Prediction of recurrence in Ta urothelial cell carcinoma by real-time quantitative PCR analysis: a microarray validation study.

Accurate prediction of tumor recurrence in patients with superficial urothelial cell carcinoma (UCC) might result in a significant reduction of invasive follow-up cystoscopies. A recent study identified a panel of 26 genes from a large cDNA microarray analysis of bladder tumors that discriminated between early- and late-recurring patients with superficial Ta tumors (Dyrskjøt et al., Nat Genet 2003;33:90-6). We aimed to validate this panel of genes in 44 primary Ta UCCs (23 and 21 tumors from patients with short or prolonged recurrence-free periods, respectively), by real-time quantitative PCR. Statistical analysis showed marginal significant different mRNA expression levels between the 2 patient groups. To evaluate a supplementary effect of genes for the identification of patients with short or prolonged recurrence-free intervals, forward logistic regression analysis was applied. This revealed that a combination of the expression profiles of the genes HNRPK, LTB4DH and ANP32B resulted in the best performance, although the combination only marginally increased the predictive value of HNRPK alone. Comparing the receiver-operating-characteristic curves for HNRPK expression among patients with short or prolonged recurrence-free periods, revealed an area under the curve of 0.696 (95% CI, 0.537-0.855). Using the median HNRPK expression level as cut-off, a sensitivity of 69.6% and a specificity of 71.4% were obtained for the identification of patients with short or prolonged recurrence-free periods, respectively. In conclusion, we were not able to confirm the microarray gene expression pattern of the 26 genes shown by Dyrskjøt et al. The discovery of accurate recurrence predictive markers, therefore, remains a challenge.