Introduction

Low-grade non-muscle-invasive bladder cancer (LG-NMIBC) is usually indolent, with a considerable risk of recurrence within 3 years (30–40%) but a low risk of progression (1–4%)1,2. Nevertheless, standard management often involves repeated transurethral resections of bladder tumour (TURBT), cystoscopies and intravesical therapies, resulting in cumulative physical, psychological and financial burdens for patients and health care systems alike3, frequently at the expense of quality of life4. Active surveillance (AS) offers a risk-adapted alternative strategy to immediate intervention in these patients.

In carefully selected individuals, AS might reduce overtreatment without compromising oncological safety, and concomitantly preserve future definitive treatment options, consistent with the established role of this approach in the management of patients with prostate cancer and small renal masses5. Reflecting this perspective, experts have advocated for de-intensified management in LG-NMIBC6. Current guidelines recognize AS as a valid option, but provide limited practical guidance7,8. Moreover, heterogeneity in inclusion criteria, follow-up protocols and outcome definitions across studies continues to impede standardization and broader implementation of AS9,10,11.

To address these challenges, we conducted a Delphi consensus with an international panel of bladder cancer experts, aiming to develop consensus-based statements on the use of AS in NMIBC, both for clinical practice and for future trials.

Methods

A modified five-step Delphi process was conducted in accordance with Accurate Consensus Reporting Document (ACCORD) guidelines (Supplementary Table 1) to develop consensus statements on AS in NMIBC. English was the official language throughout the entire Delphi process.

During phase 1, a steering committee of six international bladder cancer experts (R.C., L.S.M., B.P., A.M.K., R.H., P.G.) defined the scope and drafted an initial set of 36 statements across four domains: terminology (n = 5), inclusion and exclusion criteria (n = 21), AS follow-up monitoring (n = 4) and exit criteria (n = 6). Initial statements were developed after a targeted literature review and iterative discussion within the steering committee, referencing crucial guidelines and prior AS studies6.

Steering Committee members did not have the right to vote during the voting session. During phase 2, the statements were distributed through a web-based platform (Welphi, Lisbon, Portugal) to a multidisciplinary international panel of 70 invited experts, including urologists, pathologists, radiologists and oncologists, selected from the European Association of Urology (EAU) Guidelines Panels for NMIBC and muscle-invasive and metastatic bladder cancer; the International Bladder Cancer Group; and the EAU Young Academic Urologists (YAU) Urothelial Cancer Working Group. Participants rated each statement on a nine-point Likert scale (1 = strongly disagree, 9 = strongly agree) and could provide comments. An ‘unable to answer’ option enabled participants to abstain from statements outside their expertise. Participants who had not provided a response received a reminder e-mail at 14 and 30 days after the initial invitation. Responses were collected anonymously to minimize group influence; participants could access only aggregated feedback between rounds.

In phase 3, the steering committee reviewed all responses. Consensus agreement was defined a priori as ≥70% of participants scoring a statement 7–9 and <15% scoring 1–3, whereas consensus disagreement required ≥70% of participants scoring a statement 1–3 and <15% scoring 7–9 (ref. 12). Statements not meeting predefined thresholds for consensus were revised — based on participant feedback — to improve clarity, remove redundancy and resolve thematic overlap, and carried forward to a second round.

During phase 4, all first-round participants were invited to re-rate revised statements, with the opportunity to view both their initial responses and the overall distribution of scores.

In phase 5, second-round responses were analysed, and statements were classified as achieving consensus agreement, consensus disagreement or equivocal (if neither threshold was met). Final results were then synthesized and reported.

Results from consensus rounds

Of the 70 invited experts, 51 (73%) completed the initial Delphi round (July 2024), and 47 (92% of round-one participants) completed the second round (November 2024). The panel consisted predominantly of urologists (88% in round one, 89% in round two), with additional participation from oncologists (6% and 4%), two pathologists and one radiologist. Most participants were based in Europe (72% in both rounds), with further representation from North America (23%) and Asia (4%).

In the first round, 15 statements met the predefined consensus threshold (Supplementary Table 2A). Of the remaining 21 statements, eight were excluded by the steering committee owing to inconsistencies with items already accepted. The remaining 13 statements were revised in response to participant feedback and rephrased for clarity. These statements were resubmitted in the second round, resulting in five additional statements reaching consensus (Supplementary Table 2B), bringing the total number of consensus statements to 20 (Box 1 and Fig. 1).

Fig. 1: Visual summary of the statements that reached consensus in both Delphi rounds.
Fig. 1: Visual summary of the statements that reached consensus in both Delphi rounds.
Full size image

Schematic overview of the proposed active surveillance pathway, integrating the statements that achieved consensus and summarizing the crucial steps in patient selection, follow-up monitoring and triggers for intervention. LG-NMIBC, low-grade non-muscle-invasive bladder cancer; LG-UTUC, low-grade upper-tract urothelial carcinoma; OF, office fulguration; TURBT, transurethral resection of bladder tumour; WL, white light.

Panel recommendations and discussion

This international Delphi consensus established expert agreement across the crucial domains of AS for LG-NMIBC, including terminology, patient selection, follow-up monitoring and triggers for intervention. To our knowledge, this Consensus statement provides the first expert-derived framework to support the clinical implementation of this strategy, while also offering a foundation to harmonize contemporary practice and guide the design of future prospective studies.

Terminology

To promote consistent use of AS in LG-NMIBC, the panel agreed on some definitions aimed at establishing a clear framework distinguishing AS from immediate or subsequent treatment pathways (Box 1). Active surveillance is defined as a monitoring strategy for presumed LG-NMIBC with the aim of avoiding or deferring active treatment.

Immediate treatment is defined as a treatment with tumour destruction intent, without any intended period of surveillance.

Subsequent treatment refers to active treatment for bladder cancer after exit from an AS period.

The adoption of this precise terminology is crucial for minimizing heterogeneity in reporting outcomes, particularly by establishing a clear distinction between the initial procedure and any subsequent treatment pathway initiated following an AS exit event. This clarity is essential for accurate trial reporting and consistent clinical management.

Consensus was also achieved regarding the definition of disease progression from a pathological perspective. Specifically, pathological progression was defined as any upgrading or upstaging. Major guidelines, such as those from the EAU, often define progression more strictly — typically as development of muscle-invasive bladder cancer or metastases, but these events are rare in this patient population7. Thus, the panel used a more pragmatic definition encompassing any tumour upgrading or upstaging, which aligns with the criteria used by the International Bladder Cancer Group13. This broader definition is more sensitive than the one used in guidelines for capturing clinically significant events during AS.

Conversely, consensus on the definition of clinical progression was not reached. Although a majority of participants (74%) supported a definition based on measurable increases in tumour number or size, substantial disagreement (>15%) prevented formal adoption. This divergence reflects the variability observed in other cancers for which AS is an established management option. For instance, in prostate cancer AS protocols, histopathological confirmation is typically required to prompt treatment14, whereas for small renal masses, the decision to start treatment primarily relies on tumour size15. These differences highlight the challenge of developing a universal definition of clinical progression in NMIBC. Nevertheless, the panel identified clinical progression in LG-NMIBC as a crucial outcome requiring further refinement and standardization in future AS trials.

Exclusion and inclusion criteria

Based on expert consensus, AS should primarily be considered for asymptomatic patients with a presumed low-grade papillary recurrence following an initial diagnosis of LG-NMIBC as defined according to the WHO 2004 and 2022 classification (Box 1). In this context, the term ‘presumed’ refers to the cystoscopic appearance of a papillary low-grade lesion and also to the patient’s clinical history, including previous low-grade disease. This concept is further supported by results from published series showing that ~90% of recurrent papillary tumours following prior low-grade Ta disease and managed with AS are confirmed as low-grade at subsequent TURBT, supporting the reliability of this approach9. Furthermore, although visual cystoscopic assessment is not definitive, evidence shows that experienced urologists can predict low-grade disease with high accuracy (~86%)16. Additionally, restricting AS to asymptomatic patients excludes those with confounding symptoms such as gross haematuria or lower urinary tract symptoms, which warrant immediate investigation.

The consensus panel also supported extended eligibility for AS in two specific scenarios: patients with a presumed low-grade bladder recurrence after a previous diagnosis of low-grade upper tract urothelial carcinoma; and those with a presumed low-grade recurrence after completing a full course of adjuvant intravesical chemotherapy (including maintenance).

Last, the panel strongly emphasized that the initiation of AS requires comprehensive informed consent in the context of shared decision-making. Patients must receive clear counselling on the risks and benefits of AS versus immediate treatment (which might include office biopsy and fulguration). This essential discussion should cover the intensive monitoring schedule, the risk of tumour progression, the potential need for subsequent treatment and the substantial benefit of avoiding or deferring surgical intervention.

Tumour characteristics

The expert panel reached a consensus that AS should be restricted to papillary lesions ≤1 cm in diameter (Box 1). Consequently, patients with flat lesions must be excluded, owing to the risk of carcinoma in situ. The 1-cm cut-off aligns with most published AS protocols9,10,16. Larger thresholds (such as 3 cm) are generally used in standard NMIBC risk stratification, but surveillance of these lesions is considered unsafe owing to the risk of progression and the potential loss of eligibility for endoscopic ablation or fulguration. This smaller limit of 1 cm provides a crucial safety margin. Endoscopic size estimation might be guided by the cystoscope diameter (16 Ch = 5.4 mm) or by using a visual field 4 mm as reference points17. Notably, the ≤1-cm size criterion reflects the thresholds used in available AS studies rather than a validated safety cut-off and might be refined as prospective data on larger low-grade lesions emerge.

Unsolved criteria

Although the consensus panel identified criteria forming the core of patient selection, no consensus was reached on the number of lesions, as no specific cut-off (such as ≤3 or ≤5 lesions) achieved consensus. This evidence suggests that a fixed number of lesions is less important than a thorough assessment of factors such as tumour respectability, anatomical location, recurrence patterns, and patient comorbidities. However, clinicians should consider that a high tumour burden might complicate or prevent effective treatment (such as office fulguration) if the patient subsequently exits surveillance.

The inclusion of patients with a history of high-grade or T1 disease remains controversial. In a study in which patients with pTa–pT1, G1–G2 disease underwent AS, most progression events occurred in patients with T1 G2 tumours10. The “Bladder Italian Active Surveillance” Protocol, which enrolled patients with recurrent low-grade NMIBC receiving AS, also included a small subset of patients with prior T1a disease, but the outcomes were not clearly reported9. Considering these findings and the demonstrated risk of progression in this cohort, the possibility of including patients with a history of a high-risk tumour did not achieve consensus.

Active surveillance follow-up strategy

Defining an appropriate follow-up strategy is crucial for the safe implementation of AS. This topic generated the greatest variation in expert opinion, reflecting the need to balance oncological safety with the core AS goal of de-intensifying the treatment burden, which includes a parsimonious approach to surveillance visits.

Despite this variability, the panel achieved consensus on a minimum standard: AS follow-up monitoring should include at least white-light cystoscopy and urine cytology (Box 1). The inclusion of urine cytology constitutes a conservative safety baseline within an AS protocol. Routine cytology is not universally recommended for the follow-up monitoring of known LG-NMIBC owing to the low likelihood of high-grade recurrence7. Nonetheless, the panel favoured cytology as a non-invasive precaution. However, routine cytology at each cystoscopy, as included in some AS protocols, seems excessive and of limited clinical utility9, especially after the adoption of the Paris System, 2nd Edition, for reporting urinary cytology, which prioritizes the detection of high-grade urothelial carcinoma and, therefore, has limited sensitivity for low-grade tumours that typically affect patients receiving AS. Furthermore, urinary biomarkers might have a crucial role in future AS follow-up strategies, particularly for the early detection of high-grade disease. Biomarkers were not included in the present consensus owing to the lack of guideline endorsement and standardized evidence, but the evaluation and integration of biomarkers in this setting is a crucial focus of our ongoing and future research efforts.

No consensus was reached on the optimal frequency or duration of surveillance. This situation reflects a major evidence gap rather than a failure of consensus. Existing AS protocols vary widely18: most protocols recommend cystoscopy every 3 months for the first 2 years, followed by 6-month intervals thereafter10,16, whereas others propose a combined strategy of cystoscopy and cytology every 3 months during the first year, then every 6 months9. Early protocols were developed when AS was still investigational and required stringent monitoring to ensure safety9,10,19. As evidence increasingly supports the safety of AS, future research should focus on developing and validating risk-adapted, evidence-based follow-up models that reduce patient and health care burden without compromising oncological safety.

Last, no consensus statements were formulated regarding AS discontinuation, as currently available evidence does not define clear criteria for safely stopping surveillance in patients with long-term stable low-grade disease.

Exit criteria

The panel achieved strong consensus on specific triggers requiring immediate discontinuation of AS and transition to active treatment (Box 1).

AS should be discontinued if urinary cytology is positive or if high-grade urothelial carcinoma is suspected, as defined by The Paris System for Reporting Urinary Cytology, 2nd Edition. Furthermore, patients who develop new features that would have led to AS exclusion at baseline should also be considered for subsequent active treatment after an informed discussion. Additionally, the panel agreed on several observable changes during cystoscopy or in clinical presentation that warrant intervention. Discontinuation of AS is also recommended if a patient experiences substantial or rapid growth in one or more lesions, an increase in the total tumour burden to more than five lesions, or the onset of clinically significant gross haematuria (excluding mild, self-resolving or occasional episodes). Last, to ensure a smooth transition back to an active treatment pathway, the panel recommended that treatment options after exiting AS include a formal TURBT or office-based management (such as cold-cup biopsy and fulguration), depending on the clinical scenario and the patient preference. Emerging intravesical chemoablative systems, such as UGN-102 (mitomycin gel) and TAR-210, are likely to have an important role in patients exiting AS. However, these approaches had not yet been approved at the time of voting and, therefore, were not included in the consensus.

These well-defined exit criteria collectively form a robust framework that enables clinicians to offer AS confidently, knowing that clear and actionable triggers are in place to ensure patient safety.

Strengths and limitations

A crucial strength of this consensus was the high engagement of participants: the dropout between rounds was minimal, supporting the robustness and consistency of responses. Moreover, the panel was drawn from leading international guideline groups and collaborative networks, enhancing the credibility and applicability of the results.

Nevertheless, the Delphi method reflects structured expert opinion rather than empirical evidence and is inherently shaped by panel composition. Our panel was predominantly composed of urologists, which mirrors real-world practice but might have limited multidisciplinary perspectives. Moreover, geographic representation was not global, with most participants based in Europe and, to a lesser extent, in North America. This Delphi consensus should be regarded as a starting point for further research rather than providing definitive answers, offering a structured foundation to guide future studies and to standardize AS in LG-NMIBC.

Conclusions

AS has emerged as a promising de-intensification strategy for selected patients with LG-NMIBC, with the potential to reduce overtreatment and treatment-related burden without compromising oncological safety. However, AS broader implementation has been hampered by the lack of standardized criteria for patient selection, surveillance and intervention. In this context, this international Delphi consensus provides expert-based recommendations for the structured implementation of AS in LG-NMIBC. Consensus was achieved on crucial elements, including terminology, eligibility criteria and treatment triggers, whereas areas of disagreement such as follow-up monitoring highlight priorities for future research. These findings offer a framework to guide both clinical practice and the design of AS trials.