SIU 2017: Predicting Tumour Recurrence in Patients Treated with Bacillus Calmette-Guérin
Their study population was a prospective collection of patients with intermediate or high-risk NMIBC scheduled for induction BCG therapy. They obtained urine testing for Urovysion FISH analysis at the following time points: t0 (before BCG), t1 (before the 6th instillation), t2 (3 months post induction during the first cystoscopy). They underwent routine follow-up to identify recurrences.
They then assesses the ability of a positive FISH at each of those time points to identify early recurrence. Positive FISH at the t0 time point was not associated with any separation on the KM curve, and was not predictive of early recurrence. However, for the t1 and especially the t2 time point, the curves separated early. However, it is important to note that it only predicted recurrence during follow-up – not necessarily high-grade recurrence considered BCG failure. This was not specified in the analysis.
They then provided key markers of the test at each time point: specificity, sensitivity, PPV, NPV and accuracy. The utility at the t0 time point was very low. The t2 time point however demonstrated a NPV of 85% and specificity of 84%. Its positive predictive value and sensitivity were much lower.
While the authors focus on the benefit of a positive test at 3 months, I actually thought the negative test at 3 months was more useful. As the sensitivity and PPV were low (59% and 56%), but the specificity and NPV were quite high (84 and 85%), that is where the value lies.
Further work is warranted.
Presented by: Esmée Liem
Written by: Thenappan Chandrasekar, MD, Clinical Fellow, University of Toronto, twitter: @tchandra_uromd, at the 37th Congress of Société Internationale d’Urologie - October 19-22, 2017- Lisbon, Portugal