Definition	TNM*
I	Limited to kidney; 7 cm or less.	T1, N0, M0
II	Limited to kidney; more than 7 cm.	T2, N0, M0
III	Extends into major veins or perinephric tissues but not beyond Gerota's fascia.	T1, N1, M0
		T2, N1, M0
		T3, N0, N1, M0
IV	Metastases or local invasion beyond Gerota's fascia.	Any T4,
		any N2,
		any M1

TNM = tumor, node, metastasis.

• T3a directly invades the adrenal gland or perirenal and/or renal sinus fat; T3b grossly extends into the renal vein, its segmental branches, or the vena cava below the diaphragm; T3c invades the wall of the vena cava or extends into the vena cava above the diaphragm. N1 is defined as metastasis to a single regional lymph node irrespective of laterality, and N2 is any lymph node involvement beyond N1.

SOURCE: Guinan et al. (39). With permission. Table 29-2 The UCLA Integrated Staging System

UISS	1997 TNM Stage	Fuhrman's Grade	Zubrod Performance Status
I	I	1, 2	0
II	I	1, 2	1 or more
	I	3,4	Any
	II	Any	Any
	III	Any	0
	III	1	1 or more
III	III	2–4	1 or more
	IV	1, 2	0
IV	IV	3, 4	0
		1, 3	1 or more
V	IV	4	1 or more

TNM = tumor, node, metastasis; UCLA = University of California Los Angeles; UISS = UCLA Integrated Staging System. SOURCE: Zisman et al. (40). With permission. The UISS has five tiers (I to V) that closely follow the TNM stages (I to IV), with a fifth category composed of stage IV/grade 4 tumors in patients with Zubrod PS 1 (i.e., all the highest risk factors). In a recent update, the projected 5-year survival from nephrectomy was 94, 67, 39, 23, and 0% for UISS categories I, II, III, IV, and V, respectively (41). In a subsequent publication, Zisman et al. (42) presented another iteration of the UISS, segregating patients into low-, intermediate-, and high-risk groups with or without metastases (Fig. 29-2). Figure 29-2 Decision box defining risk groups for patients with nonmetastatic A. or metastatic B. RCC at the time of nephrectomy. [From Zisman et al. (42). With permission.] The survival curves clearly separate according to risk group and presence or absence of established metastases except that the nonmetastatic/high-risk and metastatic/low-risk groups were similar (Fig. 29-3). Figure 29-3 Disease-specific survival comparison for risk groups segregated according to metastatic or nonmetastatic RCC at the time of nephrectomy. [From Zisman et al. (42). With permission.] The nonmetastatic groups were well distributed on each curve, whereas the majority (78%) of metastaticpatients were clustered in the intermediate-risk group. This study illustrates the heterogeneity seen in nonmetastatic patients as well as the need to consider additional prognostic variables in patients presenting with metastaticRCC. Motzer and others (43) at MSKCC identified five prognostic factors predictive of shorter survival in patients with metastaticRCC: low PS (Karnofsky <80%), high LDH (>1.5 times normal), anemia, hypercalcemia (corrected >10 mg/dL), and absence of prior nephrectomy. In a cohort of 670 patients with median survival duration of 10 months, 22% of patients had three or more risk factors (poor risk), 53% had 1 or 2 (intermediate risk), and 25% had none (favorable risk). The median survival for each group was 4, 10, and 20 months, respectively, with significant differences also seen in 1-, 2-, and 3-year survival. It is important to note that "favorable-risk" does not mean "good prognosis," as these patients still have a high mortality. The MSKCC system based on these five parameters gives reproducible results and can greatly facilitate the design and interpretation of clinical trials by allowing appropriate comparisons between similar groups of patients. Risk-factor analysis does not, however, predict response to systemic therapy (44), nor does response to systemic therapy predict survival [with the notable and rare exception of a complete response (CR) to high-dose IL-2] (45). Other frequently cited risk factors include number of metastatic sites, non–clear cell histology (46), extrapulmonary versus lung-only metastases, presence of constitutional symptoms, and interval from nephrectomy to recurrence less than 1 year (47). Tumors with papillary histology can have a very indolent course, but non–clear cell histology has also been associated with a lower probability of response to systemic therapy (46). These and other considerations can be used when formulating the clinical impression for an individual patient, but they are neither as objective nor as reliable for predicting median survival as the UCLA or MSKCC classification. With regard to the latter, the absence of prior nephrectomy is becoming less significant because of the increasingly accepted practice of performing nephrectomies in patients with metastatic disease (47).

Molecular Markers

A very small subset of patients, limited almost exclusively to PS 0 and clear cell histology, achieve CR to high-dose IL-2 with median response duration greater than 10 years (48). Another subset of patients with RCC enjoy long-term survival with supportive care or no treatment at all because their disease follows an indolent course. Prospective identification of these and other distinct groups on the basis of molecular markers would allow treatment to be individualized for maximum benefit.

The VHL gene product regulates the hypoxia-induced pathway and is commonly mutated in clear cell RCC. The hypoxia-induced pathway leads to the activation of survival genes that mediate glucose transport, proliferation, angiogenesis, and pH regulation and is also implicated in other tumor types (Fig. 29-4).

Figure 29-4

A schematic representation of the hypoxia-induced pathway. A. VHL facilitates degradation of HIF under normoxic conditions. B. The loss of VHL in RCC allows HIF to accumulate, as it normally would under hypoxic conditions.

In a Japanese study of 187 patients receiving nephrectomy for clear cell RCC, mutation or hypermeth-ylation of the VHL gene was found in 58% of tumors and was associated with significantly improved disease-free and cancer-specific survival in patients with organ-confined tumors (n = 134), but not in those with stage IV disease at the time of nephrectomy (n = 53) (49). VHL abnormalities probably occur early in the development of clear cell RCC and are a potential marker for identifying the subset of patients who benefit from surgery.

Loss of VHL allows accumulation of HIF-1 protein, activation of the HIF-1 complex, and HIF-1–induced expression of genes, including VEGF and carbonic anhydrase. Carbonic anhydrase IX (CA IX), a hypoxia-inducible enzyme that controls cellular pH, is highly expressed in RCC but absent in most normal tissues. A study (50) of CA IX expression in 321 clear cell RCC nephrectomy specimens showed expression in 94%, most of which were highly positive (>90% of cells). Expression was lower in metastases than in primary tumors. Low or absent CA IX was an independent predictor of death from cancer in patients who had metastases at the time of nephrectomy (n = 140), but not for those with organ-confined (i.e. curable) disease. Studies are currently under way to determine whether CA IX expression can be used to predict responses to systemic therapy such as IL-2. Progress in the characterization of this and other molecular markers may also reveal new opportunities for targeted therapy.

Control of the Primary Tumor

Radical Nephrectomy

Radical nephrectomy was shown by Robson and colleagues in the 1960s to provide a survival advantage for patients with RCC. Eighty-six patients followed for a minimum of 3 years were found to have 5-year survival of 66, 64, 42, and 11% for stages I, II, III, and IV disease, respectively, which was an improvement over other forms of surgery performed at the time (51). Nearly two-thirds of the patients in this cohort were found to have locally advanced disease. Radical nephrectomy has thus become the "gold standard" therapy for RCC, and, as defined by Robson, includes early vascular ligation, extrafascial dissection of the kidney (outside Gerota's fascia), en bloc resection of the adrenal gland, and extensive lymphadenectomy from the crus of the diaphragm to the aortic bifurcation.

The majority of surgical outcomes research since the initial reports by Robson (52) and Skinner (53) has focused on reducing the morbidity of surgery and limiting the amount of extirpation. As such, the roles of routine adrenalectomy and lymphadenectomy have been redefined, and nephron-sparing surgery (partial nephrectomy) is now standard practice for selected patients. Lap-aroscopic radical nephrectomy (LRN) has been introduced as a new form of treatment for low-to-intermediate clinical stage tumors. Other minimally invasive therapies are being implemented for small, incidentally detected renal tumors.

The Role of Adrenalectomy

The overall incidence of adrenal metastasis is generally less than 5%. Sagalowsky et al. (54) reviewed 695 cases of RCC and found adrenal involvement in 4.3%. The risk of adrenal metastasis in that study correlated with left-sided, large, upper-pole, and high-stage tumors and in most cases was associated with additional extrarenal sites of dissemination. Shalev et al. (55) reported an overall 3.7% rate of adrenal involvement in 285 patients undergoing radical nephrectomy with concomitant adrenalectomy. In contradistinction to the data by Sagalowski et al., Kozak et al. (56) found no association between location of the renal tumor and occurrence of adrenal metastasis, and no curative benefit was afforded by adrenalectomy in their series of 225 patients.

Contemporary radiographic imaging has been found to be a very sensitive tool for the detection of adrenal metastases. In a prospective analysis of 100 patients by Kletscher et al. (57), of 4 abnormal adrenals visible on CT scan, 2 harbored metastatic disease while the other 2 contained benign disease (57). In the analysis of the UCLA data, Tsui et al. (58) reported 99.4% negative predictive value of preoperative CT scan. The overall adrenal involvement rate was 5.7%, but it was highly stage-dependent, with T1 to T2, T3, and T4 tumors associated with a 0.6, 7.8, and 40% rate of adrenal involvement, respectively. In practice, an adrenal-sparing radical nephrectomy appears appropriate in the setting of a normal-appearing adrenal on CT scan for a clinical T1 to T2 tumor that does not abut or encroach on the adrenal gland.

The Role of Lymphadenectomy

The overall incidence of lymph node metastases in patients with radiographically localized disease ranges from 3 to 15% and correlates with the number of lymph nodes removed (59–61), tumor stage, and tumor grade (60,62). A phase III randomized trial conducted by the European Organization for Research and Treatment (EORTC) comparing complications of radical nephrectomy with (N = 389) and without (N = 383) lymphadenectomy in patients with resectable localized disease found no difference in complication rates (63).

At present, lymphadenectomy is a safe adjunct to radical nephrectomy and is primarily used for staging. Extensive lymph node metastasis appears to be an independent predictor of survival (64). In a large series from UCLA, N0, M1 patients were more likely to respond meaningfully to systemic immunotherapy compared with N+, M1 patients (65), a phenomenon also observed by others (64). Select patients with minimal regional lymph node disease and no other sites of metastasis represent a subgroup who may benefit from more extensive lymph node dissection (66). Lymph nodes less than 2 cm on CT scan or MRI are not always malignant and may be enlarged because of inflammation (67). Thus, visible enlargement on preoperative imaging or palpable enlargement at surgery warrants resection for diagnostic purposes, but the therapeutic benefit of surgical debulking is unclear (68).

Nephron-Sparing Surgery (Partial Nephrectomy)

The ability to perform partial nephrectomy, or nephron-sparing surgery (NSS), has been facilitated by advances in technology as well as an improved understanding of the biology of RCC. Palmer and Swanson suggested in 1978 that partial nephrectomy, when technically feasible, be considered in the setting of a solitary kidney (69). Controversy over elective indications for partial nephrectomy (i.e., a small tumor with a normal contralateral kidney) persisted. Recently, several long-term studies have shown equivalent survival of patients undergoing NSS versus radical nephrectomy for a unilateral renal tumor <4 cm in size (70–72). Recurrence-free survival after 5 and 10 years was shown to be comparable to radical nephrectomy in this group. The risk of recurrence increases with larger tumors, bilateral tumors, multifocality, symptoms, and certain histologies such as papillary RCC (73).

Contemporary indications for NSS are listed in Table 29-3.

Table 29-3 Indications for Nephron-Sparing Surgery

Elective

 Single renal tumor <4 cm, normal contralateral kidney 

Absolute

 Bilateral renal tumors 
 Systemic condition threatening renal function (e.g., diabetes mellitus) 
 Local condition threatening contralateral kidney (e.g., stone disease, renal artery stenosis) 
 Chronic renal insufficiency 
 Solitary functioning kidney 

Recent data have also shown a significant improvement in quality of life for patients undergoing NSS versus radical nephrectomy (74–76), in addition to a lower long-term risk of renal failure (77). Historically, a 1-cm margin of tissue was deemed necessary during partial nephrectomy. However, recent data have shown that an exact width of margin is unnecessary as long as an actual margin is present (78–80).

Laparoscopic Radical Nephrectomy

The first laparoscopic radical nephrectomy (LRN) was reported in 1991 (81). LRN was slow to be accepted into the practice of urologic oncology because of concerns regarding cancer spillage, port site seeding, and other oncologic issues. Since then, three 5-year follow up studies, one of which was multi-institutional, have shown equivalent survival of patients undergoing LRN versus open radical nephrectomy (Table 29-4) (82–84).

Table 29-4 Results of Laparoscopic and Open Radical Nephrectomy

Study Author (reference)	Number of Patients	Tumor	Stage	Follow-Up (months)	5-Year DFS (%)*	5-Year OS (%)*
Saika (82)	Open 68	T1	65	87	94
	Lap 168	T1	40	91	94
Chan (83)	Open 54	T1–2	44	86	75
	Lap 67	T1–2	36	95	86
Portis (84)	Open 69	T1–2	69	91	89
	Lap 64	T1–2	54	92	81

DFS = disease-free survival; OS = overall survival.

• p = not significant for all.

The well-documented benefits of LRN include reduced blood loss, less pain and narcotic analgesic requirement, quicker ambulation, faster resumption of oral diet, shorter hospitalization, and more rapid convalescence. Most clinical T1 to T2 and some T3 tumors are amenable to laparoscopic excision (Fig. 29-5). Figure 29-5 A laparascopic nephrectomy specimen shows a small tumor invading the renal sinus. (Reprinted by permission from Surena Matin, MD.) Complication rates are equivalent to—if not better than—those after open surgery, with an overall average of approximately 12 to 15% (85,86). Operative time is equivalent to open surgery in experienced hands, and the cost is less (87). LRN is becoming an accepted standard for the majority of clinically localized RCCs (Table 29-5). Table 29-5 Indications and Contraindications for Laparoscopic Radical Nephrectomy Indications Clinical stage T1–T2 Tumor size <14 cm M+ disease fulfilling other criteria, with good performance status Relative contraindications Renal veintumor thrombus Extensive perinephric extension Prior renalsurgery Hilar adenopathy Absolute contraindications High anesthetic risk Abdominal or systemic sepsis Uncorrectable coagulopathy Tumor thrombus within vena cava Extensive hilar or interaortocaval adenopathy Invasion of surrounding organs LRN is performed by transperitoneal, retroperitoneal, and hand-assisted laparoscopic approaches (88, 89). Hand-assisted laparoscopicnephrectomy (HALN) was developed as a consequence of the need to perform intact specimen extraction and to improve the learning curve of laparoscopy by allowing the surgeon's hand to assist with surgery. The primary advantage of HALN compared to conventional LRN appears to be a shorter operative time (90), and both have similar advantages compared to open nephrectomy (91). For patients with metastatic disease, the shorter recovery time after laparascopic nephrectomy could also allow the administration of systemic therapy sooner than is possible in patients undergoing open surgery (92). Some urologists have proposed performing specimen morcellation during conventional (not hand-assisted) LRN because of a perceived advantage from a patient morbidity viewpoint. However, specimen morcellation has not consistently been shown to cause less pain, and the only accepted advantage is cosmesis. Morcellation precludes accurate pathologic staging and postoperative risk stratification, surgical margin assessment, and enrollment in risk-specific adjuvanttrials; it also risks injury to abdominal viscera as well as tumor spillage. Specimen morcellation is not performed at M.D. Anderson Cancer Center (MDACC). Our own data show that 26% of patients with clinical T1 to T2 tumors are found, on pathologic evaluation of the intact specimen, to harbor high-risk disease warranting a more intensive follow-up regimen or enrollment into an adjuvanttrial (93). Port site metastasis is an unusual event that appears to occur with the same frequency as implantation into an open incision and likely reflects the biology of disease (94). Carbon dioxide pneumoperitoneum does not facilitate tumor seeding (95). In total, four cases of port site metastasis after LRN for RCC have been reported, including one instance after HALN (96,97).

Minimally Invasive Nephron-Sparing Approaches

The next evolution of kidney surgery is the integration of minimally invasive techniques with nephron-sparing surgery. The increasing detection of small, incidentally discovered tumors, advancement of minimally invasive techniques, application of new technologies, and an improved understanding of the natural history of RCC have all made this possible. This is a rapidly evolving field for which little multi-institutional data and virtually no long-term oncologic data are available. Presently utilized clinical modalities are briefly discussed below.

Laparoscopic partial nephrectomy (LPN) is an effective, minimally invasive nephron-sparing approach for select small renal tumors. It largely adheres to principles of open surgery, including hilar clamping, sharp dissection of the renal tumor with a margin of normal tissue, suture ligation of the pelvicalyceal collecting system, ligation of central renal vessels, and parenchymal reconstruction (98). However, it remains a challenging surgical endeavor that requires extensive laparoscopic experience and careful patient selection.

Minimally invasive nephron-sparing modalities, including laparoscopic or percutaneous energy-ablative therapies [cryoablation and radiofrequency ablation (RFA)], are usually reserved for small (<4 cm) tumors and patients at higher operative risk. The single greatest advantage of a percutaneous approach is the ease of repeat therapy and the minimal alteration of patient homeostasis. Percutaneous renal cryoablation or RFA may be employed under magnetic resonance imaging (MRI) or computed tomography (CT) guidance, respectively, for tumors that are posteriorly or laterally located. Lesions that are anterior or in close approximation to the ureter, bowel, or other critical structures are best approached laparoscopically. Small exophytic tumors are probably treated just as well with either approach, while larger (>3.5 cm) and more central lesions have a higher rate of incomplete therapy with RFA.

Postoperative care for cryoablation or RFA requires a short (24-h) period of observation, and both procedures have a favorable adverse outcome profile. After cryoablation, a 3-year recurrence rate of approximately 4% has been reported (99). More incomplete treatments have been reported with RFA than with cryoablation, but recent data have identified selection factors associated with improved outcomes, including exophytic location and small size (100). All energy-ablative therapies require intensive postoperative follow-up imaging to document effective treatment, as there is no pathologic confirmation of cure or any long-term data. Our current regimen at MDACC is to obtain a renal protocol CT scan at 1, 3, 6, 12, and 24 months after the procedure and for as long as a nonenhancing lesion is visualized (Fig. 29-6).

Figure 29-6

Computed tomography of a RCC metastasis to the contralateral kidney before A. and after B. radiofrequency ablation.

Cytoreductive Nephrectomy

The role of initial cytoreductive radical nephrectomy as part of a multidisciplinary treatment approach for patients with metastatic RCC has remained a subject of much controversy in the urologic literature. At issue is the question of whether surgical control of the primary can be performed safely and also provide benefit in patient outcome. Proponents of performing cytoreductive nephrectomy prior to systemic therapy have suggested that upfront surgery can palliate local symptoms, improve quality of life, and possibly improve the response to systemic therapy and overall survival. With the increase in popularity of laparosocopic surgery, which can be applied in the cytoreductive setting, there is even more impetus to incorporate cytoreduction in the overall treatment approach, given the quick recovery and minimal morbidity associated with the procedure (92). Opponents to upfront surgery argued that there were insufficient data to support routine application of cytoreductive nephrectomy in the metastatic setting, and—in addition to the morbidity, mortality, and cost considerations, that systemic therapy would be delayed by the surgery and its potential complications.

Prior to the recent completion and publication of two large phase III trials, the evidence regarding efficacy for this treatment paradigm consisted of several single-institution series (101–108). A summary of the literature regarding cytoreductive nephrectomy can be found in Table 29-6.

Table 29-6 Summary of Results from Single-Institution Series of Cytoreductive Nephrectomy

Author (reference)	Institution	N	Mortality (%)	Did not Receive Therapy Post-Op(%)	Response Rate (%)
Rackley (101)	Cleveland Clinic	37	2.7	22	8.1
Wolf (102)	UCSF	23	0	26	13
Bennett (103)	Albert Einstein	30	17	77	13
Franklin (104)	UCLA	63	0	12	34
Walther (105)	NCI	195	1.0	38	18
Fallick (106)	Tufts	28	3.6	7.1	39
Levy (107)	MDACC	66	3.0	18	NR
Wood (108)	MDACC	126	1.6	4.0	NR

MDACC = M.D. Anderson Cancer Center; NCI = National Cancer Institute; NR = not reported; UCLA = University of California, Los Angeles; UCSF = University of California, San Francisco. These reports have been a subject of controversy in the urologic literature due to concerns about patient-selection bias and the retrospective nature of the studies. To that end, many oncologists have been critical of implementing cytoreductivenephrectomy in the treatment of metastaticRCC until the recent publication of two phase III trials, performed simultaneously in the United States and Europe, which demonstrated a survival benefit for patients who underwent the procedure prior to systemic therapy. The EORTC Genitourinary Group trial 30947 was a prospective randomized trial that included 85 patients with metastaticRCC; they were randomized to either immunotherapy alone with interferon- 2b (IFN-) or nephrectomy followed by IFN- (109). There were 42 patients in the surgery-plus-immunotherapy arm and 43 patients in the immunotherapy-alone arm. There were no perioperative deaths and six perioperative complications, with only one surgical patient not receiving immunotherapy in the postoperative period. The study demonstrated no difference in overall response (complete plus partial) to IFN- between the two groups (surgery 19% versus control 12%, p = 0.38), but the surgery group had improved median survival (17 versus 7 months, p = 0.03) and time to progression (5 versus 3 months, p = 0.04). The Southwest Oncology Group (SWOG) trial 8949 also examined the role of cytoreductivenephrectomy in 241 patients with metastaticRCC treated with IFN- (110). There were 121 patients in the arm receiving IFN- alone and 120 in the arm receiving nephrectomy plus IFN-. One perioperative death occurred; there were also 5 major and 16 mild-to-moderate surgical complications; but, impressively, 98% of the patients who underwent surgery were able to recover and go on to receive IFN-. The mean time from surgery to initiation of systemic therapy was 19.9 days, with three responses seen in each group, and an overall survival difference favoring patients who underwent cytoreductivenephrectomy (median 12.5 versus 8.1 months, p = 0.012). Of note, there were more patients with impaired PS (SWOG score = 1) in the control arm (58.1 versus 45.0%, p = 0.04). Because the groups were not balanced for PS, the investigators performed a proportional-hazards analysis and found no significant interaction between treatment group and PS. Furthermore, there was a survival advantage favoring surgery for both PS subgroups (6.9 versus 4.8 months for PS 1; 17.4 versus 11.7 months for PS 0). Both the SWOG and EORTC trials demonstrate a survival advantage for patients undergoing nephrectomy prior to the initiation of systemic therapy, but some have questioned the choice of IFN- because of its modest response rate. To that end, the UCLA group retrospectively applied the SWOG study entry criteria to a population of patients in their RCC database who had received cytoreductivenephrectomy followed by IL-2 (111). The 89 patients thus identified had a median survival of 16.7 months, or 4 months longer than the median survival reported for the surgery group in the SWOG trial. They argue that cytoreductivenephrectomy followed by systemic therapy with IL-2 may enhance the survival benefit beyond that achieved with IFN-, although no randomized trial has yet been performed. Based on these data, many urologic oncologists are now performing cytoreductive nephrectomies on selected patients with metastaticRCC, particularly those with good PS, absence of sarcomatoid histology, and without central nervous system or bone metastases. The criteria used at MDACC to determine the appropriateness of cytoreductivenephrectomy are summarized in Table 29-7. Table 29-7 Patient-Selection Criteria for CytoreductiveNephrectomy Favorable Good performance status Good operative risk Symptoms from primary tumor Clear cell histology Will likely receive systemic therapy Unfavorable Poor performance status (other than primary tumor symptoms) Poor operative risk Symptomatic or rapidly progressive metastases Diffuse bone metastases Uncontrolled CNS metastases* Constitutional symptoms Hypercalcemia Sarcomatoid histology CNS = central nervous system.

• Patients with CNS metastases can be considered after resection of solitary or oligometastases and a period of observed stability.

One of the important issues that has come to light is the wholesale application of cytoreductivenephrectomy to patients regardless of their suitability for this procedure. In the larger U.S. study, patients with a SWOG PS of 1 (Zubrod 1) who underwent cytoreductivenephrectomy demonstrated a clinically insignificant benefit, with median overall survival increased from approximately 4 months for those without nephrectomy to 6 months with the procedure (110). These data once again underscore the importance of patient selection in considering nephrectomy for those with metastatic disease.

Adjuvant Therapy

Currently, no effective adjuvant therapy exists for patients who undergo nephrectomy for locally advanced RCC. Risk stratification for this patient population is determined by a variety of factors implicated in RCC tumor biology, including tumor stage, Fuhrman's grade, and PS (112,113). Ideally, an effective adjuvant therapy should be relatively nontoxic, demonstrate some efficacy in the metastatic setting, and be easily administered in the outpatient setting. Furthermore, it should be tested against the standard of care in clinical trials prior to being widely incorporated into the practice patterns of oncologists. Currently, with the exception of a recently published study demonstrating a progression-free survival advantage for patients receiving vaccine therapy (114), proven effective adjuvant therapy for RCC does not exist. Results are anticipated from several other ongoing adjuvant trials, which may redefine the role of adjuvant therapy in RCC.

A variety of treatment modalities have been proposed and tested as adjuvant or neoadjuvant therapy, including chemotherapy, vaccines, antiangiogenic therapy, and local therapy to the primary tumor prior to extirpation (e.g., embolization). A review of this literature demonstrates both the difficulty investigators have had in uniformly defining the target population at high risk and the inability to identify effective therapy despite promising data in the metastatic setting.

The hormone medroxyprogesterone acetate was one of the first agents examined as an adjuvant in a phase III multicenter trial in RCC (115). The treatment group demonstrated no significant survival benefit over the control (observation) group, and patients in the treatment group experienced substantial treatment-related toxicity.

Immunotherapy is a logical choice as an adjuvant strategy in RCC, since IL-2 and IFN- have demonstrated some efficacy in the metastatic setting (116). Unfortunately, even the limited benefits of immunotherapy for metastatic disease have not been realized in the adjuvant setting. As shown in Table 29-8, numerous trials have examined the role of IFN- and IL-2 as adjuvant therapy in patients with RCC at high risk of relapse (117–123).

Table 29-8 Summary of Adjuvant Immunotherapy Trials

Author	(reference)	Agent -@ : alpha	Result
Prummer	(117)	Interferon-@	No benefit
Migliari	(118)	Interferon-@/vinblastine	No benefit
Messing	(119)	Interferon-@	No benefit
Basting	(120)	Interferon-@	No benefit
Jeon	(121)	Interferon-@/vinblastine	No benefit
Pizzocaro	(122)	Interferon-@	No benefit
Clark	(123)	High-dose intravenous IL-2	No benefit

IL-2 = interleukin-2.

To date, no trial utilizing adjuvant immunotherapy has demonstrated a survival benefit, and there has been significant toxicity, thus further dampening enthusiasm for this approach.

Chemotherapeutic regimens have also been examined in the adjuvant setting. In a prospective, single-arm phase II study, vinblastine, doxorubicin, and tegafur were administered to patients starting 2 weeks after surgery (124). The survival of these patients was then compared to that of 60 historical controls. An apparent survival benefit was seen by this comparison. Criticisms of this study, however, include the lack of a prospective control arm, the significant toxicity of the chemotherapy, and that there were patients in the treatment group who would not be deemed high risk by published criteria. As a consequence, there has been little to no enthusiasm for this treatment strategy in the oncologic community.

Angiogenesis inhibitors are theoretically attractive adjuvant agents in high-risk RCC. RCC tumors are highly vascular and angiogenic pathways appear central to RCC biology. Potentially useful agents are in clinical development. Thalidomide has demonstrated evidence of antitumor activity in metastatic RCC (see Table 29-8), with angiogenesis inhibition as one of its possible mechanisms (125–128). In an ongoing randomized phase II study at MDACC, eligible patients are restaged following nephrectomy and are randomized to either observation alone or treatment with thalidomide 300 mg daily for 2 years. The endpoint of the trial is disease-free survival, and preliminary results regarding safety and toxic-ity of the treatment are encouraging.

Vaccine strategies have also been explored as potential adjuvant therapies for RCC. Initial studies in tumor vaccination for RCC have examined the role of whole-tumor cell lysates harvested from radical nephrectomy specimens at the time of surgery (114,129,130). These lysates are administered postoperatively at defined intervals and patients are followed for evidence of disease recurrence and survival. In Germany, Repmann et al. performed a randomized study involving 558 patients at 55 institutions (114). Before surgery, all patients were centrally randomized to receive autologous renal tumor cell vaccine (six intradermal applications at 4-week intervals postoperatively) or no adjuvant treatment. A total of 379 patients were assessable for the intention-to-treat analysis. At 5-year and 70-month follow-up, the hazard ratios for tumor progression were 1.58 (95% CI [confidence interval] 1.05 to 2.37) and 1.59 (1.07 to 2.36), respectively, in favor of the vaccine group (p = 0.0204, log-rank test). Five-year and 70-month progression-free survival rates were 77.4 and 72%, respectively, in the vaccine group and 67.8 and 59.3%, respectively, in the control group. This is the first randomized prospective study demonstrating a progression-free survival benefit in the adjuvant setting; it provides a clear impetus for further research in this area.

Another vaccine strategy that has been applied in a variety of solid tumors including RCC is the heat-shock protein 96 vaccine (131). Heat-shock proteins are intracellular molecular chaperones that bind peptide fragments, which represent the antigenic complement of the cell. These proteins, with their associated peptides, can be highly purified from excised surgical specimens and utilized for vaccine preparation. Preclinical data with heat-shock protein 96 vaccines suggest potent effects both in eradicating existing tumors and in the adjuvant setting (132,133). Currently, there is an ongoing multicenter phase III study of adjuvant heat-shock protein 96 vaccines for patients with RCC at high risk of relapse following nephrectomy. This vaccination strategy is associated with virtually no toxicity and there is a high success rate in generating vaccine from even small quantities of harvested tumor.

Others have examined the role of autologous vaccination through preoperative tumor destruction in situ (using energy ablation or embolization) as a mechanism for preventing disease recurrence in patients at high-risk for relapse. In a nonrandomized study out of Poland, 474 patients who underwent nephrectomy for locally advanced RCC were examined for evidence of disease recurrence following nephrectomy (134). Of these patients, 118 underwent preoperative embolization. The reasons for embolization in each case are not elucidated in the manuscript, but interestingly the authors noted a significant improvement in survival in the patients who underwent primary tumor embolization prior to nephrectomy as compared with those who underwent nephrectomy alone. Some have suggested that the process of embolization results in the release of tumor antigens from the primary tumor, which subsequently activate the host immune system to eradicate residual disease following nephrectomy. Similarly, a recent case report demonstrated complete regression of metastatic RCC following the radiofrequency ablation of a primary tumor (135). This again suggests that local destruction of the primary tumor may result in the release of antigens into the circulation, acting as an autologous vaccine.

While there remains no generally accepted adjuvant therapy for RCC, several promising strategies have recently been published; these are in current clinical trials or on the near horizon. The recent publication of a positive adjuvant vaccination study (114) will undoubtedly stimulate more research in this area. With further understanding of the biology of renal cell carcinogenesis and progression, real strides in therapeutic development can occur that will significantly affect the unacceptably high rate of morbidity and mortality associated with this disease.

Management of Metastatic Renal Cell Carcinoma

A critically important question for clinicians is how to integrate systemic and local therapies in the management of metastatic RCC. As outlined below, current systemic therapies by themselves have a minimal impact on standard measures of outcome. Most treatments result in response rates of less than 20%, a level considered unacceptable in many other cancers. It is known that the primary RCC is reluctant to metastasize and grows into sizes in excess of 5 cm before metastatic risk is consistently elevated. Once a renal tumor becomes larger, a change in the tumor or host environment likely occurs that permits metastases to form. What these changes are is the subject of ongoing research at MDACC and other institutions. These observations raise several questions. How stable is the switch to a metastatic phenotype, and can we revert the metastatic phenotype back to a nonmetastatic phenotype through the application of appropriately targeted systemic therapy? In addition, how important is the integration of surgery into the management of these patients? Recent publications indicate that cytoreductive nephrectomy benefits patients with metastatic disease. There are nonrandomized data suggesting the same is true for metastasectomy. If surgery is an essential component of the management of metastatic RCC, what is the appropriate timing of these procedures?

The following sections summarize some of the treatments applied to RCC in the past 20 years and then move into a description of novel targeted therapies currently in clinical trials. Various local therapies being applied to the management of metastatic RCC are then described. The section concludes with some case discussions of specific strategies applied to patients at MDACC.

Immunotherapy

Clear cell RCC is considered an immunogenic tumor. This assumption is based on the observation of a 4% spontaneous regression rate in metastatic lesions (136), the presence of tumor infiltrating lymphocytes in tumor specimens, and well-documented responses to therapies presumed to act via stimulation of the immune system. The following paragraphs apply mainly to patients with clear cell RCC, as very little data exist supporting the use of immunotherapy in variant histology RCC.

The genes for IL-2 and IFN- 2a and 2b were cloned in 1983 and 1981 respectively (137,138); when large-scale production was achieved, both were used in clinical trials for patients with RCC. Response rates of 10 to 25% were described for IFN- and even higher for IL-2 (Table 29-9).

Table 29-9 Active Treatment Regimens in Metastastic Renal Cell Carcinoma

Author	(reference)	N	Drug(s)	Dose, Route, Schedule	Responses (%)
Stadler	(161)	55	Capecitabine	830 mg/m2 PO bid days 1–21	8 (15)
			Gemcitabine	1000 mg/m2 IV days 1, 8, 15, cycle every 28 days
Rini	(160)	39	Fluorouracil	150 mg/m2/day IV, continuous infusion days 1–21	7 (17)
			Gemcitabine	600 mg/m2 IV days 1, 8, 15, cycle every 28 days
Yang	(48)	156	IL-2, high dose	720,000 U/kg IV q 8 h, up to 5 days, cycle approximately every 2 weeks	33 (21)*
Yang	(48)	94	IL-2, low dose	Week 1: 250,000 U/kg SC daily for 5 days	9 (10)*
				Weeks 2–6: 125,000 U/kg SC daily for 5 days, cycle every 8 weeks
Medical Research Council	(144)	117	Interferon-2b	10 million units SC 3 days/week (M-W-F)	16 (14)
Motzer	(148)	145	Interferon-2a	9 million units SC daily	9 (6)
Eisen	(128)	18	Thalidomide	100 mg PO daily	3 (16)
Yang	(162)	39	Bevacizumab	10 mg/kg IV q14 days	4 (10)

IL-2 = interleukin-2.

• p = 0.033.

IL-2 was originally administered in the context of extracorporeal stimulation of autologous patient leukocytes, generating so-called lymphokine-activated killer (LAK) cells (139), later defined as a natural-killer cell variant, which were reinfused in conjunction with high-dose administration of IL-2. In addition, efforts to isolate and expand tumor-infiltrating lymphocytes (TIL), followed by reinfusion in conjunction with high-dose IL-2 were also undertaken. Both LAK-cell and TIL therapy were found to have equivalent efficacy to simple administration of high-dose IL-2 alone (140,141). These exceedingly complex and expensive undertakings were subsequently, in large part, abandoned. High-dose inpatient IL-2 was FDA-approved in 1992 on the basis of a compilation of 255 patients from a number of phase II trials, with results demonstrating a response rate of 15% and a fairly durable complete response rate of 5% (142). Two randomized studies assessing the effect of IFN- therapy on survival were published in 1999 (143,144). The Medical Research Council RenalCancer Collaborators reported on 335 patients randomized to IFN- or medroxyprogesterone acetate, demonstrating a median survival advantage of 2.5 months for the IFN group (144). Similar results were found by Pyrhonen et al. in a smaller study, with a 7-month survival improvement for patients treated with IFN plus vinblastine versus those treated with vinblastine alone (143). Efforts to improve on these results involved the combination of IL-2, IFN-, and 5-fluorouracil (5-FU) (145–147). Early enthusiasm for this combination therapy was dampened as more rigorous testing demonstrated response rates for the combination therapies that were no higher than those of cytokine monotherapy (145,148). To further confuse matters, a recent report once again demonstrated a progression-free and overall survival advantage for patients who received combination chemoimmunotherapy as opposed to monotherapy (149). Disease heterogeneity is probably responsible for some of the inconsistencies noted in these trials. The quest for improved immunologic response has led to the testing of a number of vaccine strategies and to the application of nonmyeloablative allogeneic transplantation to metastaticRCC. The published studies on vaccine therapy in metastaticRCC fail to demonstrate convincing efficacy (150–152), including a study employing dendritic cell therapy published in 2000, which has recently been retracted (152,153). Nonmyeloablative transplantation demonstrated promise in a National Cancer Institute study (154). An analysis of studies published by other centers shows an aggregate response rate of approximately 30%, with treatment-related mortality averaging 20% (155–159). In addition to the mortality risk, long-term consequences of graft-versus-host disease and other morbidities are associated with transplantation. A recent study by Ueno et al. from MDACC demonstrates that some form of response occurred in 7 of 14 evaluable patients with RCC; death from transplant-related causes occurred in 5 (156). The results of these transplant trials are important for the biological insights they provide, but they do not yet support its routine use.

Cytotoxic Chemotherapy

Until the year 2000, no chemotherapy was considered effective in metastatic RCC. Although vinblastine has been used in the past, more recently it figured as a de facto negative control in a study assessing the efficacy of IFN therapy (143). In June 2000, Rini et al. published a phase II study of gemcitabine and continuous infusion 5-FU, demonstrating a 17% overall response rate (see Table 29-8) (160). No data exist documenting the survival rates or the response durability with this combination. The combination of gemcitabine and capecitabine is currently being evaluated in a number of studies around the country and may find a role as a valuable salvage regimen in patients with RCC (161).

Targeted Therapy and New Agents

A major effort is now under way in RCC and other cancers to precisely target aberrant cellular function and achieve either cytostatic or cytocidal effects on the tumor cells. It is important that clinical research using these purportedly targeted agents be performed in a fashion that permits us to learn how they work at the cellular level. Only with this knowledge can the subsequent generation of clinical trials be appropriately designed and effective combination therapies developed.

Got To Kidney Cancer MD Anderson (Oncology) Part II

Disclaimer

E&OE - Errors & Omissions Excepted