RESUMO
OBJECTIVE: To evaluate urinary Aurora A Kinase (AURKA) mRNA expression as a diagnostic biomarker for urothelial bladder cancer (UBC). METHODS: One hundred and eighty-eight urine samples from patients with UBC (n = 122) and controls with hematuria (n = 66) were investigated. AURKA expression was quantified using real-time PCR and compared with voided urinary cytology. Associations with stage and grade were assessed. The area under curve was used to quantify the predictive accuracy (PA). RESULTS: The sensitivity and the specificity of AURKA for UBC were 83.6 and 65.2 %, respectively (PA = 74.4 %). Among those with detectable AURKA, the quantity of expression was similar in cases and controls. Compared with Ta, tumors staged T1 and T2 showed a 9.31-fold and 4.78-fold increased AURKA expression (p = 0.034), respectively. Further, high-grade tumors showed 5.33-fold higher expression levels than low-grade tumors (p = 0.031). AURKA and urinary cytology showed similar overall PA for UBC detection (74.4 vs. 72.1 %, p = 0.588). For low-grade tumors, AURKA was more accurate (72.5 vs. 59.0 %, p = 0.004), while cytology was more accurate for high-grade lesions (76.8 vs. 89.1 %, p = 0.011). CONCLUSIONS: In patients with hematuria, AURKA is associated with the presence and grade of UBC, suggesting a role as diagnostic and prognostic biomarker. As AURKA is more accurate in low-grade tumors but less accurate in high-grade tumors than urinary cytology, both could be complementary in detecting UBC.
