Login to Access Video or Poster Abstract: MP03-20
Sources of Funding: None

Introduction

Apparent diffusion coefficient (ADC) grade is increasingly reported in prostate magnetic resonance imaging (MRI). We evaluated the usefulness and limitations of ADC grade for peripheral zone (PZ) and transitional zone (TZ).

Methods

We reviewed the data of 455 consecutive men who underwent 3.0 Tesla diffusion-weighted MRI before radical prostatectomy from January 2015 to May 2016. The level of suspicion from the ADC map was graded using the 5-grade Likert scale. Patients with a Likert grade ≥ 3 were included in the final analysis. The MRI index lesion was defined as the lesion with highest grade or the largest lesion when multiple lesions had an identical grade. After radical prostatectomy, topographic analyses were performed on the intraprostatic location of tumor foci at each specimen. The pathologic index tumor was defined as the tumor with the highest Gleason score or largest tumor when multiple foci had an identical GS. We matched the location between MRI index lesions and pathologic index tumors. The concordance rate was compared between the PZ and TZ using chi-square tests.

Results

In 455 patients with prostate cancer, 350 (77%) had suspicious MRI index lesions (ADC grade ≥ 3). Of the 350 lesions, 58% were seen in the PZ and 42% in the TZ. The overall concordance rate was gradually increased from ADC grade 3 to 5 (52%, 72%, and 86%) and biopsy Gleason score 6 to ≥ 8 (68%, 77%, and 80%). The overall concordance rate in the PZ was significantly higher than in the TZ (82% vs. 67% p < 0.01). The concordance rate in the PZ was higher than in the TZ among the subgroup of patients with ADC grade 5 (91% vs. 76%, p = 0.007). However, the rate was similar among the subgroup of patients with ADC grade 3 (50% vs. 54%, p = 0.78).

Conclusions

ADC grade from diffusion-weighted MRI is more useful in men with PZ tumor than in those with TZ tumor. Especially, ADC grade 5 in the PZ showed higher accuracy for the detection of index tumor.

Funding

None