PI-RADS v2 score may be as a predictor of bone metastasis in prostate cancer.

BACKGROUND

To explore the independent influencing factors of PCa bone metastasis and evaluate the role of prostate imaging.

METHODS

The clinic data such as age, prostate volume, tPSA and fPSA . Univariate and multivariate analyses were performed to investigate the independent influencing point of the risk factors. Nomogram and ROC curve were generated to establish the prediction model. The calibration curve, leave-one-out cross validation and independent external validation were performed to evaluate the prediction model.

RESULTS

This study enrolled 325 newly diagnosed PCa patients at two hospitals. Univariate and multivariate analyses showed only tPSA , cTx, ALP, and PI-RADS v2 score were the independent influencing factors of PCa bone metastasis. The cut-off points of PI-RADS v2 score to distinguish bone metastasis was 5. The nomogram was established with a sensitivity of 81.3% and a specificity of 74.5% to predict the probability of PCa bone metastasis. The calibration curve and ROC curve displayed a good value of the prediction model. Leave-one-outcross validation showed the prediction model could classify 79.8% cases accurately. External data validation displayed sensitivity of 78.4% and a specificity of 79.1%.

CONCLUSIONS

PI-RADS v2 score could predict PCa bone metastasis, the prediction model may help discovered PCa bone metastasis.