1. Study population
From February 2009 to November 2012,
48 consecutive patients suspected of having a local recurrence of PCa,
were prospectively evaluated with MRI.
22/48 patients (age 67,86 ± 5,87 years),
in whom the presence/absence of local recurrence was established by biopsy or clinical follow-up (C-Choline PET/CT and PSA),
were identified.
The following information were collected: PSA values,
digital rectal examination (DRE) and TRUS findings closest to the time of MRI.
All patients who had undergone a biopsy within 90 days of MRI were identified.
Patients with negative findings at MRI were followed up with C-Choline PET/TC for a maximum of 12 months.
2. MRI protocol
MRI examinations were performed with a 1.5 T whole body scanner and a pelvic phased-array surface coil combined with a disposable prostate endorectal coil.
After local three-plane,
morphological study of the post-prostatectomy bed was obtained with Fast Relaxation Fast Spin Echo (FRFSE) T2-weighted sequences in the sagittal,
axial and coronal planes.
FSE T1-weighted sequence in axial plane was performed.
The DWI acquisition was added to the morphological evaluation and before the injection of contrast medium,
in the axial plane,
using a spin-echo single-shot echo-planar imaging (SSEPI) sequence,
with two b-values (0 and 600 s/mm2) and the following parameters: TE 80 ms,
TR 4000 ms,
field of view (FOV) 200 x200 mm,
slice thickness4 mm,
intersection gap0.4 mm,
matrix size 256 x 256,
number of excitations (Nex) 6,
number of slices 15.
Pre-contrast axial FRFSE T2-weighted sequence was used to match the slices of the DWI sequence.
DCE-MRI was obtained using three-dimensional Spoiled Gradient Recalled (SPGR) T1-weighted sequence during the intravenous injection of a paramagnetic contrast agent,
at flow rate of 3-4 ml/sec followed by 15 ml of saline solution.
Twenty three-dimensional data sets,
two before and eighteen after contrast administration,
were acquired with 19 seconds temporal resolution and a total duration of 6-8 minutes.
3. Imaging analysis
MR images were assessed by means of consensus between two expert radiologists,
who were blinded to all patients' information other than knowing that each patient had a suspected Pca recurrence.
Vesico-urethral anastomosis was evaluated on T2-,
DWI- and DCE-MRI sequences .
On both T2-weighted and DWI sequences,
a slightly hyperintense focal area or nodule was considered suspicious of recurrence; hypointense areas or nodules were considered fibrotic tissue.
DCE-MRI images were processed on an independent workstation with a dedicated software.
Regions of interest (ROIs) positioned on the suspected areas were used to obtain Time-Intensity (T/I) curves.
By using this analysis,
color-coded images obtained were superimposed to T2-weighted images.
Nodular red pixel clusters or pathological contrast enhancement were defined as a suspicious local recurrence.
The patterns of T/I curves were categorized into three groups (Figure 1): type 1,
a rapid initial upslope and a rapid washout; type 2,
a rapid initial upslope and a plateau or slow washout; type 3,
a rapid or slow initial upslope and a continuous increase.
4. Statistical analysis
We assessed sensitivity,
specificity,
positive predictive value (PPV) and negative predictive value (NPV) of T2-weighted,
DCE-MRI and DWI with Receiver Operating Characteristic (ROC) analysis using clinical and histopathologic findings as the standard of reference.