papillary urothelial neoplasm of low malignant potential.
Bladder cancer is the fourth most common cancer in men in the USA and the eight most common cancer in women, with an estimated 57 400 cases being diagnosed in 2003, resulting in 25 100 deaths . Commonly accepted risk factors for TCC of the bladder include cigarette smoking, occupational exposure to aniline dyes, benzidene compounds, analgesic abuse (phenacetin) and chronic irritation such as indwelling catheters. Most cases of bladder cancer are superficial at the time of diagnosis (stage Ta–T1). The recurrence rate of superficial tumours can be as high as 70%, with 10–15% progressing to muscle-invasive disease. The risk of progression is directly related to tumour grade and stage.
GENETICS AND DIAGNOSIS
Several tumour-specific genetic abnormalities have been identified with various stages of TCC. Abnormalities on chromosome 9q [2,3] have been reported quite consistently in papillary tumours and with dysplasia. Loss of 5q, 11p, 17p (which contains the p53 gene) and p53 mutations have been noted with in tumours which have progressed to muscle-invasive disease .
The traditional methods of diagnosing TCC have been urine cytology and cystoscopy. Voided urine cytology has a high specificity of 90–100%, but the sensitivity is low, particularly for low-grade tumours (Table 1) [4,5]. New urine-based markers are being developed to enable earlier and easier detection of TCC. Most of these markers are aimed at improving the sensitivity of voided urine cytology while maintaining the same high specificity. The UroVysion test (Vysis Inc., Downers Grove, IL), commonly referred to as the ‘FISH’ analysis, is aimed at detecting and quantifying the number of copies of chromosomes 3, 7, 17 and 9p21 in exfoliated bladder cancer cells in the urine by fluorescence in situ hybridization. A positive UroVysion test is defined as at least five cells with gains of at least two chromosomes, and has a sensitivity of 81% and specificity of 96%, which is better than that of markers like the bladder tumour antigen (BTA) test, telomerase assay and haemoglobin dipstick . Commercially available markers such as the BTA detect and measure levels of a complement-related protein in the urine . The NMP22 test measures urinary levels of a nuclear matrix protein (NUMA) that is present in most cancer cells. Both of these tests have a high sensitivity for detecting low-grade bladder tumours but their specificity is lower than that of urine cytology . Other promising markers include survivin, telomerase detected by the TRAP assay, urinary levels of cytokeratin (CYFRA 21–1 and 18–8), hyaluronic acid and hyaluronidase. A comparison of the sensitivity and specificity of these tests is listed in Table 1. The combined hyaluronic acid/hyaluronidase is particularly promising as it has a high sensitivity and allows stage discrimination, because hyaluronic acid is more sensitive for low-grade and hyaluronidase more sensitive for high-grade tumours . BLCA-4, a bladder cancer-specific nuclear matrix protein, has a sensitivity of 96% and a specificity of 100% which, although reduced to 81% in patients with spinal cord injury and other bladder conditions, is still higher than that of most other urine markers . Survivin, an anti-apoptosis protein, can be detected at higher levels in the urine of patients with bladder cancer; it has a sensitivity of 100% and specificity of 95% for detecting either new or recurrent cancers . Other markers which are aimed at predicting the prognosis, e.g. endothelial growth factor, p53 and Her 2-neu, have also been described as being useful in bladder cancer but these markers are more applicable to invasive disease.
Table 1. A comparison of the sensitivity and specificity of various urine-based markers for bladder cancer (adapted with permission from [4,5]
One of the more common problems affecting pathological staging of TURBT specimens is the absence of muscle in the resected specimen. This precludes an accurate assessment of muscle-invasive disease. Repeat TURBT to obtain adequate muscle from the base of the previously resected area is required in these cases. Repeat TURBT can result in a higher stage assignment in 29% of patients, 22% of patients with T2 disease might be pT0, and management could be changed in 33%. Random bladder biopsies are taken routinely in patients, particularly if there is suspicion of carcinoma in situ (CIS). In cases with subepithelial invasion it is necessary to distinguish between the muscularis mucosae of the lamina propria and the true muscularis propria. Stage subclassification of T1 disease has been proposed, as stage T1a with tumour extending into the superficial lamina propria above the muscularis mucosae, and stage T1b with tumour extending into the deep lamina propria beyond the muscularis mucosae . There is evidence to suggest that this pathological classification has prognostic implications, with the T1b patients having disease behaving more like that in those with muscle-invasive disease . The presence of vascular invasion in TURBT specimens can also be an ominous sign, increasing the likelihood of disease progression .
The management and monitoring of superficial bladder cancer is largely tied to stage and grade . Most papillary neoplasms, particularly those thought to be benign (papillary urothelial neoplasm of low malignant potential, PUNLMP), or recurrent, documented Ta tumours can be managed conservatively with fulguration, which can be undertaken using a flexible cystoscope in the office. We use this approach in older patients, especially if they have several comorbidities. Almost all tumours, including the PUNLMPs, merit TUR at the time of initial presentation for accurate diagnosis and staging. Intravesical instillation of mitomycin C immediately after TUR can result in significantly lower rates of recurrence for almost all superficial tumours, and currently is our standard practice . Early recurrence rates can be decreased by half (34% vs 16%) after immediate intravesical mitomycin C instillation, while overall recurrence rates are reduced to a lesser extent (54% vs 41%) . Intravesical thiotepa or adriamycin have also been used for this purpose.
Well accepted indications for adjuvant intravesical therapy of superficial bladder cancer include the presence of high-grade disease, multifocal disease, recurrent tumours, CIS, involvement of prostatic ducts and urethral epithelium, and for immediate prophylaxis after resection. There is some controversy about the benefit of intravesical therapy in patients with high-grade stage T1 disease. The incidence of understaging in this group of patients can be as high as 40%, leading many urologists to favour radical cystectomy to optimize the chances for cure. This is another area where re-staging TURBT may have some value. Some reports indicate that adjuvant intravesical therapy after TURBT provides no long-term benefit in avoiding disease progression , while others suggest that it may improve long-term survival . We routinely treat patients with at least one full induction course of intravesical BCG, particularly if there is concomitant CIS. We are more inclined to proceed with cystectomy if the disease recurs after one course of BCG in high-grade T1 disease. Currently the agents most commonly used for intravesical therapy include BCG and mitomycin C. BCG appears to be slightly more effective for treating CIS, while both may have comparable efficacy in all other indications (Fig. 1) . The precise mechanism of action of BCG is uncertain although it is clear that BCG is an immunostimulant. BCG is thought to attach to cells through fibronectin, after which it induces a cellular immune response, with elaboration of several cytokines including interleukin-2, -8 and interferon-γ. These agents are thought to be the mediators of the cytotoxic response. It is possible to salvage patients who have failed or are intolerant to BCG therapy with mitomycin C, and vice versa. Other chemotherapeutic agents that have been used for intravesical therapy include thiotepa, doxorubicin and epirubicin. Mitomycin C appears to be one of the most effective chemotherapeutic agents for intravesical chemotherapy. None of the chemotherapeutic agents used for primary induction therapy have shown any ability to reduce long-term recurrence rates, affect disease progression or improve survival. The efficacy of intravesical mitomycin C can be optimized by administering a dose of 40 mg in 10 mL of sterile water while ensuring that the bladder is empty, the urine is alkaline, and limiting oral hydration. This approach improved the recurrence-free rate at 5 years from 24.6% to 41%, and increased the interval to tumour recurrence from 11.9 to 29 months .
While there are data from randomized studies to support the use of maintenance intravesical BCG in patients to prolong the duration of the response, this can be a cause of substantial toxicity . Even in the studies used to determine the effectiveness of maintenance BCG, the withdrawal rates caused by toxicity were high, with only 16% of 243 patients completing all the eight scheduled courses of intravesical BCG . The optimum maintenance regimen of intravesical BCG is also unclear. The regimen used by Lamm et al. included once weekly × 3 weeks at 3 months after completing an induction course over 6 weeks. This was followed by a weekly dose for 3 weeks at 6 months after completing the induction course, and then a 3-weekly course every 6 months thereafter for 3 years. A recent meta-analysis of data from several randomized trials suggests that intravesical BCG can reduce the risk of progression to T2 disease by ≈ 4% compared with TURBT alone or resection followed by another intravesical therapy. The incidence of progression was 6.4% for papillary tumours and 13.9% for CIS.
Major side-effects of intravesical chemotherapeutic agents such as mitomycin C and doxorubicin are restricted to chemical cystitis. Thiotepa can be absorbed systemically, resulting in myelosuppression. Intravesical BCG can cause short-term cystitis and irritative voiding symptoms in 90% of patients, and this may be necessary for effective action of the agent. More seriously it can cause systemic ‘BCG-osis’ and sepsis, which has resulted in several deaths. Localized BCG infection of the testes and epididymis can occur, requiring systemic BCG therapy including isoniazid, rifampicin, vitamin B6 (pyridoxine) and ethambutol. Systemic symptoms such as fever require suspending treatment temporarily and starting isoniazid therapy if the fever is >38.5 °C. In severe cases which are unresponsive to antituberculous therapy, steroids may be needed. The response or presence of residual disease is typically re-assessed 3 months after a course of intravesical BCG but it is possible that the 3-month evaluation can be eliminated, especially in patients with normal cystoscopy and/or cytology. The efficacy of intravesical BCG may be inhibited by antibiotic therapy with fluoroquinolones, gentamicin or doxycyclines. However, the concerns about the negative effect of concomitant fluoroquinolone administration on BCG viability has not been confirmed in laboratory studies, and indeed in vitro studies show direct cytotoxic effects of fluoroquinolones on TCC cells, and might thus synergistically enhance the cytotoxic effect of BCG .
NEWER AGENTS FOR INTRAVESICAL THERAPY
Salvage regimens to rescue patients in whom intravesical BCG has failed include valrubicin, BCG and interferon-α2b, and gemcitabine. Valrubicin, a drug related to doxorubicin, is approved for use by the USA Food and Drug Administration in the salvage setting after BCG failure, yielding a response rate of 21% at a 30-month median follow-up . In the trial which was the basis for the approval of the drug, 51% of the 79 patients who developed recurrent disease went on to cystectomy. This drug is temporarily unavailable in the USA and the durability of response is so low as to limit its utility.
Interferon-α has a cytotoxic effect on bladder cancer cells which is mediated by macrophage stimulation, cytokine release, enhanced natural killer cell, T and B lymphocyte activity, and a direct antiproliferative effect. Studies of interferon-α as a single agent against bladder cancer refractory to BCG yielded disappointing results. Interferon alone has been used intravesically; doses of 50–1000 MIU of interferon-α have been used, with modest response rates. CIS may not respond as well as papillary tumours. The toxicity is low and consists mainly of influenza-like symptoms. However, when combined with BCG, interferon-α appears to have synergistic activity, and this enhanced effect allows for a reduction in the dose of BCG used, from a third to as low as a 10th of the standard dose. This can limit BCG-related toxicity without compromising efficacy and allow its use even in patients who were previously intolerant to BCG. Using this rationale and a combination of reduced-dose BCG with 50 MIU interferon-α2b, O’Donnell et al. were able to achieve a 53% disease-free rate at 2 years, with 55% of patients who were considered cystectomy candidates surviving with an intact bladder. Randomized phase III trials of the combination of intravesical BCG and interferon-α2b vs BCG alone are in progress. The combined therapy using 50–100 MIU of interferon has shown some efficacy even in aggressive (T1G3 and CIS), BCG-refractory superficial disease. However, a longer term follow-up is needed to assess the durability of the response and effect on disease progression.
Gemcitabine used in a twice-weekly regimen for 6 weeks with a 1-week break after the first 3 weeks has shown good tolerability and may hold promise as a salvage intravesical agent for BCG failures . In that study, 39% of the patients had a complete response, with an additional 22% having negative bladder biopsy but positive cytology. Phase II and phase III trials are currently underway to determine the efficacy of intravesical gemcitabine.
Mycobacterial cell wall extract from Mycobacterium phlei, administered intravesically as an induction regimen for 6 weeks followed by monthly maintenance doses, was effective in patients with CIS. The disease-free rate was 41% at 1 year after initiating the therapy, with minimal toxicity . The cell wall extract can also be used immediately after TURBT. If the therapeutic benefit is confirmed this agent may become the least toxic approach and may also be synergistic with cytokines.
The management of superficial bladder cancer continues to develop. New diagnostic and therapeutic alternatives are emerging rapidly. The availability of many intravesical agents has indeed provided urologists with a better arsenal to treat superficial bladder cancer with a bladder-sparing approach, even in those patients with aggressive superficial disease. However, bladder-sparing approaches to superficial bladder cancer should be advocated judiciously and in the appropriate patient. A risk-based approach is essential and limits must be established for the extent to which a conservative treatment is pursued. The enthusiasm spurred by the obvious benefits of intravesical therapy combined with changes in practice patterns and reimbursement rates, particularly in the USA, may have led to the progressive trend of urologists being less inclined to offer aggressive therapy such as cystectomy even in patients with high-risk superficial disease and in whom intravesical therapy has failed on many occasions . This trend is of concern, given the evidence that earlier cystectomy in patients with high-risk superficial disease is associated with better survival . It is hoped that with the availability of more accurate diagnostic and prognostic markers, we can be optimistic that a more efficient, risk-based strategy will emerge in the near future. The availability of better markers may also allow new and improved monitoring strategies to be devised for bladder cancer.