6.2 EAU Risk Stratification for NMIBC

6.2 EAU Risk Stratification for NMIBC

Accurate risk stratification in non-muscle-invasive bladder cancer (NMIBC) is the cornerstone of clinical decision-making. Because NMIBC encompasses a biologically heterogeneous spectrum of diseases—ranging from indolent, non-invasive papillary lesions with negligible progression risk to aggressive, poorly differentiated neoplasms that rapidly progress to life-threatening muscle-invasive bladder cancer (MIBC)—stratification dictates the frequency of cystoscopic surveillance, the duration of adjuvant intravesical therapy, and the timing of radical cystectomy.

---

1. Evolution of Prognostic Models: EORTC, CUETO, & the 2021+ EAU System

Historically, risk assessment relied on scoring models developed by cooperative oncology groups:

• EORTC Risk Tables (2006): Developed by the European Organisation for Research and Treatment of Cancer based on 2,596 patients treated primarily with adjuvant intravesical chemotherapy (not maintenance BCG). Weighted scores were assigned to six variables: number of tumors, tumor diameter (≥ 3 cm), prior recurrence rate, T category (Ta vs T1), concurrent CIS, and WHO 1973 grade (G1, G2, G3). While effective at predicting short- and long-term recurrence, the EORTC tables overestimate progression risk in individual patients treated with modern maintenance BCG.
• CUETO Scoring Model (2009): Developed by the Spanish Oncology Group based on 1,062 patients treated with 6-month maintenance BCG. Its seven inputs are sex, age, prior recurrence status, tumour number, T category, WHO 1973 grade and associated CIS; tumour diameter is not a CUETO input. It predicts recurrence and progression in this BCG-treated setting, with lower recurrence estimates than EORTC 2006.
• The Updated 2021+ EAU NMIBC Risk Calculator: To reconcile discrepancies between the WHO 1973 and WHO 2004/2016 grading schemes and provide contemporary risk estimates, the EAU Non-Muscle Invasive Bladder Cancer Guidelines Panel developed an updated individual prognostic calculator. Derived from 3,401 primary patients receiving TURBT with or without chemotherapy, it excluded adjuvant BCG and predicts progression, not recurrence. It defines Low, Intermediate, High and Very High Risk groups; literature-based additions address subgroups absent from the derivation cohort.

---

2. Key Prognostic Variables

Predictors differ by model and endpoint. Stage, grade, CIS, number, size and age predict progression in the EAU 2021 model; recurrence history is important in recurrence models and management of recurrent disease:

1. Pathological T Stage (Ta vs T1): T1 tumors have breached the basement membrane to invade the fibrovascular lamina propria. T1 tumors harbor an inherently higher risk of understaging, micrometastasis, and muscle invasion compared to Ta tumors.
2. Histological Grade (WHO 1973 G1–G3 and WHO 2004/2016 LG/HG): Use the actual grading system; G2 can be LG or HG. Typical molecular associations are FGFR3/PIK3CA alterations in low-grade papillary disease and TP53/RB1 alterations in high-grade disease. These associations do not replace pathological grading.
3. Presence of Carcinoma in Situ (CIS): Concurrent or primary flat CIS dramatically escalates progression risk across all stage categories.
4. Tumour Size: Larger lesions increase risk; the EAU additional-factor threshold is >3 cm, whereas the EORTC 2006 score uses ≥3 cm. Use the cutoff belonging to the chosen model.
5. Tumour Multiplicity (> 1 vs Solitary): Multifocality reflects widespread polyclonal urothelial instability (field defect), predisposing to intravesical recurrence.
6. Prior Recurrence Rate: Frequent recurrence (>1/year) predicts further recurrence and influences treatment. It does not by itself prove high-grade progression; this is not an input to the primary-patient EAU 2021 progression model.
7. Patient Age (> 70 years): Advanced age is an independent predictor of disease-specific mortality and disease progression.
8. Lymphovascular Invasion (LVI): Microscopic presence of malignant cells within endothelium-lined lymphatic or blood vessels in the lamina propria signifies high metastatic potential.

---

3. Detailed EAU Risk Group Definitions (2021+ Guidelines)

Use the full combination of stage, grade, CIS and additional factors rather than one isolated feature. The three additional factors are age >70, multiple papillary tumours and diameter>3 cm. Either grading system can be used; retain the distinctions between G2/LG and G2/HG when both are available.

Low-Risk NMIBC

• A primary, single TaT1 LG/G1 tumour <3 cm, without CIS, in a patient ≤70 years; or
• A primary Ta LG/G1 tumour without CIS with at most one additional factor.
• PUNLMP is included within LG for classification; it is not automatically low risk regardless of other features.
• Action: After initial TURBT, offer a safe immediate single chemotherapy instillation when eligible. Perform repeat resection when independently indicated. Routine maintenance is unnecessary. Low-risk follow-up is cystoscopy at 3 and 12 months, then annually to five years.

Intermediate-Risk NMIBC

• No CIS, and criteria for neither low, high nor very high risk.
• Examples include recurrent small Ta LG disease and Ta HG without CIS with fewer than two additional factors. A primary solitary large Ta LG tumour can still be low risk if size is its only additional factor.
• Action: Induction intravesical chemotherapy, with maintenance if selected for no longer than one year, or one year full-dose BCG. A single immediate instillation is useful in selected eligible papillary tumours, not mandatory for every intermediate-risk patient.

High-Risk NMIBC

• T1 HG/G3 without CIS, unless very high risk; CIS, unless very high risk.
• Without CIS: Ta LG/G2 or T1 G1 with all three factors; Ta HG/G3 or T1 LG with at least two; T1 G2 with at least one.
• Pure primary CIS is assigned high risk from literature, not a validated numerical prediction from the derivation model.
• Action: Repeat TURBT for the indications in section 6.1, then full-dose BCG for one to three years; discuss immediate cystectomy. High-risk category alone does not mandate repeat TURBT or exactly three years of BCG for everyone.

Very High-Risk NMIBC

The stage/grade combinations in EAU Table 6.2 are:

1. Ta HG/G3 with CIS and all three factors.
2. T1 G2 with CIS and at least two factors.
3. T1 HG/G3 with CIS and at least one factor.
4. T1 HG/G3 without CIS and all three factors. Literature-based additions include LVI, CIS in the prostatic urethra, and certain aggressive urothelial subtypes, including pure micropapillary, plasmacytoid, sarcomatoid, nested and neuroendocrine patterns. Report subtype and extent for specialist review.

Prostatic stromal invasion is not NMIBC: direct extension from the bladder is T4a. Distinguish it from epithelial prostatic urethral CIS. BCG-unresponsive disease is a separate treatment-failure category requiring its own definition; not every recurrence after BCG qualifies.

Action: Discuss immediate radical cystectomy with pelvic lymphadenectomy, balancing cancer control against morbidity and quality of life. Full-dose BCG for 1–3 years remains an option in selected patients, particularly if cystectomy is declined or unsuitable. BCG-treated progression risks can be lower than the untreated-model estimates; do not present 40–50% as every patient’s outcome.

---

4. Clinical Significance & Patterns of Carcinoma in Situ (CIS)

Carcinoma in situ represents a flat, non-invasive intraurothelial neoplasm composed of cytologically anaplastic cells with high mitotic activity and prominent nucleoli. It is classified clinically into three patterns:

• Primary CIS: Isolated CIS presenting in the bladder without prior or concurrent exophytic papillary tumors. Patients frequently present with severe, refractory irritative lower urinary tract symptoms (dysuria, frequency, nocturia) mimicking bacterial cystitis but with negative urine cultures.
• Secondary CIS: CIS detected during follow-up in a patient with a prior history of papillary urothelial carcinoma.
• Concurrent CIS: CIS identified in mucosal biopsies adjacent to or distant from an overt exophytic papillary tumor (Ta or T1).

Diagnostic and Therapeutic Implications of CIS

1. Cytology Correlation: Because CIS cells lack cohesive intercellular desmosomes, they readily exfoliate into the urinary stream. Cytology is useful for HG disease/CIS, but sensitivity is variable (EAU reports 28–100% for CIS); a negative result does not exclude it.
2. Field Defect: CIS is frequently multifocal and patchy, supporting targeted biopsy of abnormal urothelium. With positive cytology and no visible bladder tumour, assess upper tracts and prostatic urethra and use bladder mapping or PDD-guided biopsies.
3. Prognostic Impact: Concurrent CIS increases progression risk; the assigned category depends on the complete stage/grade/factor combination. In T1 HG/G3, concurrent bladder CIS defines very high risk when at least one additional clinical factor is present; the full EAU combination matters.
4. Treatment Sensitivity: BCG produces high complete-response rates and is superior to chemotherapy alone for CIS. EAU reports retrospective complete responses of 72–93% with BCG versus 48% with chemotherapy; durability varies, and even complete responders need surveillance.

---

5. EAU NMIBC Risk Group Classification Summary Table

The following progression estimates use WHO 2004/2016 categories in the EAU 2021 model. They describe primary patients not immediately treated with BCG and exclude recurrent disease, primary/prostatic urethral CIS, LVI and subtype cases. They are not recurrence rates or individualized post-BCG predictions.

The WHO 1973 model gives different estimates (for example 44% five-year VHR progression). EORTC/CUETO, selected for the treatment setting, address recurrence. A risk-group label must not erase the population and treatment assumptions behind its numerical estimate.