Role of Bladder EpiCheck^® in Bladder cancer: from primary diagnosis to surveillance

Abstract

Background

Cystoscopy remains the gold standard for initial diagnostic assessment and follow-up in bladder cancer (BC) patients. Novel diagnostic tools and biomarkers are needed to minimize reliance on cystoscopies whenever possible. Bladder EpiCheck^® (BE) represents one such approach with encouraging results. However, its diagnostic performance at primary BC diagnosis and its effectiveness across distinct non-muscle-invasive BC (NMIBC) risk groups during follow-up have not yet been described.

Methods

Prospective case-control study between July 2018 and December 2024 including two cohorts: cohort 1 comprised individuals during primary diagnosis of BC (n = 254); cohort 2 included BC patients in follow-up for disease recurrence (n = 138). Overall, 415 voided urine samples were collected (n = 254 for cohort 1; n = 161 for cohort 2) and analyzed using BE test (Nucleix Ltd). Urine cytology was conducted whenever sufficient residual urine was available. Diagnostic performance of both techniques in detecting BC was compared.

Results

Overall, BE sensitivity (SN) for detecting primary BC and recurrent NMIBC was 73% and 55%, respectively. When limited to high-grade (HG) tumours, BE SN increased to 92% and 87%, respectively. BE showed comparable specificity (SP) to cytology in primary (99% vs. 100%) and in recurrent tumours (91% vs. 100%). Diagnostic performance of BE was significantly better than that of cytology in primary (p = 0.013) and recurrent patients (p = 0.023). BE and cytology SN increased progressively across higher NMIBC risk groups. However, BE SN ranged from 47% in low-risk to 93% in very-high risk NMIBC patients while cytology SN ranged from 0% to 79%. Strikingly, EpiScore values were higher in those BC patients in follow-up who progressed into a muscle-invasive disease (p < 0.001), and patients with an EpiScore ≥ 90 were more likely to present high/very-high risk NMIBC or MIBC (OR = 1.93; 95% CI: 1.01–3.67; p = 0.046), ≥T1 stage (OR = 2.00; 95% CI: 1.09–3.71; p = 0.026) and HG tumours (OR = 2.59; 95% CI: 1.27–5.29; p = 0.008).

Conclusions

During initial diagnosis of BC, BE showed higher SN with similar SP compared with urine cytology. Diagnostic performance of BE increased gradually across EAU prognostic NMIBC factor risk classification, outperforming cytology in all risk groups. Moreover, EpiScore values are able to identify high risk BC patients, ≥T1 tumours and HG tumours.

Supplementary Information

The online version contains supplementary material available at https://doi.org/10.1186/s12967-026-08249-0.

Background

White light cystoscopy remains the gold standard for diagnosing and monitoring bladder cancer (BC), despite being a high-cost invasive procedure [1] with a sensitivity (SN) around 71% [2–4]. Urine cytology is commonly used in combination with cystoscopy due to its non-invasive nature and high specificity (SP) (90–96%). However, its diagnostic performance is limited by its low SN (approximately 40%), with particularly poor performance in the detection of low-grade (LG) tumours, where sensitivity rarely exceeds 16%. Furthermore, its accuracy is influenced by inter-observer variability [5, 6].

Several urinary biomarker tests have been developed to improve BC detection and reduce the number of invasive cystoscopies [7]. Among them, Bladder EpiCheck^® (BE; Nucleix Ltd., Rehovot, Israel) is a urine assay that utilizes quantitative real time PCR to analyse DNA methylation changes in a panel of 15 genomic biomarkers implicated in BC [8]. Several authors [8–10] have reported that the SN of BE is higher than that of urinary cytology for detecting recurrences in patients in follow-up for BC. Furthermore, the utility of BE has also been described for the detection of upper tract urothelial carcinoma [11, 12] and for predicting response to Bacillus Calmette-Guérin (BCG) therapy [13]. However, there is no data regarding BE usefulness in the primary diagnosis of BC in patients with a suspicion of malignancy. In addition, although it has been shown that BE performance is higher in high-grade (HG) disease, it remains unclear its effectiveness across the distinct European Association of Urology (EAU) non-muscle-invasive bladder cancer (NMIBC) risk groups [1, 14].

The objective of this study was to comprehensively assess the diagnostic performance and clinical applicability of BE in the initial evaluation of patients with suspected BC, and to further delineate its performance across the four EAU NMIBC risk groups.

Methods

Patients and samples

Prospective case-control study including 392 individuals (415 spontaneously voided urine samples) between July 2018 and December 2024. Cases were defined as patients in whom a bladder lesion was identified during cystoscopic examination performed by an expert urologist at our clinic. Controls were defined as individuals in whom no lesion was detected during diagnostic evaluation at the same clinic.

Individuals were divided into two cohorts; cohort 1 (n = 254 urine samples) comprises newly diagnosed BC patients (primary BC patients; n = 146) and individuals with clinical suspicion of BC but without evidence of BC (n = 108); cohort 2 (n = 161 urine samples) comprises NMIBC patients in follow up with a recurrent bladder tumour (n = 116 samples) and patients with non-recurrent NMIBC in follow-up (n = 45) (Fig. 1). In all tumour cases, diagnosis was histologically confirmed following transurethral resection of the bladder tumour (TURBT) (Table 1). After TURBT, eligible patients were treated with adjuvant intravesical treatment according to current clinical guidelines [1]. In cohort 1, 44 (30%) cases received Bacillus Calmette-Guérin adjuvant treatment. In cohort 2, 15 (16%) cases and 10 (22%) non-recurrent BC patients in follow-up received Bacillus Calmette-Guérin treatment and 2 (2%) cases and 2 (4%) non-recurrent BC patients in follow-up received mitomycin C treatment. Importantly, inclusion in cohort 1 did not preclude patients from also being included in cohort 2. In this sense, seven cases were included in both cohorts with urine samples collected at initial diagnosis and at recurrence.

Fig. 1.

Flowchart of both cohorts. Describes sample size for primary BC patients (cohort 1) and BC patients in follow-up (cohort 2). Abbreviations: BC: Bladder cancer; BE: Bladder EpiCheck

Table 1.

Demographic and clinicopathological features of (A) Primary BC patients and in (B) BC patients in follow-up

A) Primary BC (cohort 1)	Histologically confirmed BC patients	Patients with symptoms compatible with BC
Patients, n	146	108
Urine samples, n	146	108
Gender, women (%)	26 (18)	44 (41)
Age, median (range)	74 (42-95)	62 (28-94)
Medical Consultation, n (%)
Macrohaematuria	86 (59)	37 (34)
Microhaematuria	7 (5)	24 (22)
Imaging detected lesion	39 (27)	2 (2)
Urolithiasis	4 (3)	29 (27)
Urinary tract infection	8 (5)	0 (0)
BPH	0 (0)	10 (9)
Cystitis	0 (0)	3 (3)
Other	2 (1)	3 (3)
T stage, n (%)
PUNLMP	1 (1)	-
Ta	77 (53)	-
Tis	2 (1)	-
T1	41 (28)	-
T2	25 (17)	-
Grade, n (%)
Low-Grade	51 (35)	-
High-Grade	95 (65)	-
EAU risk group, n (%)
Low	47 (32)	-
Intermediate	16 (11)	-
High	39 (27)	-
Very High	19 (13)	-
MIBC	25 (17)	-
Clinical NMIBC risk factors
Tumour size, > 3 cm (%)	32 (26)	-
Number of tumours, n (%)
Single	94 (77)	-
2-7	19 (16)	-
≥8	8 (7)	-
Concomitant CIS, yes (%)	26 (21)	-
B) BC patients in follow-up (cohort 2)	Recurrent patients	Non-recurrent patients
Patients, n	93	45
Urine samples, n	116	45
Gender, women (%)	22 (19)	10 (22)
Age of diagnosis, median (range)	79 (55-94)	71 (47-90)
T Stage (last positive TURB), n (%)
PUNLMP	4 (4)	0 (0)
Ta	78 (67)	30 (67)
Tis	6 (5)	3 (7)
T1	19 (16)	10 (22)
T2	9 (8)	0 (0)
Not available	0 (0)	2 (4)
Grade (last TURB), n (%)
Low-Grade	67 (58)	22 (49)
High-Grade	49 (42)	23 (51)
EAU risk group, n (%)
Low	-	7 (16)
Intermediate	74 (64)	20 (44)
High	27 (23)	18 (40)
Very High	6 (5)	0 (0)
MIBC	9 (8)	0 (0)
Clinical NMIBC risk factors
Tumour size (last TURB), > 3 cm (%)	5 (5)	9 (20)
Number of tumours (last TURB), n (%)
Single	55 (51)	40 (89)
2-7	46 (43)	3 (7)
≥8	8 (7)	2 (4)
Concomitant CIS (last (TURB), yes (%)	16 (15)	4 (9)

Abbreviations: BC: bladder cancer; BCG: Bacillus Calmette-Guerin; BPH: benign prostatic hyperplasia; EAU: European Association of Urology; MIBC: muscle-invasive bladder cancer; PUNLMP: papillary urothelial neoplasm of low malignant potential; Tis: carcinoma in situ; TURB: transurethral resection of the bladder

Spontaneously voided urine samples were collected from both cohorts prior to cystoscopy or TURBT. Those patients with any other concomitant neoplasm or pelvic radiation were excluded from the study. NMIBC patients were stratified based on risk of recurrence and progression according to the EAU guidelines [1]. Tumour grade was scored according to the World Health Organization 2004/2016 grading system [15]. Follow-up period for cases was defined as the time from the tumour diagnosis or recurrence to the last follow-up visit. Follow-up period for controls was defined as the time from sample collection to the last check-up visit at our center. Recurrence was defined as any BC event between the time of sample collection and 18 months post-collection, regardless of grade or stage. Progression was defined as a higher T stage or disease dissemination occurring within 18 months of sample collection.

Bladder EpiCheck assay

BE test was performed in all samples following manufacturers’ instructions. Briefly, urine samples were centrifuged twice at 1,000 x g for 10 min to obtain cell pellet. DNA was extracted and digested using a methylation sensitive restriction enzyme that cleaves DNA if it is unmethylated. Samples were prepared for quantitative PCR. Results were analysed using proprietary EpiCheck software. BE test provides a quantitative score (EpiScore) ranging between 0 and 100 [8]. An EpiScore ≥ 60 indicated a positive result following manufacturer’s instruction.

Urine cytology

Urine cytology was conducted whenever sufficient residual urine was available. Urine samples were processed by cytocentrifugation and stained with Papanicolaou’s stain. Slides were evaluated by a genitourinary dedicated and expert pathologist who was blinded to the patient’s clinical history. Cytological diagnoses were classified according to The Paris System (TPS 2.0) for Reporting Urinary Cytology [16].

Statistical analysis

Descriptive statistics were used to summarize the study population characteristics. The diagnostic performance of BE, cytology and their combination (classification as positive when either test resulted positive) was assessed and compared using McNemar test. We were not able to estimate positive and negative predictive values due to the case-control design of the study. Chi squared test was used to determine differences in BE status according to tumour stage, histological grade or EAU NMIBC risk group. EpiScore value was compared across different NMIBC risk groups using U Mann Whitney and Kruskal Wallis test. A Receiver Operating Characteristic (ROC) curve was generated to determine the optimal EpiScore cut off value for detecting HG tumours in BC patients with a positive BE result. The cut off with the highest Youden Index was selected for analysis. Statistical analysis was performed using SPSS version 29.0 and R version 4.2.0. Statistical significance was set at α = 0.05.

Results

Patients’ characteristics

The clinicopathological characteristics of both cohorts is shown in Table 1.

Median follow-up of cohort 1 was 51 months (range = 2–76 months). During this follow-up period, 51 patients (35%) presented at least one recurrence, with 37 (25%) occurring within the first 18 months following sample collection. Median time to recurrence was 12 months (range = 2–58 months). Six patients (4%) presented progression within the first 18 months following sample collection. Median time of progression was 7 months (range = 3–9 months). Forty patients (27%) died during follow-up with 5 (13%) of them due to NMIBC. Median cancer specific death time was 14 months (range = 6–62 months).

Median follow-up of cohort 2 was 50 months (range = 1–75 months). During this follow-up period, 64 patients (55%) experienced at least one recurrence, of whom 47 (41%) did so within the first 18 months after sample collection. Median time to recurrence was 9 months (range = 3–53 months). Thirteen patients (15%) presented progression within the first 18 months following sample collection, with a median time to progression of 6 months (range = 3–14 months). Thirty-eight patients (40%) died during follow-up with 5 (13%) of them from NMIBC. Median cancer specific death time was 15 months (range = 1–44 months).

BE status and urine cytology results

In the primary BC cohort, BE and urine cytology results were obtained for 254 and 177 samples, respectively. BE was positive in 107 (42%) and urine cytology results in primary BC patients were the following: 98 (55%) were negative for HGUC, 14 (8%) were atypical urothelial cells, 14 (8%) were SHGUC and 51 (29%) were positive for HGUC (Fig. 2). Among the 177 samples with both BE and urine cytology results, 48 (27%) were positive on both tests, and all these patients had BC, 49 samples (28%) were positive on either test, with 48 (98%) of these patients presenting a tumour. In the remaining 80 samples (45%) who were negative on both tests, cystoscopy detected a lesion in 25 cases (31%).

Fig. 2.

BE status and urine cytology. (A) Describes BC patients from cohort 1 (Primary BC cohort; n = 254) and (B) patients from cohort 2 (Recurrent BC cohort; n = 161). Abbreviations: BC: Bladder cancer; FOD: Free of disease

In BC patients in follow-up (cohort 2), BE and urine cytology results were obtained for 161 and 144 samples, respectively. BE was positive in 68 (42%) and urine cytology results were the following: 95 (66%) were negative for HGUC, 16 (11%) were atypical urothelial cells, 10 (6%) were SHGUC and 23 (16%) were positive for HGUC. Among the 144 samples with both BE and urine cytology results, 22 (15%) were positive on both tests, and all these patients had BC, 41 samples (28%) were positive on either test, with 37 (90%) of these patients presenting a tumour. In the remaining 81 samples (56%) who were negative on both tests, cystoscopy detected a lesion in 46 cases (57%). Of note, proportion of positive test results in both BE and urine cytology significantly increased in patients with higher tumour stages (Additional File 1; Additional File 2, Table S1).

Diagnostic performance of BE and urine cytology

The diagnostic performance of BE and urine cytology in cohort 1 is depicted in Table 2 (Additional File 2, Tables S2-S19; Additional File 3). Overall, BE showed a SN of 73% for primary BC diagnosis, increasing to 92% in HG disease, with a SP of 99%. In comparison, urine cytology exhibited significantly lower overall and HG disease SN (42%, p < 0.001 and 60%, p < 0.001, respectively). Notably, BE SP was comparable to that of cytology (99% vs. 100%, p = NA). Interestingly, the only false positive BE case in this cohort had an EpiScore of 63, very close to the BE positivity cutoff. Urine cytology in this patient was negative. Notably, the patient presented haematuria during cystoscopy, and bladder wall thickening was detected on a CT scan both before and after urine sampling. Despite these findings, no tissue biopsy was obtained to further characterize the bladder wall thickening.

Table 2.

Diagnostic performance of BE status, cytology and the combined test

		Primary BC (cohort 1)			BC patients in follow-up (cohort 2)
		BE	Cytology	Combined test	BE	Cytology	Combined test
N samples		254	177	177	161	144	144
AUC (95% CI)		0.858 (0.811–0.906)	0.711 (0.637–0.784)	0.888 (0.838–0.937)	0.731 (0.652–0.811)	0.610 (0.515–0.705)	0.730 (0.644–0.815)
Sensitivity %, (95% CI)
	Overall	73 (65–79)	42 (34–51)	79 (76–89)	55 (46–64)	22 (15–31)	56 (47–65)
	HG	92 (84–96)	60 (49–70)	94 (97–100)	87 (74–94)	44 (30–59)	88 (74–95)
Specificity %, (95% CI)		99 (95–100)	100 (94–100)	98 (91–100)	91 (79–96)	100 (91–100)	90 (76–96)

Abbreviations: AUC: Area under the curve; BC: Bladder Cancer; BE: Bladder EpiCheck; HG: high-grade

In cohort 2, BE showed significantly higher overall (55% vs. 22%, p < 0.001) and HG disease SN (87% vs. 44%, p < 0.001) compared with urine cytology (Table 2). Moreover, BE SP was comparable to that of cytology (91% vs. 100%, p = NA). Of note, BE performance was similar when considering only patients with available cytology (Additional File 2; Table S20).

Diagnostic performance of BE combined with urine cytology

BE and urine cytology results were combined, classifying a case as positive if either test yielded a positive result (Table 2). This combined approach increased overall SN to 79% for the detection of primary tumours and 56% for detection of recurrent disease. For HG tumours, sensitivity rose to 94% and 88%, respectively. These improvements were statistically significant compared with urine cytology alone (p < 0.001 for overall primary BC and HG primary BC; p < 0.001 for overall recurrent BC and HG recurrent BC), but not when compared with BE alone (p = 0.25 for overall primary BC and HG primary BC; p = 1 for overall recurrent BC and HG recurrent BC). SP remained comparable to that of either test alone in both primary and recurrent BC patients (98% and 90%, respectively).

Diagnostic performance of BE and cytology across EAU prognostic factor risk groups

The entire study cohort was analysed to assess diagnostic performance across the EAU NMIBC prognostic factor risk groups and MIBC. However, only patients with available results for both BE and urine cytology were included in this analysis (n = 321; cases, n = 226; patients with no evidence of disease, n = 95). SN of BE, urine cytology, and the combined test progressively increased across the EAU NMIBC risk groups and muscle-invasive bladder cancer (MIBC) (Table 3). BE achieved an SN of 47% in low-risk and 92% in very high-risk patients, while urine cytology ranged from 0% to 79%. The combined test showed SN values ranging from 47% in low-risk to 96% in very high-risk BC patients. Both BE and the combined test exhibited statistically higher SN than that of urine cytology in all categories except in very-high risk group, with no significant differences between BE alone and the combined test (Additional File 4). SP was comparable among the three approaches (97% for BE, 100% for urine cytology and 95% for the combined test, p = NA).

Table 3.

Diagnostic performance of BE and cytology according to NMIBC EAU-defined risk groups and MIBC

	BE	Cytology	Combined test
N samples	321	321	321
Sensitivity %, (95% CI)
Overall	67 (61–74)	33 (27–39)	69 (62–74)
EAU risk group
Low	47 (31–63)	0 (0–11)	47 (31–64)
Intermediate	42 (32–53)	12 (6–20)	42 (32–53)
High	87 (76–93)	43 (32–56)	88 (78–94)
Very High	92 (75–98)	79 (60–91)	96 (80–99)
MIBC patients	91 (76–97)	63 (45–77)	97 (84–99)
Specificity %, (95% CI)	95 (88–98)	100 (96–100)	95 (88–98)

Abbreviations: BC: Bladder Cancer; BE: Bladder EpiCheck; EAU: European Association of Urology; MIBC: muscle-invasive bladder cancer

EpiScore analysis across EAU prognostic risk groups

Quantitative analysis of BE results using EpiScore values was performed in each of the NMIBC EAU risk groups. EpiScore differed significantly across all risk categories (p < 0.001), with patients in the high/very high-risk NMIBC and MIBC groups exhibiting higher EpiScore values than those in the low/intermediate NMIBC risk groups (Fig. 3). Importantly, patients who progressed to MIBC within 18 months of sample collection exhibited significantly higher EpiScore values than those who did not progress progression (median: 89 vs. 73; p = 0.002). Additionally, in the subset of NMIBC patients with positive BE results (n = 170), the relationship between EpiScore and prognostic factors was further evaluated. Patients with an EpiScore ≥ 90 were 1.93 times more likely to have a high/very-high NMIBC or MIBC (95% CI: 1.009–3.673; p = 0.046), twice as likely to present with a ≥T1 tumour (95% CI: 1.085–3.705; p = 0.026) and 2.59 times more likely to have a HG tumour (95% CI: 1.269–5.296; p = 0.008) (Fig. 4).

Fig. 3.

BE EpiScore according to NMIBC EAU risk group and MIBC. Median (black line) and interquartile range (grey lines) for each group are depicted. P values for paired comparisons: low vs. intermediate NMIBC groups (p = 0.651), high vs. very high NMIBC groups (p = 0.303), very high NMIBC vs. MIBC groups (p = 0.522). In this analysis all patients with BC in cohorts 1 and 2 were included (n = 262). Statistical significance: ns (p > 0.05), *** (p ≤ 0.001). Abbreviations: BC: Bladder cancer; EAU: European Association of Urology; MIBC: Muscle-invasive bladder cancer; NMIBC: Non-muscle-invasive bladder cancer

Fig. 4.

EpiScore values from BE positive BC patients. According to (A) NMIBC EAU risk group/MIBC, (B) tumour stage and (C) histological grade. In this analysis all patients with BC and positive BE in cohorts 1 and 2 were included (n = 170). Statistical significance: * (p < 0.05), ** (p ≤ 0.01). Abbreviations: BE: Bladder EpiCheck; EAU: European Association of Urology; HG: high-grade; LG: low-grade; MIBC: Muscle-invasive bladder cancer; NMIBC: Non-muscle-invasive bladder cancer; PUNLMP: papillary urothelial neoplasm of low malignant potential; Tis: carcinoma in situ

Discussion

BE is a novel, commercially available urinary biomarker test that, across published series, has demonstrated SN ranging from 62% to 90% and SP from 82% to 93% for the detection of recurrences in NMIBC patients during surveillance [1, 9, 10, 17–21].

Compared with urine cytology, BE has shown superior SN while maintaining comparable SP [22, 23]. Indeed, its potential role in the surveillance of NMIBC patients is already acknowledged in the EAU guidelines, which highlight its high SN to detect tumor recurrence, particularly in HG disease, and very high NPV to lay the foundation for its future implementation in NMIBC surveillance [1]. Notably, BE has already demonstrated that can help to reduce the number of cystoscopies during follow-up of NMIBC [7]. Furthermore, when used as a stand-alone assay for NMIBC surveillance, its performance was comparable to that of urine marker–based follow-up in the Prospective Randomized UroFollow Trial, which achieved sensitivity equivalent to the standard of care [24].

Despite the good performance of BE in NMIBC follow-up setting, evidence regarding its use at the time of primary BC diagnosis remains limited [7]. To the best of our knowledge, this is the first study that analyses BE performance at the time of initial diagnosis of BC. When assessing patients with BC-related symptomatology, test SN is critical to avoid missed diagnoses. At this early stage, BE showed high SN to detect bladder tumours (73%). Moreover, a high SP is also required to identify those patients who truly require further invasive diagnostic procedures. In this sense, BE showed SP values comparable to that of cytology. Importantly, this high BE diagnostic accuracy at primary BC diagnosis (AUC 0.858), exceeds both our own cytology results (AUC 0.731) and previously reported BE performance values in the follow-up setting [18, 25, 26]. Compared to other commercially available urine biomarkers for the primary diagnosis of BC, BE demonstrated superior SN compared with protein- and cell-based biomarkers and comparable SN to RNA-based biomarkers, with higher SP [27–29]. Despite the favorable performance of BE relative to other urine biomarkers, it remains uncertain whether its SN is sufficiently high to support routine clinical application at the time of primary diagnosis of BC. A comparative study between BE combined with cystoscopy and the current primary diagnostic standard is warranted to further assess the potential role of BE in this context.

In the context of NMIBC follow-up, we confirm BE diagnostic performance identified by previous studies [9, 10, 20, 30–33]. For example, our BE SN for detecting HG BC reached 87%, closely aligning with the mean SN of 87.9% reported by Caño Velasco et al., who analysed 15 studies encompassing a total of 3,761 patients [25]. Moreover, a recent meta-analysis by Fleshner et al. evaluated BE performance specifically during NMIBC surveillance [18], reporting an overall SN of 73%, which increased to 91% for HG NMIBC, results closely aligned with our own findings. The overall SP reported by these authors was 85%, slightly lower than the one found in this study (91%), while both remained below that of cytology [17, 34]. Interestingly, it has been reported that the SP of urine biomarkers is often lower in recurrent than in primary tumours [18], which is consistent with the findings of our study. Moreover, some BE false-positive results have been described as anticipatory positive effect that may predict later recurrences. Fleshner et al. reported that patients with positive BE and negative cystoscopy had five times higher risk of recurrence compared to those with negative BE [18]. In line with this, Mariappan et al. recently demonstrated that when BE-positive but cystoscopy-negative patients undergo biopsies under photodynamic diagnosis, recurrences missed by white-light cystoscopy can be detected, thereby increasing the specificity of BE from 87% to 93% [19]. This anticipatory positive effect has also been observed with other urinary biomarkers, such as Xpert Bladder Cancer Monitor, which it is described that can detect recurrences up to one year before histological confirmation [35]. Patients with a positive Xpert result and negative cystoscopy were with 2.7 times more likely to have recurrence than patients with a negative Xpert and negative cystoscopy result [36]. Considering this, urinary biomarkers including BE may improve follow-up by enabling early detection of recurrences in patients with negative cystoscopy.

On the other hand, the presence of a HG tumour is one of the most important prognostic factors for progression to MIBC [37, 38]. This underscores the importance of ensuring that diagnostic and surveillance methods perform optimally in detecting such HG lesions. In this context, several nomograms such as the updated EAU NMIBC [1] and the CUETO scoring system [39], incorporate histological grade as a key prognostic variable to stratify patients by their risk of progression to muscle-invasive disease. We observed that SN of both BE and cytology increased progressively across the EAU NMIBC risk groups. Moreover, BE significantly outperformed urine cytology in all but the very high-risk group, suggesting that BE could potentially replace cytology in this setting. This is consistent with previous studies indicating that BE performs better in patients with more aggressive tumours [17, 22, 40]. Despite outperforming urine cytology, BE SN remained moderate in low and intermediate risk groups, a limitation clinicians should consider when monitoring lower risk BC patients.

Given the marked variability of BE performance across specific risk groups, and considering the findings of Pierconti et al., who reported a strong correlation between quantitative results and HG tumours (OR = 10.45; p < 0.001) [40, 41], we analysed EpiScore levels to further assess tumour aggressiveness. Consistently with Pierconti et al., we also found an association between high EpiScore and high/very high risk NMIBC and MIBC patients. On top of that, EpiScore of ≥ 90 was associated with higher risk groups and with HG tumours. Moreover, previous studies have shown that methylation defects accumulate with increasing tumour stages and during disease progression in BC, suggesting their potential role as prognostic biomarkers [42, 43]. Our findings indicate that EpiScore may help identify these methylation defects and BE could provide greater prognostic value in daily clinical practice than previously anticipated if EpiScore is taken into account. For instance, a positive BE result with a EpiScore greater than 90 may identify patients at higher risk of aggressive disease, thereby supporting escalation of follow-up and treatment; conversely, a positive BE with an EpiScore below 90 may suggest a lower-risk profile. Such stratification could support physicians in tailoring disease management and treatment recommendations. However, these findings should be interpreted as hypothesis-generating, and additional studies designed to evaluate the association between higher EpiScore values and more aggressive tumours are warranted.

With the aim to increase assay performance, we combined the results of BE and cytology. The combined test performed statistically better than cytology alone but not better than BE alone in both primary and recurrent BC patients, indicating that cytology does not enhance diagnostic performance when used alongside with BE. Trenti et al. reported similar findings in a cohort of 243 NMIBC patients under follow-up. They found that combining BE with cytology could theoretically reduce the number of follow-up cystoscopies, but no significant improvement was observed compared to BE alone [22]. These results suggest that BE alone would be sufficient to achieve optimal diagnostic performance in both baseline and follow-up settings. The combination of both BE and cytology was also assessed when accounting for different risk groups, but its diagnostic performance did not differ from BE alone, supporting the notion that BE provides comparable diagnostic effectiveness to the combination of both tests, even when accounting for risk stratification.

Our study is not free of limitations. Firstly, as this was a case-control study, the diagnostic performance of the tests may have been overestimated. Moreover, urine cytology results were not available for all patients within the studied cohort, which may introduce selection bias and reduce statistical power. Regarding potential bias, we observed that the clinical characteristics were largely similar between patients with available cytology data and those without, suggesting that the impact of this limitation is likely to be minimal. A study including consecutive patients with suspicion of BC and consecutive patients in follow-up analysing both BE and cytology for all patients would allow to calculate predictive values for BE during initial diagnosis and follow-up [2, 44].

Conclusions

In the initial diagnosis of BC, BE demonstrated superior sensitivity and comparable specificity to that of urine cytology, and comparable sensitivity to cystoscopy supporting its potential use as a primary screening tool for patients with suspected BC. In addition, we show that diagnostic performance of BE improved gradually across EAU prognostic NMIBC factor risk groups reaching a sensitivity of 92% in very-high risk BC patients, outperforming cytology in all risk groups. Moreover, EpiScore values are able to discriminate high risk BC patients, ≥T1 tumours and HG tumours.

Supplementary Information

Below is the link to the electronic supplementary material.

Abbreviations

BC

Bladder cancer

BCG

Bacillus Calmette-Guérin

BE

Bladder EpiCheck^®

EAU

European Association of Urology

HG

High-grade

HGUC

High-grade urothelial carcinoma

MIBC

Muscle-invasive bladder cancer

NMIBC

Non-muscle-invasive bladder cancer

SHGUC

Suspicious for high-grade urothelial carcinoma

SN

Sensitivity

SP

Specificity

TURBT

Transurethral resection of the bladder tumour

Author contributions

LM and AA conceived and designed the study and were responsible for coordinating and supervising the research. Material preparation, data collection, and data analysis were carried out by MF, MI, FR, EG, and PG. The first draft of the manuscript was written by MF. All authors contributed to the revision of the manuscript and approved the final published version.

Funding

This research was partially supported by Nucleix, by Instituto de Salud Carlos III (ISCIII) through the projects “FI25/00165” and "PI24/01553", and co-funded by the European Union and private funding grants to A. Alcaraz and L. Mengual. The sponsors played no direct role in the study.

Data availability

The data presented in this study are available on reasonable request form the corresponding author.

Declarations

Ethics approval and consent to participate

All methods were carried out in accordance with relevant guidelines and regulations. All participants provided written informed consent (HCB/2013/8753) before being included in this study. The study methodologies conformed to the standards set by Declaration of Helsinki and were approved by the Clinical Research Ethics Committee of the Hospital Clinic of Barcelona (HCB/2018/0746; HCB2021/1168; HCB/2022/0557).

Consent for publication

Not applicable.

Competing interests

Lourdes Mengual certifies that all conflicts of interest, including specific financial interests and relationships and affiliations relevant to the subject matter or materials discussed in the manuscript (e.g., employment/affiliation, grants or funding, consultancies, honoraria, stock ownership or options, expert testimony, royalties, or patents filed, received, or pending), are the following: L. Mengual is a consultant for Nucleix. A. Antonio Alcaraz has received honorarium as speaker or part of advisory boards from the following companies: Astellas, Janssen, Ipsen, Bayer, Sanofi, Olympus and Intuitive Surgical. The remaining authors have nothing to disclose.