Prostate cancer is the second leading cause of cancer-related death and the most frequently diagnosed male malignant disease in the Nordic countries [
The aim of the present study was to correlate tumor perfusion parameters derived from DCE-MRI and clinical prognostic factors and further to explore if we can separate very early tumors from relatively advanced ones with DCE-MRI-derived parameters for decision making in early stage prostate cancer.
Seventy-one consecutive patients with histologically proven prostate adenocarcinoma were enrolled in our prospective clinical trials to develop hypofractionated image-guided and intensity-modulated radical radiotherapy. The study identifier at www.ClinicalTrials.gov is
Transrectal ultrasound-guided biopsy (12 cores, 6 on each lobe) was performed in each patient. Six biopsy cores were embedded in one paraffin block. Pathology was reviewed and graded according to the Gleason system. Major criteria include an infiltrative glandular growth pattern and an absence of basal cells and nuclear atypia in the form of nucleomegaly and nucleolomegaly. The diagnosis was based on the microscopic appearance of slides stained using haematoxylin and eosin. In difficult cases, basal cell absence has been confirmed by immunohistochemical stains for basal cell markers.
Multiparametric MR imaging was acquired using a 3 Tesla MR System (Siemens Trio-Tim, Erlangen, Germany) with a combination of 6-channel body matrix coil and 6 elements of 24-channel spine matrix coil positioned around the pelvis to cover the prostate. Tri-planar T2-weighted turbo spin echo images from below the prostatic apex to above the seminal vesicles were obtained. DWI was acquired with a single-shot echoplanar sequence on the axial plane using three
Sequence parameters for 3T multiparametric MRI with the body and spine matrix combination coil system.
Sequence | Pulse sequence | TR (msec) | TE (msec) | FA (°) | FOV (mm) | ACQmatrix | Slice/gap (mm) |
---|---|---|---|---|---|---|---|
Axial DWI, |
SE-EPI | 3800 | 77 | 90 | 221 × 260 | 102 × 160 | 3.6/0 |
Axial T2W | TSE | 4000 | 100 | 90 | 200 × 200 | 288 × 320 | 3/0.6 |
Sagittal T2W | TSE | 5000 | 100 | 90 | 200 × 200 | 288 × 320 | 3/0.6 |
Coronal T2W | TSE | 5000 | 100 | 90 | 200 × 200 | 288 × 320 | 3/0.6 |
Axial 3D |
FLASH GRE | 4.9 | 1.7 | 2 and 13 | 260 × 260 | 138 × 192 | 3/0 |
Axial 3D DCE | FLASH GRE | 4.9 | 1.7 | 12 | 260 × 260 | 138 × 192 | 3.6/0 |
SE, spin echo; EPI, echo planar imaging; TSE, turbo spin echo; FLASH, fast low angle shot; GRE, gradient recalled echo; TR, repetition time; TE, echo time; FA, flip angle; ACQ matrix, acquisition matrix.
All MR images were reviewed and analyzed on a syngo Multimodality Workplace (Siemens Healthcare). Voxelwise MRI signal enhancement time curves were fitted according to a pharmacokinetic model using Tissue 4D software (Siemens Healthcare). First, a motion correction has been performed, which registered all volumes of the time series to a user-selected reference volume to reduce the effect of patient and physiological motion during the DCE image acquisition. After the registration of the morphological image and the precontrast image, an oval-shaped or irregular-shaped region of interest (ROI) was drawn on the prostate cancer foci. ROIs were drawn in early enhancing region of DCE-MRI and with the DWI b800, ADC map, and T2-weighted image as references. T1 map calculation of precontrast was a prerequisite for pharmacokinetic modeling. T1 fitting was restricted to pixels with values above a noise level value (>20), and the respective values were automatically calculated by the system as a function of the entered contrast agents. For the Tofts modeling [
The ADC value of each identified tumor lesion was measured directly on the parametric ADC maps. The ADC map was reviewed simultaneously with the corresponding high
A prostate cancer was defined on each MRI as follows: a hypointense region relative to the adjacent parenchyma on T2-weighted image; a region with a low ADC value relative to the adjacent parenchyma on the ADC map; and a region with early wash-in and wash-out of contrast medium relative to the adjacent parenchyma on DCE-MRI. Precontrast T1-weighted images were used to identify postbiopsy hemorrhage (as an area with high signal intensity) to rule out false-positive findings.
Statistical analysis was performed with SPSS (version 23.0, SPSS Inc., Chicago, Illinois, USA). A two-sided nonparametric Mann–Whitney U test was used to compare the patients age, PSA, tumor size, ADC,
No suspicious lesion was found on MRI in 7 out of the 71 patients with a biopsy proven prostate cancer; two patients had no DCE images due to allergy to the contrast agent. Sixty-nine lesions were detected in the prostate of the remaining 62 patients (age: mean ± SD: 70 ± 5 years, range from 60 to 79 years). Ten patients had clinical stage T1c and 52 had T2 (16 in T2a, 8 in T2b, and 28 in T2c) tumors according to TNM classification for prostate cancer. The serum PSA value (mean ± SD) was 9.5 ± 3.7 ng/mL, with the range from 3.4 to 19.1 ng/mL.
There were 19 patients with a Gleason score 3 + 3, 41 with a Gleason score 3 + 4, and 2 with a Gleason score 4 + 3 tumor.
None of the measured parameters, including patients’ age, serum PSA, and DWI- and DCE-MRI-derived parameters, were different between Gleason score 3 + 3 and 3 + 4 tumor groups.
The majority of the tumors were in the peripheral zone (52, 75%), and the other 17 tumors were in the transitional zone.
There was no significant difference of the patients’ age, serum PSA, tumor ADC, The size of peripheral zone tumors (lesion number, Ve was lower in the peripheral zone tumors (lesion number,
Comparison of the 62 patients with peripheral and transitional zone prostate cancer (46 versus 16): age, tumor size, and DWI- and DCE-derived tumor parameters.
Total |
Peripheral |
Transitional |
| |
---|---|---|---|---|
Age (years) | 70 ± 5 | 70 ± 5 | 70 ± 4 | 0.974 |
PSA (ng/mL) | 9.5 ± 3.7 | 9.7 ± 3.9 | 9.1 ± 3.3 | 0.552 |
Area of tumor (cm2) | 0.74 ± 0.47 | 0.68 ± 0.41 | 0.93 ± 0.59 |
|
ADC (×10−3 mm2/s) | 0.87 ± 0.16 | 0.89 ± 0.17 | 0.82 ± 0.13 | 0.259 |
|
0.15 ± 0.05 | 0.15 ± 0.05 | 0.14 ± 0.06 | 0.743 |
|
0.57 ± 0.22 | 0.59 ± 0.21 | 0.49 ± 0.24 |
|
Ve | 0.28 ± 0.08 | 0.27 ± 0.08 | 0.32 ± 0.07 |
|
iAUC (mmoL/L/min) | 16.70 ± 5.69 | 17.26 ± 5.51 | 15.86 ± 6.21 | 0.626 |
PSA, prostate-specific antigen; ADC, apparent diffusion coefficient;
Prostate cancer showed earlier and more pronounced enhancement than surrounding normal prostate tissue (example Figure
Transverse prostate MR images from a 69-year-old male patient with biopsy proven prostate cancer (Gleason score 3 + 4 and serum PSA 6.6 ng/mL): (a) T2-weighted image showing in the transitional zone a hypointense area without clear border; (b) ADC map: transitional zone hypointense region with a clear border, with ADC value of 0.75 × 10−3 mm2/s; (c) T1-weighted image early enhancement map: the enhanced region of interest 1 (ROI1, red line) corresponds to the tumor, and ROI 2 (green line) was selected from normal prostate tissue as healthy control; (d)
Comparison of tumor
There were no significant differences of the serum PSA levels between clinical stage T1c (
Serum PSA correlated with both tumor
Correlations between serum PSA and DCE-MRI-derived tumor parameters in the 62 patients with prostate cancer. (a) Serum PSA correlated with tumor
No correlation was found between serum PSA and tumor ADC value.
A reliable diagnostic test should be able to provide an early prostate cancer diagnosis and minimize the amount of unnecessary biopsies or treatments. From this perspective, morphological MRI is a good candidate for prostate cancer investigation as it provides high-contrast and high-resolution images of the prostate. However, no single MRI sequence is sufficient to characterize prostate cancer. Each of the functional MR components has clinical advantages and limitations. Early promising data suggest that MP-MRI, which is performed concurrently with anatomical and functional techniques, is the most sensitive and specific imaging tool for lesion detection, characterization, and staging of prostate cancer [
DCE-MRI can be used to assess noninvasively the functional aspects of microcirculation of tissues. DCE-MRI relies on the fact that a bolus of contrast agent passing through the capillary bed is transiently confined within the vascular space before passing rapidly into the extravascular extracellular space at a rate determined by the permeability of the microvessels, their surface area, and blood flow [
Serum PSA is elevated as a result of disruption of the prostatic architecture in the presence of prostate disease and injury, and PSA screening helps to diagnose prostate cancer earlier, at lower clinical stages and with lower Gleason score [
The Gleason score reflects the tumor aggressiveness and is an important predictor of outcome in patients with prostate cancer [
Our study has a few limitations: firstly, the MRI was performed after biopsy. We were not sure, if the tumor ADC value and DCE parameters had been measured at the biopsy sites. Secondly, we were unable to evaluate the correlation between MRIs and histopathological features accurately because we did not obtain surgical specimens. There have been concerns about the probability of undergrading prostate cancer by biopsy due to tumor heterogeneity. Thirdly, all patients underwent needle biopsies before MRI examinations, implying that hemorrhagic or inflammatory changes caused by this procedure might have affected the MRIs. However, we excluded visible bleeding with the help of precontrast T1-weighted images, and the time interval between biopsy and MRI was long (6–10 weeks) enough for biopsy wound healing.
In conclusion, the present study has confirmed that DCE-MRI is a promising biomarker that reflects the microcirculation of prostate cancer. DCE-MRI-derived quantitative parameters in combination with clinical prognostic factors may provide an effective pretreatment diagnosis modality for early prostate cancer, especially for those with negative biopsy.
The data are available from the Medical Imaging Center of Tampere University Hospital.
The authors have no conflicts of interest to declare.
This project was supported by the Competitive State Research Financing of the Expert Responsibility Area of Tampere University Hospital, Seppo Nieminen Fund (Grant no. 150613), and Pirkko Kellokumpu-Lehtinen (Grant nos. 9R019 and 9S021). Xingchen Wu was supported by the Finnish Cultural Foundation, Pirkanmaa Regional Fund, and the Finnish Medical Foundation. The authors would like to thank research coordinator Irja Kolehmainen for her contributions.