Delay in the Surgical Treatment of Bladder Cancer and Survival: Systematic Review of the Literature

Transcription

1 european urology 50 (2006) available at journal homepage: Review Bladder Cancer Delay in the Surgical Treatment of Bladder Cancer and Survival: Systematic Review of the Literature Nader M. Fahmy a, Salaheddin Mahmud a,b, Armen G. Aprikian a, * a Department of Surgery (Urology), McGill University, Montreal, Quebec, Canada b Department of Community Heath Sciences, University of Manitoba, Winnipeg, Manitoba, Canada Article info Article history: Accepted May 30, 2006 Published online ahead of print on June 13, 2006 Keywords: Bladder cancer Treatment delay Cystectomy Outcome assessment Abstract Objectives: Eighty per cent of the newly diagnosed invasive bladder tumours are invasive from the outset. Half of these patients already have occult distant metastases reflecting the rapid nature of progression. The aim of the current study was to review the literature to determine if delay in cystectomy leads to worse prognosis and to determine if a possible cutoff point for delay exists, after which a worse outcome would be expected. Methods: We performed a systematic review of publications indexed in Medline and other scientific databases by analyzing types and causes of delay in performing radical cystectomy. Information on the impact of such delays on tumour recurrence and survival was collected and summarized. Papers that described only delay without any outcome correlation were excluded from the study. Results: A total of 13 papers published from 1965 to 2006 were included in this study. Three (23%) papers did not find any correlation between pretreatment delays and survival. Two (15%) papers reported a trend towards worse survival with delay. Eight (62%) papers documented significant association between delay and worse prognosis. Delay influenced survival as an independent variable in two (25%) of these eight papers. In the remaining six (75%) manuscripts, delay was significantly associated with a higher pathologic stage. Conclusions: Although studies on bladder cancer failed to show a linear relationship between delay and prognosis, the majority confirmed that delays are associated with worse outcome. Studies suggested a window of opportunity of less than 12 weeks from diagnosis of invasive disease to radical cystectomy. # 2006 European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. Division of Urology, McGill University, Montreal General Hospital, L8-309, 1650 Cedar Avenue, Montreal, Quebec, Canada, H3G-1A4. Tel ; Fax: address: armen.aprikian@muhc.mcgill.ca (A.G. Aprikian) /$ see back matter # 2006 European Association of Urology. Published by Elsevier B.V. All rights reserved. doi: /j.eururo

2 european urology 50 (2006) Introduction Bladder cancer is the sixth most common noncutaneous cancer in Canada. It is the 8th most common cause of cancer death in males and the 13th in females [1]. While only 20% of muscle invasive bladder tumours present initially as superficial disease and progress with time into invasive tumours, the majority (about 80%) are already invasive tumours at initial presentation. Furthermore, about 50% of bladder tumour patients have occult distant metastases at the time of presentation [2]. These statistics demonstrate the aggressive nature of bladder cancer. Radical cystectomy with urinary diversion remains the gold standard for the treatment of invasive bladder tumours. It has been suggested that delaying radical cystectomy in organ-confined disease is associated with poorer survival [3]. Even though it is well accepted that treatment should be instituted once a diagnosis of cancer has been made, several factors play a role in delaying such treatment. Some of these factors are linked to the health care system [4], while others are patient related [5,6]. The question of whether there is a window of opportunity for the treatment of invasive bladder cancer remains unanswered. We conducted a systematic review of the scientific literature to (1) determine if delay in cystectomy leads to a worse prognosis and (2) determine if a possible cutoff point for delay exists, after which a worse outcome would be expected. 2. Methods We identified relevant studies and abstracts by searching PubMed and Ovid gateway for studies published before January We also searched the Web of Science and the Cochrane Collaboration Controlled Trials Register. We used the following search terms: bladder, urological tumours, survival, death and/or delay. In addition, we screened the bibliographies of identified publications for additional citations. Studies were included if they met the following criteria: (1) The article describes a delay in treatment of bladder cancer by radical cystectomy, and (2) the article includes information on the effect of delay on prognosis. For publications studying the delay in relation to multiple management modalities, if shown, only data concerning radical cystectomy were included. We excluded published abstracts and papers published in languages other than English. Since the goal of this study was to assess association of delay with prognosis, all papers that described only delay without any outcome correlations were excluded. Relevant data were extracted into custom-made spreadsheets. Given the heterogeneity of the reported study populations and the differences in the way delay was described in each study, we could not perform a meta-analysis. Studies were divided on the basis of delay type into the following groups: 1. Delay A = onset of complaints to first general practitioner (GP) referral (Patient and GP delays) 2. Delay B = first GP referral to first hospital appointment (Hospital and Urologist delays) 3. Delay C = first hospital appointment to transurethral resection of bladder tumor (TURBT)/first treatment (Diagnostic delay) 4. Delay D = TURBT to cystectomy/definitive treatment (Treatment delay) 5. Delay E = B + C (Hospital diagnostic delay) 6. Delay F = B + C + D (Hospital treatment delay) The following reported end points were summarized: 1. Pathologic stage 2. Five-year progression-free survival 3. Five-year cancer-specific survival 4. Five-year overall survival 3. Results Results of our search yielded 23 articles, 2 editorial comments [7,8], and 2 letters to the editors [9,10]. Ten papers were excluded from our study, because they did not examine prognosis in relation to delay. The remaining 13 articles [5,6,11 21] described investigations conducted between 1958 and 2002, and were published from 1965 to They featured a total of 7700 patients aged 22 to 92 years. There were no randomized controlled trials. Nine (69%) of these studies were European, three (23%) North American, and one (8%) Japanese. Baseline patient characteristics from the 13 included studies are shown in Tables 1 and 2. The different types of delay investigated with their corresponding minimum and maximum reported values are shown in Fig. 1 and Table 3. There was only one prospective study (Wallace et al. [20]); the remaining 12 papers were retrospective studies. Data used in these studies were obtained from billing databases, cancer registries, and death registries or collected from doctor offices and hospitals. Sample sizes ranged from 50 [14] to 3000 patients [21]. Reported mean follow-up ranged from 33.9 [6] to 50 months [14]. Delay was reported as a categorical variable by all studies. Only one manuscript [18] studied delay as a continuous variable, but there was no significant association with survival. The presence of associated comorbidity has been reported by only five studies [5,6,11,13,15] and has ranged from 15.8% [6] to 25% [13] of their studied cohorts.

3 1178 european urology 50 (2006) Table 1 Baseline characteristics of papers studying delay Authors Liedberg [16] Chahal [11] Wallace [20] Gulliford [13] Mommsen [19] Wallace [21] Pub. year Country Sweden UK UK UK Denmark UK Source CR/DR. 12H/2RT/CR MDs/CR 71H/11RT/CR UH IU/CR Study period Effective sample size Mean age (yr) N/A Male 75% 54% 74% 100% 78% N/A Pathologic T stage T1 T4 T1 T4 Ta T4 Ta T2 T1 T4 T1 T2 Pathologic N stage Nx N0, N+ Nx Nx Nx Nx Comorbidity N/A 20.8% N/A 25% N/A N/A Treatment RC, TURBT RC RC, TURBT RC, TURBT ERT + RC N/A ERT, Chemo ERT, Chemo. TURBT Survival Cancer sp. Overall Overall Overall Overall Overall CR: cancer registry, Chemo.: chemotherapy, DR: death registry, ERT: radiation therapy, H: hospitals, IU: Institute of Urology, MDs: doctors, N/A: not available, RC: radical cystectomy, RT: radiotherapy unit, sp.: specific, TURBT: transurethral resection of bladder tumour, UH: university hospital. Table 2 Baseline characteristics of papers studying treatment delay Authors Mahmud [18] Liedberg [17] May [5] Sanchez-Ortiz [6] Chang [12] Hara [14] Hautmann [15] Pub. year Country Canada Sweden Germany USA USA Japan Germany Source PBD UH/DR 2 UH UH UH UH UH Study period Effective sample size Mean age (yr) 68 a 66 a Male N/A 83% 75% 79% 79% 84% 100% Mean follow-up 78 a N/A N/A (mo) Pathologic T stage Tx T0 T4b T2 T4 T2 T4 T2 T4 T2 T3 T2 T4 Pathologic N stage Nx N0, N+ N0, N+ N0, N+ N0, N+ N0, N+ N0 Comorbidity N/A N/A 21% 15.8% N/A N/A N/A Treatment RC RC + ERT + RC RC RC RC RC RC Survival Overall Cancer sp. Prog. free Cancer sp. N/A Cancer sp. Cancer sp. DR: death registry, ERT: radiation therapy, H: hospitals, IU: Institute of Urology, MDs: doctors, N/A: not available, PBD: provincial billing database, prog.: progression, RC: radical cystectomy, sp.: specific, UH: university hospital. a Median Outcome of delays Delay A (Onset of complaints to first GP referral) The reported median delay was 2 weeks in one study [20]. Patients who had delays longer than 2 weeks had a 5% increase in the incidence of stages pt2 4 ( p = 0.04) [20]. A 34% increase in the incidence of gross hematuria was associated with delayed patients ( p < 0.05). However, the presence of hematuria and its severity did not correlate with survival [20]. After adjustment for case severity, studies revealed a trend towards better survival with delays shorter than 2 weeks ( p = 0.06) [13,20]. This early referral was more advantageous in patients with superficial and less-invasive stages [20] Delay B (First GP referral to first hospital appointment) Median delays ranged from 2 [13] to 4 weeks [20]. There was a 26% increase in the incidence of gross hematuria among the patients who had delays of less than 3 weeks ( p < 0.05) [20]. After adjustment, these shorter delays translated into worse survival for all stages ( p = ) [20]. Patients assigned for admission by their referral doctor had less delays and worse outcome. After adjustment for case severity, this association was not significant ( p = 0.586) [13] Delay C (First hospital appointment to first TURBT treatment) Median delays ranged from 3 [20] to 4.3 weeks [11]. Among the patients who had delays longer than 3

4 european urology 50 (2006) Fig. 1 Definition and duration of reported delays among papers studying delays. * The shaded boxes reveal the different studied delays by each manuscript. The upper portion refers to the start point used for delay calculation, while the lower portion describes the end point of the studied delay. The median, mean, and range of delay reported by each study are also shown. For example, the first studied delay by Chahal et al. was from 1st GP referral to 1st hospital appointment. This delay had a median of 27 days, a mean of 36 days, and a range of days. weeks, there was a 40% increase in the incidence of gross hematuria ( p < ) and a 5% decrease in the incidence of large tumours >2 cm (p = 0.04) [20]. It was reported that patients who had shorter delays, who were admitted after the first consultation, and/ or who had their first cystoscopy performed by a consultant rather than by a trainee had a significantly worse prognosis. However, after adjusting for case severity, this association became nonsignificant [13]. Multivariate analyses showed that this type of delay had no independent effect on survival [13,20] Delay D (TURBT to Cystectomy/definitive treatment) (Tables 2 4) Reasons for delays were reported by only two studies [5,6]. The majority of treatment delays were due to patients appeal for a second or third opinion. Other less-significant causes in order of frequency were clearance from other comorbid diseases, an attempted less-invasive procedure (e.g., TURBT) and, lastly, patients personal choice. Reported median delays ranged from 4.7 [18] to 13.7 weeks [11], whereas mean delay ranged from 7.9 [6] to 66 weeks [15]. All studies have chosen a 3-month period (12 weeks, 90 days) as a cutoff for delay analysis (Table 3). Hara et al. [14] were the first to report a 90-day cutoff for delayed versus nondelayed groups. Others [5,6,12,18] have similarly divided their patients according to a cutoff of 12 weeks (3 months). The reason for choosing a 12- week cutoff was explained only by Sanchez-Ortiz et al. [6] who, because of the low number of patients, chose to combine patient delays with similar hazard ratios into two groups, resulting in <12 weeks and >12 weeks. Patients who were delayed longer than 3 months had a mean age that was significantly higher by 4 [5] to 8 years [14] than the nondelayed group ( p 0.038). It was also shown that patients delayed more than 12 weeks had a 17% [12] to 36% [6] significantly higher incidence of non organ-confined disease (>pt3 and/or N+) ( p < 0.01). When studying Table 3 The reported median, mean and range of treatment delays in days Authors Mahmud [18] Liedberg [17] May [5] Sanchez-Ortiz [6] Chang [12] Hara [14] Hautmann [15] Pub. year Country Canada Sweden Germany USA USA Japan Germany Median N/A 42 N/A N/A Mean N/A N/A Min. N/A Max. N/A Max.: maximum, Min.: minimum, N/A: not available.

5 1180 european urology 50 (2006) Table 4 Reported outcomes of treatment delays >3 months Authors Mahmud [18] May [5] Sanchez Ortiz * [6] Chang [12] Hara [14] Pub. Year Country Canada Germany USA USA Japan pt stage N/A 17% more pt4 ** N/A 27% more pt3 ** 6% more pt3 ** 5% more pt4 ** p = p = 0.01 NS LNI N/A 9% more N+ ** 4% more pn+ ** 17% more pn+ ** 10% more pn+ ** NS p = 0.04 NS pv+ N/A 9% more pv+ ** N/A 27% more pv+ ** NS p < 0.05 NOC N/A 36% more NOC ** N/A p < 0.01 Diversion type a N/A 20% more conduit ** N/A N/A 53% more conduit ** p = p < Independent effects of delay a. 5-yr prog. free survival N/A HR = 1.62 ( ) N/A N/A NS b p = b. 5-yr cancer sp. survival N/A N/A HR = 1.93 ( ) N/A NS b p = 0.05 c. 5-yr overall survival HR = 1.2 (1 1.5) N/A N/A N/A NS b p = 0.05 HR: hazard ratio, LNI: lymph node invasiveness, N/A: not available, NOC: non organ-confined, pn+: positive nodes, prog.: progression, pv+: vascular invasion, sp.: specific. * Delayed group, n = 19. ** than non-delayed group. a Conduit diversions independently correlated with worse cancer specific survival [15]. b Data were not shown. pathologic stage and LN invasiveness separately, a trend towards higher pt stage with longer delays was identified [5,12,14,15]. Studies have shown that patients delayed more than 3 months had a 27% significant increase in pt3 [12] and a 17% increase in pt4 [5] ( p < 0.01) (Table 4). Furthermore, delays correlated with at least a 4% [6] increase in the incidence of lymph node metastases in four studies [5,6,12,14], but this association was significant in only two studies [6,12] ( p < 0.01). Similarly, there was a trend towards higher incidence of vascular invasion among the delayed groups in two studies [5,14], but this trend was statistically significant in only one study, which showed a 27% increase ( p < 0.05) [14]. One study did not observe a correlation of delay with either pt stage or LN invasiveness [17]. Of the five manuscripts [5,6,14,15,17] that investigated cancer-specific mortality, two studies [5,6] showed a trend towards worse survival with a delay of >3 months (12 weeks), but this trend was statistically significant in only one study [6]. Reported adjusted hazard ratios (HRs) of these two papers were as follows; HR = 1.93 ( ) p =0.05 [6] and HR = 1.62 ( ) p = 0.06[5]. Similarly, the paper that investigated the multivariate effect of delay on overall survival has shown that patients delayed >12 weeks were associated with a 20% increase in their risk of dying. HR = 1.2 (1 1.5) p = 0.05[18] Delay E: B + C (Hospital diagnostic delay) Median delay ranged from 48 [13] to 68 days [20]. Patients who had shorter delays showed a trend towards worse survival. After adjusting for case severity, this trend was statistically significant in only one study ( p = 0.01) [20] Delay F: B + C + D (Hospital treatment delay) Multivariate analyses have shown that the odds for delaying hospital treatment for patients aged >65 years were reported to be 2.41 ( ) p < [11]. The odds for delaying treatment for patients with grade 2 tumours was significantly greater than that of patients with grade 3 tumours: OR = 1.78 ( ) p = 0.02 [11]. At 1 year, patients delayed >12 weeks had a statistically better prognosis. At 5 years, there was a 17% survival advantage among the delayed group ( p = 0.09) [9,11]. 4. Discussion Invasive bladder cancers have been shown to possess an aggressive nature. It is intuitively evident

6 european urology 50 (2006) that delaying treatment for such cancers can result in worse outcome. The goal of this study was to perform a systematic review of the literature to identify if pretreatment delays do in fact correlate with poorer prognosis and, if so, to attempt to determine if there is a cutoff for a reasonable delay after which a worse outcome can be predicted. Reported median hospital delays, which partially include urologist delays, doubled in 11 years, going from 14 days [13] in 1991 to 28 days [20] in Reported median treatment delays seemed to decrease in 10 years from 1991 to 2001, going from 26 [13] to 20 days [20], after which a trend towards an increase in the median delay was noticed, reaching 33 days in 2006 [18]. Median total delays seemed to also increase in 10 years, increasing from 45 days in 1983 [19] to 144 days in 2003 [16]. The ideal method to test the outcomes of delay on survival is to have randomized controlled studies. However since it is unethical to delay patients for purpose of research, one has to rely on observational studies that may have inherent major selection biases. However, if several observational studies demonstrate similar results, then perhaps some conclusions can be reasonably drawn. Review of the literature has revealed vast variations among the published studies. These variations included inconsistent studied cohorts, diverse durations and types of delays, variable treatments, dissimilar investigated end points, and use of different statistical methods. Authors describing delays from onset of complaints to first GP referral might be influenced by patient factors or assumptions. Furthermore, most papers did not differentiate between patient and GP delay because of practical difficulties in measuring these delays. Furthermore, some studies have excluded females from their cohorts [13,15]. Others papers [9,11,13,16,17,19 21] not only combined the effects of delay on both superficial and invasive tumours, but also did not differentiate between patients who already had invasive tumours at initial presentation from those who presented initially as superficial disease and then progressed to invasive [22]. In addition, bladder cancer treatment also varied among studies. Although most of the articles reported the outcome of radical cystectomy alone, a couple [13,20] combined delay outcomes for several administered treatments like TURBT, chemotherapy, and/or radiation therapy whether given alone or in combination. Others have integrated neoadjuvant radiation treatments [16,19]. Even survival definition varied among studies, with some reporting overall survival [11,13,18 21], while others expressed cancer-specific mortality [5,6,14 17]. On the basis of these significant variations, it was not possible to reach a consistent definition for delays. The only seeming consensus emerged among five (71%) [5,6,12,14,18] of seven manuscripts [5,6,12,14,15,17,18] that investigated treatment delays. These studies characterized a cutoff of 12 weeks from TURBT to radical cystectomy as the delay after which a worse prognosis could be expected. Results of these studies have shown that delay can alter survival either independently from any other variable or through a significant relation with an intermediate event. For example, treatment delays longer than 3 months were significantly associated with higher incidence of non organ-confined disease [6,15], with more stages pt3 [6,12] or pt4 [5], LNmetastases[6,12], and/or vascular invasion [14]. Furthermore, several studies [5,14,15] have shown that patients who had longer treatment delays had significantly more ileal conduit diversions, which have been shown to be independent factors correlating with worse survival [15]. Even after adjusting for other confounders, multivariate analyses have shown that treatment delays of >12 weeks significantly correlated as an independent variable with higher mortality [6,18]. Only one study [17] reported that the length of treatment delays had no influence on outcome. Conversely, the apparently paradoxical worse prognosis associated with shorter variable types of delays reported by few studies [13,16,20,21] could be explained by a selection bias, that is, patients with higher-stage disease could have been rushed to receive faster treatment and ultimately died earlier than those who performed generally better and had relatively longer delays. Causes for delays were reported by only five studies [5,6,18,19,21] and ranged from patient to provider factors. Some of these delays can be reduced by (1) increasing the level of awareness of patients about early symptoms of bladder cancer such as hematuria, (2) stressing the importance of early referral by GPs to urologists for patients, especially elderly females with hematuria and/or persistent cystitis [19,21,23], (3) initiating hematuria clinics, which has been shown to reduce patient and GP delays [23 26] but not to result in reduction of overall mortality [24], and (4) addressing financial organizational issues behind system and/or provider delays [18,23,27]. Finally, It should be noted that delay is not the only prognostic factor. There are other important

7 1182 european urology 50 (2006) predictors of survival including clinical and surgical factors. 5. Conclusion It is wise to encourage treatment immediately after the diagnosis of invasive bladder cancer is reached. Unfortunately, this is not always possible because of several factors including patient- or health carerelated issues. Although studies have not been able to show a direct linear relationship between delay and prognosis, these studies still suggest the possibility of a safe window of opportunity between TURBT and radical cystectomy. Despite the fact that there is probably no absolute safe period of delay, it is reasonable to postulate that, since a significantly worse prognosis is observed when patients are delayed longer than 3 months, a safe period of delay should be significantly shorter than 3 months. References [1] Canadian Cancer Society, National Cancer Institute of Canada, Statistics Canada, Provincial/Territorial Cancer Registries, Public Health Agency of Canada. Canadian cancer statistics; [2] Raghavan D, Quinn D, Skinner DG, Stein JP. Surgery and adjunctive chemotherapy for invasive bladder cancer. Surg Oncol 2002;11: [3] Chang SS, Cookson MS. Radical cystectomy for bladder cancer: the case for early intervention. Urol Clin North Am 2005;32: [4] Simunovic M, Gagliardi A, McCready D, Coates A, Levine M, DePetrillo D. A snapshot of waiting times for cancer surgery provided by surgeons affiliated with regional cancer centres in Ontario. CMAJ 2001;165: [5] May M, Nitzke T, Helke C, Vogler H, Hoschke B. Significance of the time period between diagnosis of muscle invasion and radical cystectomy with regard to the prognosis of transitional cell carcinoma of the urothelium in the bladder. Scand J Urol Nephrol 2004;38: [6] Sanchez-Ortiz RF, Huang WC, Mick R, Van Arsdalen KN, Wein AJ, Malkowicz SB. An interval longer than 12 weeks between the diagnosis of muscle invasion and cystectomy is associated with worse outcome in bladder carcinoma. J Urol 2003;169:110 5, discussion 115. [7] Cookson M. RE: treatment delay and prognosis in invasive bladder cancer [editorial comment]. J Urol 2005;174:1781. [8] Smith Jr JA. RE: an interval longer than 12 weeks between the diagnosis of muscle invasion and cystectomy is associated with worse outcome in bladder carcinoma [editorial comment]. J Urol 2003;169:115. [9] Chahal R, Harrison SC. RE: an interval longer than 12 weeks between the diagnosis of muscle invasion and cystectomy is associated with worse outcome in bladder carcinoma [letter to editor]. J Urol 2003;170:1327, author reply [10] Erickson D, Kunselman A, Bentley C. RE: an interval longer than 12 weeks between the diagnosis of muscle invasion and cystectomy is associated with worse outcome in bladder carcinoma [letter to editor]. J Urol 2003;170: , author reply [11] Chahal R, Sundaram SK, Iddenden R, Forman DF, Weston PM, Harrison SC. A study of the morbidity, mortality and long-term survival following radical cystectomy and radical radiotherapy in the treatment of invasive bladder cancer in Yorkshire. Eur Urol 2003;43: [12] Chang SS, Hassan JM, Cookson MS, Wells N, Smith Jr JA. Delaying radical cystectomy for muscle invasive bladder cancer results in worse pathological stage. J Urol 2003; 170: [13] Gulliford MC, Petruckevitch A, Burney PG. Survival with bladder cancer, evaluation of delay in treatment, type of surgeon, and modality of treatment. BMJ 1991;303: [14] Hara I, Miyake H, Hara S, et al. Optimal timing of radical cystectomy for patients with invasive transitional cell carcinoma of the bladder. Jpn J Clin Oncol 2002;32:14 8. [15] Hautmann RE, Paiss T. Does the option of the ileal neobladder stimulate patient and physician decision toward earlier cystectomy? J Urol 1998;159: [16] Liedberg F, Anderson H, Mansson A, Mansson W. Diagnostic delay and prognosis in invasive bladder cancer. Scand J Urol Nephrol 2003;37: [17] Liedberg F, Anderson H, Mansson W. Treatment delay and prognosis in invasive bladder cancer. J Urol 2005; 174: , discussion [18] Mahmud SM, Fong B, Fahmy N, Tanguay S, Aprikian AG. Effect of preoperative delay on survival in patients with bladder cancer undergoing cystectomy in quebec: a population based study. J Urol 2006;175: [19] Mommsen S, Aagaard J, Sell A. Presenting symptoms, treatment delay and survival in bladder cancer. Scand J Urol Nephrol 1983;17: [20] Wallace DM, Bryan RT, Dunn JA, Begum G, Bathers S. Delay and survival in bladder cancer. BJU Int 2002;89: [21] Wallace DM, Harris DL. Delay in treating bladder tumours. Lancet 1965;19: [22] Hornak M, Bardos A. Biologic characteristics of invasive bladder cancer: is there a difference between primary and progressive tumours? Rozhl Chir 2004;83: [23] Dickinson AJ, Howe K, Bedford C, Sanders T, Prentice A, Sibley GN. A retrospective study of the investigation and management of muscle-invasive bladder cancer in the South West Region. Br J Urol 1996;77:70 5. [24] Stower MJ. Delays in diagnosing and treating bladder cancer. Br Med J (Clin Res Ed) 1988;296: [25] Turner AG, Hendry WF, Williams GB, Wallace DM. A haematuria diagnostic service. Br Med J 1977;2: [26] Paul AB, Collie DA, Wild SR, Chisholm GD. An integrated haematuria clinic. Br J Clin Pract 1993;47: [27] Bishop MC. The dangers of a long urological waiting list. Br J Urol 1990;65: