Analysis of bladder cancer subtypes in neurogenic bladder tumors.

INTRODUCTION: To establish if the validated tumor biomarkers of luminal and basal bladder cancers in non neuro-urological patients are applicable to a neuro-urological population.

MATERIALS AND METHODS: We retrieved bladder cancer samples from neuro-urological patients (n = 20) and non-neurological controls (n = 40). The expression of GATA3 and CK5/6 was analyzed using immunohistochemistry of microarray tissue sections. We also assessed the correlation between previous biomarker expression, gender, age, tumor stage (non-muscle-invasive bladder cancer (NMIBC)/muscle-invasive bladder cancer (MIBC)), squamous-cell differentiation and basal/luminal subtypes using Pearson's correlation coefficient (r).

RESULTS: Mean age at diagnosis of bladder cancer in neuro-urological patients was 53.2 years (min 41-max 73). MIBC was found in 13 neuro-urological patients (65%). The luminal subtype was identified in 7 samples (35%, all urothelial differentiation). The basal subtype was found in 13 samples (65%): 12 squamous-cell and 1 sarcomatoid differentiation. GATA3 and CK5/6 were expressed in 6 (30%) neuro-urological patients. A significant positive correlation was found between GATA3 expression and the luminal subtype (p = 0.00001, r = 0.5676). This was not the case with the neuro-urological status (r = -0.307). A poor correlation was found between CK5/6 expression and the neuro-urological status (r = 0.471 and -0.471), squamous-cell differentiation (r = 0.092), tumor stage NMIBC/MIBC (r = -0.118 and 0.118) and basal/luminal subtypes (r = -0.157 and 0.194).

CONCLUSION: In summary, the expression of GATA3 and CK5/6 could not differentiate the different subtypes of bladder cancer in neuro-urological patients. This implies that their specific histopathological signature is distinct from non neuro-urological patients. Additional pathways may be involved to explain their urothelial carcinogenesis mechanism.