EORTC 30911

Clinical question: In intermediate/high-risk Ta/T1 non-muscle-invasive bladder cancer, does intravesical BCG improve outcomes versus epirubicin?

Design: Randomized phase III trial; 837 eligible patients; epirubicin (279), BCG (281), BCG+isoniazid (277). Median follow-up 9.2 years.

Recurrence

Recurrence occurred in 147/279 (52.7%) with epirubicin versus 103/281 (36.7%) with BCG and 110/277 (39.7%) with BCG+INH.

Using BCG alone versus epirubicin, observed absolute recurrence reduction ≈ 16.0 percentage points; approximate NNT ≈ 7 over the observed long-term follow-up. Overall combined BCG vs epirubicin HR for recurrence 0.62, P<.001.

Other outcomes

Combined BCG reduced distant metastasis risk (HR 0.55, P=.046), all-cause mortality (HR 0.76, P=.023) and bladder-cancer mortality (HR 0.47, P=.026). Progression was not significantly different (HR 0.84, P=.55).

Progression: epirubicin 24/279 (8.6%) vs BCG 19/281 (6.8%) vs BCG+INH 23/277 (8.3%).

NNT/NNH

NNT ≈ 7 is shown for recurrence using the observed recurrence proportions and a common follow-up population. No NNH is assigned because toxicity was not reported as one single clinically coherent binary endpoint suitable for a robust long-term NNH.

Methodology

Randomized three-arm design. The principal long-term efficacy analysis compared the two BCG groups combined with epirubicin. Intermediate- and high-risk patients were included; the benefit was not confined to the highest-risk subgroup.

Our take

The important Absolute Effect is not merely the recurrence HR of 0.62: roughly 16 fewer recurrences per 100 treated with BCG rather than epirubicin were observed over long-term follow-up. BCG also improved distant metastasis and survival endpoints, although progression was not significantly different. This established BCG as more than a recurrence-suppressing treatment.

Primary source

Sylvester et al., European Urology