Objectives: Patients with non-muscle invasive bladder cancer (NMIBC) face a high risk of recurrence and progression after transurethral resection, making adjuvant therapies necessary. Intravesical device-assisted therapies, such as radiofrequency-induced thermochemotherapy (RITE), have shown promise in enhancing the effectiveness of intravesical chemotherapies. This study aimed to evaluate oncological outcomes in patients with NMIBC treated with RITE in a real-world setting, encompassing those unresponsive to prior Bacillus Calmette-Guérin (BCG) or intravesical chemotherapy, as well as those who declined or were ineligible for BCG or radical cystectomy (RC).

Methods: A retrospective multicenter analysis of patients treated with RITE for NMIBC between 2015 and 2024 was performed. Co-primary endpoints of the study were intravesical recurrence free survival (RFS) and high-grade intravesical recurrence free survival (HG-RFS). Secondary endpoints included radical cystectomy-free survival (RC-FS), overall survival (OS), cancer-specific survival (CSS), and adverse events (AEs).

Results: Fifty-nine consecutive patients were included in the final analyses. Overall, 12 (20%) and 45 (76%) patients failed previous intravesical chemotherapy, and immunotherapy with BCG, respectively. The 24-months RFS and HG-RFS following RITE were 68.6% (95% CI: 0.568, 0.828) and 74.8% (95% CI: 0.632, 0.885). RC-FS at 24 months was 93.8% (95% CI: 0.872, 1.000). The OS probability at 24 months was 91%, with a CSS of 97.8%. Most common AEs were dysuria and urgency in 27 (45.7%) patients, treatment limiting bladder spasms in 11 (19%) patients, low bladder compliance in 11 (19%) patients and urethral strictures in 5 (8%) patients.

Conclusion: In our analyses, RITE resulted in notable antitumor activity and allows for the avoidance of more aggressive and quality-of-life-limiting therapies, such as radical cystectomy. RITE might be considered as a second-line bladder-sparing option in patients failing previous intravesical therapies. Long-term follow-up and larger-scale data are required to validate our findings.