Urothelial tumours arise from the transitional cell epithelium of the urinary tract. Bladder cancer accounts for over 95 per cent of such cancers and can present a formidable challenge to the urologist. It may also cause a patient almost more distress than any other tumour.

There are approximately 45 000 new cases of bladder cancer diagnosed each year in the United States of America and 10 000 patients die of the disease each year. Two-thirds of the patients are male, making it the fourth most common cause of cancer death in men after lung, colorectal, and prostate cancers. In England and Wales about 4500 patients die annually of the disease. In all countries the incidence has been increasing at a rate of about 1 per cent per annum since the 1950s. Most patients are elderly, 75 per cent being over 60 years, and 45 per cent over 70 years.

Bladder cancer was first reported to be associated with certain occupations in 1895, when Rehn noted three cases in dye workers. It is now regarded as a prescribed industrial disease in occupations in which chemicals of the arylhydrocarbon group, including benzidine, &agr;- and &bgr;-naphthylamine, aminobiphenyl, and o-toluidine, are used. These substances were formally used in the paints, dyestuffs, rubber, cable, and chemical industries but none have been used since the mid-1960s. As the latent period between exposure and development of the disease averages 25 years sporadic cases are still seen. In the majority of patients the exact causative agent is unclear, but an average consumption of 15 cigarettes a day doubles the risk of developing the disease: about 40 per cent of patients with bladder cancer are smokers. Other recognized causes include chronic over-ingestion of phenacetin-containing analgesics, treatment with the cytotoxic agent cyclophosphamide, and pelvic irradiation. Worldwide bilharzia, caused by infestation of the bladder with Schistosoma haematobium, remains an important aetiological factor. In Egypt 80 per cent of bladder cancers are associated with bilharzia, and in these instances the tumour is squamous cell in origin.

Benign urothelial tumours are extremely rare. An inverted papilloma is occasionally seen in patients with bladder outflow obstruction. It is a benign proliferative tumour most commonly located at the trigone or around the bladder neck.

Malignant tumours are of three main histological types. Transitional tumours account for 95 per cent of urothelial tumours in the western world. Squamous tumours account for only 1 per cent of urothelial tumours in the United Kingdom, but for 80 per cent of bladder cancers in Egypt. These arise in patches of the bladder which have undergone squamous metaplasia as a result of infestation with S. haematobium. These bilharzial tumours are nodular exophytic lesions that are usually well differentiated and carry a low risk of lymph node and distant metastases. In the western world squamous cancers are most often caused by irritation of the mucosa from calculi or indwelling catheters. About 5 per cent of paraplegics develop a squamous carcinoma. Adenocarcinoma is extremely rare, accounting for less than 0.5 per cent of bladder cancers in the United Kingdom. These are associated with persistent urachal tissue or ectopia vesicae.

Transitional cell carcinomas are morphologically one of three types. Papillary tumours (Fig. 1) 1603 account for 70 per cent of bladder cancers. These tend to be pedunculated and grow into the lumen of the bladder. They are usually superficial and well differentiated. Twenty per cent of bladder cancers are solid tumours (Fig. 2) 1604. These may appear sessile or solid with ulceration, necrosis, and calcification. They tend to be invasive lesions and are less well differentiated than papillary lesions. Carcinoma in situ appears as flat, red, velvety patches in the bladder and may be difficult to distinguish from the changes associated with chronic infection or inflammation. Histologically, it is a poorly differentiated transitional cell carcinoma confined to the mucosa. In 90 per cent of patients with carcinoma in situ cytological examination of the urine will reveal malignant cells. The behaviour of carcinoma in situ is difficult to predict, but an important prognostic feature is the presence of irritative voiding symptoms. Only 10 per cent of patients without symptoms show progression, compared to 65 per cent of symptomatic patients.

Sixty per cent of cancers arise on the lateral walls of the bladder and about 30 per cent occur around the base or trigone. Tumours are graded well differentiated (G1), moderately differentiated (G2), or poorly differentiated (G3), according to the degree of cellular anaplasia. There is a strong correlation between tumour stage and grade. Most well differentiated tumours are superficial and most poorly differentiated tumours are invasive. Bladder cancers are staged according to either the TNM classification or the Jewett-Strong-Marshall system ( Table 1 471 and Fig. 3 1605). Tumour stages which have been determined pathologically are prefixed ‘p’ (e.g. T1 becomes pT1).

Tumour stage, tumour grade, tumour size, vascular invasion, and the presence of associated carcinoma in situ are the most frequently used prognostic parameters in superficial bladder cancer. However, some superficial tumours remain localized whereas others that appear identical recur or become invasive. There are few adequate explanations for why some tumours respond to treatment whereas others do not, and why some tumours which initially respond to treatment later relapse. A number of cellular and molecular biological techniques have been used in an attempt to answer these questions.

Loss of expression of blood group antigens on the surface of bladder cancer cells correlates with a poor prognosis, stage for stage, when compared to tumours that continue to express these antigens. Expression of the Thomsen-Friedenreich (T) antigen also correlates with invasion and a poor prognosis. There is a strong correlation between the content of DNA and the stage and grade of tumours. Progression and recurrence appear to correlate with aneuploidy and a high proliferative rate. The development of bladder cancer may be the result of expression of transforming genes (oncogenes) or the result of the loss of expression of tumour suppressor genes. Ras oncogenes have been identified in bladder cancers and correlate with high tumour grade.

Altered expression of the tumour suppressor genes p53 and retinoblastoma gene (RB) has also been shown to correlate with invasiveness. These genetic changes may operate through altered expression of growth factors. Epidermal growth factor receptor level correlates closely with tumour stage, grade, and reduced time to progression.

Painless haematuria is the predominant symptom in over 80 per cent of patients; up to 10 per cent present with clot retention. The symptom complex of bladder irritability with dysuria, frequency, and urgency is seen in 20 per cent of patients: these patients often have invasive lesions or carcinoma in situ. Other patients may present with recurrent urinary tract infections, loin pain secondary to an obstructed ureter, sterile pyuria, or as an incidental finding at cystoscopy. Increasing numbers of patients are identified as a result of routine ‘dipstick’ testing of urine for haematuria: 10 per cent of these patients will have a bladder cancer.

Clinical examination is usually normal, although an abdominal or pelvic mass may be palpable in those with advanced disease.

The clinical problem is usually that of unexplained haematuria. All such patients should undergo urine culture, examination of the urine for malignant cells, cystoscopy, and an intravenous urogram. Cystoscopy is conveniently performed using the flexible cystoscope under local anaesthesia. The intravenous urogram may show a filling defect within the bladder (Fig. 4) 1606, but is used principally to exclude a source of bleeding from the upper urinary tract. Urinary cytology is particularly helpful when carcinoma in situ is suspected: 90 per cent of specimens will be positive. By contrast, cytological examination of the urine will be negative in up to 50 per cent of patients with well differentiated tumours. Trans-abdominal ultrasound of the bladder is a sensitive technique for identifying papillary tumours larger than 1 cm in diameter (Fig. 5) 1607 but is a poor technique for identifying flat bladder cancers.

Patients with tumours which are shown to be superficial on histological examination need no further staging investigations. In patients with invasive disease it is necessary to differentiate between those with localized extension of the tumour and those with advanced pelvic or disseminated disease. This is best achieved by CT (Fig. 6) 1608 which is an excellent technique for judging extravesical extension and offers a definite advantage over clinical staging by bimanual examination, which may understage up to half of all tumours. CT scanning is less effective at differentiating between T2 and T3 lesions. No further advantage is offered by magnetic resonance imaging.

Pelvic lymphadenectomy provides the ‘gold standard’ for pelvic nodal staging and is the only technique for distinguishing pathological from reactive change. Lymphadenectomy is rarely performed, however, other than at the time of cystectomy. In the future, laparoscopic lymphadenectomy may play an increasing role in the staging of patients prior to definitive treatment.

All patients with bladder tumours should initially undergo transurethral resection of the tumour under general or regional anaesthesia. A bimanual examination of the bladder should be performed both before and after the resection in order to stage the tumour clinically. Separate samples of the tumour and its muscular base should be sent to the laboratory to aid the detection of muscular invasion. Random biopsies of apparently normal bladder distant from the tumour should be taken for assessment of associated carcinoma in situ. Further management is then largely based on the stage and grade of the tumour.

This accounts for about 10 per cent of symptomatic bladder cancers. If untreated 50 to 80 per cent of these tumours will progress to invasive disease. The initial treatment is cystoscopic resection of all macrosopic disease. Localized carcinoma in situ may need no further treatment: patients need regular examination of their urine for malignant cells and cystoscopic surveillance.

Widespread or recurrent localized carcinoma in situ can be treated by intravesical instillations of Bacillus Calmette-Guerin (BCG), given weekly for 6 weeks, with a further 3-week course after 3 months. This treatment is highly effective and has superseded treatment with other intravesical agents such as thiotepa, epodyl, and mitomycin C. Side-effects of BCG treatment are common: about one-third of patients experience an increase in frequency and dysuria. This usually improves spontaneously, but may be helped by oxybutynin 5 mg three times daily by mouth. Five per cent of patients develop signs of systemic BCG infection and require antituberculous medication. The treatment produces resolution of symptoms and complete regression of the tumour for more than 2 years in 70 per cent of patients, and 50 per cent remain free at 5 years. The role of serial treatments beyond the first 6 months, in an attempt to prevent recurrence of carcinoma in situ, is unclear. The mechanism of action of BCG is thought to be immunological rather than directly cytotoxic, and as such represents a landmark in immunotherapy for cancer. Patients who fail to respond to BCG or who relapse on treatment should be offered cystectomy.

Seventy per cent of bladder cancers are superficial to the muscle layers of the bladder. The majority are well differentiated. Around 60 per cent of patients develop a recurrence and 20 per cent develop invasive disease, making cystoscopic surveillance essential. Cystoscopies should be performed 3-monthly in the first year, 6-monthly in the second year, and yearly thereafter.

The findings at the initial 3-month cystoscopy give a useful guide to prognosis. Eighty per cent of those in whom no recurrence is identified at 3 months have no further recurrence. By contrast, only 10 per cent of those patients noted to have a recurrence at 3 months have no further recurrence. Recurrences should always be biopsied as 20 per cent of superficial tumours progress to become invasive.

These comprise 35 per cent of all bladder cancers and they are usually well differentiated (G1). Only 5 per cent of pTa tumours progress to invasive disease and the 10-year survival rate is 90 to 95 per cent.

These comprise 30 per cent of bladder cancers. The outlook is less favourable than for pTa tumours with a 10-year survival rate of 60 per cent. More of these tumours are graded G2 or G3. The pT1 G3 tumour is particularly sinister: up to 50 per cent progress to invasive disease. In patients with pT1 G3 tumours a more aggressive approach to treatment, with adjuvant intravesical BCG or cystectomy, is probably appropriate.

Patients with multiple superficial tumours, tumours that have recurred at each regular cystoscopy, or with carcinoma in situ adjacent to the primary tumour require adjuvant intravesical therapy after transurethral resection. Current evidence suggests that BCG is the agent of choice. The treatment regimen for BCG in superficial bladder tumours is the same as that used for carcinoma in situ. The recurrence rate at 2 years can be reduced from around 60 per cent with no treatment to around 20 per cent after treatment with BCG.

Patients who fail to respond to BCG or who develop recurrences after treatment can be offered cystectomy. Superficial bladder tumours do not respond to external beam radiotherapy.

Invasive bladder cancer carries a poor prognosis, conventional treatment offering a 5-year survival of 50 per cent or less. At centres in Europe attempts at cure have centred largely on bladder preservation, whereas in the United States of America bladder reconstruction has found favour. The goal in bladder preservation is eradication of the cancer with maintenance of satisfactory native bladder function. Only if the cancer is not cured is excisional surgery considered. The reconstructive approach involves cystectomy, the urine flow being diverted or a neobladder being constructed.

Transurethral resection alone is inadequate treatment for most invasive bladder cancers, although some patients with T2 tumours can be cured. Overall, 5-year survival of patients with T2 tumours treated by transurethral resection alone is about 40 per cent. These poor results may be the result of inadequate clearance, understaging of tumours, or early blood-borne spread of tumour. In view of these disappointing results most urologists treat T2 tumours more aggressively.

Radiotherapy has been the mainstay of the treatment of invasive bladder cancer in Europe. However only 25 to 30 per cent of patients are cured by radiotherapy. Patients with less advanced tumours fare better: 5-year survival averages 40 per cent in patients with T2 and 35 per cent in those with T3 tumours. The prognosis in patients with T4 lesions treated by radiotherapy is dismal, with only 5 per cent surviving for 5 years.

Radical external beam radiotherapy comprises approximately 55 Gy given in 35 fractions over 7 weeks. Side-effects are common, about 75 per cent of patients experiencing worsening frequency and dysuria. Persistent disabling radiation cystitis occurs in about 5 per cent of patients. Following a course of radiotherapy patients require regular cystoscopic surveillance. If recurrent tumour is identified a salvage cystectomy can be considered, although this is not feasible in many patients because of debility or the advanced nature of the tumour. In those proceeding to cystectomy the overall 5-year survival is around 40 per cent. In those patients in whom no residual tumour is found in the resected specimen 5-year survival improves to 70 per cent. Five-year survival is 50 per cent if superficial tumour is found in the specimen and only 20 per cent if invasive tumour is found.

In view of the disappointing results of radiotherapy in the treatment of invasive bladder cancer surgeons, particularly in the United States of America, have increasingly turned to cystectomy as primary treatment. In men, the standard operation is radical cystoprostatectomy, including pelvic lymphadenectomy. Urethrectomy is only performed if biopsies confirm involvement of the prostatic urethra. Recently, nerve-sparing cystoprostatectomy has been described: this has the advantage that sexual function is likely to be preserved. It remains to be seen whether the long-term results of this operation match those of the standard radical cystoprostatectomy.

In women the bladder is removed en bloc with the uterus, fallopian tubes, ovaries, and anterior wall of the vagina. A urethrectomy is performed routinely as over one-third of women with invasive bladder cancer have urethral disease.

There are essentially two approaches to dealing with the urinary stream.

Traditionally, an isolated segment of ileum is used as a urinary conduit to an abdominal wall stoma. Urine drains continuously and is collected in a bag attached to the skin around the stoma. More recently continent urinary diversions have been described. In these procedures an intra-abdominal pouch is fashioned from ileum (the Koch pouch) or the ileocaecal segment (the Indiana pouch). The pouch can then be catheterized intermittently by the patient through the abdominal stoma.

A neobladder is fashioned by forming a pouch of ileum or sigmoid which can then be sutured to the internal urethral meatus. The patient is continent because the external urinary sphincter is preserved. Patients either void spontaneously by use of the Valsalva manoeuvre or catheterize the pouch intermittently via the urethra. A major advantage is the avoidance of an abdominal stoma. The best 5-year survival rates that might be expected following radical cystectomy for invasive bladder cancer are 60 to 70 per cent for T2 tumours, 50 per cent for T3a tumours, 25 per cent for T3b tumours, and only 5 to 10 per cent for T4 tumours. Superficially these results appear superior to those that can be achieved with radiotherapy, particularly with less advanced tumours. However, it is very difficult to make direct comparisons between surgically staged cases and clinically staged cases. Clinical staging undoubtedly understages many T2 tumours.

The generally poor results of both radiotherapy and surgery in the treatment of invasive bladder cancer are explained by the presence of unrecognized distant disease at the time of the initial treatment. Accordingly urologists are turning increasingly towards chemotherapy in the treatment of invasive bladder cancer.

Single agent treatment has so far proved disappointing but more recently the introduction of multiagent regimens has brought fresh hope. In particular methotrexate, vinblastine, Adriamycin, and cisplatin shows great promise. Such treatment prior to cystectomy can lead to down-staging of the tumour in 65 per cent of patients. Twenty-five per cent of those proceeding to a cystectomy have no evidence of tumour within the resected specimen.

It is at present unclear which patients should be treated with chemotherapy, whether all patients treated with chemotherapy should then proceed to cystectomy, or whether chemotherapy should be offered only on a selective basis following cystectomy.

The treatment of disseminated bladder cancer is essentially palliative as patients have a median survival of only 6 months. Multiagent chemotherapy as above can double the median survival, and about 15 per cent of patients survive 2 years, but it carries a significant morbidity in patients who may already be very weak.

Urothelial tumours of the upper urinary tract are rare, accounting for only 5 per cent of all urothelial tumours. In England and Wales around 800 cases are diagnosed each year. Urothelial tumours of the renal pelvis are four times more common than those of the ureter and account for 5 per cent of all renal tumours. Men are twice as susceptible as women and the peak incidence is in the seventh decade.

Aetiological factors are similar to those in bladder cancer: cigarette smoking accounts for about 50 per cent of cases. Seventy per cent of patients with a analgesic induced nephropathy develop a transitional cell carcinoma of the upper urinary tract. In the Balkans one-half of all renal tumours are urothelial and are associated with Balkan nephropathy.

Ninety per cent of upper tract urothelial tumours are transitional cell carcinomas. In 2 to 5 per cent of patients the tumours are bilateral.

Upper tract urothelial tumours are associated with carcinoma of the bladder: 3 per cent of patients with a bladder cancer develop a tumour of the upper urinary tract at some stage, and 50 per cent of patients with an upper tract urothelial tumour develop bladder cancer.

Transitional cell carcinoma of the upper urinary tract may spread by direct invasion, or by invasion of lymphatics, blood vessels, or adjacent mucosa. Mucosal invasion probably accounts for the high incidence (30 per cent) of stump recurrences following conservative surgery.

The system of staging is similar to that used for bladder cancer (Table 2) 472. Stage and grade are closely correlated. Median survival is 6 years in those with low grade tumours and 1 year in those with high grade tumours. Median survival is 8 years in those with low stage tumours (Ta or T1), but only 1 year in those with advanced disease (T2–4).

Seventy-five per cent of patients present with haematuria, usually throughout the stream. Patients may occasionally pass a clot cast of the ureter. One-third of patients have loin pain due to obstruction of a ureter.

Since the clinical problem is usually that of unexplained haematuria, patients undergo the same investigations as those with suspected bladder cancer. The intravenous urogram will usually show a filling defect in the renal pelvis or ureter. The differential diagnosis of the filling defect includes overlying bowel gas, radiolucent stone, blood clot, or a sloughed papilla. If doubt exists as to the nature of the filling defect a retrograde ureterogram can be performed.

Optional investigations which may be useful include analysis of ureteral urine samples for malignant cells, and brush biopsy of a suspected tumour. Ureteroscopy and biopsy is used increasingly in the diagnosis of upper urinary tract tumours but should probably be reserved for those patients in whom there is diagnostic doubt or in whom the management might be altered by the findings. By providing tissue for histological examination ureteroscopic biopsy may allow detection of patients with low stage or grade lesions suitable for conservative surgery.

Computed tomography provides the best non-invasive means of staging the tumour preoperatively.

The traditional treatment for upper tract urothelial tumours is nephroureterectomy with excision of a cuff of bladder around the ureteric orifice. The justification for this radical treatment has been the reported high incidence (30–75 per cent) of ureteral recurrences in patients treated by more conservative surgery. In patients treated by nephroureterectomy 5-year survival is around 90 per cent for stage Ta, 45 per cent for stage T1, 23 per cent for stages T2/3, and 0 per cent for stage T4.

More conservative surgery is increasingly finding favour for patients with tumours of low grade or stage, in whom the results of local excision are equivalent to those achieved with nephroureterectomy. Conservative surgery is technically easier for tumours of the ureter than for tumours of the renal pelvis. Conservative surgery is particularly suitable in patients with a tumour in a solitary kidney or with bilateral tumours. Ureteroscopic excision or ablation may have an increasing role in the patients selected for conservative surgery.

The radiotherapy and chemotherapy regimens for the treatment of urothelial tumours of the upper urinary tract are similar to those used in bladder cancer. Each may have some benefit but a definitive role has yet to be established for either.

Birch BRP, Harland SJ. The pT1 G3 bladder tumour. Br J Urol, 1989; 64: 109–16.

Bloom HJ, Hendry WF, Wallace DM, Skeet RG. Treatment of T3 bladder cancer: a controlled trial of preoperative radiotherapy and radical cystectomy versus radical radiotherapy. Br J Urol, 1982; 54: 136–51.

Fitzpatrick J. Superficial bladder cancer—natural history, evaluation, and management. AUA Update, 1989, 8: lesson 11.

Herr H. Intravesical therapy for superficial bladder cancer. AUA Update, 1989; 8: Lesson 12.

Koch NG, Nilsson AE, Nilsson LO, Norten LJ, Philipson BM. Urinary diversion via a continent ileal reservoir. J Urol, 1982; 728: 469–75.

Raghavan D, Shipley W, Garnick M, Russell P, Richie J. Biology and management of bladder cancer. N Engl J Med, 1990; 322: 1129–38.

Sagalowsky A. Continent urinary diversion. AUA Update, 1987; 6: Lessons 2 and 3.

Sarosdy MF, Lamm DL. Long-term results of intravesical BCG therapy for superficial bladder cancer. J Urol, 1989; 142: 719–22.

Solsona E, Borra II, Ricos JV, Monros JL, Dumont R. Feasibility of Transurethral resection for muscle infiltrating carcinoma of the bladder: a prospective study. J Urol, 1992; 147: 1513–5.

Sternberg C, Yagoda A, Scher H, Watson R, Herr H, Morse M. M-VAC (methotrexate, vinblastine, adriamycin, cisplatin) for advanced transitional cell carcinoma of the uroephithelium. J Urol, 1988; 139: 461–9.

Thompson I, Fair W. The epidemiology of bladder cancer. AUA Update, 1989; 8: lesson 27.