Narrow band imaging for detecting residual/recurrent cancerous tissue during second transurethral resection of newly diagnosed non-muscle-invasive high-grade bladder cancer


Angelo Naselli, Largo Rosanna Benzi 10, Genoa, 16132, Italy.


Study Type – Diagnostic (case series)
Level of Evidence 4


To determine if narrow-band imaging (NBI) can be used to detect high-grade cancerous lesions missed with the white light at the time of a second transurethral resection (TUR) of high-grade non-muscle-invasive bladder cancer (NMIBC).


Consecutive patients with newly diagnosed high-grade NMIBC were enrolled in a prospective observational study. Patients with incomplete resection or absence of muscle tissue in the specimen were excluded. About 1 month after the first TUR, NBI cold-cup biopsies were taken from areas suspicious for urothelial cancer at the end of an extensive white-light second TUR protocol including: (i) resection of the scar of the primary tumour; (ii) resection of any overt or suspected urothelial lesions; and (iii) six random cold-cup biopsies of healthy mucosa.


In 2008, 47 consecutive patients were recruited after giving written consent (median age 62 years, range 49–83, 39 men and eight women). Nine patients (19%) had macroscopic or microscopic high-grade NMI urothelial cancer, whereas one was reassessed as having muscle-invasive disease at the white-light second TUR plus the six random biopsies. NBI biopsies were taken in 40 of the 47 patients and detected six more patients with high-grade cancerous tissue (13%). In all 16 of the 47 patients (34%) were found to have residual/recurrent cancer using our extensive protocol of second TUR followed by NBI biopsies.


Adding NBI biopsies at the end of an extensive second TUR protocol in patients with newly diagnosed high-grade NMIBC can lead to the identification of patients with otherwise missed high-grade residual/recurrent urothelial carcinoma.


narrow-band imaging


non-muscle-invasive bladder cancer


carcinoma in situ


transurethral resection


5-aminolaevulinic acid.


Narrow-band imaging (NBI) is an optical image-enhancement technology that narrows the bandwidth of the light output from the endoscopy system to 415 nm and 540 nm. At this range the light is strongly absorbed by haemoglobin, and the visibility of surface capillaries and blood vessels in the submucosa is enhanced. It has advantages over conventional white-light flexible endoscopy. In gastrointestinal endoscopy, it allowed a more accurate prediction of Barrett’s oesophagus [1] and improved the ability to distinguish neoplastic and non-neoplastic colorectal polyps in screening colonoscopy [2]. It has also been used to significantly improve the diagnostic accuracy, sensitivity and negative predictive value in detecting squamous cell carcinoma of the head and neck in high-risk patients [3]. In urology, NBI flexible cystoscopy allowed the detection of urothelial lesions missed with white light in patients with known bladder cancer [4] and improved the diagnosis of recurrent non-muscle-invasive (NMI) bladder tumours, especially of carcinoma in situ (CIS) [5].

Therefore we postulated that NBI could have a role in the second transurethral resection (TUR) of newly diagnosed high-grade NMIBC. The second TUR is intended mainly to confirm the stage of the primary cancer by resecting the scar of the first TUR [6]. In addiction, all macroscopic overt or suspected lesions are resected and random biopsies of healthy mucosa are taken. Thus the second TUR leads to a more complete extirpation of bladder lesions [7]. Consequently it has a favourable effect on the response to the local therapy, recurrence and progression rates [7,8]. Moreover, concomitant CIS or high-grade T1 residual/recurrent cancer detected by the second TUR are recognized to be the most powerful predictors of progression [9,10]. The aim of the present study was to determine if the NBI technology could identify patients with high-grade cancerous urothelial lesions not detected by white light at the time of the second TUR.


The study was conducted in accordance with Good Clinical Practice and the Declaration of Helsinki 1964, including the most recent amendment (Edinburgh, Scotland, 2000) and after approval of the local medical ethical committee. All patients were aged >18 years and women who were pregnant, breast feeding or not on adequate contraceptive measures were excluded. All patients provided written informed consent before entry to the study.

Consecutive patients with newly diagnosed high-grade NMIBC were enrolled in this prospective observational study to determine whether NBI cold-cup biopsies could identify patients with high-grade residual/recurrent cancer undetected by an extensive white-light second TUR protocol. We decided to exclude patients with incomplete resection or absence of muscle tissue in the specimen, to form an homogeneous cohort; these cases are almost always found to have residual disease and therefore are not expected to gain any advantage from additional procedures like NBI biopsies.

A second TUR protocol (re-staging TUR) was scheduled at ≈1 month after the diagnostic endoscopy; the procedure was routinely performed under general or spinal anaesthesia, using a 26 F resectoscope (Exera II, Olympus Italia Srl, Milan, Italy). The camera head of the system was modified by dedicated engineers at Olympus Italia; the resectoscope is a prototype that allows an NBI view with the rigid rod-lens technology. The re-staging TUR began with an accurate inspection of the bladder with the white light. The scar of the first procedure was resected and any macroscopic overt or suspicious lesion found was excised. The standard procedure was completed by random cold-cup biopsies of the posterior, anterior, left and right wall of the bladder, and of the trigone (six biopsies for each patient). After accurate fulguration of any source of bleeding, the system was switched to NBI by pushing a button, and areas suspicious for tumours were biopsied.

All the specimens were analysed by a pathologist unaware of the mode of identification of the single lesion (by white light or NBI). Staging was in accordance to the TNM classification (2002) and grading by the WHO 2004 classification. The NBI was expected to increase the detection of high-grade microscopic lesions in the study group, and in particular, it was estimated that with NBI ≥10% more patients should be found with an otherwise undetected high-grade lesion.


In 2008, 47 consecutive patients were recruited after giving written consent (median age 62 years, range 49–83; 39 men and eight women). Regarding the characteristics of the primary tumour, six (13%) patients had a Ta high-grade bladder cancer, and 37 (79%) T1 and four (8%) a pure CIS. Concomitant CIS was detected in one patient with a Ta tumour and in six with a T1 tumour (16%). The tumour was single in 14 cases (30%) and multiple in 33 (70%).

In all, 16 of the 47 patients (34%) had residual/recurrent cancer identified using the extensive protocol of second TUR followed by NBI biopsies. Of these 16, only one was re-staged to T2 disease. At the white-light inspection of the second TUR, six patients (13%) had one or more papillary recurrent tumours, for a total of nine lesions detected. In three patients the recurrence was localized at the site of the primary lesion (6%), in two elsewhere (4%), and in the remaining one simultaneously at the site of the primary lesion and elsewhere. Nine patients (19%) had one or more suspicious areas outside the scar of the primary lesion, for a total of 14 areas detected. Finally, 282 random biopsies from healthy mucosa were taken, six for each patient.

The pathological examination of the recurrent papillary lesion showed Ta low-grade cancer in three patients, Ta high-grade in one and T1 high-grade in two. Excluding the four cases with a visible tumour at the site of the primary lesion, the resection of the scar revealed cancer in four more patients (8%). The stage was, respectively, CIS in one patient, Ta high-grade in two and T2 in the remaining one. The 14 suspicious areas found at the white-light inspection were positive in two (14%) and revealed CIS in one patient and micropapillary tumour in another one. In all, 14 random biopsies were positive (5%). The random biopsies showed cancer in six patients (13%), CIS in two, Ta high-grade in two, T1 high-grade in one and severe dysplasia in the remaining one. In all, nine patients (19%) were found with macroscopic or microscopic NMI high-grade cancer, whereas one was reassessed as having muscle-invasive disease by the white-light second TUR plus the six random biopsies.

After switching the system to NBI, 40 patients showed one or more areas suspicious for tumour. In all, 72 NBI biopsies were taken, of which 11 were positive (15%). The NBI biopsies showed tumour in nine patients, i.e. five with CIS, three with Ta high-grade cancer and one with micropapillary cancer. Of the nine patients, four had cancer in the specimens obtained from the white-light procedure, comprising one Ta low-grade, two Ta high-grade and one T2. Overall NBI detected five more patients with residual cancer, but one with a Ta low-grade recurrence detected during the white-light inspection was found to have a concomitant CIS detected by NBI. Therefore six more patients were detected with high-grade residual cancer by the NBI biopsies (13%).


The standard method used to diagnose, treat and monitor patients with NMIBC is white-light imaging cystoscopy and TUR. However, it is well known that this technology fails to detect small papillary and subtle flat CIS cancers [11]. Undetected tumours will later appear as a recurrence, and some might become invasive, highlighting the need to develop alternative endoscopic methods to detect bladder lesions. A more complete identification might at the least render the bladder more receptive to successful intravesical therapy [8], and more importantly, might allow for greater detection of high-grade/-stage tumours that could alter treatment decisions [12].

To this purpose, rigid and flexible fluorescence cystoscopy using 5-aminolaevulinic acid (5-ALA) or its hexyl ester have been investigated extensively, and recognized to improve significantly the detection of NMI papillary and flat bladder lesions over standard white-light cystoscopy. This translates into a more complete resection, reducing the recurrence rate of NMI tumours [13] and in more appropriate treatments [14]. However, the technique has some limits. The 5-ALA solution is instilled before the procedure, and the retention time after instillation must be ≥1 h. Often the instillation is insufficient to provide enough 5-ALA exposure. The use of the white-light mode should be rapid and at the lowest intensity before switching to fluorescence mode. The photobleaching phenomenon can be very rapid, especially in flat areas, thus limiting the time of the procedure. Patients with porphyria, allergy to 5-ALA or similar compounds cannot be instilled with 5-ALA.

NBI is a alternative promising optical enhancement technology, providing enhanced images of capillaries in the mucosal surface and detailed mucosal texture by irradiating with two narrow wave bands (390–445 nm/530–550 nm) which are strongly absorbed by circulating haemoglobin. Narrow-band light does not reach the lower layers of tissue and is absorbed well by the blood vessels. As a result, the areas with light reflection, mucous membrane, and the areas with no reflection, blood vessels, can be distinguished and displayed at higher contrast. The main advantages are that NBI has no contraindications for the patient, it requires no instillation or disposable items, it can be used an unlimited number of times for each procedure, and it costs little, because it consists of a software modification to the light output and to the camera. NBI has been tested in many trials of gastrointestinal and oropharyngeal endoscopy [1–3], where it showed a greater ability to detect early flat cancerous lesions than with white light. Therefore it might be an attractive tool in the endoscopic management of NMIBC. At the NBI inspection the aspect of a bladder lesion is dark black against an almost white normal mucosa, whereas the lesion appears red against a pink normal mucosa on white-light examination. The colour contrast of a lesion is enhanced using NBI; one report showing the feasibility of NBI flexible cystoscopy was published recently [4]. Objectively, NBI detected 15 additional urothelial carcinomas in 12 of 29 patients (41%). The main criticism was that increasing the detection rate of low-grade urothelial cancer does significantly affect disease management. However, another within-patient prospective study showed that NBI detected extra papillary tumours or more extensive CIS in ≈100 patients found to have a bladder cancer recurrence [5]. Particularly for detecting CIS, the sensitivity of white light and NBI cystoscopy was 83% vs 100%[5]. Therefore we hypothesized that NBI could have a role in the detection of microscopic lesion in the TUR re-staging of high-grade NMIBC. The presence of macroscopic/microscopic residual/recurrent cancer in those patients has a major influence on the management and prognosis of bladder cancer [9]. However, the probability of missing a microscopic tumour is significant by standard white-light endoscopy and can lead to an increased risk of progression and death from bladder cancer [11]. Thus we designed a protocol in which each patient with newly diagnosed high-grade NMIBC had an accurate re-staging TUR followed by NBI bladder inspection, and eventually biopsies. The resection of the scar of the first TUR was routine; in cases with a multiple primary tumour, the specimens of each scar, always including muscle tissue, were collected separately for pathological examination. All visible lesions and suspicious areas outside the scar were resected and analysed sequentially. Finally, six random biopsies of healthy mucosa were taken. At this point, after an accurate fulguration of any source of bleeding, the re-staging procedure was stopped. No macroscopic overt or suspicious cancerous lesion was yet visible at white-light inspection and random biopsies mainly intended to detect CIS were taken. Given the ability of NBI to detect lesions missed with white light [4,5], the system was switched to NBI; areas with increased vasculature, characterized by a more intense dark green to black aspect, were found in 40 of 47 patients and were biopsied; 11 of 72 NBI biopsies were positive. The sensitivity of the procedure was therefore 15%, as obtained from NBI in a bladder just after an extensive resection protocol and multiple biopsies, thus adding more value. With the 11 positive biopsies, nine patients were identified by NBI with microscopic high-grade cancer. Of these nine, six had a high-grade cancer not detected by white light.

In conclusion, adding NBI biopsies at the end of an extensive second TUR protocol adds no significant morbidity or costs to the procedure, but can identify patients otherwise missed with high-grade residual/recurrent urothelial carcinoma. However, further studies comparing fluorescence and NBI TUR as ‘stand-alone’ procedures are needed to determine the real effect of such technologies on the management of bladder cancer.


The Olympus Exera II system was kindly donated by Olympus Italia Srl.


None declared.