Patient characteristics correlate with diagnostic performance of photodynamic diagnostic assisted transurethral resection of bladder tumors: A retrospective, single-center study

Shuhei Suzuki,Yoshiyuki Nagumo,Atsushi Ikeda,Kosuke Kojo,Satoshi Nitta,Ichiro Chihara,Masanobu Shiga,Takashi Kawahara,Shuya Kandori,Akio Hoshi,Hiromitsu Negoro,Bryan J Mathis,Hiroyuki Nishiyama,Bryan J. Mathis
DOI: https://doi.org/10.1016/j.pdpdt.2024.104052
IF: 3.577
2024-03-20
Photodiagnosis and Photodynamic Therapy
Abstract:Background : Identification of patient subclasses that correlate with the diagnostic performance of photodynamic diagnostic (PDD)-assisted transurethral resection of bladder tumors (TURBT) may improve outcomes. Methods : Data were extracted from patients that underwent PDD-assisted TURBT at the University of Tsukuba Hospital between 2018-2023. Sensitivity and specificity were evaluated based on PDD findings (excluding WL findings) and pathology results. Cluster analysis using uniform manifold approximation and projection and k-means methods was performed, focusing on patients with malignant lesions. Results : A total of 267 patients and 2082 specimens were extracted. Sensitivity was lowest with regard to BCG treatment (53.7%), followed by flat lesions (57.2%), urine cytology class ≥ III (62.9%), and recurrent tumors (64.5%). In the cluster analysis of 231 patients with malignant lesions, two showed lower sensitivity: Cluster 3 (62.4%), consisting of patients with recurrent tumors and post-BCG treatment, and Cluster 4 (55.7%), consisting of patients with primary tumors and urine cytology class ≥ III. Clusters 1 and 2, consisting of patients without BCG treatment and patients with lower urine cytology classes, exhibited higher sensitivities (94.4% and 87.7%). Among all clusters, Cluster 4 had the highest proportion of specimens which were negative for both PDD and white light (WL) findings but actually had malignant lesions (20.8%). Conclusions : PDD-assisted TURBT sensitivity was lower in subclasses after BCG treatment or with cytology class III or higher. Random biopsy for PDD/WL double-negative lesions may improve diagnostic accuracy in these subclasses.
oncology
What problem does this paper attempt to address?