The prognostic value, sensitivity, and specificity of multiparametric magnetic resonance imaging before salvage radiotherapy for prostate cancer

Aim: To determine the operational characteristics of pelvic magnetic resonance imaging (MRI) prior to salvage radiation therapy (SRT) for biochemically recurrent prostate cancer following radical prostatectomy.

Methods and materials: We reviewed the medical records of 386 patients who underwent MRI prior to SRT. We assessed associations of pre-SRT MRI findings with biochemical recurrence (BCR), distant metastasis (DM), prostate cancer-specific mortality (PCSM), and salvage androgen deprivation therapy (ADT) use following SRT. The sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV) of MRI for detecting local recurrence were also calculated.

Results: Pre-SRT MRI was positive for local recurrence in 216 patients (56%), indeterminate in 46 (12%), and negative in 124 (32%). On univariate analysis, BCR following SRT was significantly less likely for patients with positive (HR: 0.58, 95% CI: 0.42-0.8) or indeterminate (HR: 0.6: 0.36-1) MRI findings, compared to patients with negative imaging (p = 0.003). These associations remained significant on multivariate analysis (p < 0.05) and across pre-SRT PSA groups. For the entire cohort, the sensitivity of MRI for local recurrence was 61.0% (53.5-68.1%), specificity 60.0% (44.3-73.0%), PPV 86.1% (78.9-91.5%) and NPV 27.6% (19.0-37.5%). Sensitivity of MRI was better in men with higher pre-SRT PSA (80.0% for PSA > 1.0), and specificity was improved with lower pre-SRT PSA (73.9% for PSA 0.1-0.5).

Conclusions: Positive or indeterminate MRI findings prior to SRT were associated with improved biochemical control following SRT, across PSA levels. The sensitivity and specificity of MRI for local recurrence were 61% and 58.7%, respectively.