Tumours of the kidney include those of the parenchyma and urothelial tumours in the renal pelvis. Parenchymal tumours include renal cell carcinoma, nephroblastoma, oncocytoma, angiomyolipoma, and sarcoma. Renal tumours comprise 2 per cent of all malignant tumours. For example, there were 3036 registrations of malignant tumours of the kidney (International Classification of Diseases 189) in adults in England and Wales in 1986 and these were the tenth most common tumours in men.

This tumour has previously been called Grawitz tumour, hypernephroma, adenocarcinoma and clear cell carcinoma, but these terms should be abandoned in favour of the above.

Renal cell carcinoma is unusual below the age of 40. It presents at a median age of 65 and is twice as common in men as in women. The incidence is increasing in both sexes, particularly in those over the age of 65 years. Scandinavia has a high incidence; incidence is lower in Western Europe and North America, and even lower in Japan.

The aetiology of renal cell carcinoma remains uncertain, although there are a number of known associations, the most clear of which are genetic. In some patients there is a familial history of renal cell carcinoma and 70 per cent of patients with von Hippel-Lindau disease develop renal cell carcinoma by the age of 60 years. There is an early age of onset of renal cell carcinoma in these patients compared to sporadic cases, and a tendency towards bilateral and multicentric tumours. A number of oncogenes have been studied but none is peculiar to renal cell carcinoma. Genes on the short arm of chromosome 3 are implicated in both inherited and sporadic renal cell carcinoma. In the future this may be useful, for example, in allowing early diagnosis of von Hippel-Lindau disease. When the actual genetic defect is identified perhaps even preventative gene therapy might be possible.

An increased risk of renal cell carcinoma is associated with smoking and obesity. Associations with renal cell carcinoma have been found in coke-oven workers and those with occupational exposure to cadmium or petrochemicals.

Renal cell carcinoma may present because of direct, lymphatic, or haematogenous spread, or with paraneoplastic syndromes. Direct spread occurs through perirenal fat and Gerota's fascia into surrounding structures. The incidence of lymphatic spread is 30 per cent with involvement of the renal hilar, para-aortic, paracaval, and mediastinal nodes. Venous disease occurs in up to 50 per cent of patients; disease spreads beyond the renal vein and up to the right atrium in 5 per cent. Distant metastases are seen in 20 to 30 per cent of patients at presentation, most often in the lung or mediastinum, but also in bone, liver, brain, and skin.

Renal cell carcinoma is staged to predict prognosis. The tumour–node–metastasis (TNM) system (Table 1) 468 or, particularly in North America, Robson's modification of Flocks and Kaddesky's system (Table 2) 469 is used.

Renal cell carcinoma is usually a solitary, well-demarcated mass containing yellow or brown tumour with haemorrhage and necrosis (Fig. 1) 1599. Cysts are common.

The tumour often appears to be demarcated, although there are usually numerous regions of microscopic tumour infiltration into the surrounding normal kidney. Three cell types are found: clear cells, granule cells, and spindle cells. These may produce different microscopic variants of the tumour such as alveolar, trabecular, tubular, or papillary. The typical tumour is tubuloalveolar with clear and granule cells. Immunopathological studies suggest that renal cell carcinomas arise from proximal tubule cells.

Only 10 per cent of renal cell carcinomas present with the classical triad of haematuria, loin pain, and a palpable mass, although 80 per cent of patients have one or more of these features. The primary tumour may also cause microscopic haematuria. Left renal vein disease in men may cause a rapid-onset varicocele. Systemic effects such as hypertension, weight loss, fever, and night sweats are common. An increasing number of tumours are being detected by chance during ultrasound, CT, or magnetic resonance examinations performed for other reasons.

One-third of patients present with a non-urological feature, so-called paraneoplastic syndromes. These are early manifestations of tumours and most resolve after nephrectomy. Non-specific syndromes and specific endocrine syndromes are recognized.

Anaemia occurs in 30 per cent of patients, with or without haematuria. A raised erythrocyte sedimentation rate occurs in over 50 per cent of patients. Increased or decreased platelet production and a variety of coagulopathies can occur. Thirty per cent of patients have abnormal liver function tests, and diffuse hepatomegaly is common, although it is rarely associated with hepatic metastases. Neuropathies, myopathies, vasculitis, amyloidosis, raised &agr;-fetoprotein levels, and disordered iron metabolism may also occur.

Renal cell carcinomas may produce renin and this can cause hypertension. Sixty per cent of patients with renal cell carcinoma have elevated erythropoietin levels, although only 10 per cent have erythrocytosis. Renal cell carcinomas may produce ectopic hormones such as parathormone, ACTH, prolactin, human chorionic gonadotrophin, and glucagon.

There are over 20 cases in the literature of regression of histologically-proven renal cell carcinoma and the overall incidence of spontaneous regression is about 0.5 per cent.

This has two purposes: first to demonstrate a solid mass in the renal substance and second to stage the disease and assess its effects.

An intravenous urogram will demonstrate a renal mass and contralateral renal function (Fig. 2) 1600. Ultrasound will show whether the lesion is cystic or solid. Unless a cystic lesion is obviously a simple cyst, it may be aspirated under local anaesthesia with ultrasound guidance. This should detect those ‘complex cysts’ which are in fact tumours. If the lesion is solid the liver, renal vein, and inferior vena cava should be examined to exclude the presence of hepatic metastases and spread via the renal vein. If there is still doubt, either CT or magnetic resonance imaging may confirm a solid mass and give some indication of its nature and extent (Fig 3) 1601. Solid renal lesions can be biopsied under radiological guidance, but most lesions are removed surgically because of the high likelihood of renal cell carcinoma, so biopsy is seldom performed. If there is uncertainty about spread via the renal vein, vena cavography is the definitive study.

Investigation to stage and assess the effects of the disease seeks to detect metastases and haematological or biochemical abnormalities: this allows treatment to be planned. The liver is examined by ultrasound or CT imaging. Plain chest radiography is required if CT or magnetic resonance imaging of the chest was not performed when examining the primary tumour. Bone scintigraphy is seldom undertaken in the United Kingdom, and a blood count and biochemical profile usually complete the patient's assessment.

Radical nephrectomy is now the standard operation for renal cell carcinoma. In the presence of metastases it may be performed to control local symptoms such as pain or haematuria. It entails early ligation of the renal artery, then the renal vein, and removal en bloc of Gerota's fascia (with its contents, which include the kidney, the adrenal, and the perinephric fat), the upper ureter, and the renal vessels. This procedure superseded simple nephrectomy in the 1960s when staging of renal cell carcinomas demonstrated that spread into the perinephric fat worsened prognosis. No evidence has subsequently shown that radical nephrectomy confers a survival advantage, compared to simple nephrectomy. Radical nephrectomy is usually performed transperitoneally, to allow better access to the renal vessels, but thoracoabdominal or flank incisions, or a lumbotomy can also be used.

Radical nephrectomy with abdominal lymphadenectomy has been advocated because 10 to 15 per cent of patients without identified distant metastases have lymph node metastases. However, there is no evidence to confirm a survival advantage of this more extensive procedure. The improved staging provided by lymphadenectomy confers no benefit because there is no effective therapy for metastatic disease at the present time.

Up to 10 per cent of patients have vena caval tumour thrombus, although 50 per cent of them have metastases. In the absence of metastases, removal of the thrombus can be followed by prolonged survival, although the operative mortality rate is 5 to 10 per cent. Infradiaphragmatic thrombus can be removed after proximal caval control. Cardiopulmonary bypass and hypothermic circulatory arrest has been used to facilitate removal of supradiaphragmatic thrombus.

Partial nephrectomy may be desirable in patients with a solitary kidney, with bilateral tumours, with familial renal cell carcinoma or von Hippel-Lindau disease, in whom metachronous bilateral tumours are likely, and in the increasing number of patients whose lesions are detected incidentally by radiology. In all these groups, survival following partial nephrectomy is at least as good as that following radical nephrectomy, although local recurrence rates may be increased. Although renal cell carcinomas often seem to have a pseudocapsule and to be amenable to enucleation, a rim of normal-looking kidney should be removed because microscopic infiltration is very common. So-called bench surgery, whereby the kidney is excised, the tumour is removed from the kidney ex vivo, and the kidney is then autotransplanted back into the patient, has not proved to be useful except in the case of a tumour in a solitary kidney, largely because renal function in the operated kidney is transiently impaired after surgery. Some 10 to 20 per cent of kidneys with renal cell carcinomas have satellite lesions, a potential problem following conservative surgery.

The renal artery can be occluded with a variety of materials introduced via a percutaneous arterial catheter. The most appropriate role for embolization is in patients with advanced disease and a symptomatic primary tumour. Embolization may relieve symptoms with less morbidity than that associated with a nephrectomy. Embolization can also reduce the vascularity of very large tumours before surgery, and when followed by surgery in patients with multiple metastases, may produce regression of metastases.

Neither pre- nor postoperative radiotherapy improves survival.

Renal cell carcinoma is unusual in that resection of solitary metastases (particularly pulmonary or cerebral) can be followed by long-term survival on occasions. Radiotherapy relieves pain due to skeletal metastases in up to 80 per cent of patients and internal fixation can stabilize long bones affected by metastatic disease.

Disseminated disease has been treated with many chemotherapeutic and immunotherapeutic agents. The incidence of spontaneous regression suggests that the immune system is involved in the balance between host and disease and suggested that immunotherapy might be effective against metastatic disease. Unfortunately, no form of systemic treatment improves survival.

Progestins, androgens, anti-oestrogens, steroids, and over 35 cytotoxic agents, either singly or in combination, have been used to treat disseminated disease, but none has consistently produced a response rate of over 10 per cent.

The explosion of molecular biological technology over the past decade has greatly increased our knowledge of many aspects of the biology of renal cell carcinoma and has suggested new treatment options for metastatic disease.

Over 1000 patients have been treated with recombinant interferons, with an overall response rate of 13 per cent with a typical duration of 6 to 10 months. Toxicity is usually mild or moderate.

Interleukins are lymphokines produced by T lymphocytes which enhance the killing of tumour cells. Interleukin-2 monotherapy for metastatic renal cell carcinoma has been used widely and shows a response rate of about 18 per cent lasting for up to 34 months. The toxicity of interleukin-2 is considerable, renal failure and cardiovascular failure being the most serious problems. There is an associated mortality rate of 1 to 2 per cent.

Interferons and interleukin-2 have been used in combination and the early results suggest that they exhibit synergy, with response rates of about 30 per cent and median response durations of over 19 months. Another approach is adoptive immunotherapy, which involves administration of cells which have antitumour activity: initial reports support its further assessment.

The prognosis of renal cell carcinoma is difficult to predict because its natural history is variable. Very large primary tumours with venous extensions sometimes have no associated metastases, but metastases have been reported in patients with renal cell carcinomas smaller than 2 cm. The majority of patients with localized disease who are not treated die of their disease, and patients who are apparently tumour free may develop metastases and die more than 20 years after nephrectomy. However, the incidence of spontaneous regression of renal cell carcinoma is second only to that of melanoma and some patients with untreated advanced disease survive for 5 years or more.

Robson identified five prognostic factors which are the basis of current classifications: distant metastases, regional lymph node involvement, venous extension, local spread, and tumour grade. Despite extensive basic scientific research, no other factors have been shown to have better prognostic significance.

In view of the unpredictable natural history of renal cell carcinoma, the results of treatment should be interpreted with caution in respect of the length of monitoring after putatively curative surgery, the method and accuracy of staging, and the inclusion of cause-specific death rates. Some of these factors may explain part of the variation in survival in series of nominally similar patients. In patients treated by radical nephrectomy alone, overall 5- and 10-year survival rates range between 38 and 59 per cent and 27 and 40 per cent, respectively. Good survival rates at 5 years would be 65 per cent for those with stage I disease, 55 per cent at stage II, and 50 per cent of those with stage III disease with local recurrence rates of less than 10 per cent. Patients with stage II and III disease without palpable nodal metastases may benefit from a lymph node dissection. Patients who have venous extension, but no lymphatic or distant spread, have 5-year survival rates of 40 to 50 per cent following radical nephrectomy and thrombus removal.

Survival rates of 80 to 90 per cent have been reported following conservative surgery of stage I tumours, particularly those detected incidentally: this contrasts with 65 per cent survival following radical nephrectomy for stage I lesions. Lead- and length-time bias may account for some of this difference. This suggests that the results of any screening programme for renal cell carcinoma will appear excellent and this may wrongly be ascribed to screening rather than to a biological head-start.

About 10 per cent of those with stage IV disease survive for 1 year following nephrectomy, and the median survival is about 4 months. If there are solitary or few pulmonary metastases, 5-year survival rates of 25 to 35 per cent have been reported following lung resections. Patients with metachronous metastases have a better prognosis than those with synchronous metastases, particularly if the disease-free interval is more than 24 months.

Between 5 and 10 per cent of solid renal masses are either oncocytomas or angiomyolipomas. Oncocytoma is a benign neoplasm of distal tubular origin and angiomyolipoma is a hamartoma. The affected kidney is often removed for presumed renal cell carcinoma: this is probably overtreatment for oncocytoma or angiomyolipoma. These tumours are often asymptomatic, and are likely to be detected more often incidentally with the increasing use of abdominal ultrasound, computed tomography, and magnetic resonance imaging. This may lead to nephrectomies which would not otherwise have been performed and which may be unnecessary.

Oncocytomas are twice as common in men as in women and are diagnosed at a median age of 60. They are usually well-encapsulated and rarely penetrate the renal capsule. Oncocytomas are often spherical and are a homogeneous brown colour, although larger lesions may have a characteristic central stellate scar. Distinction from oncocytic renal cell carcinoma may be difficult pathologically.

Eighty per cent of angiomyolipomas occur in women, usually around the age of 60. Eighty per cent are associated with tuberous sclerosis. They contain muscle, vascular tissue, and fat and are well circumscribed but not encapsulated. They vary in colour from yellowish to pinkish-grey, depending on composition. Both oncocytomas and angiomyolipomas can be bilateral and multicentric and can coexist with renal cell carcinoma. Rarely, an angiolipoma metastasizes.

Oncocytomas are seldom symptomatic whereas angiomyolipomas may present with flank pain, a mass, or microscopic haematuria. Angiomyolipomas may rupture and present acutely with retroperitoneal haemorrhage.

Both lesions may be diagnosed correctly on radiology in both symptomatic and asymptomatic patients. This may allow alternative treatments to routine nephrectomy for solid renal masses, which were previously all thought to be renal cell carcinoma.

Patients with small peripheral lesions, particularly if asymptomatic, may only require periodic radiological review. Those with larger, symptomatic, bilateral, multicentric, or indeterminate lesions may be treated by partial nephrectomy, or enucleation for angiomyolipoma. Deaths apparently from oncocytoma may actually be caused by oncocytic renal cell carcinoma.

About 1 per cent of renal tumours are sarcomas and their incidence increases with age. Distinction from renal cell carcinoma is usually only possible histopathologically. All forms of sarcoma can occur in the kidney, but leiomyosarcoma accounts for 60 per cent. Leiomyosarcoma of the kidney tends to compress the kidney, rather than invade it and it metastasizes early and widely. Treatment is by radical nephrectomy. Survival may be increased after chemotherapy.

Nephroblastoma (Wilms' tumour) is the most common intra-abdominal tumour of children and accounts for 10 per cent of childhood malignancy. The incidence is 1/10 000 live births or seven per million children under the age of 15 years. There were 54 registrations of kidney and unspecified urinary tract tumours (International Classification of Diseases 189) in children in England and Wales in 1986. The peak incidence is between 2 and 4 years; few cases occur after the age of 7.

About 1 per cent of children with nephroblastoma have associated congenital anomalies. A genetic influence in the aetiology of nephroblastoma has been borne out by detailed study of its genetics and molecular biology. Nephroblastoma was the model for Knudson's two-hit hypothesis which proposes that two separate factors act to cause the disease: cases with an associated genetic predisposition occur at earlier ages and are more commonly bilateral than sporadic cases. Genetic defects were localized to chromosome 11, where a tumour-suppressor gene has been identified, which is normally expressed in the early fetal kidney. Loss of this gene, Knudson's first hit, then allows tumour development in the presence of any other predisposing factor, the second hit. Renal abnormalities such as nephroblastomatosis are known to predispose to nephroblastoma.

Like renal cell carcinoma, nephroblastoma may present because of direct, lymphatic, and haematogenous spread. Direct spread occurs through perirenal fat and Gerota's fascia into surrounding structures. Renal hilar, paraaortic, paracaval, and mediastinal nodes may become involved. Venous involvement occurs in about 5 per cent of cases, with intracardiac extension in less than 1 per cent. Distant metastases occur most often in the lung, but can occur in liver or other sites.

The United Kingdom Medical Research Council, the American National Wilms' Tumor Study Group (NWTSG) and the International Society for Paediatric Oncology (SIOP) have undertaken clinical trials in nephroblastoma and have provided a great deal of vital information about the disease and its treatment. The NWTSG has developed a staging system (Table 3) 470.

Most primaries are more than 10 cm across and weigh over 500 g. The tumour is usually soft and friable with areas of necrosis, haemorrhage, and cyst formation. There is usually a well-defined pseudocapsule.

The tumour is of renal origin and contains elements of blastema, tubules, and stroma. Heterologous tissue elements such as muscle and bone may also be present. Tumours are divided into those with and without anaplastic histology: anaplasia correlates with metastatic spread, relapse after treatment, and survival.

Nephroblastoma usually presents as a mass in the flank, with abdominal pain and haematuria in about 25 per cent of patients and fever, anorexia, and vomiting in about 5 per cent. Anomalies associated with nephroblastoma include aniridia, hemihypertrophy, Beckwith–Wiedemann syndrome, and external genitalia deformities.

Ultrasound is usually the initial investigation, and will confirm that a solid mass arises from the kidney. The other kidney is examined, the liver is checked for metastases, and the renal vein and inferior vena cava are imaged to exclude venous extensions. An intravenous urogram gives more anatomical detail and an indication of renal function, although a renogram may be required. A chest radiograph, blood count, and biochemical profile are performed. Staging, however, requires information from surgery and histology.

In contrast to renal cell carcinoma, radiotherapy and chemotherapy have much to offer after nephrectomy in patients with nephroblastoma. Survival from nephroblastoma has improved dramatically in recent decades with improved surgical care and refinement of radiotherapy and chemotherapy.

Radical nephrectomy is performed through an anterior transperitoneal approach. The objective is complete tumour removal without spillage, which doubles the relapse rate. The opposite kidney is mobilized, inspected and palpated to exclude a bilateral tumour: this occurs in about 5 per cent of patients and one-third are not detected preoperatively. The usual approach to bilateral tumours is a radical nephrectomy on the side of the larger tumour and removal of as much of the tumour as possible on the other side. This may necessitate a contralateral nephrectomy and dialysis followed by transplantation.

There is debate about the renal-hyperperfusion syndrome which follows nephrectomy and contralateral partial nephrectomy, and which can lead to chronic deterioration in renal function. This effect is stimulating a renal-sparing approach to surgery, particularly as effective chemotherapy becomes available.

Extensive lymphadenectomy is not advocated, but there is debate about the role of lymph node sampling to improve staging.

Preoperative chemotherapy is given when venous tumour thrombus is found at diagnosis, and the thrombus may well recede. Removal of intracardiac tumour may require cardiopulmonary bypass.

In contrast to renal cell carcinoma, nephroblastoma is sensitive to both radiotherapy and cytotoxic drugs. Numerous studies have investigated how these treatments can maximize benefit and minimize side-effects. Complications of radiotherapy in children with nephroblastoma include second tumours, scoliosis, and radiation enteritis.

The SIOP studies investigated the use of preoperative radiotherapy and chemotherapy to downstage tumours and reduce the need for postoperative radiotherapy. In the NWTSG studies, however, preoperative treatment is avoided on the grounds that staging is more accurate in the untreated patient and that up to 8 per cent of children prove, after nephrectomy, to have some other condition.

The SIOP studies showed that preoperative vincristine and actinomycin D were as successful as preoperative radiotherapy in downstaging and reducing surgical tumour spillage from rupture. Later studies have shown that postoperative chemotherapy reduces relapse and that lung metastases can be treated effectively by chemotherapy and salvage lung resections.

The NWTSG studies showed that anaplasia predicts poor outcome and the need for intensive treatment. Combination chemotherapy was shown to be more effective than single agents. Postoperative radiotherapy in addition to chemotherapy is unnecessary in children younger than 2 years, those with stage I tumours, and those with stage II tumours and favourable histology. The addition of doxorubicin to vincristine and actinomycin D enabled the dose of postoperative radiotherapy to be halved in children with stage III disease and favourable histology. The latest NWTSG study is assessing the use of shorter chemotherapy regimens and attempting to reduce the toxicity of doxorubicin.

Recurrence after treatment is difficult to manage and a multidisciplinary approach is most successful. Ifosfamide, cisplatin, and etoposide all have activity against nephroblastoma and are used to treat recurrence.

The most important prognostic factors are anaplastic histology and stage. Four-year survival rates in the NWTSG 3 study for children with favourable histology were stage I, 97 per cent; stage II, 94 per cent; stage III, 88 per cent; and stage IV, 82 per cent. For children with unfavourable histology, the 4-year survival rates were stage I, 89 per cent; stages II–IV, 54 per cent.

Belldegrun A, Figlin R, Danella J, deKernion J. Immunotherapy for renal cell carcinoma. Semin Urol, 1992; 10: 23–7.

Brewster SF, Gingell JC, Brown KW. Tumour suppressor genes in urinary tract oncology. Br J Urol, 1992; 70: 585–90.

Lemerle J, Tournade MF, Habrand JL. The treatment of Wilms' tumor: an update. Prog Clin Biol Res, 1989; 303: 629–33.

Licht MR, Novick AC. Nephron sparing surgery for renal cell carcinoma. J Urol, 1993; 149: 1–7.

Marshall FF. Surgery of renal cell carcinoma with inferior vena caval involvement. Semin Urol, 1989; 7: 186–90.

Montie JE. Lymphadenectomy for renal cell carcinoma. Semin Urol, 1989; 7: 181–5.

National Wilms' Tumor Study Committee. Wilms' tumor: status report, 1990. J Clin Oncol, 1991; 9: 877–87.

Ramsay J. Immunotherapy and chemotherapy for carcinoma of the kidney. Br J Urol, 1992; 70: 465–8.

Robson CJ, Churchill BM, Anderson W. The results of radical nephrectomy for renal cell carcinoma. J Urol, 1969; 101: 297–301.

Stenzl A, deKernion JB. The natural history of renal cell carcinoma. Semin Urol, 1989; 7: 144–8.

Webber BL, Parham DM, Drake LG, Wilimas JA. Renal tumors in childhood. Pathol Annu, 1992; 1: 191–232.

Wood DJ, Herr HW. The evolving role of surgery in the management of renal cell carcinoma. Semin Urol, 1989; 7: 172–80.