The Role of Dynamic Contrast Enhanced Magnetic Resonance Imaging in Evaluating Prostate Adenocarcinoma: A Partially-Blinded Retrospective Study of a Prostatectomy Patient Cohort With Whole Gland Histopathology Correlation and Application of PI-RADS or TNM Staging

Abstract

Background: Dynamic contrast-enhanced (DCE) magnetic resonance imaging (MRI) in the current Prostate Imaging-Reporting and Data System version 2.1 (PI-RADS v2.1) is considered optional, with primary scoring based on T2-weighted imaging (T2WI) and diffusion weighted imaging (DWI). Our study is designed to assess the relative contribution of DCE MRI in a patient-cohort with whole mount prostate histopathology and spatially-mapped prostate adenocarcinoma (PCa) for reference.

Methods: We performed a partially-blinded retrospective review of 47 prostatectomy patients with recent multi-parametric MRI (mpMRI). Scans included T2WI, DWI with apparent diffusion coefficient (ADC) mapping, and DCE imaging. Lesion conspicuity was scored on a 10-point scale with ≥ 6 considered "positive," and image quality was assessed on a 4-point scale for each sequence. The diagnostic contribution of DCE images was evaluated on a 4-point scale. The mpMRI studies were assigned PI-RADS scores and tumor, node, metastasis (TNM) T-stage with blinded comparison to spatially-mapped whole-mount pathology. Results were compared to the prospective clinical reports, which used standardized PI-RADS templates that emphasize T2WI, DWI and ADC.

Results: Per lesion sensitivity for PCa was 93.5%, 82.6%, 63.0%, and 58.7% on T2WI, DCE, ADC and DWI, respectively. Mean lesion conspicuity was 8.5, 7.9, 6.2, and 6.1, on T2W, DCE, ADC and DWI, respectively. The higher values on T2WI and DCE imaging were not significantly different from each other but were both significantly different from DWI and ADC (p < 0.001). DCE scans were determined to have a marked diagnostic contribution in 83% of patients, with the most common diagnostic yield being detection of contralateral peripheral zone tumor or delineating presence/absence of extra-prostatic extension (EPE), contributing to more accurate PCa staging by PI-RADS or TNM, as compared to histopathology.

Conclusion: We demonstrate that DCE may contribute to lesion detection and local staging as compared to T2WI plus DWI-ADC alone and that lesion conspicuity using DCE is markedly improved as compared to DWI-ADC. These findings support modification of PI-RADS v2.1 to include use of DCE acquisitions and that a TNM staging is feasible on mpMRI as compared to surgical pathology.

Keywords: ADC; DCE; DWI; MRI; PI‐RADS; T2W; conspicuity; prostate cancer.

Figure 1

Study Flowchart. mpMRI = multiparametric magnetic resonance imaging.

Figure 2

(A) Per lesion sensitivity of MRI sequences compared to whole mount pathology with percentages given above the bars and (B) Box plot of the tumor conspicuity scores for each of the evaluated MRI sequences. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 3

Relative contribution of DCE‐T1W MRI in lesion detection. 4 = Essential, 3 = Increased Confidence, 2 = No Added Value, 1 = No DCE. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 4

PCa multi‐focal scattered lesions in the bilateral peripheral zones (PZs) most conspicuous on T2W and DCE‐T1W scans in a 65‐year‐old male with rising PSA. Axial T2W images (A), DWI with b‐value of 1600 (B), corresponding ADC map (C), and arterial phase DCE‐T1W images (D). Histopathological specimen showing multifocal postero‐lateral PZ, with corresponding high magnification of regional highest‐grade tumor (E). Spatially corresponding tumor foci (Gleason score 4 + 4 on the right, and 3 + 4 on the left) are marked by yellow arrows (A–D) and black rectangles (E). Focal and nonuniform T2 signal hypo‐intensities are evident at the left posterolateral and right lateral PZ (A). DWI and ADC sequences provide lower lesion conspicuity with indistinct signal changes throughout the PZ (B and C). The mean ADC value for the lesion on the left is 1064 mm²/s and 1186 mm²/s on the right (C, arrows), which was interpreted as benign (PI‐RADS 2) on the prospective clinical radiology report. Combined with the T2W features, the prospective report provided a PI‐RADS score of up to 3, based on T2W images. DCE‐T1W imaging, however, shows distinct arterial enhancement (D, yellow arrows), in keeping with malignancy. A 1.7 cm confluent area of T2 hypo‐intensity (white dotted oval, A and D) was also prospectively reported in the central gland with features of broad internal capsular abutment and possible extra‐capsular extension and categorized as a PI‐RADS 5 lesion. However, this region was not associated with diffusion restriction (B, C) or early arterial enhancement (D), and, in retrospect, in keeping with central zone (CZ) stromal hypertrophy. Postsurgical pathology revealed left and right PZ acinar adenocarcinoma with a Gleason score of 3 + 4 = 7 and 4 + 4 = 8, respectively, corresponding to T2W and DCE‐T1W imaging (A and D). No tumor was identified in the CZ, with histological features of stromal hypertrophy. [Color figure can be viewed at wileyonlinelibrary.com]

Figure 5

Image quality of sequences analyzed. Image quality is scored from 1 to 4. The relative quality score of T2W images greater than for DWI (p = 0.0043), ADC (p = 0.0007) or DCE T1W (p < 0.0001). A higher proportion of DCE scans were given a score of 3 due to the lack of fat suppression in a subset of cases (Fat‐suppression is the current recommended standard of practice by the ACR PI‐RADS documentation). [Color figure can be viewed at wileyonlinelibrary.com]