Original Article
Predictive implications of genomic microarray for the diagnosis of prostate cancer metastasis following radical surgery
Abstract
Background: The application of using the appropriate genetic markers for predicting the risk of a prostate cancer (PCa) metastasis in patients would be beneficial; however, there is a large amount of irrelevant genomic information which negatively influences this predictive clinical accuracy.
Methods: We selected 10 candidate genes based on the microarray dataset of the two single-patient derived PCa tumor lines with different metastatic potentials. PCa patients who underwent a radical surgery in the MSKCC cohort, were enrolled to evaluate the predictive performance of the single gene and gene combination.
Results: All of the included candidate genes were significantly associated with a PCa metastasis, and some were also related with prostate specific antigen (PSA), Gleason score, biochemical recurrence (BCR), and death. Multivariate logistic regression analysis was performed, and it showed that three candidates, COL1A1, SEMA3C and SLC22A3, may be associated with the metastasis of PCa; however, there was no significant independent predictive gene found, based on Bonferroni correction. In addition, The area under the curve (AUC) of the combination group (AUC =0.972, sensitivity =100.00%, specificity =90.16%) was significantly higher than any candidate gene alone (P<0.05).
Conclusions: Our findings suggest that the genomic microarray tests could assist in the determination of which PCa patients are at a high-risk for metastasis following a radical surgery. It also would help urologists to determine the optimal treatment for PCa patients.
Methods: We selected 10 candidate genes based on the microarray dataset of the two single-patient derived PCa tumor lines with different metastatic potentials. PCa patients who underwent a radical surgery in the MSKCC cohort, were enrolled to evaluate the predictive performance of the single gene and gene combination.
Results: All of the included candidate genes were significantly associated with a PCa metastasis, and some were also related with prostate specific antigen (PSA), Gleason score, biochemical recurrence (BCR), and death. Multivariate logistic regression analysis was performed, and it showed that three candidates, COL1A1, SEMA3C and SLC22A3, may be associated with the metastasis of PCa; however, there was no significant independent predictive gene found, based on Bonferroni correction. In addition, The area under the curve (AUC) of the combination group (AUC =0.972, sensitivity =100.00%, specificity =90.16%) was significantly higher than any candidate gene alone (P<0.05).
Conclusions: Our findings suggest that the genomic microarray tests could assist in the determination of which PCa patients are at a high-risk for metastasis following a radical surgery. It also would help urologists to determine the optimal treatment for PCa patients.
