Reduced bladder tumour recurrence rate associated with narrow-band imaging surveillance cystoscopy


Harry W. Herr, 1275 York Avenue, New York, NY 10021, USA.


Study Type – Diagnostic (exploratory cohort)
Level of Evidence 2b

What’s known on the subject? and What does the study add?

Narrow-band imaging cystoscopy is a new imaging modality developed to enhance conventional standard white-light cystoscopy to evaluate bladder tumors. The current paper suggests that fulguration of low-risk papillary bladder tumours using NBI cystoscopy results in fewer subsequent tumour recurrences than fulguration using standard cystoscopy. How, or if, NBI cystoscopy will become integrated into routine management of non-invasive bladder tumours remains for further study.


To evaluate frequency of recurrences among patients with papillary bladder tumours followed sequentially with conventional white-light (WLI) cystoscopy and narrow-band imaging (NBI) cystoscopy.


A cohort of 126 patients with recurrent low-grade papillary bladder tumours were followed every 6 months for 3 years by conventional WLI cystoscopy, and then over the next 3 consecutive years by NBI cystoscopy.

Recurrent tumours detected were treated by outpatient fulguration or transurethral resection.

We compared the tumour recurrence rate during follow-up with WLI and NBI cystoscopy, using patients as their own controls.


Of the 126 patients, 94% had tumour recurrences during WLI cystoscopy vs 62% during NBI cystoscopy.

The mean number of recurrent tumours was 5.2 with WLI cystoscopy vs 2.8 with NBI cystoscopy, and the median recurrence-free survival time was 13 vs 29 months (P= 0.001).


Compared with follow-up with WLI cystoscopy, NBI cystoscopy was associated with fewer patients having tumour recurrences, fewer numbers of recurrent tumours, and a longer recurrence-free survival time.


narrow-band imaging


transurethral resection




In a previous report, we showed in a cohort of patients with non-invasive low-grade papillary bladder tumours that surveillance cystoscopy at 6-month intervals coupled with outpatient fulguration controls recurrent tumours and reduces the frequency of transurethral resections (TURs) of tumour [1]. The patients were followed and treated by conventional white-light (WLI) endoscopy up to 2006. Since then, they have been followed by both WLI and narrow-band imaging (NBI) cystoscopy. We previously showed that NBI cystoscopy enhances detection of recurrent bladder tumours [2]. What is not known is whether improved detection facilitates better local control, resulting in fewer tumour recurrences. In the same cohort of patients, we compare, in the present study, tumour recurrence using WLI cystoscopy with that using NBI cystoscopy, with patients as their own controls.


Our original report involved 215 low-risk patients (low-grade papillary tumours), 143 (67%) of whom had multiple tumour recurrence detected and treated over 5 years using WLI cystoscopy [1]. Seventeen patients progressed in grade, leaving 126 patients who were followed by WLI cystoscopy up to July 2006, when NBI became available. Of these, 119 patients had multiple tumour recurrences between July 2003 and July 2006, and at least one recurrence detected and treated using NBI cystoscopy between July and December 2006. Since then and until December 2009, all patients have been followed with both WLI and NBI cystoscopy at 6-month intervals; these 126 patients are the focus of the present study.

In the same patients, who were followed for 3 years before July 2006 (WLI), and for 3 years after December 2006 (NBI), we compared the number of patients with tumour recurrence, the number of tumour recurrences and the recurrence-free survival, using patients as their own controls. Over the past 3 years, patients underwent outpatient flexible WLI cystoscopy first, then switched to NBI cystoscopy to inspect the bladder, using the same instrument. Recurrent tumours were either fulgurated using the NBI-equipped digital flexible cystoscope, as previously described [3], or patients had same-day TUR using a NBI camera head. Perioperative or adjuvant chemotherapy was not used.

We deliberately selected patients who had tumour recurrences during follow-up with WLI cystoscopy, to compare recurrence rates of subsequent tumours in the same patients undergoing surveillance and local treatment, as necessary, with NBI cystoscopy. We also evaluated tumour endpoints over similar time intervals – 3 years before compared with 3 years after NBI follow-up cystoscopy was started.

Descriptive statistics were used to evaluate endpoints. Kaplan–Meier curves were constructed for recurrence-free survival and compared using the log-rank test. All tests are two-sided. These patients are enrolled in a prospective, regularly updated registered database to evaluate NBI cystoscopy in the management of tumours not invading bladder muscle, and the study was approved by the institutional review board.


Tumour recurrence rates during the 3 years before and the 3 years after 2006 were compared among 126 patients with a history of frequent tumour recurrences, including 119 who had one or more tumours detected and treated during the last 6 months of 2006, when NBI cystoscopy became available. Fifteen patients (13%) had clusters of papillary tumours detected by NBI cystoscopy, which on initial inspection were not seen on WLI cystoscopy. None of the patients progressed in grade or stage.

Of 126 patients followed between July 2003 and July 2006, 119 (94%) had multiple (> one) recurrences detected and treated using WLI cystoscopy, and they had at least one tumour detected and treated using NBI cystoscopy between July 2006 and December 2006. In the ensuing 3 years, from December 2006 to January 2010, when 126 patients were followed by NBI cystoscopy, 78 (62%) had tumour recurrence, and 48 patients (38%) remained recurrence-free. The median (95% CI) recurrence-free survival time on WLI surveillance was 13 (11.6–14) months, compared with 29 (26–32) months on NBI cystoscopy (Figs 1, 2, P= 0.001). Among patients with recurrent tumours, recurrence-free survival time was 11 (9–13) months before vs 18 (13–22) months during NBI follow-up.

Figure 1.

Three-year recurrence-free survival of 126 patients followed by WLI cystoscopy; 119 patients had tumour recurrence; median (95% CI) recurrence-free survival time was 13 (11.6–14).

Figure 2.

Three-year recurrence-free survival of 126 patients followed by NBI cystoscopy; 78 patients had tumour recurrence; median (95% CI) recurrence-free survival time was 29 (26–32) months.

The mean (±SD) number of recurrent tumours during WLI follow-up for all patients was 5.2 (±3.7) compared with 2.8 (±3.4) during the subsequent 3 years with NBI cystoscopy (P= 0.001, paired samples test). For the patients with recurrent tumours, the mean (±SD) number of tumours was 6.5 (±3.9) before vs 4.5 (±3.4) after NBI became routine (P= 0.01). In all, 102 (81%) patients had fewer tumours detected during NBI follow-up than during WLI cystoscopy, 11 (9%) had a similar number, and 13 patients (10%) had more tumours detected by NBI cytoscopy; 57 (73%) patients were treated by fulguration alone, and 21 (27%) patients required one or more TURs (one TUR in 15 patients, two TURs in four patients, and three TURs in two patients). During follow-up with WLI cystoscopy, 50 (64%) patients required one or more TURs.


The major finding of the present study is that, compared with conventional WLI cystoscopy, the same patients had no or fewer tumour recurrences and longer recurrence-free intervals after they began to be followed and treated by NBI cystoscopy. For patients with tumour recurrence during NBI follow-up, the number of tumours was slightly fewer than detected by WLI cystoscopy, suggesting that small papillary tumours, or clusters of tumours, were easily detected and more efficiently destroyed by NBI cystoscopy than by WLI cystoscopy. Further, during follow-up with NBI cystoscopy, most patients (73%) had recurrent tumours treated by fulguration alone, and less than a third required a TUR, compared with 64% of the 119 patients who needed at least one TUR to control recurrent tumours during follow-up with WLI cystoscopy.

The present study has limitations. It is not randomized. Instead we chose to compare recurrence rates among patients deliberately selected because they have frequently recurring tumours, and followed them in an identical manner every 6 months sequentially over similar time periods before and after NBI surveillance cystoscopy was initiated. Papillary tumours are better seen and defined by NBI, however, which subjectively makes them easier to fulgurate completely, and randomized trials show that enhanced detection using fluorescence cystoscopy reduces subsequent bladder tumour recurrence [4].

An alternative explanation for our results is the natural history of low-grade papillary bladder tumours. With prolonged follow-up, such tumours tend to recur less frequently, especially after 5 and 10 years [5]. However, patients who have multiple tumour recurrences, as with our patient population, continue to have frequent recurrence, making it less likely that natural history explains the subsequent long-term reduced recurrence rate. The only change in management strategy for our patients was the introduction of NBI cystoscopy and treatment.

In conclusion, patients with frequently recurring low-grade papillary bladder tumours, when followed and treated by WLI cystoscopy, had fewer subsequent recurrences and longer recurrence-free intervals when they were followed and treated using NBI cystoscopy.


None declared.