Bioinformatics analysis shows association between a gene expression signature and biochemical relapse after radical prostatectomy in patients with prostate cancer.

VAZQUEZ ELBA; COTIGNOLA JAVIER; BRANDANI, JAVIER NAHUEL; BIZZOTO, JUAN; ABBATE, MERCEDES; GUERON GERALDINE

Mar del Plata

Congreso; LXIII Reunión Científica de la sociedad Argentina de Investigación Clínica; 2018

SAIC

Prostate cancer (PCa) is the second most incident type of cancer in men and the fifth leading causeof cancer-related deaths worldwide. Surgical resection of the entire gland by radicalprostatectomy is one therapeutic option with curative purposes for men with organ confined PCa.However, 30?40% of patients will have a biochemical relapse (BCR) after surgery, which mayindicate clinical recurrence of an aggressive disease. Current clinical indicators of PCa recurrenceafter radical prostatectomy have limited sensitivity and specificity. Thus, the ability to establish therisk of progression after surgery is of high importance for patient management and to avoidovertreatment. We used transcriptome data from the dataset GSE54460 publicly available at GEO-NCBI to study differential gene expression between patients with and without BCR post radicalprostatectomy. We obtained a list of 1021 genes differentially expressed (adjusted p40% of the subsamples. The selected genes were: CRISP3, APOE, CELand DDTL; and they all were significantly overexpressed in the tumors of patients who relapsed.The analysis of BCR-free survival showed significant poorer survival for patients with high-tumoralexpression of these genes. Moreover, the risk of BCR is >2-fold for patients with high-tumoralexpression compared to patients with low expression. The analysis of the combined expressionshowed a HR=15 (95%CI=3.5-63.9) for patients with high-tumoral expression for two or moregenes. In conclusion, the expression signature of these genes might have the potential to detectaggressive tumors and predict BCR at the time of radical prostatectomy.