Preliminary assessment of magnetic resonance spectroscopic imaging in predicting treatment outcome in patients with prostate cancer at high risk for relapse.
Academic Article
Overview
abstract
The purpose of the study was to determine whether 3D proton magnetic resonance spectroscopic imaging (MRSI) can predict treatment outcome in high risk patients with prostate cancer. Endorectal magnetic resonance imaging (MRI) and 1H-MRSI were performed in 16 patients with prostate cancer who were considered high risk because of clinical stage T3-4, Gleason score>/=8, and/or prostate-specific antigen (PSA) level>20 ng/mL. Patients were treated with chemotherapy/hormone therapy, underwent radical prostatectomy (RP) or radiation therapy, and were followed for PSA relapse (follow-up, 19-43 months). The ratio of choline plus creatine to citrate was used to localize peripheral zone cancer. An MRSI risk score on a scale of 0-3 was derived from the volume and degree of metabolic abnormality. Magnetic resonance spectroscopic imaging risk score, MRI tumor/node (TN) stage, clinical stage, Gleason score, and PSA were used as predictors of pathologic stage in patients treated with RP (n=10) and PSA relapse in all patients. Magnetic resonance imaging TN stage (P<0.01) and MRSI risk score (P<0.05) correlated with pathologic stage, but clinical stage did not (P=0.35). Magnetic resonance imaging TN stage was the only significant predictor of PSA relapse in the univariate analysis (P<0.05). Although the MRSI risk score did not reach significance (P=0.13), 6 patients with a score<0.9 were relapse-free, whereas 7 of 10 patients with a score>0.9 relapsed. Magnetic resonance imaging and MRSI risk assessments agreed in 15 of 16 patients. These preliminary results suggest that tumor metabolic assessment may indicate treatment outcome in high-risk patients with prostate cancer. Although MRSI did not provide added prognostic value to MRI in this small number of patients, MRSI might increase the confidence of the clinician in assessing risk on MRI by contributing supporting metabolic data.
