• 2022-05
  • 2022-04
  • 2021-03
  • 2020-08
  • 2020-07
  • 2020-03
  • 2019-11
  • 2019-10
  • 2019-09
  • 2019-08
  • 2019-07
  • br Material and methods br Patient characteristics br


    2. Material and methods
    2.1. Patient characteristics
    Forty-three consecutive patients who had undergone radical pros-tatectomy at the Urology Department of our institution between 10/ 2010 and 02/2013 were included in this retrospective analysis. Patients initially presented with a history of elevated serum prostatespecific antigen (PSA) levels and positive findings from standard biopsies (guided by transrectal ultrasound, TRUS) suggesting surgical treatment. Prior to surgery, all patients had an IRB-approved, multiparametric MRI (mpMRI) of the prostate. MRI data were correlated with whole-mount histology slides of prostatectomy specimens. Exclusion criteria were patients with general contraindications to either MRI (e.g. pacemaker) or gadolinium-based contrast agents as well as patients with prostate biopsies performed less than 3 weeks before MRI to minimize chances of hemorrhage-related image artifacts. Two patients with effective le-sion diameters under 3 mm and one patient with incomplete histo-pathological workup were also excluded from analysis.
    2.2. Magnetic resonance imaging
    Presurgical mpMRI was performed in a 3 T MRI system (Magnetom Trio, Siemens Healthcare, Erlangen, Germany) following ESUR
    Table 1
    Imaging parameters of multiparametric prostate MRI.
    guidelines [3,14]. Signals were acquired with pelvic and spine array coils combined with an endorectal coil (eCoil, Medrad, Pittsburg, PA). The endorectal coil was filled with 30–40 ml of perfluorocarbon solu-tion (perfluorooctyl bromide, ABCR GmbH, Karlsruhe, Germany) to minimize susceptibility artifacts. All patients received an intravenous injection of either 40 mg butylscopolamine (Buscopan, Boehringer In-gelheim, Germany) or 1 mg glucagon (Glucagen, Nordisk, Gentofte, Denmark) to reduce bowel peristalsis. The MRI protocol included T2-weighted imaging (T2w), diffusion-weighted imaging (DWI) and dy-namic contrast-enhanced imaging (DCE) following Tigecycline injection of 15–20 ml of contrast agent (Gadovist, Bayer Healthcare, Berlin, Ger-many). Perfusion curves were analyzed with a standard application (Mean Curve, Siemens Healthcare, Erlangen, Germany). An overview of the main imaging parameters is given in Table 1.
    2.3. Histopathological work up
    Prostatectomy specimens were processed immediately after radical resection and fixed in 10% neutral buffered formalin for about a week. The seminal vesicles were separated, and the prostate was sliced into 4 to 5-mm thick step sections, perpendicular to the posterior, rectal sur-face of the gland. After paraffin embedding and hematoxylin-eosin staining, histology slices (typically 10 per prostate) were evaluated by a senior pathologist (15 years of experience in urogenital pathology), who was blinded to MRI. Tumor foci were outlined on the microscopy slides. Lesions were annotated and characterized according to the Gleason grading system [15]. Low-grade and high-grade PCa were de-fined as Gleason scores ≤ 7a = 3 + 4 and ≥ 7b = 4 + 3, respectively. All specimens were staged according to the TNM classification [16]. Only tumors with diameters of at least 3 mm were considered for fur-ther evaluation. Findings like nodular prostatic hyperplasia, prostatitis and normal prostate tissue were summarized as benign findings.
    Two radiologists, reader R1 with 5 years' and reader R2 with 4 years' experience in prostate mpMRI independently reviewed the MRI data with a standard application (Syngo Plaza, Siemens Healthcare). Both were blinded to clinical and histopathological findings and were allowed to define up to four prostate lesions per patient. Lesion loca-tions were reported using a 16-region PI-RADS v1 scheme [3]. Lesion diameters were measured in three orthogonal views and an effective diameter was computed as geometric mean. Each reader applied both v1 and v2 in random order with at least 10 days between readings to minimize the chance of remembering findings. All lesions with an overall PI-RADS scores ≥ 4 were considered to be suspicious for PCa.
    PI-RADS v1 assessment involved scores from 1 to 5 (highly likely to be benign or malignant, respectively) for each modality (T2w, DWI and DCE). The DCE score rates the presence of focal or asymmetric contrast enhancement and the shape of the signal intensity-time curve. Individual scores were added and converted to an overall PI-RADS score between 1 and 5, as published elsewhere [17]. Two patients had in-complete DCE data and, therefore, PI-RADS scores 1–5 were assigned for sum scores of 2, 3–4, 5–6, 7–8 and 9–10, respectively.
    PI-RADS v2 assessment followed a zone-specific model with
    Note: DWI-diffusion weighted imaging; DCE-dynamic contrast enhanced; TSE-turbo spin echo; SS-EPI-single-shot echo planar imaging; SPGR-spoiled gradient echo; ST-slice thickness; SG-slice gap; IPR-in-plane resolution; TR-repetition time; TE-echo time; FOV-field of view; FA-flip angle.