Is there a learning curve for photodynamic diagnosis of bladder cancer with hexaminolevulinate hydrochloride?

INTRODUCTION: To assess the learning curve for fluorescence cystoscopy using hexaminolevulinate hydrochloride (HAL) in patients with bladder cancer.
MATERIAL AND METHODS: Fifty patients underwent bladder instillation with HAL. Two senior residents inspected separately the bladder using white light cystoscopy, followed by fluorescence cystoscopy and mapped the lesions. An experienced with photodynamic diagnosis (PDD) urologist also performed both cystoscopies, mapped, resected or cold biopsied suspect lesions under the supervision of another experienced urologist. To evaluate the learning curve, patients were divided into five subgroups, including group 1 (patients 1-10), group 2 (11-20), group 3 (21-30), group 4 (31-40) and group 5 (41-50). The kappa statistics was calculated to assess interobserver agreement between the physicians and the false positive rates of urologists and residents were also compared.
RESULTS: Histologically verified tumors were diagnosed in 103 of 142 lesions identified by PDD. The interobserver agreement between urologists and residents was moderate, moderate, good, excellent, and excellent for group 1, 2, 3, 4, and 5, respectively. Both residents had increased false positive rates compared to urologists in all subgroups of patients but this difference did not reach statistical significance. In addition, false positive rate of residents was declining as the number of procedures was increasing.
CONCLUSIONS: Our data suggest that 20 cases of HAL PDD are required to achieve a good interobserver agreement between inexperienced and experienced operator, and excellent agreement is achieved after 30 cases. The false positive rate of inexperienced operators was comparable to the experts and showed a gradual decrease.