Urothelial carcinoma (UC), also termed transitional cell carcinoma, arising from urinary bladder is the second most frequently occurring genitourinary malignancy after prostate cancer, and is one of the main causes of cancer death in Korea [1]. Bladder UC presents as superficial (i.e., non-muscle-invasive), muscle-invasive (MIBC), or metastatic disease. In patients with MIBC, cisplatin-based neoadjuvant chemotherapy (NAC) followed by radical cystectomy and bilateral pelvic lymphadenectomy has been considered the standard treatment [2]. Studies have revealed that NAC results in substantial tumor downstaging (15%-30% pathologic complete response [pCR]), and thus prolonged survival in patients with MIBC [3-5].

Although radical cystectomy remains the standard of care for MIBC, some patients are not candidates for curative surgery due to advanced age, poor performance status, or multiple comorbidities. In addition, radical surgery and urinary diversion significantly affect patient quality of life [6]. Thus, bladder preservation chemoradiotherapy (CRT) may be offered to non-surgical candidates who are medically unfit to undergo radical surgery, or to those who refuse bladder removal. Although bladder preservation strategies involving radiotherapy with or without chemotherapy have often been perceived to result in inferior overall survival (OS) compared with radical cystectomy, no randomized trials support this bias. Five-year disease-free survival (DFS) rates of 61% to 71%, and 5-year OS rates ranging from 30% to 70% have been achieved in appropriately selected patients with modern bladder preservation approaches [7,8]. Treatment decisions should be based on patient preference and an understanding of the quality of life issues associated with radical surgery.

With the introduction of a multidisciplinary team approach in the treatment of MIBC, treatment decision-making has become more complex given the multiple specialties involved in delivering different components of treatment. Since 2017, we performed two prospective, open-label clinical trials involving NAC for patients with MIBC [9,10]. In these two consecutive trials, patients who achieved a clinical complete response (cCR) were offered bladder preservation CRT. Because of the long-term follow-up and the increased number of patients, the present combined analyses could be useful to identify factors predicting durable DFS, and to provide an evidence-based recommendation for the selection of treatment strategy for MIBC according to the cCR after NAC.

The present study was performed using data obtained from two consecutive phase II trials evaluating NAC for MIBC: gemcitabine and cisplatin (GC; Clinicaltrials.gov, NCT030-61630) [9] and nivolumab plus GC (ONO-4538-X41; CRIS. nih.go.kr, KCT0003804) [10]. Both prospective trials were conducted in accordance with Good Clinical Practice and the Declaration of Helsinki. Study protocols were approved by the Samsung Medical Center (Seoul, Korea) Institutional Review Board, and all patients provided written informed consent before study entry. Detailed study designs and results have been published previously [9,10], and only data regarding disease recurrence and survival were updated at the time of the present analysis. In both trials, eligible patients with T2-T4aN0M0 MIBC received NAC with GC±nivolumab. Sixty-six patients who achieved cCR post-NAC were eligible to receive bladder preservation CRT. A cCR was evaluated following NAC, and defined as no tumors visible either on imaging studies or on urine cytology and cystoscopy. However, an additional 10 patients without cCR were included because they refused surgery and underwent CRT as an alternative option.

An extensive literature search of clinical trial results, guidelines, and real-world data was performed to identify criteria commonly used to predict higher risk of recurrence and the development of metastases. Endpoints included DFS and metastasis-free survival (MFS). Disease recurrence was defined as no evidence of disease both in the urinary tract (i.e., bladder and upper tract) and distant organs or lymph nodes, and was determined by follow-up imaging studies, urine cytology, and cystoscopy. MFS was defined as the time to the development of metastatic disease. For the present analysis, starting point of survival was set as the date of commencing CRT. Survival curves and their 95% confidence intervals (CIs) were presented according to the Kaplan-Meier method. Analyses for predictive DFS and MFS were performed using the Cox regression model. Variables used to identify predictive factors included age (> vs. < median), sex (male vs. female), T category at the time of study entry (T2 vs. T3/T4a), grade (World Health Organization 2 vs. 3), previous history of intravesical Bacillus Calmette-Guerin (BCG), presence of carcinoma in situ (CIS), multiple primary tumors, status of large (< 5 vs. > 5 cm), presence of hydronephrosis, Eastern Cooperative Oncology Group performance status (0 vs. 1), study treatment (with nivolumab vs. without), and the achievement of cCR following neoadjuvant chemotherapy. Interactions between these variables were represented by products of the corresponding variables. Non-significant variables were dropped one at a time, beginning with the least significant. Statistical tests were performed using the R package (https://www.r-project.org), and the cutoff date for the present analysis was November 2023.

A total of 76 patients treated with NAC followed by bladder preservation CRT were included (Fig. 1). Median time from study enrollment to data cutoff date was 65 months (range, 62 to 68 months) in the GC trial [9], and 44 months (range, 42 to 47 months) in ONO-4538-X41 [10]. Baseline patient characteristics are summarized in Table 1. The median age was 67 years and most (83%) were male. Sixty-six patients (87%) achieved cCR following NAC.

With a median follow-up of 64 months, bladder preservation CRT following NAC was generally well-tolerated, with low urinary tract symptoms being the most common late complication. Severe (grade 3 or 4) complications associated with CRT including intractable bleeding or recurrent infections were noted in eight patients. Disease recurrence was recorded in 43 patients (57%): intravesical only (n=20), metastatic only (n=16), and both (n=7). In 20 patients with intravesical recurrence only (Ta/CIS, n=11; T1, n=8; T2, n=2), transurethral resection and BCG treatment was given to 17 patients. Salvage cystectomy was performed in 10 patients. The most commonly observed metastatic site was the lymph nodes (n=20), followed by lung (n=7), bone (n=5), and liver (n=4). The median DFS was 46.3 months (95% CI, 25.1 to 67.5 months), and the median MFS was not reached (Fig. 2). At 5 years after starting CRT, the DFS and MFS rates were calculated to be 38% and 70%, respectively.

Baseline factors that showed individual predictive value in univariate models were used to examine their joint value in a multivariate model. We found a significant interaction between T category and cCR, suggesting the effect of NAC was higher in T2 than in T3/T4a tumors. In univariate analysis, T category, no BCG history, and cCR were significantly associated with DFS. Multivariate model revealed that DFS was independently associated with cCR (hazard ratio, 0.465; 95% CI, 0.222 to 0.976; p=0.043). In contrast, MFS was not affected by any of the baseline factors (Table 2).

The present exploratory post hoc analyses of our two prospective clinical trials demonstrated that bladder preservation CRT in MIBC patients who achieved cCR after NAC is feasible, and provides durable DFS and MFS. With a novel treatment strategy of bladder preservation CRT according to cCR, the 5-year DFS and MFS rates were 38% and 70%, respectively. In 26% of patients with intravesical-only recurrence, the disease could be controlled with transurethral resection or salvage cystectomy. Multivariate analyses revealed that the ideal candidate for bladder preservation CRT is the patient with a cCR following NAC.

Although cisplatin-based NAC followed by radical cystectomy is the current standard of care for patients with MIBC [2-4], the major drawback associated with radical cystectomy is significant morbidities that have profound impact on patient quality of life [6,11]. As noted in the present study, most patients are elderly, which can increase rates of postoperative complications or even impact eligibility to undergo radical surgery. Therefore, a major motivation for bladder preservation following NAC is to avoid complications caused by the removal of bladder and urinary diversion. For example, a pooled analysis of the Radiation Therapy Oncology Group (RTOG) studies involving bladder preservation CRT showed cCR in 69% of patients and low rates of muscleinvasive local failure (14% in 10 years) [12]. If the prognostic value of cCR can be reliably validated as a promising alternative endpoint, accurate determination of cCR would provide a paradigm shift in bladder preservation approaches for the management of MIBC [13]. Unfortunately, studies have demonstrated that cCR after NAC did not necessarily predict pCR based on cystectomy specimens [14,15]. Even with post-NAC transurethral resection and molecular biomarkers [16,17], the false downstaging rate reflects the continuing clinical dilemma of the inability to accurately restage tumor status after NAC. However, these two trials (active surveillance in RETAIN [16] and nivolumab maintenance in HCRN GU16-256 [17]) did not involve local treatment (i.e., CRT) in patients who achieved cCR. The need for salvage cystectomy in a subset of patients with intravesical recurrence raises the need for CRT to further optimize the likelihood of bladder preservation. In addition, patient selection is important for long-term DFS and/or MFS in bladder preservation approaches. Ideal candidates for CRT are patients with small, single tumors that can be completely resected, no CIS, and no hydronephrosis [2]. Perhaps one of the most interesting findings in the present analysis was the observation of similar DFS and MFS irrespective of baseline factors. Although it requires a validation in further studies, if a patient achieves cCR post-NAC, bladder preservation CRT can be considered a reasonable alternative to radical cystectomy.

The limitations of the present study include the nature of post hoc exploratory analyses, and possible selection bias as the two trials had no comparative arms. It is also unclear how the addition of immune checkpoint inhibitors to NAC, either concurrently and/or as maintenance therapy, could impact the rate of cCR [18]. Additionally, some patients with MIBC are cisplatin-ineligible; that population is frequently excluded from clinical trials. More than a few phase III clinical trials are under way in an MIBC neoadjuvant setting, both in cisplatin-eligible (NIAGARA, NCT03732677; ENER-GIZE, NCT03661320, Keynote 866, NCT03924856) and cisplatinineligible (Keynote 905, NCT03924895; VOLGA, NCT04960- 709) patients. For patients who choose bladder preservation and those who are not eligible for cystectomy, trials comparing immune checkpoint inhibitor combinations and CRT (SunRISe-2, NCT04658862) also are ongoing. Questions raised here and in other contemporary trials of MIBC should be addressed by these ongoing trials.

In conclusion, the findings obtained here demonstrate the feasibility and efficacy of the strategy of NAC followed by selective bladder preservation CRT according to the cCR. Although current assessment of cCR lacks validation, and NAC followed by radical cystectomy is still the standard of care in MIBC, bladder preservation CRT may also yield acceptable oncologic outcomes. Considering the substantial perioperative morbidity and the significant effect on patient quality of life, the results from ongoing and future clinical trials will be useful in decision-making among MIBC patients.