• 3T
• diffusion-weighted imaging
• lymph node metastases
• magnetic resonance imaging
• prostate cancer

PURPOSE: To evaluate the role of MRI Diffusion-Weighted Imaging (DWI) to detect pelvic lymph node metastases in patients with prostate cancer (PC) candidate to radical prostatectomy and extended pelvic lymph node dissection.
MATERIALS AND METHODS: From June 2011 to March 2012, 5 patients with PC (high or intermediate risk patients), were scanned before surgical treatment using a 3T MRI scanner and a 8 channel phased-array surface body coil. Imaging protocol included T2w FSE, T1w FSE and DWI sequence (b-values: 0, 500, 800, 1000 and 1500 s/mm^2).
The appearance of benign and metastatic lymph nodes on the FSE MR images was analyzed by two observers in conference.
The measurement of the ADC value was performed, in the following nodal stations: external iliac proximal and distal, internal iliac proximal and distal and obturatory, each on both right and left sides.
RESULTS: A total of 84 lymph nodes were removed during surgery and histologically analysed. The smallest metastatic lymph node detected by this method measured 4mm on its short axis. The appearance of benign and metastatic lymph nodes on the FSE MR images was documented in terms of short axis, the long to short axis ratio, node contour and intranodal heterogeneity signal intensity, in all ten nodal stations. For each of these parameters a Grading Score system was assigned using a two-point-level score and the grade system was obtained by adding the point-level obtained for each of these 4 parameters. The Grading Score ranged from 4, indicator of benign nature, to 8, with 8 having the worst score indicator of malignant nature. The mean Grading Score was 6,46 ± 0,42 in the nodal metastatic group and 5,02 ± 0,59 in the nodal non-metastatic group ( P < 0.0001). A Grading Score > 4 was considered suspicious for malignancy.
With the threshold score of 4, the lymph node station-based sensitivity, specificity, positive and negative predictive values and diagnostic accuracy for FSE-MRI analysis, were 100%, 19%, 29%, 100% and 39%, respectively.
Mean ADC value was 0.796 ± 0,09 × 10^–3 mm^2/s in the nodal metastatic group and 1,17 ± 0,25 × 10^–3 mm^2/s in the nodal non-metastatic group ( P = 0.0008).
The ADC cut-off value, obtained by the ROC curve was 0.91 × 10^–3 mm^2/s.
The lymph node station-based sensitivity, specificity, PPV, NPV and diagnostic accuracy were 100%, 95,2%, 87,5%, 100% and 96,4%, respectively.
CONCLUSIONS: Our preliminary data seem to suggest that DW-MRI of lymph nodes can now be performed as part of a primary tumour staging without significantly increasing the imaging time. This unique modality can help to distinguish benign from malignant lymph nodes and that it is more accurate than FSE-MRI evaluation alone. Further, large scale studies are certainly needed to confirm our initial results.