Renal cell carcinoma

Current Problems in

Cancer Volume 21

Number4

July/August 1997

Renal Cell Carcinoma

Robert J. Motzer, MD

Paul Russo, MD

Genitourinary Oncology Service Division of Solid Tumor Oncology Department of Medicine Memorial Sloan-Kettering Cancer Center New York, New York, and Department of Medicine Cornell University Medical College New York, New York

Urology Service Department of Surgery Memorial Sloan-Kettering Cancer Center New York, New York, and Department of Urology Cornell University Medical College New York, New York

David M. Nanus, MD

William J. Berg, MD

Genitourinary Oncology Service Division of Solid Tumor Oncology Department of Medicine Memorial Sloan-Kettering Cancer Center New York, New York, and Department of Medicine Cornell University Medical College New York, New York

Genitourinary Oncology Service Division of Solid Tumor Oncology Department of Medicine Memorial Sloan-Kettering Cancer Center New York, New York, and Department of Medicine Cornell University Medical College New York, New York

Mosby nr~ ATimes Mirror t~ Company

Current Problems in

Cancer Renal Cell Carcinoma

Fo rewo rd

189

Abstract

191

Introduction

192

Incidence and Mortality

192

Causes and Risk Factors

193 193 193 194 194 194 194

Environmental Factors Tobacco Use Obesity Hypertension and Antihypertension Medication Other Factors Hereditary Disease

Clinical Presentation and Diagnosis Clinical Presentation Paraneoplastic Syndromes Metastasis Incidental Tumor Detection Diagnosis Pathologic Classification Evaluation of Renal Mass Determination of Extent of Disease Assessment of Tumor Stage Surgery Radical Nephrectomy Partial Nephrectomy Involvement of Vena Cava Metastatic Disease Relapse, Adjuvant Therapy, and Follow-up Curr Probl Cancer, July/August1997

196 196 196 197 198 198 198 199 2O0 201 202 202 2O4 2O5 206 207 187

SystemicTherapy for Advanced Disease Chemotherapy and Hormone Therapy Modulation of Drug Resistance Treatment With Biological Response Modifiers Interferon Interleukin-2 Combination Therapy Observation Alone

Tumor Biology Tumor Antigen Expression Molecular Genetic Abnormalities Oncogenes, Tumor-Suppressor Genes, and Growth Factors

208 208 211 213 213 214 215 217 218 218 219 219

Conclusions

220

References

221

188

Curt Probl Cancer, July/August 1997

Foreword In this issue of Current Problems in Cancer, Dr. Motzer and his colleagues, from Memorial Sloan-Kettering Cancer Center, have presented a thorough review of renal cell carcinoma. The topic is a timely one because of the interest in the biologic therapy of this tumor. The clinical characteristics of this tumor, including the various paraneoplastic syndromes, are thoroughly discussed. The authors present a careful review of the surgery for localized disease and for the resection of solitary metastases. Systemic therapy of locally advanced and metastatic tumor is reviewed also, and the results of such treatments are put into perspective. It is clear that systemic therapy as we know it today has very modest activity and is of debatable clinical benefit for most patients. The authors conclude with a discussion of recent observations of the biology and molecular genetics of renal cell carcinoma. It is clear that improvements in therapy of this resistant disease will require more knowledge of fundamental biology and participation by patients and physicians in well-designed, innovative clinical trials testing hypotheses so generated.

Stephen D. Williams, MD Editor-in-Chief

Curr Probl Cancer, July/August 1997

189

Robert J. Motzer earned his degree in medicine at the University of Michigan Medical School in 1981. He served his internship and residency in medicine at North Shore University Hospital, Manhasset, New York, and at MemorialSIoan-Kettering Cancer Center, New York. Dr. Motzer was a fellow in medical oncology/hematology and a research fellow at Memorial Sloan-Kettering Cancer Center from 1984 through 1987. Since 1987 he has held various positions at Michigan State University, East Lansing; Memorial Sloan-Kettering Cancer Center, New York; and Cornell University Medical College, New York. Dr. Motzer is board certified by the American Board of Internal Medicine and holds certification in medical oncology. He is coauthor of 72 articles, 13 book chapters, and 50 abstracts.

Paul Russo earned his medical degree at the College of Physicians and Surgeons Columbia University in 1979. His postdoctoral training includes an internship in general surgery and three residencies in general and urologic surgery at Washington University Medical Center, St. Louis, and fellowships at Memorial Sloan-Kettering Cancer Center in New York. Dr. Russo is a member of the American Medical Association and the American Urological Association and a fellow of the American College of Surgeons. He has coauthored 92 published papers and abstracts.

David M. Nanus earned his Doctorate of Medicine at the University of Health Sciences, The Chicago Medical School, in 1982. His postdoctoral training includes an internship in internal medicine at University of Wisconsin Hospitals & Clinics, Madison; a residency in internal medicine at Albert Einstein College of Medicine, Bronx; a clinical fellowship in medical oncology at Memorial Sloan-Kettering Cancer Center; and a research fellowship at the Laboratory of Mammalian Cell Transformation, Sloan-Kettering Institute for Cancer Research in New York City. Dr. Nanus is a member of the American: Association for the Advancement of Science, American Association for Cancer Research, American Society of Clinical Oncology, American Urological Association, International Society for the Study of Comparative Oncology, New York Academy of Sciences, and the Society for Basic Urological Research. He is the coauthor of 37 peer-reviewed papers, 7 reviews and monographs, and 43 abstracts.

William J. Berg earned his MD at Cornell UniversityMedical College in New York City in 1991. His clinical training includes a position as oncology fellow at Memorial Sloan-Kettering Cancer Center and a residency in internal medicine at Yale University School of Medicine. Dr. Berg has held three research fellowships at Cornell University Medical College.

190

Curr Probl Cancer, July/August 1997

Renal Cell Carcinoma ABSTRACT.--Renal cell carcinoma (RCC) is characterized by (a) lack of early warning signs, which resuits in a high proportion of patients with metastases at the time of diagnosis; (b) protean clinical manifestations; and (c) resistance to radiotherapy and chemotherapy. The estimates of new diagnoses and deaths from kidney cancer in the United States during 1996 are 30,600 and 12,000, respectively. R C C occurs nearly twice as often in m e n as in women. The age at diagnosis is generally older than 40 years; the median age is in the midsixties. The incidence of R C C has been rising steadily. Between 1974 and 1990, there was a 38 % increase in the n u m b e r of patients who had a diagnosis of RCC. This increase was accompanied by a significant improvement in 5-year survival. Both trends are likely the result of improved diagnostic capability. Newer radiographic techniques, including ultrasonography, computed tomography, and magnetic resonance imaging, are detecting kidney tumors more frequently and at a lower disease stage, when t u m o r s can be resected for cure. Surgical treatment is the only curative therapy for localized RCC, Radical nephrectomy remains the mainstay of surgical management, but techniques are being modified. These modifications include partial nephrectomy and resection of vena caval thrombi. In highly selected cases, surgical resection of locally recurrent RCC or of disease at a solitary metastatic site is associated with long-term survival. Metastatic R C C is highly resistant to the many systemic therapies that have been extensively investigated. A minority of patients achieve complete or partial response to interferon, interleukin-2, or both. Response can be dramatic but is rarely durable. Because most patients do not achieve response, these agents are not Curr Probl Cancer, July/August 1997

191

considered effective treatments for RCC, but the response in some patients indicates the need for continued research on their use. Identification of new agents with better antitumor activity against metastases remains a high priority in clinical investigation of therapy for this refractory disease.

M

enal cell carcinoma (RCC) is characterized by lack of early warning signs, which results in a high proportion of patients with metastases at the time of diagnosis and by protean clinical manifestations. 1 This cancer is also resistant to radiotherapy and chemotherapy, but study results suggest that immunomodulation with biological response moditiers may have promise for inhibiting tumor growth. 1Surgical resection of the primary tumor in patients with localized disease remains the mainstay of therapy.1Although no single agent consistently shows a response rate of 20% or higher, interleuldn-2 (IL-2) and interferon alfa (IFN (x) have demonstrated a low but reproducible response rate of 10% to 20%, with durable (>2 years) responses of 5% or less. 1

Incidence and Mortality The term kidney cancer includes all malignant tumors originating from a primary site in the kidney. RCC is by far the most common kidney cancer, comprising 85% to 90% of these malignant tumors. Tumors of the renal pelvis account for most of the remaining cases. Wilms' tumor occurs primarily in children. In 1985, kidney cancer accounted for 1.7% of all malignant disease worldwide, and 127,000 new cases were diagnosed.2 The worldwide incidence of kidney cancer was 3.1 per 100,000 for men and 2.1 per 100,000 for women? The international variation in incidence was 20-fold for men and 10-fold for women? North America and northern, eastern, and western Europe had the highest incidence, and geographic areas of low incidence were located in Africa and Asia. 2 There were an estimated 73,000 deaths from kidney cancer worldwide in 1985, and the estimated mortality rate was 1.5 per 100,000. 3Table 1 shows the estimated 5-year survival for kidney cancer in developing countries and the former United Soviet Socialist Republic (USSR) in 1985, compared with that for the black population in the SEER (Surveillance, Epidemiology, and End Results) registry in 1980 and that reported by the Finnish Cancer Registry for cases diagnosed in 1955. 3The worldwide incidence of kidney cancer is rising, and the number of deaths in the year 2000 is expected to exceed 100,000. 3 192

Curr Probl Cancer, July/August 1997

TABLE 1. Five-year survival for kidney cancer reported from Finnish Cancer Registry in

1955, black population from SEER registry in 1980, former USSR in 1985, and developing countries in 19853,* Males Data source

SEER registry, 1980 (blacks) Finnish Cancer Registry, 1955 USSR, 1985 Developing countries, 1985

Females

(%)

(%)

49.9 15.9 47.3 31.9

57.8 24.2 45.9 33.0

*SEER = Surveillance, Epidemiology, and End Results.

The estimates of new diagnoses of kidney cancer and deaths from this disease in the United States during 1996 are 30,600 and 12,000, respectively.4 States with the highest mortality rates from 1973 through 1990 included Maine, North Dakota, Vermont, and Kansas; Utah and Hawaii had the lowest mortality rates? RCC occurred nearly twice as often in men as in women, and incidence in the United States was equivalent for whites and African Americans? The age at diagnosis was generally older than 40 years; the median age was in the midsixties. 5 The lifetime risk of a male for developing RCC was 1.2%, and the risk of death was 0.51%.5 For females, the risks of diagnosis of and death from RCC were 0.76% and 0.33%, respectively? The incidence of RCC has been rising steadily. Between 1974 and 1990, there was a 38% increase in the number of patients who had a diagnosis of RCC? This increase was accompanied by a modest but significant improvement in 5-year survival, from 52% of patients who had a diagnosis of this cancer from 1974 to 1976 to 58% of patients who had the diagnosis from 1983 to 1989. 5 Both of these trends are likely the result of improved diagnostic capability. Newer radiographic techniques, including ultrasonography, computed tomography, and magnetic resonance imaging, are detecting kidney tumors more frequently and at a lower stage of disease, when tumors can be resected for cure.

Causes and Risk Factors Environmental Factors The cause of RCC is not well understood. Study findings suggest that several environmental factors may contribute to development of RCC.An international case-control study has provided insight into the causative role of environmental factors. 6-1~ Tobacco Use. Tobacco use was found in numerous studies to increase the risk of RCC. 11,12Cigarette smoking has been estimated to contribute to as many as one third of all cases of RCC in the United States. 11,12Findings of the recent international case-control study were as follows: 13 Curr Probl Cancer, July/August 1997

193

9 No association between cigar and pipe smoking and the development of RCC was detected, but the risk of this cancer was 40% higher for current cigarette smokers than for nonsmokers. 9 The relative risk of developing RCC increased according to the number of cigarettes smoked per day and the number of years a person had smoked, 9 Cessation of smoking affected risk: persons who had not smoked for more than 15 years had 75% to 80% of the risk of current smokers. 9 There was an inverse relationship between risk for RCC and the age at which a person started to smoke cigarettes: individuals who started smoking at an older age had about two thirds the risk of those who started smoking at a younger age. Obesity. Studies have shown that obesity is also a risk factor, particularly in women. Study findings were as follows: 9 There was a linear relationship between increasing body weight and increased risk of RCC. 8,14 9 High body mass index was a risk factor for women and to a lesser extent for men? 9 Risk of RCC was three times higher for women with a weight in the top 5% than for those with a weight in the lowest 25%. 8 9 Physical exercise and height were unrelated to risk. 7 Hypertension and Antihypertension Medication. Some investigators have proposed that hypertension and use of antihypertension medications increase the risk for developing RCC. ~1,~2Results of the international study indicate that hypertension and use of diuretics and other antihypertension medications had some effect on the risk for RCC. 8Other scientists reported that analgesics such as acetaminophen, aspirin, and phenacetin did not increase the risk for RCC, even though other findings suggest that these agents increase the risk for transitional cell carcinoma originating in the renal pelvis. 6 Other Factors. Additional factors associated with the development of RCC include use of unopposed estrogens] occupational exposure to petroleum products, heavy metals, and asbestos; 9 and a high-protein diet, 8 Investigators have also reported increased risk of developing RCC for patients with (a) acquired cystic kidney disease associated with chronic renal insufficiency, 15 (b) tuberous sclerosis, 16 or (c) polycystic kidney disease."

Hereditary Disease In most cases of RCC, there is no recognized hereditary pattern. Having a first-degree family member with a diagnosis of RCC increases risk of this cancer fourfold, but this risk factor is relevant for only a small proportion of patients, ~7 Several rare familial forms of RCC are characterized by au194

Curr Probl Cancer, July/August 1997

tosomal dominant inheritance, a young age at diagnosis (third to fifth decades), and bilateral and multifocal tumors. 18,19Although these familial forms are uncommon, studies of the cytogenetics and molecular genetics of these tumors have contributed significantly to our understanding of tumor pathogenesis (see section on "Tumor Biology"). von Hippel-Lindau disease is a familial syndrome that predisposes patients to development of a variety of other neoplasms. These include retinal hemangiomas; cerebellar, brain stem, and spinal cord hemangioblastomas; renal and pancreatic cysts and malignant tumors; pheochromocytomas; and epididymal cysts. 18von Hippel-Lindau disease is diagnosed if a patient manifests two or more of these conditions or one condition in the context of a positive family history. von Hippel-Lindau disease is clinically heterogeneous. 18Hemangiomas of the retina and hemangioblastomas of the central nervous system are the most common tumors. They are present in 60% of patients with this disease. Pheochromocytoma is diagnosed in 15% of the patients with von Hippel-Lindau disease. 18Pancreatic cysts are present in nearly one half of the patients, and pancreatic carcinoma occurs in fewer than 5%. 18The median age at diagnosis of von Hippel-Lindau disease is 20 to 24 years, and the median age at death is 40 years. 18 Renal cysts are frequently multiple and bilateral in patients with von HippelLindau disease and often contain RCC. Nearly 40% of patients with this disease develop RCC, and the median age at diagnosis of RCC is 39 years, m RCC in these patients is of the clear cell type. Because these tumors are frequently bilateral and multifocal, a nephron-sparing operation (partial nephrectomy) is often advocated to preserve as much renal function as possible. Previous studies showed complications of the central nervous system to be the major cause of death in patients with von Hippel-Lindau disease; RCC was the cause of death in nearly 20% of the patients.18 With better treatment of aspects of the disease relating to the central nervous system, RCC has become the most common cause of death in patients with von Hippel-Lindau disease. Several kindreds with familial clear cell RCC have been identified in the absence of von Hippel-Lindau disease. These kindreds have a constitutional, balanced translocation between the short arm of chromosome 3 (3p) and chromosome 6 or 8.19-21A third hereditary form, which has been described in a three-generation family, is characterized b y multiple tumors with tubulopapillary histologic findings not linked to chromosomal abnormalities on 3p. 22Additional classes of inherited RCC may exist. One team of investigators described nine families having two or more members with a diagnosis of RCC. None of these patients had von Hippel-Lindau disease or translocations of chromosome a r m 3p. 23 Curr Probl Cancer, July/August 1997

195

TABLE 2. Pathologic classification of renal cell carcinoma

Cell type

Growth pattern

Clear cell

Acinar or sarcomatoid Tubulopapillary or sarcomatoid Solid, tubular, or sarcomatoid Tumor nests Tubulopapillary or sarcomatoid

Chromophilic Chromophobic Oncocytic Collecting duct

% of cases 75

Prognosis Unfavorable

~14

Favorable

5

Favorable

5 1

Favorable Unfavorable

Clinical Presentation and Diagnosis Clinical Presentation RCC is characterized by diversity in clinical presentation, sites of distant metastases, and clinical course. Consequently, it has the reputation of being an elusive disease and is considered in the differential diagnosis of many medical conditions. 24RCC has been referred to as the "internist's tumor" because systemic rather than urologic manifestations often lead to diagnosis.24 Small, localized tumors rarely produce symptoms. Therefore the diagnosis is often delayed until after the disease is either locally advanced or metastatic. Initial signs and symptoms of RCC result from a direct (local) effect of the primary tumor, distant metastases, or paraneoplastic syndromes. The most common presentations are hematuria, abdominal pain, or a palpable flank or abdominal mass (Table 2). 25The finding of hematuria warrants careful evaluation of the entire urinary tract, with consideration of RCC in the differential diagnosis. The hematuria may be gross or microscopic and is painless unless partial clotting causes obstructive symptoms. The classic triad of hematuria, abdominal pain, and a palpable mass occurs in 5% ofpatientsy and 1% to 3% of tumors are bilateral. At initial examination, 2% of patients with RCC have a varicocele.25This condition is usually on the left side because of obstruction of the testicular vein, which originates from the left renal vein.26However, varicoceles on the fight side have been reported. 26They may signify tumor extension into the inferior vena cava. Paraneoplastic Syndromes. The broad range of paraneoplastic syndromes have been reported with RCC. Nonspecific signs and symptoms include fever, anorexia, weight loss, and anemia, which are present in as many as 30% of patients at diagnosis. More specific paraneoplastic syndromes are present in fewer than 5% of patients; they include erythrocytosis, hypercalcemia, hepatic dysfunction (Stauffer's syndrome), and amyloidosis.24'25 The paraneoplastic 196

Curr Probl Cancer, July/August 1997

syndromes may be present with either localized or metastatic tumors. Thus their presence does not preclude the resection of a localized tumor. Causes of paraneoplastic syndromes associated with localized or metastatic tumors include (a) amaor production of hormone-like substances (erythropoietin27 or parathyroid hormone-related protein28) and (b) formation of an antigen-antibody complex (amyloidosis).24 IL-6, a mnltifunctional cytokine that has been reported to induce a variety of acute-phase proteins, has been proposed as a cause of fever associated with RCC. 29 Erythrocytosis has been reported in 1% to 5% of patients with RCC, but the exact incidence of elevated serum concentrations of erythropoietin and the clinical significance are unknown.24 Other known causes of elevated plasma concentrations of erythropoietin include hepatoma, renal cyst, hemangioblastoma, and conditions of hypoxia. 24Parathyroid hormone-related protein is the predominant cause of humoral hypercalcemia.24The syndrome of hypercalcemia produces symptoms similar to those of primary hyperparathyroidism. It is most common with RCC and squamous cell carcinoma originating at sites that include the lung, head and neck, or penis. Stauffer's syndrome presents a clinical picture of reversible hepatic dysfunction associated with RCC.The syndromeoccurs in the absence of hepatic metastases; it is characterized by albumin:globulin ratios less than 1, elevated blood levels of alkaline phosphatase, and elevated serum levels of haptoglobin and prothrombin. 3~ Hypochromic anemia, thrombocytosis, and an elevated erythrocyte sedimentation rate are also common features of this syndrome. Production of abnormal fibrinogen and development of dysfibrinogenemia have been reported. 31 These findings suggest that Stauffer's syndrome may encompass a spectrum of disorders. Microscopic examination of hepatic parenchyma from these patients has revealed only nonspecific hepatitis. 31 In the absence of distant metastases, the hepatic dysfunction normalizes with nephrectomy. The cause of Stauffer's syndrome is unknown, but it may be related to tumor production of a colony-stimulating factor.32 Metastasis. RCC can spread by direct extension or by metastasis via lymphatic or venous routes. Surrounding structures that may become involved include the liver, colon, adrenal gland, and pancreas. Regional, left supraclavicular, and posterior mediastinal lymph nodes are commonly involved. Involvement of the left supraclavicular and posterior mediasfinal lymph nodes reflects the cephalad drainage of the lymphatic systems. At initial examination, 25 % to 30% of patients with RCC have overt metastases. The most common sites are lung parenchyma (50% to 60%), bone (30% to 40%), liver (30% to 40%), and brain (5%). 25One feature of RCC, however, is the presence of metastases at unusual sites. Metastases from RCC can involve essentially any organ site, including the pancreas, adrenal gland, thyroid, skeletal muscle and skin, and underlying soft tissue. Curr Probl Cancer, July/August 1997

197

IncidentalTumor Detection. Use of computed tomography and ultrasonography for indications of other conditions has led to increased detection of RCC as an incidental finding) 3 This is not surprising, because two thirds of RCC detected in an autopsy series conducted from 1958 through 1969 had not been recognized clinically.34 Since use of these imaging techniques has become common, as many as 25% to 40% of patients have a diagnosis of RCC after incidental detection of a renal m a s s , 33,35,36 compared with 10% or fewer before these imaging techniques were in general use. 25'35 Most important, tumors found incidentally are typically smaller than those that produce symptoms, and "cure" is more likely to result from resection. 35-37 Renal tumors discovered as an incidental finding are confined within the renal capsule in 75% of patients and are associated with 5-year survival of 75%. 33,38,39 This detection of RCC at earlier stages has resulted in a decrease in the incidence of metastatic disease at diagnosis to less than 20% and improved 5-year survival. 4~The incidental discovery of tumors smaller than those that are symptomatic has also encouraged the more liberal use of partial nephrectomy.

Diagnosis Pathologic Classification. The majority of malignant tumors involving the kidney are RCC. However, the differential diagnosis requires consideration of squamous and transitional cell carcinoma of the renal pelvis, Wilms' tumor, sarcoma, lymphoma, and metastatic carcinoma. There are no histopathologic features distinguishing a renal adenoma from RCC. Renal tumors less than 3.0 cm in diameter at excision were previously referred to as adenomas. RCC was historically classified by cell type (clear cell, granular, spindle cell, and oncocytic carcinoma) and by growth pattern (acinar, tubulopapillary, and sarcomatoid). Because this classification provided little insight into the clinical behavior and histogenesis ofRCC, a new histopathologic classification was proposed (Table 3).41 This classification was based on morphology, histochemical staining, and data from electron microscopy. 41 It was subsequently validated by cytogenetic and molecular studies. 42-46 By the new classification, RCC has five cell types: clear cell, chromophilic, chromophobic, oncocytic, and collecting duct (Bellini's duct) carcinomas.43,47 Clear cell carcinoma comprises 75% of RCC. The tumor cells contain clear cytoplasm, consisting of lipid and glycogen. Clear cell carcinomas vary according to the degree to which the cytoplasm appears "clear" and according to the growth pattern, which is acinar, tubulopapillary, or sarcomatoid. Increasing nuclear g r a d e 43,48 or sarcomatoid pattern43,49correlates with a poorer prognosis. Clear cell carcinoma is characterized by deletion of one or both copies of chromosome arm 3p (3p-). 5~ Chromophilic carcinomas were in most cases classified previously as 198

Curr Probl Cancer, July/August 1997

TABLE 3. Symptoms at initial examination in 2366 patients with renal cell carcinoma 2s

Symptom Hematuria Abdominal pain Palpable mass Weight loss Fever Varicocele

% of patients 56 38 36 27 11 2

tubulopapillary. These tumors comprise approxirnately 14% of RCC and are often multifocal and bilateral. There may be one or two dominant masses and small tumors throughout the kidney. Chromophilic carcinomas often present as small tumors less than 3 cm in diameter. 51These tumors do not contain the 3p deletion, 52 but they exhibit monosomy Y, trisomy 7, and trisomy 17.53Because chromophilic tumors are usually low stage at initial examination, they are associated with a more favorable prognosis than clear cell carcinoma. 5~'54Clear cell and chromophilic carcinomas originate from the proximal tubule cell of the nephron, 42,55so they express cell-surface proteins normally found on proximal tubule cells. Chromophobic carcinoma comprises approximately 5% of RCC. It is characterized by a hypodiploid chromosome number 45,56and multiple chromosomal losses but does not exhibit the 3p deletion. 46Previously, the tumors were most often categorized as granular RCC. Patients with these tumors generally have a favorable prognosis. 45,57 Renal oncocytomas are uncommon tumors that are only rarely associated with metastases. 58 These tumors are characterized by a central, stellate scar. Cells are arranged in nests, cords, or tubules and do not exhibit a papillary growth pattern. Renal oncocytomas do not exhibit the 3p deletion or trisomy 7 or 17. Both chromophobic and oncocytic tumors originate from the intercalated cells of the cortical collecting duct. Collecting duct carcinomas originate in the medullary collecting duct and are rare. 43These tumors can present at any age but tend to occur in young patients. They are often metastatic at diagnosis, clinically aggressive, and associated with a poor prognosis. 59 Evaluation of Renal Mass. Intravenous urography remains a useful initial screening test for the evaluation of hematuria. It provides visualization of the collecting system to exclude obstruction caused by urolithiasis or extrinsic compression and may detect RCC or urothelial tumors. However, diagnostic imaging of the renal mass has evolved from intravenous urography and angiography to newer techniques that include ultrasonography, computed tomography, and magnetic resonance imaging. Curr Probl Cancer, July/August 1997

199

Ultrasonography and dynamic contrast-enhanced computed tomography can detect and characterize renal masses with greater accuracy than intravenous urography.6~ For tumors 3 cm in diameter or less, the sensitivities of urography, ultrasonography, and computed tomography have been quantified as 67%, 79%, and 94%, respectively. 61 Magnetic resonance imaging is also useful in detecting a renal mass when intravenous contrast medium cannot be administered because of renal insufficiency or allergy to contrast dye.62 Administration of intravenous contrast medium is essential for optimal use of computed tomography. On noncontrast computed tomography, RCC can appear hypodense, isodense, or hyperdense compared with normal renal parenchyma. Administration of intravenous contrast medium enhances visualization of RCC on computed tomography scan. Arteriography is rarely used in diagnosis of RCC. The utility of arteriography as a diagnostic tool is primarily restricted to preoperative mapping of vasculature in cases when a nephronsparing operation is planned and the tumor is centrally located. The diagnoses of simple, uncomplicated cysts, pseudotumors, and angiomyolipomas (renal hamartomas) can often be rendered on the basis of ultrasonogram or computed tomography scan, because each type of lesion has a characteristic appearance on these radiographic visualizations. 63 For example, a renal tumor that appears to contain fat on a noncontrast computed tomography scan is, with rare exception, a benign angiomyolipoma. A central stellate scar within a homogeneous, well-circumscribed, solid mass often indicates an oncocytoma on computed tomography scan. 58These findings are not specific to oncocytoma, however, and the patient should have surgical resection.58 Cystic lesions can be categorized by their appearance on computed tomography scan and can be assigned confidence levels for probability of containing malignant elements. 64 Category I lesions are uncomplicated, simple, benign cysts. Category II lesions are benign in appearance but have radiographic findings that are not characteristic of a benign cyst. Such lesions include high-density and septated cysts. Class HI lesions have some findings seen in malignant tumors, including evidence of irregular thickening of the wall or extensive calcification. Class III lesions cannot be confidently distinguished from malignant tumors. Class IV lesions are malignant tumors with cystic components. Diagnosis of any potentially malignant lesion requires examination of a biopsy specimen. When a localized tumor has characteristics suggesting RCC, tissue for diagnosis is obtained at nephrectomy; needle biopsy is not performed routinely. If the disease is not judged to be curable by excision, needle biopsy of the renal mass or of metastases is performed. Determination of Extent of Disease. RCC is associated with a spectrum of systemic manifestations, including anemia, polycythemia, hypercalcemia, weight loss, acute varicocele, and fever. Physical examination is performed 200

Curr Probl Cancer, July/August 1997

with attention to the presence of supraclavicular adenopathy, palpable abdominal mass, edema of the lower extremities, varicocele, and subcutaneous nodules. Laboratory evaluation includes a complete blood cell count, determination of serum calcium and serum creatinine levels, and liver function studies. Evaluation of the extent of disease includes a chest radiograph and abdominal computed tomography scan focusing on the lung and regional lymph nodes, the most common sites of metastatic disease. The accuracy of computed tomography in preoperative staging is approximately 90%. 65The recognized limitations are in (a) evaluation of minimally enlarged regional lymph nodes and (b) determination of the cephalad extent of vena caval tumor extension. Regional lymph nodes larger than 2 cm are nearly always involved with metastases. 65,66In contrast, enlargement of lymph nodes to 1 to 2 cm in diameter may be caused by either metastasis or reactive hyperplasia.65,~This degree of enlargement does not preclude surgical intervention and resection of an otherwise localized renal mass. In cases of suspected tumor involvement of the inferior vena cava, magnetic resonance imaging with intravenous gadolinium is strongly recommended, because this technique is superior to computed tomography for evaluating the degree of cephalad extension of tumor.67 A computed tomography scan of the chest is performed when the chest radiograph shows a nodule, the primary tumor is large, or the tumor involves the inferior vena cava.68A bone scan is not performed routinely unless the serum concentration of alkaline phosphatase is elevated or the patient complains of bone pain. 69 Computed tomography of the brain is not performed routinely unless the patient complains of headache or manifests a neurologic deficit. Assessment of Tumor Stage. The two frequently used systems for assessing tumor stage in kidney cancer are the American Joint Committee on Cancer (AJCC) staging system 7~(Table 4) and the modified Robson staging system 71 (Table 5). In the Robson system, there is limited description of regional lymph node involvement and involvement of the renal vein or inferior vena cava. The AJCC staging system is more explicit in describing the extent of local and regional disease, but it is more complicated. Cancer registry data for 3307 patients who had diagnosis of RCC in Illinois from 1975 through 1985 showed the distribution of tumor stage according to the Robson system to be as follows: stage I, 40%; stage II, 10%; stage Ill, 20%; and stage IV, 30%. 72 The most important prognostic determinant of survival is the tumor stage at initial examination. According to Robson staging, 5-year survival is 65% to 85% with stage I disease, 45% to 80% with stage II, 15% to 35% with stage IH, and 0% to 10% with stage I~. 73Other adverse prognostic indicators, which are usually associated with locally advanced or metastatic tumors, include high pathologic grade, sarcomatoid histologic findings, large tumor size (greater than 10 cm in diameter), weight loss, hypercalcemia, and elevated sedimentation rate. 74 Curr Probl Cancer, July/August 1997

201

The outcome for Robson stage III cancers differs according to whether there is vascular invasion or nodal involvement. After complete resection of tumors involving the renal vein, 5-year survival approaches 50% in the absence of regional nodal or distant metastatic disease. In contrast, fewer than one third of patients with regional nodal metastases survive 5 years. 72 The data suggest that solitary tumor invasion of the renal vein, independent of tumor size and regional nodal status, may not predict a poor outcome. In support of this hypothesis, two recent multivariate analyses have failed to show that tumor invasion of the renal vein is an independent risk factor. 75,76After complete resection of disease, recurrence in the inferior vena cava, in the absence of regional node involvement or distant metastasis, is also associated with 5- and 10year survival in the range of 50%. 77 Several features predict prognosis in patients with involvement of the inferior vena cava. Survival varies according to whether the tumor, with its thrombus, floats free within the caval lumen or directly invades the caval wall. Freefloating extension of tumor into the vena cava is associated with a 70% 5-year survival (median, 10 years), whereas direct tumor invasion of the caval wall is associated with a 25% 5-year survival (median, 1.2 years). 78 In patients with direct tumor invasion into the vena cava, 5-year survival increased to 57% if the involved segment of inferior vena cava was resected completely.78Prognosis does not appear to depend on the level of involvement of the inferior vena cava up to and including the right atrium. It is dependent, however, on the extent of disease elsewhere, including extrafascial spread, lymph node involvement, and distant metastases. With any of these three characteristics, 5-year survival is less than 20%. 78,79

Surgery Surgical treatment is the only curative therapy for localized RCC. The radical nephrectomy remains the mainstay of surgical management, but techniques are being modified. These modifications include partial nephrectomy and resection of vena caval thrombi in selected patients. In highly selected casesl surgical treatment of locally recurrent RCC or metastatic disease at a solitary site is associated with long-term survival.

Radical Nephrectomy In 1969, Robson et al. reported the results of radical nephrectomy, defined as resection of the kidney, perirenal fat, regional lymph nodes, and ipsilateral adrenai gland.7/The investigators reported overall survival of 65 % for patients with stage I and 1/tumors. More recent series have shown that the 5-year survival for patients with stage I tumors exceeds 80%. 73During the last 25 years, modifications to this approach have gained acceptance. These modifications include 202

Curr Probl Cancer, July/August 1997

TABLE 4. Classification of kidney cancer by tumor stage according to American Joint Committee on Cancer staging system 7~

Primary tumor (7") TX TO T1 T2 T3

Primary tumor cannot be assessed No evidence of primary tumor Tumor 2.5 cm or less in greatest dimension limited to the kidney Tumor more than 2.5 cm in greatest dimension limited to the kidney Tumor extends into major veins, invades adrenal gland, or perinephric tissues but not beyond Gerota's fascia T3a Tumor invades the adrenal gland or perinephric tissues but not beyond Gerota's fascia T3b Tumor grossly extends into renal vein or vena cava below the diaphragm T3c Tumor grossly invades vena cava above the diaphragm T4 Tumor invades beyond Gerota's fascia

Regional lymph nodes (N) NX NO N1 N2

Regional lymph nodes cannot be assessed No regional lymph node metastasis Metastasis in a single lymph node, 2 cm or less in greatest dimension Metastasis in a single lymph node, more than 2 cm but not more than 5 cm in greatest dimension, or multiple lymph nodes, none more than 5 cm in greatest dimension N3 Metastasis in a lymph node, more than 5 cm in greatest dimension

Distant metastasis (M) MX Presence of distant metastasis cannot be assessed MO No distant metastasis M1 Distant metastasis

Stage grouping Stage I Stage II Stage III

Stage IV

T1 T2 T1 T2 T3a T3a T3b T3b T3c T3c T4 Any T Any T Any T

NO NO N1 N1 NO N1 NO N1 NO N1 Any N N2 N3 Any N

MO MO MO MO MO MO MO MO MO MO MO MO MO M1

9 restriction of adrenalectomy to large tumors involving the entire kidney or upper pole; 9 assessment of the need for and extent of regional lymphadenectomy, particularly in cases of incidental detection of RCC; and 9 use of partial nephrectomy for small clinically localized tumors, despite the presence of a normal contralateral kidney. Controversy exists concerning the components of the radical nephrectomy, particularly with regard to the extent of regional lymph node dissecCurr Probl Cancer, July/August 1997

203

TABLE 5. Classification of kidney cancer by tumor stage according to Robson staging system 71

Extent of disease Tumor limited to kidney Tumor extending through renal capsule and confined to Gerota's fascia Tumor invading renal vein or vena cava Tumor involving lymph nodes Combination of two preceding stages Invasion of surrounding organs or distant metastases

Stage

IliA IIIB IIIC IV

tion. In 10% to 20% of patients, nodal involvement is detected at radical nephrectomy and there is no clinical evidence of distant metastases. 8~ Nearly all of these patients subsequently have relapse with distant metastases, despite lymphadenectomy.8] A complete regional lymphadenectomy allows more accurate staging, but the therapeutic value is questionable (a) because of the high probability of distant metastases when nodal involvement is found at nephrectomy and (b) because of the lack of effective adjuvant systemic therapy. Although a patient may occasionally be cured by resection of lymph node metastasis, the benefit of lymphadenectomy in most patients is limited to providing prognostic information. Ipsilateral adrenal involvement with RCC occurs in approximately 4% of patients who have nephrectomy. 84-86 In most instances this involvement is associated with direct extension of a lesion of the upper pole of the kidney, regional lymph node involvement, or distant metastases. 84-86Adjunctive adrenalectomy contributes to cure in fewer than 0.5% of patients. 84Adrenal involvement often can be identified preoperatively on computed tomography scan. 86Therefore adrenalectomy can be reserved for patients with adjacent lesions of the upper pole, findings on computed tomography scans that suggest tumor involvement, or both.

Partial Nephrectomy Accepted indications for partial nephrectomy for RCC include situations in which standard radical nephrectomy would render the patient functionally anephric, requiring dialysis. An absolute indication for partial nephrectomy is the presence of (a) bilateral tumors or (b) cancer of a solitary anatomic or functional kidney. Relative indications include a contralateral kidney threatened by disease such as hypertension or diabetes mellitus. A functioning renal remnant of at least 20% is needed to maintain adequate renal function. Study results show a range of 4% to 10% for local 204

Curr Probl Cancer, July/August 1997

disease recurrence in patients who had partial nephrectomy because there was either an absolute or a relative indication for the procedure. 87-9zMost of the patients in these series had large or multifocal tumors or both. The overall survival is similar to that for patients with comparable tumor stage who have radical nephrectomy. 93-95 The widespread use of computed tomography and ultrasonography in screening for evaluation of kidney symptoms as well as other conditions has led to increased detection of small, low-stage tumors. The increased n u m b e r of patients with small tumors who had a diagnosis of RCC and the favorable results of a nephron-sparing operation in patients with absolute and relative indications for the procedure have led to its use for an elective indication, that is, small localized tumors and a normal contralateral kidney. The major concern about this approach has been the risk of postoperative local tumor recurrence, either from unrecognized multicentricity of RCC or recurrence at the site of the resected lesion. 96For selected patients with small, single tumors, the local recurrence rate is 2% or less. The resuits of seven series comprising 280 patients who had partial nephrectomy for "elective indications" s h o w e d the local r e c u r r e n c e rate to be 1.4%. 58'87-91'97 This rate is similar to the risk of developing a tumor in the contralateral kidney after radical nephrectomy. 98 These data support the use of a nephron-sparing operation as an acceptable therapeutic approach in selected patients with small polar lesions (4 cm or less) and a normal contralateral kidney.

Involvement of Vena Cava RCC frequently invades the renal vein and can extend to the inferior vena cava, occasionally reaching the right atrium. In the absence of detectable metastases, about one half of patients whose tumors extend into the vena cava can achieve prolonged survival after resection. 26,99The prognosis for patients with tumors involving the vena cava depends on the stage of the primary tumor and whether the tumor, with its thrombus, is floating free or directly invades the caval wall. The presence of regional lymph node metastases or distant metastases is associated with a dismal prognosis, which is not altered by radical resection. Resection of vena caval RCC often requires the assistance of cardiovascular surgeons and may require the techniques of venovenous bypass or cardiopulmonary bypass with or without circulatory arrest. This approach allows excision of the tumor and associated thrombus with close visual examination of the atrium and caval lumen and decreases the risk of sudden, life-threatening hemorrhage. Excision of tumors with caval involvement is a complex operation and should be undertaken in a specialized Curr Probl Cancer, July/August 1997

205

center; even in that setting, mortality due to surgical treatment is 5% to 10%.100,101

Metastatic Disease Nephrectomy may be indicated for patients with metastatic disease in selected instances when the intent is to provide improved quality of life, such as alleviation of local symptoms. Also, resection of a solitary metastasis is indicated in selected patients; study results show 5-year survival of approximately 3 0 % . 73 Disease outcome for patients who have initial resection of the primary tumor and of a solitary metastasis at the time of initial diagnosis is generally inferior to the outcome for those who have resection of a solitary metastasis at relapse after nephrectomy.73 Nephrectomy is not indicated for the purpose of inducing spontaneous tumor regression. This conclusion is based on 1~ 9 the low incidence of spontaneous tumor regression (less than 1%); 9 the lack of evidence that removal of the primary tumor causes spontaneous regression of metastases; and 9 the morbidity and mortality associated with nephrectomy in patients with metastatic disease. There is controversy about tumor debulking before treatment with systemic immunotherapy for RCC in patients with metastases, and this approach is not considered to be general practice. Mortality due to surgical treatment is 2% to 11% for patients with metastases who have nephrectomy. In addition, nephrectomy may preclude the administration of systemic therapy. 1~176 The results of one series showed that 40% of patients who had nephrectomy did not receive immunotherapy, because of complications from the nephrectomy or clinical deterioration from progressive disease. To fully assess response and remove residual tumor, it may be preferable to perform nephrectomy after immunotherapy in highly selected patients who have achieved a major response. (See definition of major response in section on "Systemic Therapy forAdvanced Disease") The relative merits of initial versus delayed adjuvant nephrectomy in patients who receive immunotherapy need to be further delineated. Local disease recurrence is more common in patients with large, locally advanced tumors associated with regional nodal metastases. Local recurrence is not usually an isolated clinical event; it is frequently associated with distant metastases. Before the advent of computed tomography, local recurrence was detected in 5% of all patients who had radical nephrectomy. l~ With the advent of computed tomography, early detection of local recurrence has been observed in as many as 30% of patients who had radical nephrectomy and in 5% of those who had partial nephrectomy. 1~176 206

Curr Probl Cancer, July/August 1997

After a complete evaluation of the extent of disease to exclude distant metastases and in the absence of effective systemic treatment, resection of local recurrence remains the treatment of choice. This surgical procedure can be formidable because it may require en bloc resection of adjacent muscle, bowel, spleen, pancreas, or liver. A small subset of patients with solitary distant metastasis may benefit from resection of the metastatic lesion. Several series show 5-year survival of 15% to 50% after resection of solitary distant metastasis. 1~ RCC is characterized by a prolonged and somewhat unpredictable natural history, so reports of survival after resection of metastatic RCC may be attributable to the natural history of this disease. H4'1~5The impact of resection of metastasis on survival could be further assessed in a randomized trial.

Relapse, Adjuvant Therapy, and Follow-up After radical nephrectomy, 20% to 30% of patients with localized tumors have relapse. H6,H7Lung metastasis is the most common distant recurrence, occurring in 50% to 60% of patients.H6'~ 17The median time to relapse after nephrectomy is 15 to 18 months; 85% of relapses occur within 3 years. H6,117 The longer the disease-free interval between diagnosis and recognition of metastatic disease, the longer is the survival. There is no established role for adjuvant treatment after nephrectomy in patients who have complete resection of tumor. Radiation therapy after nephrectomy is not beneficial, even in patients with nodal involvement or incomplete resection of tumor. 1~8No systemic therapy has been shown to reduce the likelihood of relapse. In a recent randomized trial, treatment with IFN ~ was compared with observation alone in patients who had RCC that was assigned to the following TNM stage categories of theAJCC staging system after nephrectomy: T3, T4a, N1, N2, or N3.119 No delay in time to relapse or improvement in survival was associated with adjuvant IFN ~. Standard care remains observation after nephrectomy. The follow-up program for patients with RCC at the Memorial SloanKettering Cancer Center is based on the pattern of metastatic failure and the efficiency of tests required to diagnose relapse in patients with localized RCC who had radical nephrectomy.116 Patients with tumors that are Robson pathologic stage I, II, or IIIA are evaluated by history and physical examination every 6 months for the first year after radical nephrectomy and yearly thereafter. A chest radiograph, complete blood cell count, and hepatic function tests are obtained at each visit. Patients with no symptoms who had complete resection of node-positive disease and those whose nephrectomy specimens had positive margins on histologic review have evaluation of disease by history and physical examination, chest Curr Probl Cancer, July/August 1997

207

radiograph, complete blood cell count, and hepatic function tests. This evaluation is performed every 3 months for the first year after surgical treatment, every 6 months for the second year, and yearly thereafter. For both groups, a computed tomography scan of the abdomen is obtained at 3 to 6 months and at 1 year after surgical treatment and yearly thereafter. We avoid routine use of bone scan or brain imaging techniques; these procedures are used when they are indicated for specific symptoms or signs in the patient. Excessive use of computed tomography of the abdomen is also avoided. Satellite tumors may be clinically undetected at the time of partial nephrectomy. The incidence of satellite tumors increases with the size of the index tumor. This incidence has been assessed as approximately 10% in patients with RCC less than 5 cm in diameter. 12~The success of the search for satellite tumors depends largely on the completeness of the pathologic prosection. 121Intraoperative ultrasonography performed at the time of partial nephrectomy is useful to the surgeon in locating satellite tumor nodules and deep nonpalpable parenchymal lesions. 122'123The increased use of partial nephrectomy requires follow-up on the kidney receiving surgical treatment and more frequent postoperative monitoring by ultrasonography, computed tomography, or both at 6-month intervals for the first 3 years after surgical treatment.

SystemicTherapy for Advanced Disease Many different types of systemic therapy have been studied for potential antitumor effect against metastatic RCC. The clinical method for evaluating antitumor effect is to determine the proportion of patients who achieve major response. Major response is defined as either complete disappearance of all evidence of tumor (complete response) or reduction of more than 50% in tumor burden (partial response). Advanced RCC has high resistance to all treatment modalities that have been studied, including cytotoxic chemotherapy, hormone therapy, and treatment with biological response modifiers. No single agent consistently shows a response rate of 20% or higher, but IL-2 and IFN c~ have demonstrated a low but reproducible response rate of 10% to 20%, with durable responses of 5% o r lessf124

Chemotherapy and Hormone Therapy Investigative efforts with chemotherapeutic agents have been extensive. A review of the published literature showed that from 1975 through 1994, 80 single agents were studied in 155 trials. 125Overall, 143 partial or complete responses (4%) were achieved in 3951 patients assessable for response. 208

Curr Probl Cancer, July/August 1997

No agent was shown to achieve major response (complete or partial) in more than 20% of the patients in studies of 14 or more patients. In addition to the trials of single agents, many combinations of chemotherapy agents have been studied. Study results have not shown the antitumor activity of combined agents to be superior to that of the single agents, but the toxic effects of the combined agents were generally increased over the effect observed with single agents. Two agents reported to have minimal antitumor activity against RCC are vinblastine and floxuridine (FUDR) (Table 6). Results of early studies suggested that vinblastine had activity as a single agent; a 26% response rate was reported in a study of 135 patients. 126This finding served as the basis for inclusion of vinblastine in trials as a part of combined therapy with IFN and more recently with agents that modulate multidrug resistance. The results of more recent trials with vinblastine showed only nine responses in 135 patients (6%) who were assessable for response. A 20% response rate was reported with F U D R administered by continuous intravenous infusion according to a circadian schedule in which the dose varies according to the time of day. 127 Investigators in seven subsequent trials of FUDR administered by continuous infusion according to a circadian schedule reported response rates of 0% to 14%. 128-134Enthusiasm generated by the first trial 127prompted a randomized multicenter trial of FUDR administered by constant continuous infusion versus a circadian 14-day schedule. The preliminary report of this trial indicated that the response rate for the 82 patients who were assessable for response was 9% (95% confidence interval, 4% to 17%). 135 The rationale for the study of hormone therapy for RCC was provided by results obtained in animal models in the 1940s and by the low concentration of progesterone receptors found in human RCC. 136Renal cancers induced in the Syrian hamster model showed hormone dependence and responsiveness. 136 Bloom initially reported a 21% response rate for medroxyprogesterone acetate in treatment of RCC. 136,137 In the four trials published since 1980, the response rate declined to 5% (Table 7). 138-141Likewise, other hormonal agents have been extensively studied. Androgens (testosterone), 142,143the direct androgen antagonist fiutamide, 142 and the antiestrogens tamoxifen, 144,145nafoxidine, 146,147and toremifene 148were studied in multiple trials and found to be relatively inactive. The addition of hormone therapy to chemotherapy does not increase efficacy. This finding was evident from the results of single-arm phase n trials 149-152 and a randomized trial, which showed no increase in the response rate when medroxyprogesterone acetate combined with IFN was compared with IFN alone. ~53 In conclusion, hormone therapy has negligible antitumor activity Curr Probl Cancer, July/August 1997

209

s s -4

r O~ t-

8

C) 9~

on -

S

-0

q

0r

0

FO

1990 1990 1991 1991 1992 1993 1993

Hrushesky et al. 127 Damasce~li et al. 128 Merrouche et al. 129 Dexeus et al. ~3~ Budd et al. T M Conroy et al. ~32 Porter et al. 133 1992 1991 1993

1977 1984 1984 1985 1987 1988 1992

Hahn et al. 2~7 Kuebler et a l Y 8 Zeffron et al. 23g Tannock and Evans 240 Crivellari et at. T M Elson et a1742 Fossa et a1743

Raminski et al. ~34 Richards et al. T M Wilkinson et al. 245

Year

Reference

15 29 29

0 0 6

11 6 0 4 2 4 1

0 3 0 0 2 3 1

0 0 21

20 14 0 10 8 14 7

0 16 0 0 9 9 4

Complete or partial response No. of patients

Floxuridine + folinic acid (continuous infusion)

56 42 14 40 26 28 15

Floxuridine (circadian schedule)

10 19 10 14 21 35 26

Vinblastine

No. of patients assessable for response

TABLE 6. Results of chemotherapy for renal cell carcinoma with vinblastine and floxuridine

TABLE 7. Results of therapy for renal cell carcinoma with medroxyprogesterone or other progestins

No. of patients assessable for response

Complete or partial response

Reference

Year

Samuels et al. T M Talley et al. 246 Paine et al. 24~ Bloom 248 Wagle and Murphy 249 Werf-Messing and Gilse 250 TalleyT M Alberto and Senn 143 Tirelli et al. 2s2 Pearson et al? ~ Nakano et al. 139 Gottesman et al? 4~

1968 1969 1970 1971 1971 1971

21 8 15 60 35 21

No. of patients 3 2 2 9 6 2

14 13 13 15 17 10

1973 1974 1980 1981 1984 1985

61 17 23 9 13 18

7 0 3 0 0 0

11 0 13 0 0 0

%

against RCC; responses are rare and of short duration. In the treatment of RCC, medroxyprogesterone acetate is a component of best supportive care because it promotes appetite and weight gain. Hormonal and chemotherapy agents show little or no antitumor effect against RCC,154 and no chemotherapy agent has produced response rates that justify its use as a single agent. Thus there is a need for study of new agents for treatment of patients with RCC who have not had chemotherapy.

Modulation of Drug Resistance Multidrug resistance is a proposed mechanism of disease resistance to drugs. Multidrug resistance is associated with the MDR1 gene and its protein product P-glycoprotein. RCC originates in epithelial cells of the renal proximal tubule in more than 85% of cases. 155Phenotypic markers of proximal tubule cells are commonly found on renal carcinoma cells. This finding may explain the nearly uniform expression of P-glycoprotein on renal carcinoma cells; P-glycoprotein is normally present in the luminal membrane of the proximal tubule cell. 155Multidrug resistance was first recognized in the laboratory, where cancers induced in animal models exposed to a single drug developed a broad cross-resistance to a group of distinct cytotoxic agents. Several agents that reverse multidrug resistance have been tested in clinical studies of patients with RCC (Table 8). 156-161 In these clinical trials, the addition of cyclosporine, dipyridamole, nifedipine, tamoxifen, or quinidine to vinblastine did not result in enhanced antitumor activity. A randomized phase II trial of treatment with vinblastine alone folCurr Probl Cancer, July/August 1 9 9 7

211

TABLE 8. Results of therapy for renal cell carcinoma with agents to reverse multidrug resistance in combination with vinblastine

No. of patients assessable Agents

Reference

Year

Vinblastine + cyclosporine Vinblastine + dipyridamole Vinblastine + nifedipine

Rodenburg et al. ls8 1 9 9 1 Murphy et al. ls7 1994 Schwartsmann 1991

Vinblastine + quinidine Vinblastine + dexverapamil Vinblastine + cyclosporine or tamoxifen

Agarwala et al. r Motzer et al. ~6~ Samuels et al. ~6~

for response

CR/PR*

%

15 15 14

0/0 0/0 0/0

0 0 0

23 23 28

1/0 0/0 0/0

4 0 0

e t al, 158

1991 1995 1994

*CR = complete response; PR = partial response.

lowed by treatment with vinblastine plus either cyclosporine or tamoxifen has been completed. 161 Sixty-four patients were assessable for response to vinblastine alone, and no responses were seen. No objective response was observed in 28 patients who received agents that modulate multidrug resistance. The prototype of agents that reverse multidrug resistance is verapamil. Since 1981, this agent has been known to substantially reverse multidrug resistance in vitro. 162Treatment of patients with verapamil has been limited by the profound hypotension, negative inotropic effects, and delays in atrioventricular conduction associated with its clinical use.163-168In nearly all instances, plasma levels of this agent have been lower than those targeted for reversal of multidrug resistance. Standard verapamil is a racemic mixture of dextro (d-) and levo (l-) stereoisomers. The/-isomer slows atrioventricular node conduction to a markedly greater degree than the d-isomer (dexverapamil)J 69Because both isomers have been shown to be equally effective in reversing drug resistance in vitro, 17~ the less severe cardiovascular toxic effects of the d-isomer made this drug an attractive alternative to the standard racemic verapamil preparations for reversal of multidrug resistance. A trial of dexverapamil and vinblastine was conducted in patients with advanced RCC. 16~ Dexverapamil was generally well tolerated, but no antitumor activity was detected in 23 patients assessable for response, even though the plasma concentrations of dexverapamil and norverapamil were in the range of those that enhanced cytotoxicity in vitro. The existence of multiple, redundant mechanisms of drug resistance has been proposed as one explanation for the lack of antitumor activity observed in clinical trials using agents shown to modulate multidrug resis212

Curr Probl Cancer, July/August 1997

tance in experimental s t u d i e s . 172 The patients in these clinical trials had tumors, including RCC, that constitutively express the MDR1 gene. The lack of antitumor activity found in these trials emphasizes the need for new insight to develop methods of overcoming drug resistance in RCC.

Treatment With Biological ResponseModifiers Studies of immunotherapy for RCC were fostered by clinical observations of late relapse after nephrectomy, 173 prolonged periods of disease stabilization in the absence of systemic treatment, 173 and rare spontaneous disease regression. 174Interferons and interleukins are two categories of agents that have been studied and shown with reproducible findings to elicit tumor response. Interferon. In vitro, IFNs activate host mononuclear cells to increase expression of major histocompatibility complex molecules, and have a direct antiproliferative effect on renal tumor cells. 175 The overall response rate for treatment with IFN ~ in 1042 patients was 12%.t24 In patients treated with IFN ~, longer survival is associated with high* performance status, previous nephrectomy, and metastases predominantly in the lung. 176,177The major response rate (complete plus partial response) has been reported to be as high as 30% for patients with these characteristics. 178 The average period from start of treatment to an objective response is 3 to 4 months. However, the response to immunotherapy with IFN ~ and other biological response modifiers is characterized by slow regression of tumor masses over several months and partial response for as long as 12 months. The duration of response rarely exceeds 2 years, TM but long-term survival after treatment with IFN ~ has been reported. A dose of 5 to 20 million units/ day of recombinant IFN ~ appears to have maximal efficacy and avoids the more severe toxic effects of higher d o s e s . 179 IFN ]t has been studied less extensively, but it also shows a low degree of antitumor activity. TM Table 9 shows patient characteristics and treatment results for 159 patients treated for RCC with IFN ~ in three clinical trials at the Memorial Sloan-Kettering Cancer Center. 176In the first trial, each of 42 patients received 50 million units/m 2 intramuscularly 3 days a week. In the second trial, each of 64 patients received gradually escalating doses of IFN ~-2a from 3 to 36 million units/day subcutaneously. The third trial was randomized; 25 patients received daily IFN ~-2a alone (at the dose and schedule stated above for prior trial), and 28 were treated with daily IFN o~-2a and 0.15 mg/kg of vinblastine every 3 weeks. The overall rate for complete plus partial response was 10%, and the *Lack of symptoms related to their disease. Curr Probl Cancer, July/August 1997

213

TABLE 9. Patient characteristics and results of therapy with interferon alfa in 159 patients with renal cell carcinoma treated in three clinical trials at the Memorial Sloan-Kettering Cancer Center*

Trial I Total no. of 42 patients Median age (years) 57 No. of patients (% in parentheses) Nephrectomy 32 (76) Assessable for 39 (93) response Complete response 0 (0) Partial response 7 (18) Alive 5 (12) Disease free 0 (0)

Trial II

Trial III

Total (% in paratheses)

64

53

159 (100)

59

58

57

48 (75) 59 (92)

34 (64) 51 (96)

114 (72) 149 (94)

2 (3) 5 (8) 11 (17) 5 (8)

0 (0) 2 (4) 10 (19) 0 (0)

2 (1) 14 (9) 26 (16) 5 (3)

*The regimen in trial I was 50 million units/m 2 intramuscularly 3 days a week. The regimen in trial II was gradually escalating doses of IFN ~-2a from 3 to 36 million units/day subcutaneously. The regimen in trial III was daily IFN c(-2a alone or daily IFN c(-2a plus 0.15 mg/kg of vinblastine every 3 weeks.

median duration of response was 12 months. The median survival for all patients treated in one of these protocols was 11 months; 3% of the patients were alive at 5 years or more after treatment. A univariate statistical analysis showed that a high Karnofsky performance status, previous nephrectomy, and an interval from diagnosis to treatment of more than 1 year were significant prognostic factors for survival. 176Treatment with IFN t~2a was relatively well tolerated in an outpatient setting, and the low-dose regimen (trials II and III) was preferable to the high-dose regimen (trial I) because toxic effects were less severe. Interleukin-2. IL-2 has no direct effect on renal cancer cells in vitro; it exerts its effect on tumor growth by activation of lymphoid cells in vivo. Short intravenous infusions of high-dose IL-2 in combination with autologous lymphocytes 0ymphokine-activated killer [LAK] cells) resulted in objective responses in more than 30% of the patients in the initial study of the National Cancer Institute. 18~Three randomized trials showed that LAK cells do not add therapeutic benefit compared with the benefit from IL-2 alone. 181 A multicenter series of more than 250 patients treated with high-dose IL-2 alone demonstrated complete and partial response rates of 4% and 8%, respectively. 182 Some responses were durable, however, and the median duration of response was 23 months. 1s2 These results were achieved in a group of patients characterized by high performance status and relatively young age. The findings resulted in Federal Drug Administration approval for use of high-dose IL-2 in the treatment of RCC. Toxic effects associated with high-dose IL-2 given by bolus are related 214

Curr Probl Cancer, July/August 1997

TABLE 10. Results of therapy for advanced renal cell carcinoma with interleukin 2

Schedule High-dose bolus (inpatient) is~ Moderate-dose bolus (inpatient) ~81 Low-dose bolus (outpatient) ~s~

No. of studies 10 22 6

Total no. of patients

% with complete or partial response

537 650 104

19 15 20

to increased vascular permeability and require inpatient monitoring, often in an intensive care unit. High-dose IL-2 has been associated with a 4% treatment-related mortality m and should be offered only to patients with normal cardiac, renal, and pulmonary functions. Because of the formidable toxic effects and requirements for supportive care that are associated with this regimen, lower doses of IL-2 have been studied (Table 10). 181The results of treatment with IL-2 given at moderate doses to inpatients at the Memorial Sloan-Kettering Cancer Center are shown in Table 11. Overall, 8% of 100 patients assessable for response achieved major response. 181,183 The effect of the dose of IL-2 on response rate was summarized in a quantitative literature survey, ls~ Response rates were studied by the dose of the bolus (high, moderate, or low dose) with or without concomitant administration of LAK cells. Data were analyzed for 39 published series consisting of 1291 patients (Table 12). 181 The overall response rate was 16% (95% confidence interval, 8% to 25%). 181 For the high-dose group, the overall rate for complete plus partial response was 19% (11 studies; 95% confidence interval, 4% to 34%). For the moderate-dose group, the response rate was 15% (22 studies; 4% to 25%). For the low-dose group, the response rate was 20% (6 studies; 0% to 50%)? 81 The toxic effects of low-dose IL-2 given by bolus to outpatients are considerably less severe, and the response rates are comparable, but whether the responses are as durable as with the high-dose bolus remains to be determined. The relative efficacy of three schedules of IL-2 is being studied in a randomized trial at the National Cancer Institute, Combination Therapy. The low but reproducible response rates resulting from treatment with IFN and IL-2 have led to the study of these agents as part of combined therapy (Table 12). One combination protocol extensively studied was IFN plus vinbiastine. The results of 13 studies, including 307 patients, showed an overall 24% response rate, which appeared to be somewhat higher than that reported with IFN alone. ~24Three randomized trials have addressed the efficacy of IFN c~ plus vinblastine versus IFN o~ alone. 176'178'184One randomized trial did show a slightly higher response rate for the combination/84 but no study has shown improved surCurr Probl Cancer, July/August 1997

215

TABLE 11. Results of therapy with interleukin 2 in 105 patients treated in Memorial Sloan-Kettering Cancer Center clinical trials* Total (% in IL-2 a l o n e ~

Total no. of 36 patients Median age (years) 53 No. of patients (% in parentheses) Nephrectomy 30 (83) Assessable for 34 response Complete response 1 (3) Partial response 2 (6) Alive 9 (25) Disease free 2 (6)

IL-2 + LAK cells ~ x

IL-2 + IFN ~

parenthesis)

35

34

105

53

54

53

31 (89) 32

20 (59) 34

81 (77) 100

1 (3) 0 6 (17) 1 (3)

1 (3) 3 (9) 13 (38) 3 (9)

3 (3) 5 (5) 28 (28) 6 (6)

*IL-2 = interleukin 2; IFN = interferon; LAK = lymphokine-activated killer cells.

vival, and treatment with vinblastine resulted in greater gastrointestinal and hematologic toxic e f f e c t s . 176 Initiation of clinical trials of the combination of IL-2 and IFN o~ was supported by the results of preclinical studies showing synergistic therapeutic activity. A large number of institutions have studied this combination, with wide variation in doses, schedules, and routes of administration. A randomized phase II trial of high-dose IL-2 with IFN c~ versus highdose IL-2 alone showed no difference in response rates. 185 Moreover, in this randomized trial, toxic effects were more severe with the combination of IFN o~and IL-2 than with IL-2 alone, t85A review of 607 patients treated in 23 clinical trials showed a 19% response rate, similar to that achieved with IL-2 a l o n e . 186 The toxic effects of these two agents in combination were additive, and the authors concluded that IFN ~ plus IL-2 provided no apparent benefit compared with IL-2 alone) 86One randomized trial reported a response rate for the combination of IL-2 plus IFN oc that was higher than the response rate for either agent given alone, but the data on relative survival have not been reported. ~87 New combination regimens of IL-2, IFN ~, and fluorouracil; 188IFN plus 13-cis-refinoic acid; 189and IL-2 plus tumor-infiltrating lymphocytes tg~ show promise (Table 12) and are being studied in randomized clinical trials. A randomized trial of a u t o l y m p h o c y t e therapy compared with cimetidine showed a survival benefit for patients treated with autolymphocyte therapy. 191 This finding led to a confirmatory trial of autolymphocyte therapy compared with IL-2; the results are pending. The combination o f l F N and 13-cis-retinoic acid showed a 30% response rate in a phase II trial conducted at the Memorial Sloan-Kettering Cancer 216

Curr Probl Cancer, July/August 1997

TABLE 12. Results of therapy for renal cell carcinoma with combination regimens* No. of studies

Agents IFN + vinblastine IFN + IL-2 IFN + 5-FU or IFN + FUDR IL-2 + IFN + 5-FU IFN + 13-cis-retinoic acid

(reference) 13124 23186 8124 2253,254 1189

Total no. of patients

% with complete or partial response

307 607 145 54 43

24 19 19 48 30

*5-FU = fluorouracil; FUDR = floxuridine.

Center? 89 Three results of this study suggested that 13-cis-retinoic acid added efficacy to IFN o~: (1) The response rate was triple that previously reported from the center in trials of IFN o~alone or in combination with vinblastine (10% in 149 patients). 176 (2) Durable responses were achieved. (3) Responses were seen in sites generally considered to be resistant to IFN o~ (bone and primary tumor). Moreover, the results of in vitro studies suggested (a) that 13-cis-retinoic acid augments the antitumor effects of IFN o~in several renal cancer cell lines and (b) that the effects of retinoids in renal cancer cells may be mediated through the retinoic acid receptor 1].189A phase III randomized trial is under way to study the relative efficacy of IFN o~ plus 13-cis-retinoic acid compared with I F N 13~a l o n e . 176

The results of these recently reported studies of new combinations are promising and warrant further investigation, but caution must be exercised in interpreting these results in the context of standard care of patients with advanced RCC. Response rates vary among clinical trials of cytokines in patients with RCC, suggesting that patient selection plays an important role in the response rates in phase II trials. The relative efficacy of these combination protocols compared with that of IFN or IL-2 can only be assessed in the context of a randomized phase l]I trial.

Observation Alone Metastatic RCC is characterized by variability in clinical course and is one of the few malignant diseases for which spontaneous regressions of disease are well documentedJ74Investigators conducted a prospective study of patients in a phase 1/ clinical trial who had spontaneous regression. Twelve percent of the patients in this trial had remained free of disease progression for 12 months or more after spontaneous regressionJ 73This finding has implications for both clinical investigations and patient care. First, tumor regression or prolonged stabilization of disease after treatment with an investigational agent must be considCurr Probl Cancer, July/August 1997

217

ered in the context of the natural history of RCC. Second, responses to systemic therapies for RCC are infrequent, and all are associated with toxic effects. Therefore, patients who have no symptoms may be offered the option of observation until there is clinical evidence of disease progression or until the patient has symptoms. Then systemic treatment can be initiated, preferably in a clinical trial.

Tumor Biology Recent advances in molecular biology have improved our understanding of normal genitourinary tract development and identified numerous genetic events involved in the multistage process of malignant transformation. Many of the alterations associated with neoplasia have been correlated with other biologic characteristics such as tumor stage or differentiation status, and scientists have generated genetic models of tumorigenesis. During the past 10 years, molecular studies of RCC have begun to define the biology of renal cancer in research using both in vitro methods (human renal cell culture and renal cancer xenografts in athymic mice) and in vivo methods (animal models of RCC and models of human tumor specimens). Translating these studies into improved therapies will require an understanding of the cellular events that define the clinical course of this disease. Many questions remain unanswered. Why do some resected tumors recur quickly and others recur after many years? Why is RCC resistant to chemotherapy and radiotherapy but responsive in selected cases to immunotherapy with biological response modifiers? What is the role of the immune system in control of tumor growth?

TumorAntigen Expression Monoclonal antibodies generated against cell-surface glycoproteins, glycolipids, and blood group antigens of RCC have provided a series of phenotype markers that characterize the antigenic phenotype of RCC tumors. Many of these monoclonal antibodies also react with the proximal tubule portion of the human nephron, confirming that more than 90% of RCCs derive from epithelial cells of the proximal tubule. 192The results of recent studies have revealed that many of these kidney-related antigens are cell-surface peptidases and that their primary function in the brush border of kidney proximal tubule cells is to hydrolyze peptides and thus recycle or excrete amino acids. 193These findings suggest that these antigens also regulate autocrine and paracrine signals by proteolytic inactivation of specific mitogenic or regulatory peptides. Specific proximal tubule antigens are absent in many tumors, creating a series of molecular subtypes of RCC. For instance, lack of expression of the gpl60 glycoprotein correlates with sensitivity to IFN t~ in renal cancer cell lines and in renal cancer xenografts in athymic mice. 175To date, however, expression of any antigen has not proved to be a useful indicator of prognostic significance. 218

Curr Probl Cancer, July/August 1997

Molecular GeneticAbnormalities Classic cytogenetic analyses of kindreds with familial clear cell RCC revealed a translocation break point on the short arm of chromosome 3.2~ Subsequent molecular genetic studies of this region in both familial and sporadic RCC commonly showed loss of 3p14 through 3p26. 5~ The discovery that von Hippel-Lindau disease, which is characterized by a high incidence of bilateral RCC, was also linked to chromosome a r m 3p 196 enabled researchers to use DNA derived from affected families to identify the gene for von Hippel-Lindau disease at locus 3p25.197Analysis of the gene for this disease in sporadic clear cell RCC revealed loss of one allele in as many as 97% of tumors, 198A99with frequent point mutations, transcriptional inactivation through hypermethylation of the remaining allele, or both. 1982~ These data suggest that the gene for yon Hippel-Lindau disease is a classic tumor-suppressor gene and that inactivation contributes to tumor development. The gene for this disease is involved in the control of gene transcription. The protein product of the gene inhibits transcription elongation by binding to elongin B and C, two regulatory subunits of a transcriptional elongation factor complex. 2~176 Regulation of transcript elongation is an important control mechanism for gene expression, and it is possible that the gene for yon Hippel-Lindau disease modifies the expression of proto-oncogenes and growth-suppressor genes. The yon Hippel-Lindau messenger RNA is differentially expressed within renal tubules during kidney embryogenesis, suggesting that the gene product has a specific role in kidney development.2~ The precise role that loss of the gene for von Hippel-Lindau disease plays in the development of RCC has not been defined. However, reintroduction of the gene into RCC cells inhibits growth of the cells2~176 and tumor formation. 2~ This finding highlights the importance of the gene in control of the growth of renal cancer cells. Other regions of genetic loss on chromosome 3 have been identified in clear cell RCC. These regions include sites on chromosomes 6q, 8p, 9pq, 1 lp, 14q, 17, and 18q. 2~176 Multiple genetic losses in clear cell tumors are associated with poorer prognosis. 2~176

Oncogenes, Tumor-SuppressorGenes, and Growth Factors The development of RCC from normal renal epithelium may also involve alterations in genes with products that control cell division, including direct participation in control of the cell cycle. These genes include the retinoblastoma (Rb) gene, the p53 tumor-suppressor gene, and the r a s oncogene family, which has protein products that are involved in communicating signals from outside the cell to the nucleus (signal transduction pathway). The frequency of mutations of the p53 gene in primary RCC is low (less than 10%), 210-212e v e n though Curr Probl Cancer, July/August 1997

219

p53 protein is expressed in approximately one third of these tumors 213,214and may be associated with the development of metastatic disease. 214 These findings suggest that other mechanisms can contribute to stabilization of the p53 protein in renal cancer cells. The SDI1/WAF1/CIP1 gene (p21) is a downstream mediator of the p53 gene. Although analysis of this gene in renal cancers did not reveal any mutations,215studies of p21 protein expression showed wide variation in the level of expression in renal cancer cells, suggesting that aberrant regulation of p21 expression may have a role in development of renal cancer.215 Molecular alterations of the Rb gene are rare in renal cancers, and Rb protein is present in the majority of primary and metastatic renal tumors. 216These findings suggest that loss of Rb protein expression does not significantly contribute to malignant transformation or progression of renal cancers. Similarly, alterations of the CDKN2 gene and its protein product p 16, 217which is involved in control of the cell cycle, are infrequent. Mutations that activate the ras oncogenes are rare,218but mitogen-activated protein (MAP) kinase, a downstream mediator of this signal transduction pathway, may be activated in development of R C C . 219 Increased expression of growth factors or their receptors may also enhance RCC replication. In studies of RCC, investigators have reported overexpression of the epidermal growth factor receptor,22~ its ligand-transforming growth factor c~,22~ and insulin-like growth factor receptors. 223Angiogenic growth factors such as basic fibroblast growth factor and vascular endothelial growth factor may contribute to the hypervascularity that is characteristic of clear cell R C C 224'225and to the development of metastases. 226These factors may also have prognostic significance.2~ Elevated levels of the m y c oncogene occur in a large percentage of sporadic renal cancers227 but do not seem to be prognostic. 2% Abnormalities in known tumor-suppressor genes occur infrequently in RCC 229-233and do not appear to play a major role in its development. Recent studies have implicated cadherins234and integrins235in the pathogenesis of RCC, and newer techniques are facilitating the identification of novel genes that may be associated with this disease. Future studies should expand our understanding of the molecular events involved in the development and progression of RCC and should lead to novel and more effective treatment strategies.

Conclusions The need for continued research in RCC is highlighted by the finding that nearly one half of patients die within 5 years of diagnosis and, more specifically, that 5% to 10% of those who develop metastatic disease survive 5 years.5 Major areas of study should include tumor biology, causative factors, tumor prevention, and early detection. 220

Curr Probl Cancer, July/August 1997

Metastatic RCC is highly resistant to the many systemic therapies that have been extensively investigated. Small numbers of patients have complete or partial response to treatment with IFN, IL-2, or both. Response can be dramatic, with reduction of bulky tumors, and occasionally, it can be durable. However, because most patients do not respond and there is no clear evidence of improvement in survival, these agents by themselves should not be considered effective treatments for RCC. The response in some patients, however, indicates a need for continued research of immunotherapy. The identification of new agents with better antitumor activity against metastases remains a high priority in clinical investigation of therapy for this refractory tumor.

References 1. Motzer RJ, Bander NH, Nanus DM. Renal cell carcinoma. N Engl J Med 1996;335:86575. 2. Parkin DM, Pisani P, Ferlay J. Esfirnates of the world-wide incidence of eighteen major cancers in 1985. Int J Cancer 1993;54:594-606. 3. Pisani P, Parkin DM, Ferlay J. Estimates of the world-wide mortality from eighteen major cancers in 1985. Implications for prevention and projections of future burden. Int J Cancer 1993;5:891-903. 4. Cancer facts and figures--1996. Atlanta: American Cancer Society; 1996. 5. SEER cancer statistics review: 1973-1990. Bethesda, Maryland: National Cancer Institute, 1993; NIH publication no. 93-2789. p. XI1-XI22. 6. MeCredie M, Pommer W, McLanghlin JK, et al. Internationalrenal cell cancer study. lI. Analgesics. Int J Cancer 1995;60:345-9. 7. Linblad E Mellemgaard A, Schlehofer B, et al. International renal cell cancer study. V. Reproductive factors, gynecologic operations and exogenous hormones. Int J Cancer 1995;61:192-8. 8. Mellemgaard A, Bergstrom R, Lindblad E et al. International renal cell cancer study. III. Role of weight, height, physical activity and use of amphetamines. Int J Cancer 1995;60:350-4. 9. Mandel JS, McLaughlin JK, Schlehofer B, et al, International renal-cell cancer study. IV. Occupation. Int J Cancer 1995;61:601-5. 10. WolkA, Gridley G, Niwa S, et al. International renal cell cancer study.VII. Role of diet. Int J Cancer 1995;65:67-74. 11. McLaughlin JK, BlotWJ, Mandel JS, Schuman LM, Mehl ES, Fraumeni JF. A population-based case-control study of renal cell carcinoma. J Natl Cancer Inst 1984;72:27484. 12. Yu MC, Mack TM, Hanisch R, Cicioni C, Henderson BE. Cigarette smoking, obesity, diuretic use, and coffee consumption as risk factors for renal-cell carcinoma. J Nail Cancer Inst 1986;77:351-6. 13. McLauglin JK, Lindblad E MelemgaardA, et al. Internationalrenal cell cancer study. I. Tobacco use. Int J Cancer 1995;60:194-8. 14. Wolk A, Lindblad E Adami H. Nutrition and renal cell cancer. Cancer Causes Control 1996;7:5-I8. 15. Ishikawa I, Kovacs G. High incidence of papillary renal cell tumours in patients on chronic hemodialysis. Histopathology 1993;22:135-9. 16. Washecka R, Hanna M. Malignant renal tumors in tuberous sclerosis. Urology 1991;37:340-3. 17. Mellemgaard A, Engholm G, McLaughlin JK, Olsen JH. Risk factors for renal cell Curr Probl Cancer, July/August 1997

221

18. 19. 20. 21.

22. 23.

24. 25. 26.

27. 28.

29.

30. 31. 32.

33. 34. 35. 36. 37.

38. 39.

222

carcinoma in Denmark. I. Role of socioeconomic status, tobacco use, beverages, and family history. Cancer Causes Control 1994;5:105-13. Lamiell JM, Salazar FG, Hsia YE. von Hippel-Lindau disease affecting 43 members of a single kindred. Medicine 1989;68:1-29. Cohen AJ, Li FP, Berg S, et al. Hereditary renal cell carcinoma associated with a chromosomal translocation. N Engl J Med 1979;301:592-5. Pathak S, Strong LC, Ferrell RE, Trindade A. Familial renal cell carcinoma with a 3;11 chromosome translocation limited to tumor cells. Science 1982;217:939-41. Kovacs G, Brusa P, De Riese W. Tissue-specific expression of a constitutional 3;6 translocation: development of multiple bilateral renal-cell carcinomas. Int J Cancer 1989;43:422-7. Zbar B, Tory K, Merino M, et al. Hereditary papillary renal cell carcinoma. J Urol 1994;151:561-6. Li FP, Decker HJ, Zbar B, et al. Clinical and genetic studies of renal cell carcinomas in a family with a constitutional 3;8 translocation. Genetics of familial renal carcinoma. Ann Intern Med 1993;118:106-10. Cronin RE, Kaehny WD, Miller PD, et al. Renal cell carcinoma: unusual systemic manifestations. Medicine 1976;55:291-311. Ritchie AWS, Chisholm GD. The natural history of renal carcinoma. Semin Oncol 1983;10:390-400. Skinner DG, Colvin RB, Vermillion CD, et al. Diagnosis and management of renal cell carcinoma. A clinical and pathologic study of 309 cases. Cancer 1971;28:116577. Nseyo UO, Williams PD, Murphy GD. Clinical significance of erythropoietin levels in renal carcinoma. Urology 1986;28:301-6. Gotoh A, Kitazawa S, Mizuno Y, et al. Common expression of parathyroid hormone-related protein and no correlation of calcium level in renal cell carcinomas. Cancer 1993;71:2803-6. StadlerWM, Vogelzang NJ. Serum interleukin-6 levels in metastatic renal cell cancer: correlation with survival but not as an independent prognostic indicator. J Natl Cancer Inst 1992;84:1835-6. Boxer RJ, Waisman J, Lieber MM, Mampaso FM, Skinner DG. Non-metastatic hepatic dysfunction associated with renal carcinoma. J Urol 1978;119:468-71. Dawson NA, Barr CF, Alving BA. Acquired dysfibrinogenemia:paraneoplastic syndrome in renal cell carcinoma. Am J Med 1985;78:682-6. Chang S, Yu D, Sherwood ER, Kozlowski JM, Lee C. Inhibitory effects of suramin on a human renal cell carcinoma line, causing nephrogenic hepatic dysfunction. J Urol 1992;147:1147-50. Porena M, Vespasiani G, Rosi P, et al. Incidentally detected renal cell carcinoma: role of ultrasonography. J Clin Ultrasound 1992;20:395-400. Hellsten S, Berge T, Whelin L. Unrecognized renal cell carcinoma. Scand J Urol Nephrol 1981;15:269-72. Konnack JW, Grossman HB. Renal cell carcinoma as an incidental finding. J Urol 1985;134:1094-6. Thompson IM, Peek M. Improvement in survival of patients with renal cell carcin o m a - t h e role of the serendipitously detected tumor. J Urol 1988;140:487-90. Smith SJ, Bosniak MA, Megibow AJ, Hulnick DH, Horii SC, Raghavendra BN. Renal cell carcinoma: earlier discovery and increased detection. Radiology 1989; 170:699-703. Tsukamoto T, Kumamoto Y, Yamazaki K, et al. Clinical analysis of incidentally found renal carcinomas. Eur Urol 1991;19:109-13. Aso Y, Homma Y. A survey on incidental renal cell carcinoma in Japan. J Urol 1992; 147:340-4. Curr Probl Cancer, July/August 1997

40. Kessler O, Mukamel E, Hader H, et al. The impact of the improved diagnosis of renal cell carcinoma on the course of the disease. J Surg Oncol 1994;57:201-5. 41. Thoenes W, Storkel S, Rumpelt HJ. Histopathology and classification of renal cell tumors (adenomas, oncocytomas and carcinomas). The basic cytological and histomorphological elements and their use for diagnostics. Pathol Res Pract 1986;181:125-43. 42. van den Berg E, van der HoutAH, Oosterhuis JW, et al. Cytogenetic analysis of epithelial renal-celltumors: relationship with a new histopathologicalclassification. Int J Cancer 1993;55:223-7. 43. Weiss LM, Gelb AB, Medeiros LJ. Adult renal epithelial neoplasms. Am J Clin Pathol 1995;103:624-35. 44. Storkel S, Steart PV, Drenckhalm D, Thoenes W. The human chromophobe renal cell carcinoma: its probable relation to intercalated cells of the collecting duct. Virchows Arch B Cell Pathol 1989;56:237-45. 45. Akhtar M, Kardar H, Linjawi T, et al. Chromophobe cell carcinoma of the kidney--a clinicopathologic study of 21 cases. Am J Surg Pathol 1995;19:1245-56. 46. Speicher MR, Schoell B, du Manoir S, et al. Specific loss of chromosomes 1, 2, 6, 10, 13, 17, and 21 in chromophobe renal cell carcinomas revealed by comparative genomic hybridization. Am J Pathol 1994;145:356-64. 47. Storkel S, Vanderberg E. Morphological classification of renal cancer. World J Urol 1995;13:153-8. 48. Furhmaim SA, Lasky LC, Limas C. Prognostic significance of morphologic parameters in renal cell carcinoma. Am J Surg Pathol 1982;6:655-63. 49. Ro JY, Ayala AG, Sella A, Samuels ML, Swanson DA. Sarcomatoid renal cell carcinoma: a clinicopathologic study of 42 cases. Cancer 1987;59:516-26. 50. Presti JC Jr, Rao PH, Chen Q, et al. Histopathological, cytogenetic, and molecular characterization of renal cortical tumors. Cancer Res 1991 ;51:1544-52. 51. Mancilla-Jimenez R, Stanley RJ, Blath RA. Papillary renal cell carcinoma: a clinical, radiologic, and pathologic study of 34 cases. Cancer 1976;38:2469-80. 52. Kovacs G, Wllkens L, PappT, De RieseW. Differentiationbetween papillary and nonpapiUary renal cell carcinomas by DNA analysis. J Natl Cancer Inst 1989;7:527-30. 53. Kovacs G, Guzesi L, Emanuel A, Kung H. Cytogenetics of papillary renal cell tumors. Genes Chromosom Cancer 1991;3:249-55. 54. Sene AP, Hunt L, McMahon RF, Carroll RN. Renal cell carcinoma in patients nndergoing nephrectomy: analysis of survival and prognostic factors. Br J Urol 1992;70:12534. 55. Bander NH, Finstad CL, Cordon-Cardo C, Ramsawak RD, Vaughan ED, Whitmore WF. Analysis of a mouse monoclonal antibody that reacts with a specific region of the human proximal tubule and subsets of renal cell carcinomas. Cancer Res 1989;49:677480. 56. Kovacs A, Kovacs G. Low chromosome number in chromophobe renal cell carcinoma. Genes Chromosom Cancer 1992;4:267-8. 57. Crotty TB, Farrow GM, Lieber MM. Chromophobe cell renal carcinoma---clinicopathologic features of 50 cases. J Urol 1995;154:964-7. 58. Morra MN, Das S, Renal oncocytoma: a review of histogenesis, histopathology, diagnosis and treatment. J Urol 1993;150:295-302. 59. Kennedy SM, Merino MJ, Linehan WM, Robert JR, Robertson CN, Neumann RD. Collecting duct carcinoma of the kidney. Hum Pathol 1990;21:449-56. 60. Warshauer DM, McCarthy SM, Street L, et al. Detection of renal masses: sensitivities and specificifies o f excretory urography/linear tomography, US and CT. Radiology 1988; 169:363-5. 61. Amendola MA, Bree BL, Pollack HM, et al. Small renal cell carcinomas: resolving a diagnostic dilemma. Radiology 1988;166:637-41. Curr Probl Cancer, July/August 1997

223

62. Rofsky NM, Weinreb JC, Bosniak MA, Libes RB, Bimbaum BA. Renal lesion characterization with gadolinium-enhanced MR imaging: efficacy and safety in patients with renal insufficiency. Radiology 1991; 180:85-9. 63. Bosniak MA. The small (