O'Brien 2011.

Methods	Study design: Prospective randomised nonblinded study Statistical design: N/A Setting/Country: Multicentre/United Kingdom Dates when study was conducted: July 2000 to December 2006
Participants	Ethnicity: likely English Inclusion criteria A new clinical diagnosis of a transitional cell tumour of the upper urinary tract above the intramural ureter A negative cystoscopy for urothelial cell carcinoma of bladder within one month of the nephroureterectomy A nephroureterectomy that is planned to be performed in such a way that there is early distal ligation of the ureter prior to the mobilisation of the tumour The intramural portion of the ureter may be resected prior to the open part of the operation, taken at the open operation with a cuff of bladder or everted at the end of the operation and resected endoscopically according to surgical preference. Have a life expectancy of at least one year Informed consent to participate (written and witnessed) Adult Exclusion criteria The histology of the upper tract tumour does not confirm a transitional cell carcinoma. Stage N1 or M1 Had additional systemic chemotherapy or additional intravesical mitomycin at follow‐up cystoscopies Existing or previous urothelial cell carcinoma of bladder Children, pregnant women excluded Life expectancy less than one year Known sensitivity to mitomycin Total number of participants randomly assigned: Screened: N/A Eligible: 284 Group A (MMC instillation) Number of all participants randomly assigned: 144 Age: median: 70 years (range 44 to 87) Gender (M/F): N/R Death from UTUC: N/A; death from any cause: N/A Previous or concomitant bladder tumour: exclusion criteria Tumour stage (Ta/T1/T2/T3/T4, n, %): 28 (23.3)/40 (33.3)/19 (15.8)/29 (24.1)/2 (1.8)/not stated 2 (1.8); tumour grade (low/high, n, %): 67 (55.8)/50 (41.7)/not stated 3 (2.5) Tumour location: N/A Presence of concurrent carcinoma in situ: N/A; tumour multifocality (n, %): 15 (12.5) Bladder cuff excision method: rip and pluck, laparoscopic, open Group B (No instillation) Number of all participants randomly assigned: 140 Age: 71 years (range 36 to 90) Gender (M/F): N/R Death from UTUC: N/A; death from any cause: N/A Previous or concomitant bladder tumour: exclusion criteria Tumour stage (Ta/T1/T2/T3/T4, n, %): 45 (37.8)/26 (21.8)/13 (10.9)/28 (23.5)/2 (1.7)/not stated 5 (4.3); tumour grade (low/high, n, %): 73 (61.3)/42 (35.3)/not stated 4 (3.4) Tumour location: N/A Presence of concurrent carcinoma in situ: N/A; tumour multifocality (n, %): 15 (12.6) Bladder cuff excision method: same as group A
Interventions	Group A: Single‐dose MMC 40 mg in 40 mL of normal saline was delivered into the bladder prior to removal of the urethral catheter and was retained for 1 hour (the timing of the administration of the intravesical chemotherapy was chosen to minimise the risk of extravasation). Group B: No instillation Follow‐up: 12 months
Outcomes	Primary outcome Bladder cancer recurrence at 1 year How measured: bladder cancer recurrence: cystoscopy (visual); histologic proof of recurrence was not required Time points to measurement: cystoscopy at 3, 6, and 12 months post‐nephroureterectomy Time points to reported: N/A Secondary outcome Post‐surgical survival of participants over five years How measured: N/R Time points to measurement: not reported Time points to reported: not reported Safety outcome Adverse events How measured: not reported Time points to measurement: not reported Time points reported: not reported Subgroup Recurrence by grade, recurrence by multifocality, recurrence by stage, recurrence by method of nephroureterectomy
Funding Sources	The trial was funded through Guys and St Thomas’ Hospitals Urology Research Fund. Kyowa Hakko gave two unrestricted donations totaling £7000 to offset some administrative expenses. No payments were made to recruiting centres. None of the team at Guys Hospital had financial links with Kyowa.
Declarations of interest	None
Notes	Protocol: ISRCTN36343644 Language of publication: English
Risk of bias
Bias	Authors' judgement	Support for judgement
Random sequence generation (selection bias)	Low risk	Comment: Randomisation stated and trial author provided random sequence generation method "used ‘Tombola’ blinded selection of treatment from within the block"; therefore selection bias was considered to have low risk of bias.
Allocation concealment (selection bias)	Low risk	Quote from publication: "Randomisation was performed at Guys Hospital following the nephroureterectomy and was by means of sealed envelopes in blocks of 20". Comment: Since this study was multicentre and allocation was performed by central allocation (randomisation was performed at Guys Hospital with sealed envelopes), we assumed that this method may ensure allocation concealment.
Blinding of participants and personnel (performance bias) All outcomes	High risk	Quote from publication: "nonblinded trial" Comment: Participants and personnel were not blinded; therefore risk of performance bias was considered to be high.
Blinding of outcome assessment (detection bias) Subjective outcomes (susceptible to detection bias); time to bladder cancer recurrence, time to death from UTUC, serious and minor adverse events, disease‐specific quality of life	High risk	Quote from publication: "nonblinded trial" Comment: Participants and personnel were not blinded; therefore risk of performance bias was considered to be high.
Blinding of outcome assessment (detection bias) Objective outcomes (not susceptible to detection bias); time to death from any cause	Low risk	Comment: Objective outcomes were not likely to be affected by lack of blinding.
Incomplete outcome data (attrition bias) Time to bladder cancer recurrence	Unclear risk	Comment: 24/144 (16.6%) in intervention arm and 21/140(15%) in control arm were excluded from the analysis; owing to the moderate number of participants lost to follow‐up (> 10%), risk of attrition bias was considered to be unclear.
Incomplete outcome data (attrition bias) Time to death from UTUC	Unclear risk	Comment: This study did not address this outcome.
Incomplete outcome data (attrition bias) Serious adverse events	Unclear risk	Comment: 24/144 (16.6%) in intervention arm and 21/140(15%) in control arm were excluded from the analysis; owing to the moderate number of participants lost to follow‐up (> 10%), risk of attrition bias was considered to be unclear.
Incomplete outcome data (attrition bias) Time to death from any cause	Unclear risk	Comment: This study did not address this outcome.
Incomplete outcome data (attrition bias) Minor adverse events	Unclear risk	Comment: This study did not address this outcome.
Incomplete outcome data (attrition bias) Diseas‐specific quality of life	Unclear risk	Comment: This study did not address this outcome.
Selective reporting (reporting bias)	High risk	Comment: Protocol was provided (ISRCTN36343644) but secondary outcomes were not reported and subgroup analyses were not predefined.
Other bias	High risk	Comment: There was baseline imbalance in Ta disease and high grade tumour.

^{ALT: alanine aminotransferase
 AST: aspartate aminotransferase
 ECOG PS: Eastern Cooperative Oncology Group performance status
 F: female
 Hb: haemoglobin
 M: male
 M1: metastasis
 MMC: mitomycin
 n: number of participants
 N1: lymph node involvement
 N/A: not available
 N/R: not reported
 PLT: platelet
 THP: pirarubicin
 UTUC: upper tract urothelial carcinoma
 WBC: white blood cell}