Topical chemotherapy in human urothelial carcinoma explants: a novel translational tool for preclinical evaluation of experimental intravesical therapies.

Background: Urothelial carcinoma (UC) is associated with a high local recurrence rate despite intravesical therapy. There is a lack of representative preclinical models for standardized testing of novel experimental therapies.

Objective: To develop an ex vivo model for human UC and to evaluate its ability to generate reproducible and reliable results when testing cytotoxic agents.

Design, Setting, And Participants: Normal human urothelium (NHU) and bladder UC explants were collected from patients treated at our institution. A total of 195 surgical explants were cultured on a gelatine matrix. Tissue viability was regularly assessed using nicotinamide adenine dinucleotide (NADH) diaphorase enzymehistochemistry. Topical paclitaxel (PTX) or mitomycin C (MMC) chemotherapy was performed in a subset of 45 UC specimens.

Intervention: All patients underwent radical cystectomy (RC) or primary transurethral resection (TUR) of a bladder UC.

Measurements: Triple immunofluorescence (pan-cytokeratin [pan-CK]; 4',6-diamidin-2'-phenylindol-dihydrochloride [DAPI]; terminal deoxynucleotidyl transferase biotin-dUTP nick-end labelling [TUNEL]) and caspase-3 staining of paraffin sections was performed. Proliferation rates were assessed using Ki-67 labelling indices. Apoptosis (percent) was quantified in representative tissue areas to characterize culture stability and to assess antineoplastic effects.

Results And Limitations: No signs of necrosis and no significant changes in apoptosis were observed during the first 12 d of culture. Of all explants, 88.5% were vital after 20 d. In a highly reproducible fashion, topical chemotherapy resulted in significantly increased apoptosis (37.4% [19.0-75.0%] for PTX and 36.2% [18.8-46.7%] for MMC) compared with controls (7.5% [3.0-26.8%]; p<0.001]). No statistically significant difference was observed regarding the effects of the two chemotherapeutic agents (p=0.119).

Conclusions: The presented human ex vivo model takes UC heterogeneity into account and serves as a valuable translational tool. It offers an attractive alternative to preclinical cell line experiments or animal models and may even be used for prospective toxicity and drug efficacy tests in individual patients.