Postoperative continuous saline bladder irrigation reduces active urinary cancer cells: a prospective study in NMIBC

Abstract

Purpose

There is a lack of clinical evidence on whether further clinical strategies are needed after TURBT combined with immediate bladder instillation. This study intends to establish a reliable quantitative assay for active urinary cancer cells (AUCC) and to investigate the clinical efficacy of continuous saline bladder irrigation (CSBI) as a feasible option by analyzing the perioperative AUCC changes in TURBT.

Methods

An AUCC assay was developed and its reliability was verified by single-cell whole genome sequencing. Bladder cancer patients (N = 324) diagnosed by cystoscopy and pathologic biopsy and control individuals (N = 92) were included from 2021 to 2023 in the study. Enrolled patients with non-muscle invasive bladder cancer (NMIBC) underwent TURBT followed by immediate bladder instillation of epirubicin, after subgroups received CSBI or not, and AUCCs were tested on the first and fifth postoperative day. The patients were followed up for two years for postoperative recurrence.

Results

The AUCC assay achieved good detection accuracy, with a sensitivity of 0.821 and specificity of 0.902. AUCC increased on the first day after TURBT in combination with immediate bladder instillation, regardless of whether or not the patient received CSBI. However, AUCCs decreased more rapidly on the fifth day in patients treated with CSBI, and patients with concomitant risk factors benefited more from CSBI. The two-year follow-up results showed that high-risk patients with complex surgeries could benefit significantly from CSBI.

Conclusions

We pioneered a quantitative assay for AUCC and provided laboratory evidence that TURBT causes tumor cell dissemination and CSBI can be a further clinical strategy to reduce the risk of potential recurrence.

Supplementary information

The online version contains supplementary material available at 10.1007/s13402-025-01059-4.

Introduction

Bladder cancer, the second most common urologic tumor [1], often presents as non-muscle invasive (NMIBC) in over 70% of cases at diagnosis [2]. Standard treatment, transurethral resection of bladder tumors (TURBT), is often followed by BCG or chemotherapy instillation, but 60–70% of recurrence rates after TURBT highlight the need for improved strategies [3]. Continuous saline bladder irrigation (CSBI) is among the approaches [4, 5], though evidence for CSBI post-immediate bladder instillation is lacking.

The quest for a reliable assay to monitor tumor burden over time is essential to observe treatment effects and detect recurrence [6]. Current diagnostic methods, cystoscopy, and urine cytology have limitations including invasiveness or low sensitivity [7–9]. Non-cellular biomarkers, while advanced, do not quantify tumor burden effectively [10, 11].

Tumor cells exfoliated in urine, reflecting overall tumor burden and indicating recurrence, offer a promising biomarker. A study showed that PD–L1-positive urinary tumor cells predict response to immunotherapy and survival outcomes [12]. Microfluidics combined with surface markers have improved sensitivity, but don’t distinguish active cells amidst urine complexity [13–15].

This study introduces a telomerase reverse transcriptase (TERT)-based protocol to detect active cancer cells, validated in the lung [16], prostate [17], and glioma cancers [18], now applied to bladder cancer for active urinary cancer cells (AUCC). The prospective study with 416 subjects demonstrated high diagnostic accuracy of AUCC, which correlated with clinical characteristics. And AUCC increased rather than decreased after TURBT surgery, revealing that TURBT resulted in disseminated tumor cells that could not be eliminated by immediate bladder instillation. CSBI was found to improve the clearance of disseminated tumor cells, potentially reducing recurrence.

Methods

Study design

This prospective study cohort included patients with NMIBC (N = 324) and control individuals (N = 92) confirmed by cystoscopy and pathological biopsy at the Cancer Hospital of the Chinese Academy of Medical Sciences and China-Japan Friendship Hospital from 2021 to 2023. All subjects underwent cystoscopy and detection of AUCC. ROC curve determined the AUCC cut-off value for diagnostic assessment. Enrolled NMIBC patients underwent bipolar TURBT with saline irrigation during the procedure. Epirubicin (50 mg in 40 ml saline) was immediately instilled into the bladder after surgery and maintained for 1 h before drainage. After drug drainage, patients in the CSBI group received continuous saline bladder irrigation (ranging from 3 to 12 h, with a median irrigation time of 6 h), while patients in the non-irrigation group did not receive subsequent irrigation treatment. All patients’ AUCCs were detected respectively on the 1st and 5th postoperative days before leaving the hospital. The effect of CSBI on disseminated tumor cells was analysed concerning baseline and postoperative AUCC levels (Fig. 1). The patients were followed up for two years for postoperative recurrence. All subjects were anonymously coded and followed a double-blind method for sample collection and testing. Clinical information such as patient tumor grading or risk stratification followed NCCN guidelines (NCCN Guidelines Version 5.2024 Non-Muscle Invasive Bladder Cancer).

Fig. 1.

Study design flow chart. AUCC active urinary cancer cell, BLCA bladder cancer, TURBT transurethral resection of bladder tumor, CSBI continuous saline bladder irrigation

Cell culture and virus transfection

The human bladder cancer cell lines (J82, UM–UC-3) were purchased from the National Infrastructure of Cell Line Resource (Beijing, China) and kept by our laboratory. Cells were cultured in DMEM (Gibco) supplemented with 10% FBS(Gibco), 1% penicillin, and 1% streptomycin cultured at 37 °C in a 5% CO2 atmosphere and saturated humidity. Stably growing cells were transfected with oHSV1-hTERTp-GFP (MOI = 1) in a serum-free medium. Green fluorescence was observed under a fluorescence microscope (Leica) 24 h after transfection.

Assessment of capture efficiency and variability

J82 and UM–UC-3 cells were diluted and transfected with oHSV1-hTERTp-GFP according to a concentration gradient. The concentration gradients were 50, 100, 500, 1000, 5000, and 10000 cells, and for each concentration 3 samples were repeated. Bladder cancer tumor cells were captured by flow cytometry (BD Biosciences) after 24 h.

Urine sampling and AUCC detection

For all subjects, a minimum of 100 mL of the first urine after morning urine was collected according to the sample collection points in the study design. Urine samples were sent to the laboratory for downstream processing within 1 h of collection. After centrifugation, the samples were washed with PBS, resuspended in serum-free medium, transfected with oHSV1-hTERTp-GFP (MOI = 1), and stained with anti-CD45 antibody 24 h later. AUCC was then identified by flow cytometry (BD Biosciences). “CD45-/GFP +” cells were recorded as AUCC.

Flow imaging and single cell sorting

Urine samples of bladder cancer were treated with the AUCC detection protocol. For imaging, the treated samples were co-incubated with APC anti-human CD45 antibodies (BioLegend, clone: H130) and tested using Amnis ImageStream MK II (Luminex). The imaging data were analyzed using IDEAS 6.2. For flow sorting, the same protocol was used to separate CD45-/GFP + single-cell samples using BD FACSAria™ III (BD Biosciences) for whole genome sequencing.

Single-cell whole genome sequencing and analysis

Whole genome amplification (WGA) was performed on each CTC using REPLI-g® Single Cell kit (QIAGEN). Concentrations were measured using Qubit® dsDNA HS Assay Kits (Invitrogen) for quality control. All procedures are carried out according to the manufacturer’s instructions. The prepared library was then sequenced on the Illumina NovaSeq S4 platform at a depth of approximately 15X. GATK was used for SNP/InDel/CNV analysis, Lumpy was used for SV analysis, and Circos was used for panoramic display.

Statistical analyses

Standard software (IBM SPSS Statistics 26.0, Prism GraphPad 10.0) was used for statistical analysis of the study data. The nonparametric Mann–Whitney U test and Kruskal-Wallis test were used to compare patients’ AUCC levels. The Wilcoxon Paired Signed Rank test was used to compare the preoperative and postoperative AUCC of patients from different groups. The ROC curve was constructed based on the diagnostic efficiency of AUCC, and the AUC represented the diagnostic performance. All P-values were bilateral and P < 0.05 was considered statistically significant.

Role of the funding source

This project was supported by the Clinical and Translational Medicine Research Project of the Chinese Academy of Medical Sciences (Grant No. 2022-I2M-C&T-B-056). Funders only provided funding and had no role in the study design, data collection, data analysis, interpretation, and writing of the report.

Results

The TERT-based detection protocol accurately identifies AUCCs

To verify the applicability of TERT-based protocol for detecting live urinary cancer cells in bladder cancer patients, methodological confirmation was performed. First, after 24 h of transfection with oHSV1-hTERTp-GFP, the human bladder cancer cell line (J82, UM–UC-3) showed a high transfection efficiency and highly expressed GFP (Fig. 2A-B). The capture efficiency of the simulated samples of both cell lines reached 80–90%, and the AUCC detection of each concentration gradient conformed to a linear relationship (Fig. 2C–F). This assay protocol showed high accuracy and reproducibility when applied to bladder cancer cells.

Fig. 2.

Methodological validation of the TERT-based assay to identify AUCC. A, B Images of J82 (A) and UM–UC-3 (B) in brightfield or fluorescent field when transfected or not transfected with oHSV1-hTERTp-GFP; scale bar: 100um. C, D Linear regression (C) and recovery efficiency (D) of J82 cells assayed at different input cell volumes. E, F Linear regression (E) and recovery efficiency (F) of UM–UC-3 cells assayed at different input cell volumes. G Images of representative tumor cell, leukocyte and epithelial cell acquired by FlowSight, including brightfield, TERT-GFP (green), and CD45-APC (red) channels. AUCC was defined as GFP +/CD45 −, leukocyte as GFP −/CD45 +, and epithelial cell as GFP −/CD45 −. Scale bar: 10 μm

Next, the morphology and phenotype of patient-derived AUCCs were characterized by flow cytometry. Figure 2G shows tumor cells, normal epithelial cells, and leukocytes captured by the TERT-based assay. AUCCs (CD45 −/TERT +) were larger and highly expressed GFP (TERT +), but didn’t express the leukocyte common antigen CD45. All leukocytes highly expressed CD45, the vast majority were TERT − and very few were TERT +, probably naive leukocytes retaining proliferative activity. In contrast, the exfoliated normal epithelial cells did not express TERT or CD45. This result suggested that the AUCCs captured by this method can be transfected by oHSV1-hTERTp-GFP and highly expressed GFP.

To confirm the identity of the captured AUCCs as tumor cells, we sorted 9 GFP +/CD45 − tumor cells from 3 bladder cancer patients using flow cytometer, followed by single-cell WGS. Together, there was an abundance of variation in individual cells, and the 9 tumor cells showed high levels of CNV (Supplementary Fig. 1A–B). Notably different tumor cells from the same patient showed different levels of variants, confirming the high degree of tumor heterogeneity (Supplementary Fig. 1C–E). This confirmed the authenticity and reliability of the approach to capture tumor cells.

TERT-based AUCC assay has good diagnostic efficacy for bladder cancer

Clinical characteristics of the non-bladder cancer control (N = 92) and bladder cancer patient (N = 324) cohorts were detailed in Table 1. To evaluate the efficacy of the TERT-based AUCC protocol for detecting bladder cancer, we randomly divided the data set of 92 nonbladder cancer controls and 324 bladder cancer patients into a training set (70%) and a test set (30%) and performed a ROC analysis. ROC curve analysis showed an AUC of 0.910 (95% CI = 0.878–0.942) in the training set. The Youden index showed an optimal threshold of 2.5 AUCC/100 ml with a diagnostic sensitivity and specificity of 0.838 and 0.873 (Fig. 3A). Applying this threshold to the test set yielded an AUC of 0.904 (95% CI = 0.861–0.948), with a sensitivity of 0.812 and a specificity of 0.966 (Fig. 3B). To further validate the reliability of the diagnostic thresholds, a 10-fold cross-validation of the dataset was performed, and found that the diagnostic efficacy of the AUCC was stable in the total cohort, with a mean AUC, sensitivity, and specificity of 0.906 (95% CI = 0.878–0.934), 0.821 and 0.902, respectively (Fig. 3C and Supplementary Table 1). The diagnostic performance of AUCC increases as the risk classification of bladder cancer patients increases, and this performance is stable in the training set and test set (Fig. 3D, E and Table 2). Comparing the diagnostic performance of AUCC and traditional urine cytology in the test set, it was found that the sensitivity of AUCC was significantly better than traditional cytology, especially in low- and intermediate-risk patients (Fig. 3F and Table 2).

Table 1.

Baseline characteristics and AUCC counts of bladder cancer and control subjects

Patients variable	Bladder cancer	Controls	P
N	324	92
Age (range)	65 (28–89)	58 (26–71)	<0.001
Males, n (%)	248 (76.5)	53 (57.6)	0.010
Body mass index	24.6	24.1	0.297
Smoking history, n (%)	121 (37.3)	23 (25.0)	0.028
Bladder irritation, n (%)	155 (47.8)	8 (8.7)	<0.001
Urine
Leukocyte, n (%)	120 (37.0)	12 (13.0)	<0.001
Red blood cell, n (%)	129 (39.8)	4 (4.3)	<0.001
Recurrence, n (%)			–
Yes	116 (35.8)	–
No	208 (64.2)	–
Multifocality, n (%)
Yes	134 (41.4)	–	–
No	190 (58.6)	–
Tumor size, n (%)			–
<2 cm	195 (60.2)	–
≥2 cm	129 (39.8)	–
Submucosal invasiveness, n (%)			–
Yes	101 (31.2)	-
No	218 (67.3)	-
Unknow	5 (1.5)	–
Risk Stratification, n (%)			–
Low/ Intermediate risk	132 (40.7)	–
High risk	153 (47.2)	–
Unknow	39 (12.0)	–

Abbreviation: AUCC, active urinary cancer cell

Fig. 3.

Diagnostic efficacy of TERT-based AUCC assay for bladder cancer. A ROC curve of the AUCC assay in the training set, AUC = 0.910. Sensitivity = 0.838 and specificity = 0.873 at best threshold value. B ROC curve of the AUCC assay in the validation set, AUC = 0.904. C ROC curves of the AUCC assay in 10-fold cross-validation. D, E ROC curves and AUC of low & intermediate risk patients and high risk patients in the training set (D) and validation set (E). F The proportion of AUCC positive or negative patients among traditional cytology positive and negative patients. ROC receiver operating characteristic, AUC area under curve

Table 2.

Diagnostic performance of the TERT-based AUCC assay and traditional cytology

	Total patients (N = 324)	Low & intermediate Risk (N = 132)	High risk (N = 153)
Training set (N = 291)
AUC (95% CI)	0.910 (0.878–0.942)	0.885 (0.840–0.929)	0.939 (0.905–0.973)
Sensitivity	0.838	0.797	0.829
Specificity	0.873	0.873	0.952
Validation set (N = 125)
AUC (95% CI)	0.904 (0.861–0.948)	0.857 (0.786–0.927)	0.952 (0.907–0.996)
Sensitivity	0.812	0.729	0.896
Specificity	0.966	0.966	0.966
Traditional cytology (N = 125)
AUC (95% CI)	0.787 (0.733–0.840)	0.743 (0.662–0.825)	0.822 (0.752–0.893)
Sensitivity	0.573	0.486	0.644
Specificity	1.000	1.000	1.000

AUCC counts significantly correlate with clinical characteristics of bladder cancer

Further preoperative baseline analysis of AUCC showed a significant correlation with various clinical characteristics of bladder cancer. It revealed that AUCC levels were significantly higher in patients with submucosal infiltration than in those without (Fig. 4A), significantly higher in patients with high grade tumor than low grade (Fig. 4B), significantly higher in patients with multiple cancer foci than solitary ones (Fig. 4C), significantly higher in patients with tumor sizes ≥ 2 cm than < 2 cm (Fig. 4E), significantly higher in patients with bladder irritation signs than asymptomatic patients (Fig. 4F), and significantly higher in patients with positive urinary leukocytes/erythrocytes than negative ones (Fig. 4G, H). Furthermore, AUCC levels increased significantly with increasing risk classification (Fig. 4I), while there was no correlation with tumor recurrence or initial onset (Fig. 4D). Positive urine leukocytes/ erythrocytes may be due to tumor invasion of blood vessels leading to hematuria, as well as being the main reason for the appearance of bladder irritation. It suggested that the degree of infiltration, tumor grade, multifocality, tumor size, and hematuria are key factors in the dissemination and survival of AUCC.

Fig. 4.

AUCC counts significantly correlate with clinical characteristics of bladder cancer. A–G Correlation between preoperative baseline AUCC levels and clinical characteristics grouped by submucosal invasiveness (A), tumor grade (B), multifocality (C), recurrence (D), tumor size (E), bladder irritation (F), leukocytes (G), and erythrocytes (H). Data were tested using the non-parametric Mann–Whitney U test. I Correlation between preoperative baseline CTC levels and risk classification. Data were tested using the Kruskal-Wallis test. All data were shown as the median. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001

CSBI effectively accelerates disseminated tumor cell discharge due to TURBT

Analysis of perioperative AUCC levels in 324 bladder cancer patients who underwent TURBT revealed a significant increase in AUCC on the 1st postoperative day, suggesting that segmental resection indeed triggers the dissemination of numerous tumor cells (Fig. 5A and Supplementary Table 2). Since this method detected only viable tumor cells, it suggested that immediate postoperative bladder instillation of chemotherapeutic agents was insufficient to eliminate disseminated tumor cells. With urine flushing or CSBI, there was a decrease in AUCC on the 5th postoperative day. Although this trend was present in both groups of patients who did and did not receive CSBI, the trend of decreasing AUCC was more significant in the CSBI group of patients on the 5th postoperative day, which was significantly lower than that of the baseline level of AUCC (Fig. 5B).

Fig. 5.

CSBI effectively accelerates disseminated tumor cell discharge due to TURBT. A Changes in AUCC counts before and after TURBT in 324 NMIBC patients. B Correlation between CSBI and the changes in AUCC levels before and after TURBT. C, D Correlation between perioperative AUCC changes and CSBI in patients with submucosal invasive (C) and noninvasive (D) subgroups. E, F Correlation between perioperative AUCC changes and CSBI in patients with low grade (E) and high grade (F) subgroups. G, H) Correlation between perioperative AUCC changes and CSBI in patients with multifocal (G) and non-multifocal (H) subgroups. I, J Correlation between perioperative AUCC changes and CSBI in patients with tumor size ≥ 2 cm (I) and < 2 cm (J) subgroup. K, L Correlation between perioperative AUCC changes and CSBI in patients with bladder irritation (K) and asymptomatic (L) subgroup. M, N Correlation between perioperative AUCC changes and CSBI in patients with urine leukocyte positive (M) and negative (N) subgroup. O, P Correlation between perioperative AUCC changes and CSBI in patients with urine erythrocyte positive (O) and negative (P) subgroup. Q, R) Correlation between perioperative AUCC changes and CSBI in patients with low & intermediate risk (Q) and high risk (R) subgroups. All data were shown as median and tested using the Wilcoxon Paired Signed Rank test. *p < 0.05; **p < 0.01; ***p < 0.001; ****p < 0.0001

Given the results of the baseline analysis that the degree of infiltration, tumor grade, multifocality, tumor size, and hematuria were critical for tumor cell dissemination and survival in the urine, further critical subgroup analyses of perioperative AUCC levels were performed. The AUCC levels on the 5th postoperative day were found to be significantly lower than the baseline levels in CSBI patients with high risk, high grade, multifocal tumors, tumor sizes ≥ 2 cm, bladder irritation symptoms, and urine leukocytes/ erythrocytes positive. This difference was not significant in patients with low/intermediate risk, low grade, solitary tumors, tumors < 2 cm, asymptomatic, and urine leukocyte/ erythrocyte negative (Fig. 5E–R). Additionally, patients with TURBT benefited from CSBI regardless of the presence or absence of submucosal infiltration (Fig. 5C–D). It provided direct laboratory evidence that TURBT led to tumor cell dissemination and demonstrated that CSBI accelerated the clearance of disseminated tumor cells out of the urinary tract and reduced retention, thereby reducing the risk of colonization and recurrence.

High-risk patients with complex surgeries may benefit more from CSBI

Given that the laboratory evidence in this study demonstrated that the main beneficiaries of the flushing effect of CSBI on AUCC are high-risk patients with risk factors for recurrence, we followed up the postoperative recurrence of high-risk patients in the cohort for 2 years and ultimately obtained follow-up data for 104 high-risk patients (Supplementary Table 3). The high-risk patients in this follow-up cohort underwent postoperative adjuvant therapy with intravesical BCG or chemotherapeutic drug instillation in accordance with guidelines. Among patients who received CSBI, 81.8% (45/55) received postoperative intravesical BCG infusion and 18.2% (10/55) received intravesical chemotherapy infusion (epirubicin, pirarubicin or gemcitabine); among patients who did not receive CSBI, 83.7% (41/49) received postoperative intravesical BCG infusion and 16.3% (8/49) received intravesical chemotherapy infusion. The distribution of different adjuvant treatment regimens did not correlate with CSBI, so between-group differences due to instillation regimens can be excluded. It was found that high-risk patients who received postoperative CSBI had a trend toward better DFS within 2 years than those who did not receive CSBI, especially the long-term recurrence results over 1 year (Fig. 6A). Based on the finding in this study that surgery led to massive dissemination of AUCC, high-risk patients were further grouped according to the surgery time. It was finally found that high-risk patients with longer surgery time (surgery time ≥ 60 min) could benefit more from CSBI (Fig. 6B,C). Although no significantly statistical difference was obtained in the current results (p = 0.053) due to the small cohort of patients, adequate trends have been demonstrated that require further confirmation in expanded cohort studies.

Fig. 6.

CSBI can effectively reduce the recurrence of high-risk patients with complex surgeries. A Kaplan–Meier curves for disease-free survival between CSBI patients and non-CSBI patients among high-risk patients. B Kaplan–Meier curves for disease-free survival of CSBI patients versus non-CSBI patients in patients at high risk and with an operative time ≥ 60 min. C Kaplan–Meier curves for disease-free survival of CSBI patients versus non-CSBI patients in patients at high risk and with an operative time < 60 min. P values were calculated by the log-rank test

Discussion

Our study pioneered a quantitative assay for active urinary cancer cells (AUCC), demonstrating high diagnostic accuracy in a 416-subject cohort. The cross-validation results indicated an AUC of 0.906 for bladder cancer diagnosis, with sensitivity at 0.821 and specificity at 0.902, which was significantly better than conventional cytology. AUCC levels correlated significantly with bladder cancer recurrence risk factors, including risk classification, invasiveness, multifocality, tumor size, bladder irritation, and hematuria. Post-TURBT changes in AUCC levels revealed massive dissemination of tumor cells not eliminated by immediate instillation, providing direct evidence for recurrence due to tumor cell reimplantation. This supports the need for aggressive clinical management to reduce AUCC. In patients receiving CSBI, especially patients with recurrence risk factors, AUCC levels significantly decreased after treatment, unlike those not receiving CSBI, suggesting CSBI’s potential to reduce AUCC residence time and recurrence risk. Clinical evidence further confirmed the efficacy of CSBI, especially in high-risk patients with complex surgeries.

Advancements in urine-based liquid biopsy for bladder cancer diagnosis have been notable [19], with molecular studies involving Xpert®Bladder, Cxbladder, Urodiag, adxbladder, etc [20, 21]. The advent of advanced sequencing technologies has enabled whole genome and single-cell RNA sequencing of urinary exfoliated cells for cancer detection [22–24]. Yet, these techniques struggle to quantify tumor burden or track changes over treatment. Urinary exfoliated cells offer a more direct reflection of tumor burden, which is beneficial for surveillance, but most exfoliated cells will die quickly due to urine osmotic pressure and treatment, leaving only a small number that can remain active and colonize the bladder, causing recurrence [25]. Identifying and capturing AUCC is crucial for monitoring bladder cancer. Traditional methods combining physical properties with surface markers are limited by urine composition and cannot discern cell viability. Our assay, using oHSV1-hTERTp-GFP to identify viable tumor cells, effectively filters out dead cells, bypasses non-tumor elements, and confirms active tumor cells with high accuracy in both training and validation cohorts.

Clinical correlation analysis has shown that clinical features are significantly linked to the spread and survival of AUCC. Specifically, deeper tumor invasion and larger tumor extent are associated with higher AUCC levels, potentially indicating a higher risk of bladder cancer recurrence [26]. This correlation was not observed in previous studies [15], suggesting that it applies to active tumor cells rather than all exfoliated cells, including dead ones.

Our study is the first to report an increase in AUCC levels after TURBT, indicating that segmented resection of the procedure can lead to significant tumor cell dissemination and survival. Although tumor cell dissemination due to incomplete resection may occur, the likelihood of incomplete resection in the cohort is low given that the median time to recurrence for recurrent patients was 18 months (from 3 to 42 months), i.e., the vast majority of recurrent patients did not have incompletely resected lesions detected on their first cystoscopic follow-up within 3 months after surgery. TURBT actively fragments bladder tumors during segmented resection, theoretically causing numerous tumor cells to float in the urine, risking re-implantation into the bladder wall [25, 27, 28]. Several genomic studies of bladder cancer have found that multifocal tumors in the same patient are of monoclonal origin, which indirectly proves the theory of floating tumor cell re-implantation [29–31]. The study provides direct laboratory evidence of tumor cell dissemination and re-implantation due to TURBT, highlighting the need for improved TURBT techniques and more proactive clinical strategies to reduce AUCC. Current EAU and AUA guidelines support a single immediate instillation of chemotherapy after transurethral resection to combat spreading tumor cells [32, 33]. While this approach has been shown to reduce the risk of recurrence in patients with NMIBC [34, 35], it has also been found to lack efficacy in subgroups at high risk of recurrence (patients with multiple tumors, tumors ≥ 3 cm, T1, or highly recurrent tumors) [36]. In addition, our study showed that single immediate instillation of epirubicin after TURBT did not completely eliminate AUCC. Although the effectiveness of immediate single instillation also depends on its administration immediately after surgery, the clinical protocol at our study center is that a single immediate instillation of chemotherapeutic agents is performed immediately after surgery, usually within 1 h of the end of the surgery, so the effectiveness of a single immediate instillation can be guaranteed. In addition, some patients (105/324) had received previous perfusion therapy before enrollment and may have had post-resistant recurrence (5.9% had received intravesical BCG and 26.5% had received chemotherapeutic agents), but the significant increase of postoperative AUCC in the majority of patients except for this suggests that non-resistant tumor cells can also survive and persist in the bladder.

This study’s prospective cohort analysis first showed that CSBI can flush out disseminated tumor cells, hastening their removal, reducing dwell time in the bladder, and potentially lowering the risk of regrowth and recurrence, with enhanced benefits for high-risk patients, especially those with complex disease and longer surgery times. The effect of CSBI on preventing recurrence was comparable to immediate chemotherapy instillation, and it offered superior safety and fewer side effects, presenting as a viable alternative to mitigate tumor cell spread and recurrence risk [4, 5]. Retrospective studies reported the benefits of CSBI in improving recurrence-free survival compared to non-irrigation patients [37, 38]. Although a study reported that CSBI did not improve oncologic outcomes in patients with NMIBC treated with thulium laser en bloc resection of the bladder tumor (TmLRBT) combined with immediate bladder perfusion chemotherapy [39], TmLRBT uses complete resection, which is different from TURBT in terms of postoperative dissemination of tumor cells.

In particular, in accordance with the usual thinking that urinary tumor cells should drop significantly after surgical removal of the tumor, our findings unexpectedly demonstrated that the elimination of AUCC from the bladder was not as rapid as expected. And AUCC were significantly increased on the first postoperative day compared to the preoperative day, regardless of whether or not the patient had received CSBI. Notably, even patients who underwent CSBI had their AUCC increased more on the first postoperative day than non-CSBI patients. Based on this result, it is hypothesized that CSBI may be able to wash down tumor cells at the surgical lesion to a greater extent, accelerating their exfoliation and allowing them to float in the bladder urine. However, it was limited by the rate of expulsion of cellular components from the bladder, and could not be completely cleared on the first postoperative day, thus causing the AUCC to increase rather than decrease on the first postoperative day, and was more significant than that of the non-CSBI group. This also proved that CSBI accelerated the clearance of AUCC from the bladder, and thus on the fifth postoperative day, the decrease in AUCC was more significant in the CSBI group. Since the patients in our study center were discharged 5 days after TURBT, samples at subsequent time points could not be obtained, and therefore the exact time variation of AUCC clearance could not be known, and the above speculation has not been confirmed. Therefore, to further demonstrate its clinical effect, a 2-year follow-up of selected high-risk patients was performed in this study, and the results confirmed the trend of CSBI to reduce recurrence in the group of high-risk and surgically complex patients, although no significant difference was obtained.

This study was double-blind and did not interfere with clinical decision-making; only sample collection and testing were done during the patient’s visit. Therefore, patients undergoing CSBI were a normal clinical practice, with the main goal of reducing the risk of bleeding and avoiding ureteral blockage. The aim of this study was to investigate the role of CSBI in reducing the recurrence rate and the final clinical effect of this process. However, the final statistics showed that not intervening in the clinical decision to strictly randomize the grouping of patients on whether to undergo CSBI or not, resulted in more patients with a higher risk of bleeding in the CSBI group, as the clinic preferred to perform CSBI on patients with larger and more tumors to reduce postoperative bleeding. This may account for the trend of differences in clinical recurrence data observed only in high-risk patients. Therefore, in order to reduce the effect of bias caused by the uneven distribution of patients, subgroup analysis of patients was performed based on key clinical characteristics, which could to some extent ensure that tumor characteristics were similar in the paired subgroups. AUCC in some patients who relapsed after resistance may be resistant and not reflect the clinical benefit of immediate postoperative perfusion. In addition, the study’s validation was limited to two centers, necessitating a larger, multicenter confirmation. While perioperative AUCC analysis supports the clinical value of CSBI, long-term clinical evidence is pending from ongoing cohort monitoring.

The TERT-based AUCC assay is accurate and effective for bladder cancer screening and monitoring, with AUCC levels well correlated with clinical risk factors, indicating the potential of recurrence. The study also underscores the need for strategies to curtail AUCC spread due to the effect of tumor cell dissemination of TURBT. CSBI stands out as an effective postoperative intervention to clear disseminated AUCC.

Electronic supplementary material

Below is the link to the electronic supplementary material.

Abbreviations

TURBT

Transurethral resection of bladder tumors

AUCC

Active urinary cancer cells

CSBI

Continuous saline bladder irrigation

NMIBC

Non-muscle invasive bladder cancer

BCG

Bacillus Calmette-Guérin

TERT

Telomerase reverse transcriptase

GFP

Green fluorescent protein

WGA

Whole genome amplification

ROC

Receiver operating curve

AUC

Area under the receiver operating curve

TmLRBT

Thulium laser en bloc resection of the bladder tumor

WGS

Whole genome sequencing

SNP

Single nucleotide polymorphism

InDel

Insertion and deletion

CNV

Copy number variation

Author contributions

J.S., W.Z. and K.Z. conceived and designed the study. Q.Z. completed the methodological validation in cell lines. Q.Z., X.Y. and C.C. collected clinical samples. Q.Z. and X.Z. completed single cell and tissue sample collection, library construction and sequencing. Q.Z., P.X. and D.W. performed genomic and transcriptomic data analysis. Q.Z., Y.D., G.D. and L.W. completed clinical data collection and analysis. H.S., Y.G., L.L. and X.B. assisted with data analysis. Q.Z. and W.Z. wrote the original manuscript. S.C., K.Z. and J.S. supervised the project and revised the manuscript, and all authors contributed and provided feedback.

Funding

This project was supported by the Clinical and Translational Medicine Research Project of the Chinese Academy of Medical Sciences (Grant No. 2022-I2M–C&T–B-056). Funders only provided funding and had no role in the study design, data collection, data analysis, interpretation, and writing of the report.

Data availability

Single-cell WGS data supporting this study’s findings are available from the corresponding author upon reasonable request. No new algorithms were developed for this study. All codes generated for analysis are available.

Declarations

Ethics approval and consent to participate

The study complied with all relevant ethical regulations and was approved by the Ethical Committee of National Cancer Center/Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College (No. NGC2022C-490). All patients provided oral and written informed consent for participation and publication.

Consent for publication

Not applicable.

Competing interests

The authors declare no competing interests.