David F. Penson, Department of Urology, University of Southern California Keck School of Medicine, Norris Comprehensive Cancer Center, Los Angeles, CA 90089, USA. e-mail: email@example.com
Invasive T1 bladder cancers are potentially lethal tumours with varying degrees of aggressiveness and progression. The management of invasive tumours can be very difficult and includes bladder-sparing with transurethral resection and intravesical therapy, or a more aggressive approach with radical cystectomy. Certain clinical and pathological factors might provide some risk stratification for invasive T1 tumours, and might better direct the physician towards an earlier cystectomy for some patients. This review provides a rationale for a radical cystectomy in patients with high-risk, invasive T1 bladder cancer.
Bladder cancer is an important health issue in the USA, representing the fourth most common cancer in men and eight most common in women. In 2006, it was estimated that 61 420 new cases were diagnosed (44 690 men) with >13 000 deaths from the disease. The chance of a man or women developing bladder cancer during his/her lifetime is ≈1 in 28 and 1 in 88, respectively .
TCC is the most common bladder cancer cell type in the USA. Nearly 80% of patients present with bladder tumours confined to the mucosa or submucosa, i.e. so-called superficial (non-muscle-invasive) bladder cancers. It is commonly thought that superficial tumours are not ‘deadly’ per se, rather a ‘nuisance’ tumour characterized by two key features, i.e. disease recurrence and progression to muscle-invasive disease. Treatment goals for superficial bladder cancer therefore include: (i) preventing tumour progression (as muscle-invasive cancer is considered lethal if not properly treated); and (ii) reducing tumour recurrences and the need for additional therapies (cystoscopy, transurethral resection, TUR, intravesical therapy) and the morbidity associated with these treatments.
Superficial bladder tumours are a heterogeneous group of cancers that include: (i) papillary tumours, limited to the mucosa (Ta); (ii) high-grade and flat, confined to the epithelium (carcinoma in situ, CIS); and (iii) those that invade the submucosa or lamina propria (T1). The natural history of superficial bladder cancer is difficult to predict because the tumours are heterogeneous. Conventional histopathology (tumour grade and stage) has historically, and continues to be, the most used clinical measure to provide some degree of prognostic stratification and to direct therapy. Of superficial bladder tumours, 30% are T1 at presentation; T1 tumours are potentially lethal cancers with varying degrees of aggressiveness and progression rates (30–50%), rates which might even be higher with long-term follow-up [2–7]. The lamina propria, which lies just beneath the bladder mucosa, is rich with blood and lymphatic vessels, channels that allow for haematogenous and lymphatic progression, and metastases. Although T1 tumours are commonly referred to as ‘superficial’, this is a dangerous misnomer that underestimates the malignant and metastatic potential of these lesions. T1 tumours are invasive of the lamina propria (not muscularis propria) and by strict definition therefore, are not superficial tumours, and should not be considered in the same class of truly superficial tumours that only involve the most shallow urothelial layer, and do not invade any of the deeper structures. T1 lesions are usually high-grade (grade 3) and invade the lamina propria layer. It is more appropriate to refer to these lesions as ‘invasive’ tumours, as this better connotes the seriousness of the diagnosis and underscores the metastatic potential of these cancers. Therefore, it is important to acknowledge that these invasive T1 tumours are more than a simple ‘nuisance’ and have the potential to cause significant morbidity and mortality.
The decision to attempt bladder preservation or to use a so-called ‘early’ radical cystectomy (RC) for invasive T1 tumours is challenging. Indeed, it is currently one of the most difficult management issues in bladder cancer. Many patients with invasive T1 disease are appropriately treated with TUR followed by intravesical therapy. BCG is the most effective intravesical adjunct shown to decrease tumour recurrence, and if applied with a maintenance programme can decrease tumour progression . However, some patients with invasive T1 disease might be better treated more appropriately and aggressively with an ‘earlier’ RC . The physician must be sufficiently astute to know when to abandon a conservative bladder-sparing approach and proceed to more aggressive therapy (RC) that might improve survival at the potential cost of increased morbidity. Clearly, there is a compelling argument to take a more conservative approach whenever possible to avoid the potential morbidity and mortality associated with RC and urinary diversion. If we were able to accurately predict which patients with invasive T1 disease would progress to muscle-invasive disease, or which patients could be definitely salvaged with delayed RC, there would be no debate. Unfortunately, despite a sophisticated understanding of the molecular mechanisms and biology of bladder cancer, we remain unable to predict which patients would benefit from bladder-sparing treatment or best be suited for an earlier and more aggressive form of therapy.
Here we review the controversial topic and management of the invasive T1 bladder tumour, focusing on an aggressive surgical approach for the disease, i.e. RC. What are the clinical indications, outcomes and the optimum timing for RC in patients with this cancer? There is no level I evidence or randomized trial to unequivocally direct the physician in this decision process for invasive T1 bladder tumours. However, certain clinical and pathological features might better direct the physician in the appropriate treatment scheme for high-risk, invasive T1 bladder cancers, to maximize clinical outcome.
CONSERVATIVE THERAPY FOR HIGH-RISK T1 BLADDER CANCER
Patients with invasive T1 bladder tumours are generally managed with TUR followed by intravesical therapy. Despite the impressive short-term effect of BCG immunotherapy, in 30–50% of patients it will fail and the tumour will progress. With continued, long-term follow-up, progression rates might be even higher [6,10]. Cookson et al. reported on the natural history of 86 patients with high-risk, Ta, Tis and invasive T1 bladder cancer (median follow-up 15 years) treated with TUR alone or with intravesical BCG. Although not specific for only invasive T1 bladder tumours, over half of patients developed tumour progression and 36% ultimately required RC. Indeed, at 15 years, 34% of patients were dead from bladder cancer and only 27% were alive with a so-called intact, functional bladder. Furthermore, 21% of patients in that study developed upper tract TCC, while 19% developed tumours involving the prostatic urethra . These so-called ‘extravesical’ tumour sites are problematic recurrences, difficult to effectively treat conservatively, and pose further risk and even death to the patient.
Similar results were reported by Shahin et al., who retrospectively evaluated 153 patients with invasive T1G3 bladder cancers. There were two groups in that study, invasive T1 tumours treated with and without intravesical BCG therapy. With a median follow-up of only 5.3 years, there was no significant difference in median time to tumour recurrence (38 vs 22 months) or progression (33% vs 36%) in the treated and untreated groups, respectively. Furthermore, deferred RC was ultimately performed in 29% of the BCG treated and 31% of the TUR-only group. What is disturbingly insightful is that with only a modest follow-up, BCG did not appear to affect disease-specific survival, and that the survival curves continued to decline over time, with an estimated 30% of patients who died from progressive disease at 10 years .
STEPPING BEYOND CONSERVATIVE THERAPY FOR ‘INVASIVE’ T1 BLADDER CANCER
Bladder preservation with effective TUR and intravesical therapy is certainly preferable to exenterative surgery, assuming similar cancer-specific and overall survival can be achieved. However, the desire to preserve the bladder must be weighed against potential under-treatment and subsequent cancer-related morbidity and mortality. Several important points need to be considered when counselling the patient with invasive T1 bladder cancer who has this potentially life-threatening condition, including: (i) the significant risk that patients with invasive T1 disease are clinically understaged at the time of TUR; (ii) continued concern about tumour recurrence and progression if the native bladder remains in place; (iii) the marked improvement in surgical and anaesthetic techniques, resulting in a safer RC with acceptable long-term quality-of-life and functional outcomes and; (iv) the potential for better oncological outcomes with early RC in patients with invasive bladder cancer.
Despite the best clinical efforts with advanced histopathological evaluations and imaging techniques, significant clinical staging errors remain. Clinical staging errors range from 24 to 62% for T1 bladder tumours [11–16]. Indeed, in a large group of patients with T1 bladder cancer who had RC, 34% were pathologically understaged (only half had organ-confined tumours) and 14% had metastatic disease to the regional lymph nodes at surgery . Furthermore, patients who had RC for clinically invasive T1 disease and who were clinically understaged had a significantly worse survival (Fig. 1).
The incidence of nodal involvement in patients treated with RC for invasive T1 bladder cancers is reportedly 12–15%[11,16–20]. In one study, the number of TURs before RC for non-muscle-invasive TCC correlated with the prevalence of nodal involvement, with 8% of patients undergoing one TUR vs 24% undergoing four TURs . The authors suggested that inappropriate delays and inadequate staging of high-grade, non-muscle-invasive tumours must be avoided. Further evidence comes from Bianco et al., who reported the clinical outcomes of RC for 66 patients with invasive T1 bladder cancer. There was clinical understaging in 27%, and 12% of patients had nodal involvement. These authors commented that invasive T1 tumours have an insidious clinical course, as shown by staging discrepancies and that RC should be carried out early, rather than later, for these cancers.
Collectively, the results of these studies show the limitations of current clinical staging methods for invasive T1 bladder cancers. Such inaccuracies lead to delays in definitive therapy and ultimately undermine optimal cure rates. These data support the role of aggressive therapy (early RC) in patients with invasive bladder cancer.
CONTINUED RISK OF TUMOUR RECURRENCE AND PROGRESSION WITH BLADDER-SPARING THERAPY
Even if clinical staging is correct and the patient truly has T1 disease, the risk of tumour recurrence and progression persists if the bladder remains in place, not only in the bladder, but also in the upper tract and urethra. Many patients with invasive T1 disease have associated CIS, which predisposes to tumour recurrence in the bladder and the aforementioned ‘sanctuary’ extravesical sites. Despite intravesical therapies, the potential for tumour progression, death and treatment related side-effects remain.
The risk of extravesical recurrence after BCG therapy for patients with invasive T1 bladder cancer is significant [10,11,19–22]. Prostate involvement in these patients might not only be underestimated (25%) but also difficult to treat with conservative measures. Huguet et al. retrospectively evaluated 62 patients in whom intravesical BCG therapy failed, and who then had RC. As expected, 27% of patients were clinically understaged at surgery. The authors found that prostate involvement at the time of endoscopic staging before RC (23%) was the most significant factor associated with clinical understaging, and related to a worse survival. Herr and Donat  reported tumour involvement of the prostatic urethra in 39% of 186 men with invasive T1 and Ta bladder tumours treated with TUR and intravesical therapy. Over a third of these patients had invasive stromal tumour involvement of the prostate, which is associated with an even worse prognosis. Upper tract and prostatic tumour recurrences were also frequent, as reported by Stockle et al.. In that study, 13% of 98 patients treated with BCG for superficial TCC of the bladder developed upper tract disease and 21% prostatic tumour involvement. Similar findings were reported previously .
These observations suggest that patients with high-risk, non-muscle-invasive bladder cancer (including T1 tumours) are at risk of developing recurrences outside the bladder that involve the prostate and upper urinary tract. These patients might be better treated with RC, which removes not only the urothelium at risk in the bladder, but also the prostate and distal ureters, sites where a recurrence is potentially dangerous, difficult to detect and treat with conservative measures.
DECREASED MORBIDITY AND IMPROVED QUALITY OF LIFE WITH RC
Traditionally, providers have been reluctant to proceed with RC due to relatively high postoperative morbidity and mortality rates associated with the procedure, and concerns about the negative effect of the urinary diversion and the impact on quality of life. While these issues remain a concern, they are less so than in the past, with the current application of orthotopic neobladder reconstruction and nerve-sparing techniques. The operation is safer than it was previously and is therefore more acceptable for patients who could derive a real survival benefit. Most patients with invasive T1 disease requiring RC can safely have an orthotopic neobladder constructed, with excellent functional outcomes. Most patients with invasive T1 bladder cancer have no urethral involvement or bulky local disease which might preclude them from undergoing this procedure. Furthermore, men with invasive T1 disease might also be reasonable candidates for nerve-sparing approaches, which offers the possibility of retained erectile function. We believe that the application of orthotopic diversion and nerve-sparing techniques have improved the quality of life and arguably lessened the social impact of RC, making RC less morbid. Patients appear to adjust very well after surgery and can resume a quality of life similar to that before RC.
POTENTIAL FOR IMPROVED SURVIVAL WITH EARLY RC
In the end, the most important outcome to patients with invasive T1 bladder cancer is survival. This begs the question; does early RC improve survival in these patients? Limited data suggest the answer to this is yes. In 1987, Stockle et al. reported the survival rates of patients with invasive T1 bladder tumours who had either immediate or delayed RC. The 5-year recurrence-free survival in the immediate RC group was 90%, vs 62% in the delayed group. Although not randomized, these data suggested the potential for improved survival with an early RC.
Further support for earlier RC comes from Herr and Sogani , who reported on 90 patients with high-risk, superficial bladder cancer. In these patients, all were treated initially with TUR and BCG, and then had RC for recurrent disease during a 15–20 year follow-up. Although not specific for invasive T1 bladder cancers only, patients who had RC within 2 years of the initial BCG therapy had a 92% survival, vs only 56% survival if treated after 2 years. There was also a survival difference if RC was done within ≥1 year after BCG (75% vs 34%, respectively). When the RC was for a non-muscle-invasive (superficial) recurrence, survival was also improved (83%) compared to that in patients with a muscle-invasive recurrence (27%). Furthermore, Solsona et al. reported on 46 patients with high-risk, superficial bladder cancer (refractory to TUR and intravesical therapy) who then had RC. The progression-free survival rate was only 54%, and no different from that in patients undergoing RC for clinically muscle-invasive tumours. There was a high incidence of prostatic stromal and extravesical tumour involvement after RC in these patients with superficial bladder cancer. The authors noted that the criteria for RC were inappropriately too late for this group of patients and that the response to intravesical therapy evaluated at 3 or 6 months seems to be an important factor when considering an early RC . It is clear that we do not need to wait until a patient with a history of high-risk invasive T1 bladder cancer has muscle invasion before recommending RC; this often is too late.
WHICH PATIENTS ARE MOST LIKELY TO BENEFIT FROM EARLY RC
There is no doubt that certain patients with high-risk invasive T1 bladder tumours have a more malignant phenotype and might derive significant benefit from an aggressive early extirpative surgical approach for their disease. The challenge is to identify which patients with invasive T1 disease are at greatest risk of clinical understaging and/or disease recurrence or progression. Although we feel that RC should be discussed with all patients who present with invasive T1 disease, we think that there should be a selective approach to this application of aggressive surgery, with some patients more strongly encouraged to consider RC than others. Risk assessment is most imperative in patients with invasive T1 disease. RC is indicated in patients with uncontrollable or endoscopically unresectable T1 disease, or those who progress to muscle invasion. Otherwise, the ability to precisely identify which patients with invasive T1 bladder cancer need an earlier RC and the optimum time to proceed to RC is more challenging. Several clinical and pathological factors might help direct this decision process.
PRESENCE OF CIS
The presence of CIS is an important risk factor in patients with invasive T1 bladder tumours. This prognostic relationship was examined by Masood et al. in patients with high-grade invasive T1 tumours. The presence of CIS led to pathological upstaging in 55% of RC specimens, compared to only 6% of T1 tumour with no CIS . Furthermore, Solsona et al. reported on 191 patients with invasive T1 disease (with and without CIS) and found that patients with associated CIS who had RC were at risk of progression. Others confirmed the ominous association of CIS and invasive T1 bladder cancer at the time of RC [18,26].
To better assist in risk stratification for invasive T1 bladder tumours, microstaging, or determining the depth of invasion of the lamina propria, has been proposed. The muscularis mucosa is a pathological landmark to help determine the level of penetration of an invasive T1 bladder tumour. Although some argue that microstaging is not a reproducible measurement and that the muscularis mucosa might be difficult to locate, several reports showed a worse prognosis for T1 tumours with deeper lamina propria invasion [26–30].
Recently, Orsola et al. evaluated the depth of lamina propria invasion in 97 T1 bladder tumours. T1 subclassification or microstaging was possible in 87% of cases. The progression rate for superficially invasive (T1a) tumours was only 8%, vs 34% with deep lamina propria invasion (T1b/T1c). These findings were consistent regardless of BCG treatment. Furthermore, the presence of CIS was also associated with significantly higher progression rates in patients with invasive T1 tumours . Similar results were reported by Holmang et al., who sub-staged 121 patients with high-grade, T1 bladder tumours. Overall, 94% of patients were appropriately classified, with 58% progressing to muscle-invasive disease with deep penetration (T1b/c), vs only 36% with T1a penetration.
The suggestion that microstaging of T1 tumours provides risk stratification is not new. Although increasing evidence suggests that this pathological evaluation might better direct treatment and possibly lead to a more timely RC in patients with deeply invasive T1 tumours, the 1998 Bladder Consensus Conference Committee suggested that ‘substaging based on the relationship of tumour to muscularis mucosa should not be universally adopted or advocated by pathologists’, but recommended that pathologists should be encouraged to provide some assessment of the extent of lamina propria invasion (focal vs extensive) . A collaborative relationship between urologists and pathologists might facilitate the identification of this risk factor in most cases of T1 bladder cancer, that might in turn be used to better optimize treatment recommendations.
To improve clinical staging and to ensure complete resection of all visible disease, the concept of a repeat TUR within the first 4–6 weeks after the initial TUR has gained increasing acceptance . Data suggest that routine re-staging TUR of invasive T1 bladder tumours improves the initial response to intravesical therapy, reduces the frequency of subsequent tumour recurrences and might prevent or delay tumour progression . In one series of 71 patients with invasive T1 bladder cancer managed by an experienced urologist, 25% of cases had residual invasive T1 disease on re-TUR 4 weeks later . In addition, when muscle was not submitted in the TUR specimen, 49% of invasive T1 tumours had an incomplete resection, vs only 14% of invasive T1 tumours when muscle was identified . In a review of reported series of invasive T1 tumours assessed with a re-staging TUR, 15–53% were found to have residual invasive T1 disease and 4–29% were upstaged to muscle involvement (T2) after the re-staging TUR . These data support a re-staging TUR to optimize clinical staging and to identify which patients would be better treated with an early and more aggressive therapy.
Recently, Herr et al. evaluated the pathological findings from a re-staging TUR, to identify high-risk patients with invasive T1 bladder tumours for early tumour progression, and who might benefit from an earlier RC. In all, 352 patients had invasive T1 bladder cancer on initial TUR and were then evaluated by a second/re-staging TUR. All patients received intravesical BCG therapy. Overall, 58% of patients had residual tumour on repeat TUR and 26% had persistent residual invasive T1 disease. As expected, during the 5-year follow-up, 66% of patients had tumour recurrence, with 35% progressing in stage. Most importantly, only 19% of patients with non-T1 tumours on re-staging TUR had tumour progression, vs 82% with residual invasive T1 tumours on re-staging TUR. Re-staging TUR pathology was found to be the most significant and an independent predictor in multivariable analysis for tumour progression in these patients. The authors suggested that immediate RC is advised for patients with residual invasive T1 tumours on contemporary re-staging TUR .
These data suggest that a re-staging TUR improves the quality of the endoscopic evaluation of bladder tumours, improves clinical staging and outcomes in patients with invasive T1 disease. Furthermore, re-staging TUR might better identify high-risk patients with T1 tumour who could benefit from a more timely RC.
OTHER HISTOPATHOLOGICAL VARIABLES
There are some high-risk histopathological variables with invasive T1 bladder tumours that might provide a risk assessment and better direct appropriate therapy. Small-cell (neuroendocrine)  or micropapillary  T1 tumours are rare but highly aggressive histological variants that are best treated with an early RC and possibly adjuvant therapy. Large and multifocal invasive T1 tumours might also be best treated with an early RC. Invasive T1 tumours with obvious lymphovascular invasion  are also high-risk cancers that warrant a more aggressive treatment plan. It is emphasized that the urologist should review the pathological slides in concert with the pathologist, to identify these risk factors and thus better direct therapy.
OUTCOMES AND RESULTS OF RC FOR INVASIVE T1 BLADDER CANCER
RC has become a reasonable option in the treatment of high-risk, invasive T1 bladder cancer; it results in excellent long-term oncological and clinical outcomes. In general, advocates of this aggressive approach argue that the 80% long-term cancer-specific survival seen with RC might decrease to 50% or 60% if it is not done or is unusually delayed. Bianco et al. detailed the surgical management with RC of 318 patients with invasive T1 bladder cancer managed over a 10-year period. As expected, 27% of patients were clinically understaged at cystectomy; 12% of patients with invasive T1 tumours and CIS had pathological nodal involvement, all of whom died within 36 months. Overall, the cancer-specific survival was excellent, at 92% for patients with T1 disease. In a multivariable analysis, only the presence of nodal disease and pathological stage of the primary tumour were significant factors associated with survival. The results were similar in a report evaluating early RC for patients with high-risk superficial bladder tumours . Although not specific for only invasive T1 disease, survival was significantly better when patients had an earlier RC (1 year after BCG therapy) than RC after 1 year. Furthermore, when the RC was early for a superficial recurrence, disease-specific survival was 92% at 10 years, significantly better than a delayed RC for a superficial recurrence (56%) at 10 years . Similar findings were reported by Yiou et al., for the outcomes of three groups of patients undergoing RC; (1) for primary muscle-invasive disease; (2) those with superficial disease who progressed to muscle invasion; and (3) those with high-risk, superficial tumours in whom conservative therapy had failed. The 10-year cancer-specific survival was 48% and 47% for groups 1 and 2, respectively, and 82% for group 3. The authors commented that the prognosis of patients with superficial tumours which progressed was no different from those with muscle-invasive disease at presentation, and that early RC for high-risk superficial disease yields better results .
Although historically those patients with recurrent invasive T1 bladder cancer after treatment with BCG have had bladder-sparing protocols, there is increasing evidence to support an earlier RC in these patients. Raj et al. compared a historical cohort of 85 patients with recurrent invasive T1 bladder cancer who were treated with repeat TUR and intravesical BCG, to 65 with invasive T1 bladder cancer who had an immediate RC at the time of recurrence. At 5 years, the cumulative incidence of death from bladder cancer was 48% in the historical group vs 31% in the early RC group. Despite methodological issues, this retrospective analysis clearly supports an early RC for recurrent, invasive T1 bladder cancer after intravesical BCG. Collectively, these data support the improved oncological outcomes obtained with an appropriately timed RC for high-risk, T1 bladder cancer.
Although it is difficult to determine the ideal time to proceed with RC for patients with invasive T1 bladder cancer, it is important to understand and counsel patients about certain clinical and histopathological characteristics that might better direct treatment decisions. If a patient has an invasive T1 tumour with no high-risk features and intravesical therapy is considered, it is prudent to perform a second TUR. The risk of upstaging on a second TUR is >30% if muscle is present in the specimen, and even higher if muscle is absent. Patients with residual T1 cancer on re-staging TUR are at risk and should consider an early RC. Small-cell neuroendocrine and micropapillary invasive T1 variants are best treated aggressively with RC and even systemic chemotherapy. Patients with multifocal or extensive T1 tumours (with/without CIS) might consider an immediate RC, particularly after failing a course of intravesical BCG therapy. Invasive T1 tumours with deep penetration of the lamina propria approaching muscularis propria, those with prostatic tumour involvement and those with obvious lymphovascular invasion should also consider an early RC. Invasive T1 tumours that recur early (within 3 months) after intravesical BCG might be best treated with RC. It is clear we do not need to wait for muscle invasion before we recommend RC for patients with invasive T1 disease; this is often too late and associated with a worse prognosis. RC provides the best local control of the disease and optimal long-term oncological outcomes for invasive bladder cancer. Advances in orthotopic diversion and nerve-sparing techniques to improve the quality of life after RC have been developed. Patients must be counselled and educated early about these issues, and so should a RC be necessary it does not come as a surprise and lead to unwanted delays in a more definitive therapy.