Login to Access Video or Poster Abstract: MP38-11
Sources of Funding: This work was supported by funds from the Integrated Diagnostics Program, Department of Radiological Sciences & Pathology, David Geffen School of Medicine at UCLA

Introduction

To evaluate Prostate Cancer (CaP) tumor volume (TV) of regions of interest (ROI) on prostate multiparametric magnetic resonance imaging (MP-MRI) (TVm) with concordant foci TV on whole mount histopathology (WMHP) (TVh), stratified by Gleason score (GS) and MP-MRI Prostate Imaging Reporting and Data system version 2 (PI-RADSv2) assessment.

Methods

A HIPAA-compliant, IRB-approved study of 105 men who underwent 3Tesla (3T) prostate MP-MRI before robotic radical prostatectomy was performed. Suspicious regions of interest (ROIs) were contoured on MRI and PI-RADSv2 was assigned per suspicious ROI. TVT2, TVDCE and TVDWI was calculated by delineation of suspicious tumor on T2-weighted, Dynamic Contrast Enhancement and Diffusion Weighted Imaging pulse sequences. A 3D patient-specific mold was printed to precisely fit the excised prostate based on MRI prostate segmentation. A pathologist contoured each tumor on all WMHP slides and graded the tumor by GS system. Custom software automatically imported the annotated contoured WMHP slides, reconstructed the tumors in 3D, and calculated the TVh. Spearman correlation coefficient(r) was calculated to determine strength of association between volumes of concordant lesion foci on each MRI and WMHP. Spearman correlation and Wilcoxon rank sum test were used. P-values<0.05 were considered significant.

Results

In 105 patients, 132 CaP foci (including 103 index tumors) on WMHP matched with MRI ROIs concordantly. Of 132 CaP foci GS was ≤6(3+3) in 24 (18.2%) and ≥7(3+4) in 108 (82.8%). PI-RADS was normal or mild suspicious(1-2) in 11(8.3%), 3 in 37 (28%), 4 in 48 (34.4%), and 5 in 36(27.3%). The median (IQR) of TVh, TVT2, TVDCE and TVDWI were 1.1cc (0.5-3.1), 1.3cc (0.8-2.7), 1.6cc (1-2.8) and 1.4cc (0.6-2.7) respectively. The r between TVh and TVT2, TVDCE and TVDWI was 0.53, 0.42 and 0.42 in all tumors,and 0.52, 0.47 and 0.53 in GS≥7(3+4) tumors respectively . The r between TVh and TVT2 was 0.33, 0.55 and 0.61 in PI-RADS 3, 4 and 5 lesions retrospectively. The r between TVh and TVDCE was 0.38,0.28 and 0.67 in PI-RADS 3,4 and 5 retrospectively. The r between TVh and TVDWI was 0.26, 0.40 and 0.61 in PI-RADS 3, 4 and 5 retrospectively (p<0.001).

Conclusions

Tumor volume on pathology (TVh) correlates with corresponding tumor volumes on 3T MP-MRI pulse sequences (TVT2, TVDWI and TVDCE) and it is stronger in high grade tumors and in tumors with higher PI-RADSv2. However, tumor volume on T2 (TVT2) consistently demonstrates the strongest correlation with pathology tumor volume (TVh).

Funding

This work was supported by funds from the Integrated Diagnostics Program, Department of Radiological Sciences & Pathology, David Geffen School of Medicine at UCLA