Integration of MRI to clinical nomogram for predicting pathological stage before radical prostatectomy.

Abstract

Debate persists regarding whether MRI should be used routinely for preoperative evaluation of prostate cancer. The aim is to assess the role of prostatic magnetic resonance imaging (MRI) and other preoperative data in extra-prostatic extension (EPE) evaluation. From 2000 to 2013, 1743 patients operated for radical prostatectomy had a preoperative MRI. Age, clinical stage with digital rectal exam (DRE), PSA, prostate weight, biopsy, MRI and pathological findings of the surgical specimen were noticed. A multiparametric score of the variables independently associated with EPE was built with or without MRI on a random sample test population and internally validated. With mean age of 62.9years and mean PSA of 9.6ng/ml, the population was distributed as follows: 1424 DRE T1, 254 T2, 32 T3; on biopsy 990 Gleason score=6 and 717≥7; on MRI 1322 iT2, 290 iT3A and 131 iT3B; on prostatectomy 15 pT0, 998 pT2, 548 pT3A, 181 pT3B and 1 pT4A. In multivariate analysis, DRE, PSA, Gleason score, prostate weight and MRI were independently associated with EPE and integrated in a score with an area under curve (AUC) of 0.74 [95% CI 0.71-0.77] (0.72 without MRI, p<0.01) a positive predictive value of 61% and a negative predictive value of 74%, internally validated. The Hosmer-Lemeshow goodness-of-fit test showed good accuracy (p=0.77). Integration of MRI with clinical data for predicting pathological stage before radical prostatectomy permits to exclude accurately EPE in 74% of cases.