SEARCH

SEARCH BY CITATION

Keywords:

  • narrow-band imaging;
  • BCG;
  • bladder tumours;
  • cystoscopy

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

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

OBJECTIVE

To evaluate whether narrow-band imaging cystoscopy (NBIC) can identify bladder tumour suspected on follow-up white-light cystoscopy (WLC) after intravesical bacille Calmette-Guérin (BCG) therapy, as BCG causes an intense reaction in the bladder, appearing as red lesions on WLC, which might be residual tumour or BCG-induced inflammation.

PATIENTS AND METHODS

Sixty-one patients with high-risk non-muscle-invasive bladder tumours were evaluated 3 months after starting induction BCG therapy. All patients had abnormal erythematous lesions on WLC, suspected to be residual carcinoma in situ. After WLC, they were evaluated by NBIC, urine cytology and biopsy, followed by transurethral resection of all visible lesions.

RESULTS

Of the 61 patients, 22 (36%) had residual tumour. NBIC correctly identified tumour in 21 patients, but another 10 had unnecessary biopsy (NBIC positive, negative biopsy). Only one of 30 patients who had negative NBIC findings had tumour. NBIC outperformed urine cytology in detecting residual tumour after BCG therapy.

CONCLUSION

NBIC appears to better identify patients who have suspected residual tumour on follow-up WLC at 3 months after BCG therapy.


Abbreviations
WLC

white-light cystoscopy

NBIC

narrow-band imaging cystoscopy

TUR

transurethral resection

CIS

carcinoma in situ.

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

Intravesical BCG is the standard therapy for high-grade non-muscle-invasive bladder cancer. The response to BCG is determined by cystoscopy, urine cytology and biopsy, usually 3–6 months after the start of 6 weeks of induction BCG therapy. However, BCG causes intense inflammation in the bladder wall, making it difficult at times to visually distinguish tumour, especially carcinoma in situ (CIS) from benign erythematous lesions. The aim of this study was to investigate whether narrow-band imaging cystoscopy (NBIC) improves the evaluation of patients suspected to have residual tumour after BCG therapy on follow-up conventional white light cystoscopy (WLC).

PATIENTS AND METHODS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

Sixty-one patients had a re-staging transurethral resection (TUR) for multiple, high-grade pTa (37) or pT1 (24) tumours, and 48 (79%) had associated diffuse CIS. At 3 months after starting induction BCG therapy, patients were evaluated for the response with flexible WLC, NBIC, urine cytology and biopsy, using methods described previously [1]. All patients had flat or slightly raised red mucosal lesions in known tumour sites on WLC, suspected to be residual tumour. Switching to NBIC, capillary-rich areas that appeared brownish or black were identified as likely tumour. Mucosal lesions reflecting a greenish hue, similar to normal mucosa, were deemed to be negative for tumour. All lesions were cold-cup biopsied before TUR.

I attempted to set objective measures of WLC-positive lesions using NBIC to identify residual tumour and CIS at the first follow-up cystoscopy after BCG therapy. However, BCG and inflammatory conditions, such as interstitial cystitis, can simulate CIS on WLC. NBIC might distinguish inflammatory changes from unsuspected CIS [2]. TUR removes visible papillary tumours, and BCG causes mucosal disease to slough, exposing the submucosal vessels, which appear green on NBIC. A greenish tint against a white or red background suggests absence of surface neovascularity and healing of surrounding epithelium. Such appearance was considered negative for tumour on NBIC. Capillary-dense mucosal lesions, representing angiogenesis, are silhouetted dark brown or black, and I regarded such findings as representing positive disease on NBIC. This admittedly arbitrary classification was set before and adhered to throughout the study, to test its validity.

NBIC findings (positive or negative, for tumour) were entered into a secure database, to be correlated with biopsy results. Descriptive statistics were used to compare NBIC findings with urine cytology and the results of biopsy. All tests were two-sided, and P < 0.05 was considered to indicate significance. Each patient signed an informed consent for this ongoing prospective protocol approved by the institutional review board.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

Of the 61 patients, 22 (36%) had residual tumour, mostly CIS, and 39 (64%) had no residual tumour on biopsy, even though all had suspicious red lesions on WLC. Table 1 shows the findings at 3 months after BCG therapy, stratified by NBIC and cytology results. NBIC was positive in 31 patients; 21 (68%) had a positive biopsy for tumour, and 10 (32%) did not; these last patients had an unnecessary biopsy. Another 16 patients had negative urine cytology despite positive NBIC, and nine had tumour. NBIC was deemed negative in 30 patients, and regardless of the results of urine cytology, only one patient had a positive biopsy.

Table 1.  The findings at 3 months after BCG therapy
Cystoscopy/cytology resultNo. cases (biopsied)n or n (%)
TumourNo tumour
NBIC +, cytology +1512 3
NBIC +, cytology −16 9 7
NBIC −, cytology + 7 1 6
NBIC −, cytology −23 023 (100)
Totals6122 (36)39 (64)

The sensitivity, specificity, positive predictive value and negative predictive value of NBIC vs urine cytology in predicting a positive biopsy was 95% vs 59% (P = 0.01, Pearson’s chi-square test), 78% vs 77% (not significant), 68% vs 59% (not significant) and 97% vs 77% (P = 0.03). The Wilcoxon signed-rank test for NBIC predicting a positive biopsy gave P = 0.007, and for cytology, P = 0.4. Figures 1–3 show examples of positive and negative NBIC endoscopic images, with corresponding WLC.

image

Figure 1. Top panel shows erythema on WLC; bottom panel shows typical brown capillary pattern on NBIC, interpreted as positive for tumour. Biopsy after BCG therapy showed CIS.

Download figure to PowerPoint

image

Figure 2. WLC on the left shows bland red lesions; NBIC on the right, of the same lesion, appears densely brown. The biopsy after BCG therapy showed CIS.

Download figure to PowerPoint

image

Figure 3. The top panel shows red lesions on WLC and the bottom panels show normal green mucosa on NBIC. The biopsy after BCG therapy showed no tumour (granulomatous reaction).

Download figure to PowerPoint

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES

The major finding of this study is that NBIC outperformed urine cytology in identifying which patients had or did not have persistent tumour after BCG therapy. Although the individual favourable results are based on extensive experience using flexible NBIC to follow patients for tumour recurrence [1], they are subjective evaluations by one urologist and have not been reproduced by others. Few patients were studied, and the endpoint was simply whether NBIC could differentiate red mucosal lesions (erythema), ubiquitous on WLC, at 3 months after starting BCG therapy, as caused by BCG-induced inflammation or persistent tumour. Cystoscopy by any method remains operator-dependent, beset with subjectivity. Much more work is needed using newer imaging techniques to prospectively evaluate patients with bladder tumour.

The early response to BCG has prognostic significance. Patients with persistent tumour 3 months after starting BCG have a higher risk of progression [3] and death [4] from bladder cancer. Such patients might benefit from another TUR, more intensive BCG treatments, or even early cystectomy [5]. Patients who do not have residual disease might be spared an unnecessary biopsy or resection, and are candidates for maintenance BCG.

Some investigators have suggested that biopsy at 3 months after BCG is unnecessary if cytology is negative, despite abnormal cystoscopic findings [6]. However, among 49 patients with suspected tumour and negative cytology, we detected residual disease in 20%[7], and another study found tumour in two of seven patients with erythema and negative cytology after BCG [8]. In the present study, half the cases with positive NBIC and negative cytology were found to have persistent bladder tumours. On the other hand, among 30 patients with a positive WLC but negative NBIC findings suggesting tumour, 29 (97%) had no disease. None of the patients with erythema after BCG on WLC, but negative NBIC and negative cytology, had residual tumour.

Detecting residual tumour as early as 3 months after BCG therapy is important, especially pT1 cancer, probably leading to changes in treatment strategy. However, positive urine cytology is common 3 months after BCG in the absence of tumour, and almost all patients have erythema, raised red lesions, oedema or other abnormal findings on conventional WLC that might represent residual unresponsive tumour. Perhaps NBIC might differentiate confusing lesions seen on WLC, identify those who need more intensive therapy and help to select patients who can be spared unnecessary biopsy.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. PATIENTS AND METHODS
  5. RESULTS
  6. DISCUSSION
  7. CONFLICT OF INTEREST
  8. REFERENCES