EUROPEAN UROLOGY 60 (2011)

Transcription

1 EUROPEAN UROLOGY 60 (2011) available at journal homepage: Platinum Priority Urothelial Cancer Editorial by Marko Babjuk on pp of this issue Antegrade Perfusion With Bacillus Calmette-Guérin in Patients With Non Muscle-Invasive Urothelial Carcinoma of the Upper Urinary Tract: Who May Benefit? Gianluca Giannarini, Thomas M. Kessler, Frédéric D. Birkhäuser, George N. Thalmann, Urs E. Studer * Department of Urology, University of Berne, Inselspital, Berne, Switzerland Article info Article history: Accepted July 18, 2011 Published online ahead of print on July 27, 2011 Keywords: Urothelial carcinoma Upper urinary tract Bacillus Calmette-Guérin Follow-up studies Please visit europeanurology to read and answer questions on-line. The EU-ACME credits will then be attributed automatically. Abstract Background: There is paucity of data on bacillus Calmette-Guérin (BCG) perfusion in patients with non muscle-invasive urothelial carcinoma (NMIUC) of the upper urinary tract (UUT). Objective: To assess the long-term results of BCG perfusion in patients with UUT NMIUC in terms of efficacy and tolerability. Design, setting, and participants: Retrospective analysis of 55 consecutive patients (64 renal units [RUs]) with UUT NMIUC prospectively followed according to a standardised protocol for a median of 42 mo (range: mo). Our series includes negatively selected patients, most of whom were not eligible for radical surgery, with additional invasive urothelial carcinoma of the urinary tract in roughly one-third of the cases. Intervention: Antegrade BCG perfusion of the UUT was performed either with curative intent for carcinoma in situ (Tis; 42 RUs) or with adjuvant intent after ablation of TaT1 tumours (22 RUs). Measurements: Primary outcome measures were recurrence-free, progression-free, and nephroureterectomy-free survival. The secondary outcome measure was treatment tolerability. Results and limitations: Recurrence occurred in 30 of 64 RUs (47%), 17 of 42 (40%) with Tis and 13 of 22 (59%) with TaT1 tumours. Progression occurred in 11 of 64 RUs (17%), 2 of 42 (5%) with Tis and 9 of 22 (41%) with TaT1 tumours. Nephroureterectomy was eventually performed in 7 of 64 RUs (11%), 2 of 42 (5%) with Tis and 5 of 22 (23%) with Ta T1 tumours. Patients treated with curative intent for Tis tended to have better recurrence-free survival ( p = 2) and significantly better progression-free survival ( p < 1) and nephroureterectomy-free survival ( p = 5) compared with those treated with adjuvant intent after ablation of TaT1 tumours. Adverse events, mostly minor, occurred in a total of 11 patients (20%), with one case of fatal Escherichia coli septicaemia. Conclusions: In our patients with UUT NMIUC, antegrade BCG perfusion resulted in a high kidney-preservation rate. Patients treated with curative intent for Tis apparently benefited in terms of local disease control more than those treated with adjuvant intent after ablation of TaT1 tumours. Treatment tolerability was good. # 2011 European Association of Urology. Published by Elsevier B.V. All rights reserved. * Corresponding author. Department of Urology, University of Berne, Anna-Seiler-Haus, Inselspital, CH-3010 Berne, Switzerland. Tel ; Fax: address: urology.berne@insel.ch (U.E. Studer) $ see back matter # 2011 European Association of Urology. Published by Elsevier B.V. All rights reserved. doi: j.eururo

2 956 EUROPEAN UROLOGY 60 (2011) Introduction The standard treatment of upper urinary tract (UUT) urothelial carcinoma (UC) is nephroureterectomy with bladder cuff excision [1]. In patients with bilateral tumours, tumours in solitary kidney, and chronic renal failure as well as in those with small, non muscle-invasive, low-grade tumours and a normal contralateral UUT, a kidney-sparing treatment by open, percutaneous, or ureteroscopic approach is an alternative option with acceptable oncologic outcome when appropriately applied [1]. In recent years, the therapeutic armamentarium has been enriched by UUT perfusion with chemotherapeutic agents (mitomycin C, epirubicin, and thiotepa) and immunomodulatory substances (bacillus Calmette-Guérin [BCG] and interferon) [1,2]. These drugs can be administered either as curative treatment for carcinoma in situ (Tis) or adjuvant therapy after ablation of TaT1 tumours. Currently, BCG is the most common drug used for UUT perfusion [1]. Although the efficacy of endocavitary BCG is long established for the curative treatment of primary Tis and in the adjuvant setting after transurethral resection of high-risk non muscle-invasive tumours of the bladder, its use for the treatment of UUT tumours is still regarded as investigational. We started using BCG for UUT perfusion in 1986 [3], and we reported on the first 37 treated patients in 2002 [4]. In the present analysis, including more patients and a longer observation time, we present the long-term results of BCG perfusion in terms of efficacy and tolerability in patients with UUT non muscle-invasive UC (NMIUC). 2. Patients and methods 2.1. Patients Between January 1986 and September 2010, 55 consecutive patients with UUT NMIUC who were mostly unable, unfit, or unwilling to undergo nephroureterectomy with bladder cuff excision received BCG perfusion with either curative intent for Tis or adjuvant intent after ablation of Ta T1 tumours. Ablative treatment consisted of either percutaneous resection or ureteroscopic laser ablation of the tumour after representative biopsy had been taken. All clinical and pathologic data were evaluated retrospectively from a prospectively maintained database. BCG perfusion was performed under conditions that were imperative (bilateral tumours, tumour in a solitary kidney, chronic renal failure), relative (multiple andor severe comorbidities increasing the perioperative risk for major surgery, defined as an American Society of Anaesthesiologists score 3, or the future risk of severe chronic renal failure), or elective (normal contralateral UUT). All patients gave written informed consent. The diagnostic work-up included (1) urine cytology obtained after forced diuresis, which we stimulated with 500 ml fluid intake (40 mg oral furosemide and ambulation), to collect a sufficient number of cells exfoliating from the UUT; (2) cystourethroscopy with bladder barbotage cytology and multiple random biopsies of the bladder mucosa and prostatic urethra (alternatively, urethroscopy with urethral barbotage cytology for patients with previous radical cystectomy); (3) cytology of urine collected by selective retrograde UUT catheterisation; and (4) UUT imaging including excretoryretrograde urography, computed tomography (CT), or magnetic resonance imaging (MRI). Diagnosis of UUT UC was confirmed either histologically or cytologically. The 2002 TNM classification and the 1973 World Health Organisation system were used for stage and grade, respectively. Diagnosis of UUT Tis was made by the following criteria: (1) urothelial cells of malignant type in urine cytology after forced diuresis and selective retrograde UUT catheterisation, (2) negative bladder barbotage cytology and negative multiple random biopsies of the bladder mucosa and prostatic urethra (alternatively, negative urethral barbotage cytology), and (3) no detectable lesions on UUT imaging Treatment Our technique for BCG perfusion was described previously [3]. Briefly, with the patient in the prone position, a 10-F nephrostomy tube was placed with local anaesthesia under ultrasound control. BCG was administered through the nephrostomy tube only after exclusion of relevant bacteriuria and fluoroscopic control of unobstructed UUT outflow. A flask with 360 mg Immun BCG Pasteur F (Institut Pasteur, Paris, France) or 243 mg ImmuCyst (Sanofi Pasteur MSD AG, Baar, Switzerland) dissolved in 150 ml 0.9% saline was placed 20 cm above thekidneylevelofthesupinepatientandinfusedatarateof1mlminto have continuous perfusion of the UUT for 2 h (Fig. 1). After successful voiding and having tapped the nephrostomy tube, the patient was discharged with a prescription of antibiotics for 5 d. A treatment course consisted of six weekly perfusions. After the last perfusion, the nephrostomy tube was removed Follow-up Patients were followed according to a prospective protocol including 3-mo cystoscopy with bladder barbotage cytology (alternatively, urethral barbotage cytology for patients with previous radical cystectomy) and urine cytology after forced diuresis for the first 2 yr, then every 6 mo for the following 3 yr. UUT imaging was performed 1 yr and 3 yr after treatment or whenever clinically appropriate with excretory urography, CT, or MRI. In case of positive urine cytology after forced diuresis, cytology of urine collected by selective retrograde UUT catheterisation as well as appropriate UUT imaging was always performed. If recurrence or progression occurred, therapeutic options, such as no therapy, further kidney-sparing treatments, nephroureterectomy, or systemicchemotherapy, were then individually discussed with thepatient based on clinical status and tumour extent and applied accordingly. [(Fig._1)TD$FIG] Fig. 1 Scheme of our technique of perfusion of the upper urinary tract with bacillus Calmette-Guérin. Reproduced with permission from Elsevier [4].

3 EUROPEAN UROLOGY 60 (2011) Outcome measures and statistical analyses 3. Results Primary outcome measures were recurrence-free, progression-free, and nephroureterectomy-free survival per renal unit (RU) and cancer-specific andoverall survival per patient. Recurrence (including nonresponders) was defined as an ipsilateral UUT tumour of equal or lower stage and grade. Progression was defined as an ipsilateral UUT tumour of higher stage and or grade in any of the subsequent recurrences. Recurrence-free, progression-free, and nephroureterectomy-free survival was calculated as time from diagnosis to date of first recurrence, progression, and nephroureterectomy. Cancer-specific survival was calculated as time from diagnosis of UUT NMIUC to date of death from progressive UC elsewhere in the urinary tract. Overall survival was calculated as time from diagnosis to date of death from any cause. Live patients without recurrence or progression were censored at date of last follow-up, and patients dying before recurrence or progression were censored at time of death. The secondary outcome measure was treatment tolerability. Adverse events were classified according to the Common Terminology Criteria for Adverse Events v.4.0 [5]. The analysis was based on intention to treat. Differences in patient characteristicswere assessed with the Mann-Whitney U orchi-squaretest, as appropriate. Kaplan-Meier estimates were used to explore recurrencefree, progression-free, and nephroureterectomy-free survival as well as cancer-specificandoverall survival. The log-rank test wasused to compare survival rates between patients receiving BCG with curative intent for Tis and those receiving it with adjuvant intent afterablation of TaT1 tumours. All reported p values are two-sided, and statistical significance was set at 5. Statistical analyses were performed in collaboration with the Department of Mathematical Statistics, University of Berne, Switzerland, using a commercially available software. [(Fig._2)TD$FIG] A recurrence-free survival 1.0 progession-free survival 1.0 Median patient age was 69 yr (range: yr). All patients hadbeentreatedpreviouslyforucinthecontralateral UUT (n = 4), in the bladder (n =48), or in both (n =3). Five had undergone a contralateral nephroureterectomy due to UUT muscle-invasive UC and eight had undergone radical cystectomy due to muscle-invasive tumours at a median of 29 mo before BCG perfusion. Eight patients required radical cystectomy at a median of 22 mo after BCG perfusion. Median follow-up was 42 mo (range: mo). A total of 64 RUs were treated. BCG was administered with curative intent to 42 RUs with Tis and with adjuvant intent to 22 RUs after ablation of TaT1 tumours (in the pelviscalices in 4 RUs, in the ureter in 17 RUs, and in both in 1 RU). Nine patients received BCG perfusion of both RUs. BCG perfusion was given under imperative, relative, and elective conditions in 23 RUs (35%), 30 RUs (47%), and 11 RUs (18%), respectively. Recurrence occurred in 30 of 64 RUs (47%), 17 of 42 (40%) with Tis and 13 of 22 (59%) with TaT1 tumours. A second and third BCG course was given to eight and two RUs, respectively, due to Tis recurrence. Progression occurred in 11 of 64 RUs (17%), 2 of 42 (5%) with Tis and 9 of 22 (41%) with TaT1 tumours. Nephroureterectomy was eventually performed in 7 of 64 RUs (11%; of which only 1 had received nephroureterectomy-free survival 1.0 B cancer-specific survival 1.0 overall survival 1.0 BCG with curative intent for Tis BCG with adjuvant intent for TaT1 Fig. 2 Kaplan-Meier estimates for cumulative 5-yr (A) recurrence-free, progression-free, and nephroureterectomy-free survival and (B) cancer-specific and overall survival in patients receiving perfusion of the upper urinary tract with bacillus Calmette-Guérin with curative intent for carcinoma in situ or with adjuvant intent after ablation of TaT1 tumours. BCG = bacillus Calmette-Guérin.

4 958 [(Fig._3)TD$FIG] EUROPEAN UROLOGY 60 (2011) Fig. 3 Rate of (A) recurrence and (B) nephroureterectomy per renal unit in patients receiving perfusion of the upper urinary tract with bacillus Calmette- Guérin with curative intent for carcinoma in situ or with adjuvant intent after ablation of TaT1 tumours after stratification by cystectomy status. BCG = bacillus Calmette-Guérin. BCG perfusion under elective conditions), 2 of 42 (5%) with Tis and 5 of 22 (23%) with TaT1 tumours. Patients receiving BCG with curative intent for Tis tended to have better recurrence-free survival ( p = 2), significantly better progression-free survival ( p < 1), and significantly better nephroureterectomy-free survival ( p = 5) compared with those receiving BCG with adjuvant intent after ablation of TaT1 tumours (Fig. 2). The projected 5-yr and 10-yr recurrence-free survival for patients treated with curative intent for Tis was 57% and 49%, respectively. After stratifying patients by cystectomy status, significantly lower recurrence and nephroureterectomy rates were observed in patients with no previous or subsequent cystectomy receiving BCG with curative intent (6 of 26 and 1 of 26, respectively) compared with those treated with adjuvant intent (7 of 13 and 4 of 13, respectively) ( p < 5 for both) (Fig. 3). Of the 55 patients, 20 are alive with no evidence of disease (at a median follow-up of 41 mo), 17 died due to progressive UC of the urinary tract (at a median of 46 mo), and 18 died due to other causes (at a median of 37 mo). Among survivors, 20 RUs were treated with BCG with curative intent for Tis and 7 were treated with adjuvant intent after ablation of TaT1 tumours. Of patients dying due to progressive UC, only 5 did so possibly because of tumour progression of the RU initially treated with BCG perfusion, and all of them received BCG with adjuvant intent. Most of them (n = 12) probably died from progression of bladder cancer because they had unfavourable tumour characteristics at definitive pathology (non organ-confined tumour with or without lymph node metastases). No significant difference in cancer-specific and overall survival rates was observed between patients receiving BCG with curative intent and those treated with adjuvant intent (Fig. 2). Adverse events occurred in 11 patients (20%). Severity and type were: grade 1 fever in five patients, grade 2 storage lower urinary tract symptoms in five patients, grade 2 pericarditis in one patient, grade 3 haematuria in one patient, grade 3 infection in two patients, and grade 4 and 5 Escherichia coli septicaemia in one patient each. Scheduled treatment was prematurely interrupted in seven patients (13%), five because of adverse events and two due to other causes (persistent nephrostomy malfunction, consent withdrawal). No tumour seeding along the nephrostomy tract was observed. One patient receiving BCG after ablation of multiple Ta tumours of the renal pelvis and ureter and with preexisting moderate chronic renal failure eventually went on dialysis. 4. Discussion The first case of BCG perfusion dates back to 1985, when Herr reported on a patient with a solitary pelvic kidney who had been treated for muscle-invasive UC with wide excision

5 EUROPEAN UROLOGY 60 (2011) Table 1 Case series reporting perfusion of the upper urinary tract with bacillus Calmette-Guérin as treatment of primary carcinoma in situ y Study No. patients No. renal units Route of BCG perfusion BCG dose per single perfusion, mg Positive response * (per renal unit), n (%) UUT recurrence or progression (per patient), n (%) Follow-up, mo Sharpe et al. [8], Retrograde (76) 211 (18) 49 Yokogi et al. [9], Antegrade or retrograde (63) 05 (0) Nishino et al. [10], Retrograde (100) 06 (0) 22 Nonomura et al. [11], Retrograde (82) 29 (22) NR Okubo et al. [12], Retrograde (64) 511 (45) Irieet al. [13], Retrograde (range) 1313 (100) 19 (11) 36 Miyake et al. [14], Antegrade or retrograde (81) 316 (19) 30 Hayashida et al. [15], Antegrade or retrograde 65 (median) 1111 (100) 510 (50) 50.9 Kojima et al. [16], Retrograde (77) 311 (27) 1 76 BCG = bacillus Calmette-Guerin; NR = not reported; UUT = upper urinary tract. y Only series including a minimum of five patients are considered. * Defined as negative cytology. Meanmedian or range. of the renal pelvis and the entire ureter, followed by creation of a pyelovesical anastomosis [6]. Because of positive surgical margins for Tis, the patient underwent a course of six weekly BCG perfusions via the bladder. Urinary cytology remained negative for about 14 mo. In 1987, Smith et al. published the results on a series of nine patients treated with percutaneous resection of UUT NMIUC, of whom five received BCG perfusion in an adjuvant setting [7]. Since then, we and other institutions have reported the use of BCG perfusion, either as curative treatment for Tis [3,8 16] or as adjuvant therapy after ablation of non-muscle-invasive tumours [17]. On the whole, however, < 300 RUs have been reported to be treated with BCG perfusion, and no prospective randomised controlled trial has ever tested its efficacy. To our knowledge, we report on the largest series of patients with UUT NMIUC receiving antegrade BCG perfusion. Notably, our series includes negatively selected patients, most of whom were not eligible for radical surgery and had previously been treated for bladder or contralateral UUT UC, with 29% undergoing radical cystectomy for invasive bladder cancer at some time in their clinical course. The main finding of our analysis is that in this pooroutcome population, kidney preservation could be achieved in 89% of the cases, preventing the development or worsening of chronic renal failure or, even worse, the need for dialysis with its negative consequences. This treatment modality allows for buying time, maintaining quality of life (QoL) due to the retained kidney. This QoL aspect is also of particular importance for patients who are destined to die from metastatic UC (mostly of the bladder). Moreover, a local control of the disease prevents or at least delays the occurrence of complications of UUT tumours, such as bleeding or flank pain. Finally, the preservation of sufficient renal function is a great advantage if platinum-based chemotherapy is required in case of UC progression. Patients receiving BCG with curative intent for Tis seem to profit more than those receiving it with adjuvant intent after ablation of TaT1 tumours in terms of local control of disease and kidney preservation. After excluding subjects with poorer prognosis (ie, those undergoing cystectomy for invasive bladder UC), recurrence, progression, and nephroureterectomy rates were all significantly lower in patients receiving BCG for Tis. The survival was not very different for the two patient categories, possibly due to the overall poor prognosis of the population. The rather low number of events for each outcome measure, however, demands cautious interpretation. Comparisons between the present and other series should be attempted with prudence, given the particular negative selection of our patients and the substantial differences in BCG perfusion technique. In the literature, BCG perfusion for curative treatment of Tis is associated with a uniformly high (63 100%) positive response in terms of short-term normalisation of urinary cytology but with a recurrence or progression rate ranging from 0% to 50% (Table 1). Possible reasons for this wide dispersion are the small sample size and short-term follow-up of most series and the heterogeneity in route of perfusion, dosing, and follow-up scheme. As far as BCG perfusion after ablation of TaT1 tumours is concerned, the most recent data from the Long Island Jewish Medical Centre in New York, with the largest experience in this area, showed no significant benefit for adjuvant BCG perfusion after percutaneous resection of UTT UC in 50 RUs compared with resection alone in 39 RUs with regard to recurrence and progression rates and time to recurrence, even after stratification by tumour stage and grade [18]. This apparent lack of efficacy of adjuvant BCG perfusion after ablation of UTT NMIUC was confirmed in a subsequent nonsystematic review performed by investigators from the same institution [17]. The most threatening complications of BCG perfusion in the UUT are BCG dissemination and septicaemia due to gram-negative bacteria, with both of these occurrences being rare. In our series, adverse events developed in nearly one of five patients and were of severe grade in only two patients, with death for one; however, this latter event was sustained due to E coli, not BCG infection. In the literature, two other fatal cases have been reported, with one death due to non-bcg-related septicaemia after treatment of 50 RUs [18] and one death due to interstitial pneumonia, possibly following BCG dissemination, after treatment of 13 RUs [16]. To avoid this potentially lethal complication, we recommend that relevant bacteriuria be ruled out with nephrostomy urinalysis before each single perfusion. Minor

6 960 EUROPEAN UROLOGY 60 (2011) complications, such as transient fever and storage lower tract urinary symptoms, are more common, occurring in up to two-thirds of patients [10]. Our study is not devoid of limitations. First, it has the inherent bias of a retrospective study with no control group. However, due to the very low incidence of UUT NMIUC, the efficacy of BCG perfusion and of any treatment will be difficult to test in randomised trials; therefore, good-quality retrospective studies with homogeneous follow-up, such as the present one, are currently providing the highest level of evidence. Second, the median follow-up time of our negatively selected population is limited in absolute terms and possibly acts as a confounder of the true efficacy of BCG treatment. Third, a multivariable analysis assessing the predictors of the different survival measures was not performed. This analysis was deliberately omitted because the poor outcome of our patients, many of whom were affected by invasive UC elsewhere in the urinary tract, precludes reliable subgroup analyses. 5. Conclusions In our patients with UUT NMIUC, most of whom were not eligible for radical surgery, antegrade BCG perfusion resulted in a kidney-preservation rate of approximately 90%. Patients treated with curative intent for Tis apparently benefited in terms of local disease control more than those treated with adjuvant intent after ablation of TaT1 tumours. The projected 10-yr recurrence-free survival for patients treated with curative intent for Tis was approximately 50%. Most treatment-related adverse events were minor. A single case of fatal E coli septicaemia was observed. BCG perfusion should be considered as first-line treatment for UUT Tis before any upfront nephroureterectomy because the chances for cure are relatively high in carefully selected patients. The place for BCG perfusion after ablation of TaT1 tumours is less evident and should be evaluated on an individual basis. Treatment-related toxicity is acceptably low if relevant bacteriuria is repeatedly excluded. Author contributions: Urs E. Studer had full access to all the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Giannarini, Thalmann, Studer. Acquisition of data: Giannarini, Kessler, Birkhäuser, Thalmann. Analysis and interpretation of data: Giannarini, Kessler, Thalmann, Studer. Drafting of the manuscript: Giannarini, Kessler. Critical revision of the manuscript for important intellectual content: Thalmann, Studer. Statistical analysis: Giannarini, Kessler. Obtaining funding: None. Administrative, technical, or material support: None. Supervision: Studer. Other (specify): None. Financial disclosures: I certify that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (eg, employmentaffiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: None. FundingSupport and role of the sponsor: None. References [1] Rouprêt M, Zigeuner R, Palou J, et al. European guidelines for the diagnosis and management of upper urinary tract urothelial cell carcinomas: 2011 update. Eur Urol 2011;59: [2] O Donoghue JP, Crew JP. Adjuvant topical treatment of upper urinary tract urothelial tumours. BJU Int 2004;94: [3] Studer UE, Casanova G, Kraft R, Zingg EJ. Percutaneous bacillus Calmette-Guerin perfusion of the upper urinary tract for carcinoma in situ. J Urol 1989;142: [4] Thalmann GN, Markwalder R, Walter B, Studer UE. Long-term experience with bacillus Calmette-Guerin therapy of upper urinary tract transitional cell carcinoma in patients not eligible for surgery. J Urol 2002;168: [5] Common Terminology Criteria for Adverse Events v.4.0. National Cancer Institute Web site _ _QuickReference_5x7.pdf. Accessed May 20, [6] Herr HW. Durable response of a carcinoma in situ of the renal pelvis to topical bacillus Calmette-Guerin. J Urol 1985;134: [7] Smith AD, Orihuela E, Crowley AR. Percutaneous management of renal pelvic tumors: a treatment option in selected cases. J Urol 1987;137: [8] Sharpe JR, Duffy G, Chin JL. Intrarenal bacillus Calmette-Guerin therapy for upper urinary tract carcinoma in situ. J Urol 1993;149: [9] Yokogi H, Wada Y, Mizutani M, Igawa M, Ishibe T. Bacillus Calmette- Guérin perfusion therapy for carcinoma in situ of the upper urinary tract. Br J Urol 1996;77: [10] Nishino Y, Yamamoto N, Komeda H, Takahashi Y, Deguchi T. Bacillus Calmette-Guérin instillation treatment for carcinoma in situ of the upper urinary tract. BJU Int 2000;85: [11] Nonomura N, Ono Y, Nozawa M, et al. Bacillus Calmette-Guérin perfusion therapy for the treatment of transitional cell carcinoma on situ of the upper urinary tract. Eur Urol 2000;38: [12] Okubo K, Ichioka K, Terada N, Matsuta Y, Yoshimura K, Arai Y. Intrarenal bacillus Calmette-Guérin therapy for carcinoma in situ of the upper urinary tract: long-term follow-up and natural course in cases of failure. BJU Int 2001;88: [13] Irie A, Iwamura M, Kadowaki K, Ohkawa A, Uchida T, Baba S. Intravesical instillation of bacille Calmette-Guérin for carcinoma in situ of theurotheliuminvolving the upperurinary tract usingvesicoureteral reflux created by a double-pigtail catheter. Urology 2002;59:53 7. [14] Miyake H, Eto H, Hara S, Okada H, Kamidono S, Hara I. Clinical outcome of bacillus Calmette-Guérin perfusion therapy for carcinoma in situ of the upper urinary tract. Int J Urol 2002;9: [15] Hayashida Y, Nomata K, Noguchi M, et al. Long-term effects of bacille Calmette-Guérin perfusion therapy for treatment of transitional cell carcinoma in situ of upper urinary tract. Urology 2004;63: [16] Kojima Y, Tozawa K, Kawai N, Sasaki S, Hayashi Y, Kohri K. Long-term outcome of upper urinary tract carcinoma in situ: effectiveness of nephroureterectomy versus bacillus Calmette- Guérin therapy. Int J Urol 2006;13: [17] Rastinehad AR, Smith AD. Bacillus Calmette-Guérin for upper tract urothelial cancer: is there a role? J Endourol 2009;23: [18] Rastinehad AR, Ost MC, Vanderbrink BA, et al. A 20-year experience with percutaneous resection of upper tract transitional carcinoma: is there an oncologic benefit with adjuvant bacillus Calmette Guérin therapy? Urology 2009;73:27 31.