Arising from O'Donnell MA (2008) Does the probiotic L. casei help prevent recurrence after transurethral resection for superficial bladder cancer? Nat Clin Pract Urol 5: 526–527

We read with interest the recent commentary by O'Donnell,1 in which he questioned our published findings on the efficacy of the probiotic Lactobacillus casei for preventing recurrence of superficial bladder cancer after transurethral resection.2 He noted two major flaws in our study: a high and unbalanced withdrawal rate, and the lack of blinding. However, we do not agree that these flaws undermine our conclusions, and we think that our findings are valid for the following reasons.

As O'Donnell stated, the withdrawal rate was 3.5 times higher in the epirubicin plus L. casei group than in the epirubicin monotherapy group. Our study, however, was performed according to the intention-to-treat (ITT) principal, so the difference in withdrawal rate does not necessarily invalidate the comparison, although, as O'Donnell notes, it does increase the potential for bias. We are currently reviewing the data on withdrawals to determine whether they affected our results. Other studies that compared L. casei with placebo have demonstrated a preventive effect of L. casei against recurrence after transurethral resection of superficial bladder cancer, with similar withdrawal rates for the two interventions.3,4 Additionally, Larsson et al.5 reported that a high intake of cultured milk might lower the risk of developing bladder cancer, offering support for the conclusions of our earlier work.3,4,6

We do not think that our nonblinded study design presents a problem, because recurrence (the primary end point) was confirmed objectively by cystoscopy or urine cytology, and the positive cases were examined histologically. A previous double-blinded study has shown a preventive effect of L. casei against disease recurrence in this setting, compared with placebo.4

As mentioned in our discussion,2 it is possible that patients who received epirubicin alone began ingesting Lactobacillus products during this study. The nonblinded design might, therefore, have resulted in underestimation (but not overestimation) of the prevention of recurrence by the L. casei preparation. Regardless of this possibility, the epirubicin plus L. casei group had a significantly higher recurrence-free survival rate, and the effect of treatment remained significant on multivariate analysis.