Juergen E. Scheidler MD, Presenter: Nothing to Disclose

Markus Rechl, Abstract Co-Author: Nothing to Disclose

Christian Brinkschmidt MD, Abstract Co-Author: Nothing to Disclose

Andreas Friedrich Heuck MD, Abstract Co-Author: Nothing to Disclose

Christian Glaser MD, Abstract Co-Author: Nothing to Disclose

Studies have shown the high influence of readers' experience on the accuracy of prostate MRI. The aim of this study was to assess whether the widely reader-independent calculation of minimal ADC within the peripheral (PZ) and transitional zone (TZ) may assist in patient selection for biopsy or re-biopsy in pts with suspected prostate cancer (PC). 

After IRB approval 412 patients (pts) referred to prostate MRI were identified who fulfilled the inclusion criteria of rising PSA and unremarkable DRE/TRUS. eDWI was performed at 3T or 1.5T at b-values of 50 and 800. Min. ADC were calculated for the left/right peripheral (PZ) and transitional (TZ) zone and correlated on a side-by-side basis to 8-12 core biopsy (231 pts) or clinical follow-up (PSA reduction) of at least two years. ROC curves and post-test probabilities for a given ADC-threshold were calculated using the Bayes theorem for PZ and TZ prostate cancer (PC).

157/412 pts (234/824 prostate lobes) were diagnosed with PC. In 193 lobes tumor was present in the PZ, whereas in 41 lobes tumor was only affecting the TZ. Mean ADC±SD values for benign vs. malignant tissue were 1.60 ± 0.25 vs. 0.97 ± 0.19 x10 -3mm2/sec within the PZ (p<0.001); and 1.25 ± 0.25 vs. 0.97 ± 0.24 x10-3mm2/sec within the TZ (p<0.05). No significant difference were found between 1.5 and 3T. AUC was calculated to 0.92 for the PZ (95%CI 0.90 - 0.94) and 0.74 for the TZ (95%CI 0.70 - 0.78). Applying an ADC-threshold of 1.3 for the PZ and 1.0 for the TZ, sensitivity and specificity were 97%/72% and 95%/37%, respectively. ADC below these thresholds increases the probability for PC to 52% (PZ) and 44% (PZ), whereas a minimal ADC above these thresholds reduces the probability for PZ-PC to 1.4% and for TZ-PC to 6.8%. No tumor free TZ had an ADC below 0.75.

Diagnosis of PC based on min. ADC in eDWI assists in patient selection for biopsy. Reducing the post-test probability for PZ-PC in pts with min. ADC of >1.3 to 1.4% allow for further clinical follow-up instead of (re-)biopsy. Since the threshold based approach (ADC >1.0) is less effective (post-test probability 6.8%) for the rarer TZ-PC, additional criteria (min. benign ADC=0.75, morphology) need to be considered for diagnosis. 

eDWI prostate MRI may serve as a rule-out test prior to biopsy in patients with rising PSA and unremarkable DRE/TRUS. 

Scheidler, J, Rechl, M, Brinkschmidt, C, Heuck, A, Glaser, C, Diagnosis of Prostate Cancer in Patients with Rising PSA but Unremarkable Digital Rectal Exam/Transrectal Ultrasound: Value of Endorectal Diffusion-Weighted MR Imaging at 1.5 and 3T in a Large Patient Cohort for the Selection of Patients for Biopsy.  Radiological Society of North America 2014 Scientific Assembly and Annual Meeting, - ,Chicago IL. http://archive.rsna.org/2014/14011349.html