Pathological Analysis of False-Positive Samples in Photodynamic Diagnosis of Bladder Cancer According to a History of Intravesical Bacillus Calmette-Guérin Therapy

Abstract

Objectives

To demonstrate the diagnostic performance of photodynamic diagnosis (PDD) for bladder cancer using oral 5-aminolaevulinic acid in patients with or without a history of intravesical Bacillus Calmette-Guérin (BCG) therapy, and to investigate the potential causes of false-positive results.

Methods

This retrospective study included 110 patients with suspected non-muscle-invasive bladder cancer who underwent PDD. Among them, 80 patients had no history of BCG therapy (BCG-naïve group) and 30 had received BCG therapy (BCG group). Pathological examination of false-positive and true-negative PDD samples was performed.

Results

In the BCG-naïve group, 215 samples were PDD-positive, of which 53 (24.7%) showed no malignancy. In the BCG group, 94 samples were PDD-positive, of which 51 (54.3%) showed no malignancy. The sensitivity, specificity, positive predictive value, and negative predictive value of PDD in the BCG-naïve versus BCG group were 70.1 versus 76.8, 82.6 versus 56.0, 75.3 versus 45.7, and 78.5 versus 83.3, respectively. Only six of 100 samples (6%) with false-positive PDD results showed precancerous findings. Irrespective of whether patients had received intravesical BCG therapy, > 90% of the samples with false positivity in PDD showed reactive changes, whereas 22 of 95 samples (23.2%) with true negativity in PDD showed reactive changes (p < 0.0001).

Conclusions

Both the specificity and positive predictive value of PDD were lower in the BCG group than in the BCG-naïve group. In most false-positive PDD samples, reactive changes, rather than precancerous findings, were observed—even in the BCG-naïve group.