Neoadjuvant immunotherapy driven bladder preservation for muscle invasive bladder cancer

The standard treatment for muscle-invasive bladder cancer (MIBC) is radical cystectomy (RC) following neoadjuvant treatment. Patients lose their bladders and their functions after RC [1, 2], and urine is drained through a stoma bag on the abdominal wall, which severely impacts their quality of daily life and mental health. With the advancement of multidisciplinary treatment and the pursuit of a higher quality of life for patients, the treatment schedule for localized MIBC has gradually shifted from RC to comprehensive modes, including bladder preservation therapy [3]. For decades, the preferred bladder preservation method has been trimodal therapy (TMT), which combines maximal transurethral resection of bladder tumors (TURBT), chemotherapy, and external beam radiotherapy. However, the limitations of TMT are obvious, and its current clinical application rate is extremely low. For example, approximately 50% of patients are ineligible for cisplatin-based chemotherapy [4], and pelvic fibrosis caused by radiotherapy will increase the difficulty of radical surgery for patients who fail in bladder preservation therapy [3]. Bladder preservation combined-modality therapies (CMTs) based on neoadjuvant chemotherapy (NAC-CMT) [5, 6] and neoadjuvant immunotherapy (Neoimmu-CMT) [7-10] are widely used in clinical practice. Unlike chemotherapy, the tail effect of Neoimmu-CMT reactivates and boosts anticancer immunity in the periphery and tumor microenvironment (TME), maximizing the elimination of micrometastatic foci and circulating tumor cells that cause recurrence [11, 12]. Previously, we have systematically evaluated the efficacy and safety of various neoadjuvant treatment strategies for MIBC. We found that the overall response rate of the combined regimen with neoadjuvant immunotherapy as the core is extremely high. Specifically, the overall response rates of neoadjuvant immunotherapy combined with RC48-ADC and neoadjuvant immunotherapy combined with chemotherapy reached 75.5% and 60.2% [13, 14]. Through this series of studies, the survival of patients with MIBC has been significantly improved and guaranteed. If we can preserve the patients' bladders and their functions while treating the tumors, it will bring about a great improvement in their quality of life and mental health. Especially, many patients who achieve pathological remission after neoadjuvant treatment have a more urgent desire for bladder preservation therapy. There is still no clear specification for the comprehensive bladder preservation therapy model based on neoadjuvant treatment, but the demand for bladder preservation therapy is both urgent and substantial. Therefore, we currently face the following three key issues: ① How to carry out the comprehensive bladder preservation therapy based on neoadjuvant therapy? ② Among Neoimmu-CMT, TMT, and NAC-CMT, which therapy might show greater potential based on current evidence? ③ Which type of patients are suitable for bladder preservation therapy and can benefit from it? To support Neoimmu-CMT, we conducted a multicenter real-world cohort study via propensity score matching (PSM). We aimed to prove its feasibility and effectiveness compared to TMT or NAC-CMT and identified the appropriate candidates. Figure 1A illustrates the patient selection process. There were 97 and 36 patients in the Neoimmu-CMT and NAC-CMT groups, respectively. Table S1 shows the baseline differences between the two groups before and after PSM. Before matching, the Neoimmu-CMT group was older. After matching, both groups had similar baseline values and included 72 and 36 in the matched Neoimmu-CMT and matched NAC-CMT groups, respectively. Neoimmu-CMT resulted in significantly better disease-free survival (DFS) (hazard ratio [HR]: 2.112; 95% confidence interval [CI]: 1.247–3.576; p = 0.005) and bladder-intact (BI)-DFS (HR: 2.329; 95% CI: 1.138–4.770; p = 0.021) compared to that of NAC-CMT (Figure 1B). The 2-year BI-DFS rate for Neoimmu-CMT (90.28%, 95% CI: 82.25%–99.09%) exceeded that of the NAC-CMT group (71.59%, 95% CI: 58.06–88.27%). In contrast, overall survival (OS) did not differ significantly between the groups (HR: 0.441; 95% CI: 0.106–1.844; p = 0.262), potentially owing to the short duration of the overall follow-up. Thirty patients were assigned to the TMT group. Table S2 shows the baseline differences between the pre- and post-PSM. Before matching, differences in age, KPS, and clinical T were observed between the Neoimmu-CMT and TMT cohorts. Preliminary analysis suggested superior DFS (HR: 2.105; 95% CI: 0.944–4.695; p = 0.069) and BI-DFS (HR: 2.552; 95% CI: 0.915–7.115; p = 0.073) in Neoimmu-CMT (Figure S1). However, these differences in DFS (HR: 1.301; 95% CI: 0.711–2.382; p = 0.393) and BI-DFS (HR: 1.476; 95% CI: 0.657–3.316; p = 0.346) were no longer significant after adjusting for baseline characteristics via PSM (Figure 1C). The 2-year BI-DFS rate for Neoimmu-CMT (86.42%, 95% CI: 75.39%–99.07%) marginally exceeded that for TMT (80.89%, 95% CI: 66.56%–98.30%), though no statistical significance was found. No significant differences were observed in OS either before or after PSM. Univariate Cox analysis revealed significantly better DFS (HR: 9.105; 95% CI: 2.565–32.325, p = 0.001) and BI-DFS (HR: 15.888; 95% CI: 1.950–129.462; p = 0.010) in the clinical complete response (cCR) subgroup than in the clinical partial response (cPR) subgroup (Tables S3, S4). Additionally, the DFS (HR: 2.463; 95% CI: 0.873–6.953; p = 0.089) and BI-DFS (HR: 6.835; 95% CI: 1.619–28.859; p = 0.009) in the low clinical stage subgroup outperformed those in the high clinical stage subgroup. Further, multivariate competing risk Cox regression analysis highlighted the neoadjuvant therapy response and clinical T stage as independent prognostic factors for Neoimmu-CMT outcomes. Table S5 shows the baseline characteristics of patients in the immunotherapy group. Notably, the immunotherapeutic regimen had no statistically significant impact on bladder preservation outcomes. Regardless of whether they were combined with antibody-drug conjugate (ADC) (HR: 1.283; 95% CI: 0.224–7.359; p = 0.780) or chemotherapy (HR: 0.769; 95% CI: 0.255–2.323; p = 0.642), the DFS outcomes did not surpass those of patients receiving immunotherapy alone. This result underscores the significance of neoadjuvant immunotherapy for bladder preservation. Other conventional clinical parameters were not significant predictors of bladder preservation outcomes. As shown in Figure 2A, the tumors were divided into inflamed, excluded, and deserted phenotypes. Consistent with the results of previous studies showing that inflamed tumors were more sensitive to immunotherapy [13], patients with inflamed tumors were more likely to receive Neoimmu-CMT (Figure 2B). In addition, granzyme B (GZMB) expression was closely associated with successful bladder preservation (Figure 2C,D). However, programmed death-ligand 1 (PD-L1) expression was not associated with the Neoimmu-CMT outcomes. Therefore, the spatial distribution and functional activity of CD8 + T cells are potential biomarkers for identifying the populations likely to benefit from Neoimmu-CMT. Moreover, we assessed the potential relationship between various cell types in the TME and bladder preservation outcomes using single-cell sequencing for the first time. All cells were divided into 17 subclusters and subsequently categorized into nine cell types (Figure 2E and Figures S2, S3A). We further compared the differences in the proportion of each cell type between the groups with failed versus successful bladder preservation. The results indicated an increase in fibroblasts and a decrease in NK/T cells in the group with failed bladder preservation (Figure 2F), with minor differences in other cell types between the two groups. Based on these findings, we further reclustered T/NK cells into 15 subclusters, which were then classified into six major categories (Figure 2G, Figure S3B). Overall, CD8+ tissue-resident memory T cells were more common in the failed bladder preservation group, with no significant differences in the other cell types between the groups (Figure 2H). Cancer-associated fibroblasts (CAFs) were further divided into 16 subclusters and classified into three categories: inflammatory CAFs (iCAFs), myofibroblastic CAFs (myCAFs), and antigen-presenting CAFs (apCAFs) (Figure 2I,J). Notably, iCAFs, which are typically associated with immunosuppression, were significantly more prevalent in the group with failed bladder preservation, whereas apCAFs were more prevalent in the group with successful bladder preservation (Figure 2K). Our study is the first to reveal no significant differences in DFS, BI-DFS, or OS between Neoimmu-CMT and TMT. Consequently, the emerging Neoimmu-CMT regimen is on par with traditional TMT in terms of bladder preservation efficacy in patients with MIBC. However, Neoimmu-CMT is associated with fewer side effects compared to TMT. Moreover, benefiting from the tail effect of immunotherapy, Neoimmu-CMT demonstrates superior long-term efficacy. We also found that Neoimmu-CMT markedly surpassed traditional NAC-CMT in terms of efficacy, an advantage linked to the prolonged impact of immunotherapy. Previously, we demonstrated that neoadjuvant chemoimmunotherapy resulted in a higher response rate than chemotherapy or immunotherapy alone [13]. In this study, we found that the bladder preservation effects of regimens based on different neoadjuvant immunotherapies were similar in patients who achieved a cCR/PR. Currently, there is no established consensus on the appropriateness of Neoimmu-CMT in patients who achieve cPR [5, 15, 16]. To address this gap, we evaluated the prognostic distinctions between the cCR and cPR subgroups and found that patients who achieved a cCR were more amenable to Neoimmu-CMT than those who demonstrated cPR. Notably, there was no significant difference in OS between these subgroups, indicating that the cPR subgroup should also be considered for Neoimmu-CMT with more rigorous selection criteria. Furthermore, clinical staging serves as a critical predictive marker of the effectiveness of neoadjuvant immunotherapy in bladder cancer. Patients with lower-stage cancer were more likely to undergo bladder preservation. Biomarker analysis showed that the TME immune phenotype was closely related to bladder preservation outcomes. However, PD-L1, a clinical biomarker for immunotherapy efficacy, was unrelated to bladder preservation outcomes, which is consistent with our previous finding of no association between PD-L1 expression and neoadjuvant immunotherapy effectiveness [13]. Thus, the current clinical practice of using PD-L1 as an indicator of eligibility for immunotherapy requires further investigation in large-scale studies. This study faces limitations, including inherent bias from its retrospective design, though this was somewhat offset by employing PSM to equalize baseline characteristics across subgroups. The follow-up duration was brief and the sample size was limited, reflecting the recent introduction of neoadjuvant immunotherapy and chemoimmunotherapy, which constrains the availability of long-term data. However, our study boasts a larger sample size compared to many prior studies. Neoimmu-CMT is a novel bladder preservation model that is comparable to traditional TMT and superior to NAC-CMT. The development of Neoimmu-CMT will substantially expand its application in bladder preservation therapies. Future research should focus on the development of personalized bladder preservation therapies. The regimen for Neoimmu-CMT was tailored to the patient's clinical status and recommendations of the multidisciplinary team, potentially incorporating concurrent NAC or ADC therapies. Upon completion of neoadjuvant treatments and achievement of cCR/PR, the patients underwent bladder preservation therapy. Concurrently, data were collected from patients who underwent TMT and NAC-CMT during the same period to serve as the control group. The primary outcome measure was DFS. The secondary outcomes included BI-DFS and OS. Detailed methods are described in the Supplementary Methods. Jiao Hu: Conceptualization; methodology; software; data curation; investigation; validation; formal analysis; supervision; funding acquisition; visualization; project administration; resources; writing—original draft; writing—review and editing. Luzhe Yan: Data curation; investigation; validation; formal analysis; visualization; writing—original draft; writing—review and editing. Jinhui Liu: Data curation; investigation; validation; formal analysis; visualization; writing—original draft; writing—review and editing. Minfeng Chen: Conceptualization; methodology; software; data curation; investigation; supervision; resources. Yunbo He: Data curation; investigation; validation; formal analysis; visualization. Benyi Fan: Conceptualization; data curation; investigation; supervision; funding acquisition; resources. Bo Peng: Data curation; Investigation. Long Wang: Data curation; investigation. Weibin Hou: Data curation; investigation. Chao Li: Data curation; investigation. Bosen You: Data curation; investigation. Meng Zhang: Data curation; investigation. Wenze Li: Data curation; investigation. Jiaxing Wang: Data curation; investigation. Hongzhou Cai: Data curation; investigation. Shenglin Gao: Data curation; investigation. Yang Liu: Data curation; investigation. Dingshan Deng: Investigation; visualization. Huihuang Li: Formal analysis; visualization. Guanghui Gong: Data curation; investigation. Jiansheng Tang: Data curation; investigation. Chengyong Wang: Data curation; investigation. Xiaofeng Yang: Data curation; investigation. Liang Wei: Data curation; investigation. Guangzheng Lin: Data curation; investigation. Ruizhe Wang: Data curation; investigation. Xiao Guan: Data curation; investigation. Shiyu Tong: Data curation; investigation. Yangle Li: Data curation; investigation. Wei He: Data curation; investigation. Zhiyong Cai: Validation; software. Peihua Liu: Data curation; investigation. Yu Gan: Data curation; investigation. Yu Cui: Data curation; investigation. Yuanqing Dai: Data curation; investigation. Yi Cai: Data curation; investigation. Zefu Liu: Visualization; validation. Jiatong Xiao: Validation; visualization. Zhenyu Nie: Validation; visualization. Zhenyu Ou: Conceptualization; methodology; software; data curation; investigation; supervision; resources. Jinbo Chen: Conceptualization; investigation; funding acquisition; methodology; data curation; supervision; resources. Xi Guo: Conceptualization; methodology; investigation; resources; software. Xiongbing Zu: Conceptualization; methodology; software; supervision; funding acquisition; resources. This study was supported by the Noncommunicable Chronic Diseases-National Science and Technology Major Project (2024ZD0525700), the National Natural Science Foundation of China (81902592, 82070785, 82303760, 82373337), the China Postdoctoral Innovation Talents Support Program (BX20230431), the China Postdoctoral Science Foundation (2023M733951), Hunan Natural Science Foundation (2023JJ40946, 2024JJ2093), Hunan Provincial Key Area Research Plan (2023SK2016), Hunan Province Young Talents Program (2023RC3073, 2021RC3027), Changsha Natural Science Foundation (kq2208377), the National Clinical Research Center for Geriatric Disorders (2022LNJJ04), the Central South University Research Programme of Advanced Interdisciplinary Studies (2023QYJC029), the Health Research Project of Hunan Provincial Health Commission (202204054669), the Scientific Research Program of FuRong Laboratory (2024PT5102), and the Youth Science Foundation of Xiangya Hospital (2022Q20). This study was supported by the Bioinformatics Center of Xiangya Hospital. We acknowledge all authors participating in this study for data collection, preparation, and quality control. We thank Dr. Jianming Zeng (University of Macau), and all the members of his bioinformatics team, biotrainee, for generously sharing their experience and codes. We would like to thank Intelligent Weixi Biotechnology Co., Ltd (Changsha, China) and OE Biotech Co., Ltd (Shanghai, China) for providing single-cell RNA-seq and Dr. Yongbing Ba, Xiaohua Yao, Yao Lu, Zhiyu Tan, Junhao Yu and Hengyun Wang for assistance with bioinformatics analysis. The authors declare no conflicts of interest. This study was approved by the Ethics Committee of Xiangya Hospital (202212287). The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. Written informed consents were obtained from participants or their immediate families. All relevant data supporting the key findings of this study are available within the article and its supplementary files. The codes during the current study are available at https://github.com/jnhv/2025PSM. Due to ethical and legal restrictions, individual-level data from single-cell RNA sequencing cannot be made publicly available. All data are available upon request to the corresponding author and subject to local rules and regulations. Supplementary materials (methods, figures, tables, graphical abstract, slides, videos, Chinese translated version and update materials) may be found in the online DOI or iMeta Science http://www.imeta.science/. Figure S1. Comparisons between Neoimmu-CMT and TMT before PSM. OS: overall survival; DFS: disease-free survival; BI-DFS: bladder-intact disease-free survival. Figure S2. Expression of marker genes for nine major cell types. CD3D (NK/T cells); LYZ (myeloid cells); CD79A (B cells); MZB1 (plasma cells); TPSAB1 (mast cells); COL1A1 (fibroblasts); EPCAM (epithelial cells); PECAM1 (endothelial cells); DES (smooth muscle cells). Figure S3. Marker genes of each subcluster and cell types within T/NK cells (A) and fibroblasts (B). Table S1. Patient demographics before and after propensity score matching for Neoimmu-CMT and NAC-CMT subgroups. Table S2. Patient demographics before and after propensity score matching for Neoimmu-CMT and TMT subgroups. Table S3. Prognostic factors of DFS. Table S4. Prognostic factors of BI-DFS. Table S5. Baseline characteristics of patients in the immunotherapy group. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.