Presentation Authors: Stefano De Luca, Cristian Fiori, Matteo Manfredi, Eleonora Amante, Orbassano, Italy, Eugenio Alladio, Turin, Italy, Francesca Ragni, Giorgio Alleva, Diletta Garrou, Enrico Checcucci, Daniele Amparore, Giovanni Cattaneo, Angela Pecoraro, Roberta Aimar, Sabrina De Cillis, Orbassano, Italy, Marco Vincenti, Turin, Italy, Francesco Porpiglia*, Orbassano, Italy
Introduction: The correlation between PCa and increased levels of the most important androgens is well known. The aim of the present study was to evaluate prostate cancer (PCa)-induced deviation of a limited endogenous androgenic anabolic steroids (EAAS) profile on urine samples.
Methods: The current study enrolled 283 patients: 139 â€œhealthy subjectsâ€ or control group (âˆ¼49%) (PSA â‰¤ 4 ng/mL or PSA > 4 with at least one negative biopsy), and 144 (âˆ¼51%) individuals affected by PCa histologically proved. The individuals affected by diabetes, other carcinoma, metabolic diseases, subjected to steroidal therapy or other kind of therapy suspected to alter the urinal steroid profile (USP), i.e. anti-hypercholesterolemia treatments, were not included in the study. All samples were collected in the morning.Two multivariate classification models based on the partial least square discriminant analysis (PLS-DA) algorithm were built (in cross validation, using the venetian blind strategy and 5 deletion groups) employing the training set: model 1 including three urinary steroids (Testosterone, Etiocholanolone and 7Î²-OH-DHEA) based on the methodology of gas chromatography â€“ mass spectrometry (GC/MS) + PSA and model 2 including only USP. The performances of the models were evaluated using the test set.
Results: PLS-DA models built using the 3 steroids and PSA (model 1) and only the USP selected variables (model 2) are reported in Figure 1. Sensibility, specificity and Area Under the Curve of the ROC analysis for model 1 and model 2, were: 83%, 84%, 0.93 and 88%, 79%, 0.89, respectively.
Conclusions: This study can confirm the effectiveness of a limited and cheap USP in PCa diagnosis and the possible activation of anomalous steroid biosynthesis induced from cancer cells.