Predictive role of ADC parameters in differentiating prostate cancer with predominant Gleason pattern of 3 and 4
Type:Scientific Electronic Presentation
Authors: Lorenzo Vassallo, Mirella Fasciano, Andrea Cerutti, Carla Angela Zavattero
Keywords:ADC - Prostate cancer - MRI - Gleason Pattern

To compare the performance of mean tumor apparent diffusion coefficient (ADC T ) with that of ADC ratio (ADC T /ADC normal tissue) in differentiating Gleason pattern (GP) <= 3+4 from the more aggressive GP >= 4+3 using radical prostatectomy (RP) specimen as the reference standard and to evaluate intereader agreement.

Material and methods

We retrospectively analyzed 67 consecutive patients who underwent 1.5T Mp-MRI of the prostate followed by radical prostatectomy between January 2017 and July 2020.
For each lesion identified on MRI and histopathologically confirmed, three radiologists independently calculated mean tumor ADC(ADC T ), ADC of adjacent normal tissue (ADC N ) and ADC ratio ( ADC T /ADC N ).
For the purpose of the study GP at histology were binarized as follows: 3+3 and 3+4 were considered low grade tumor, while 4+3, 4+4+, 4+5 and 5+4 were considered high.
ADC T and ADC ratio were correlated with GP using the Spearman correlation coefficient and were compared between low grade and high grade GP using the unpaired t test. ROC curves were used to compared diagnostic accuracies of ADC T and ADC ratio in determining high prostate cancer.
Agreement between the readers was measured using the... more


38 lesions were GP <= 3+4, 29 were >=  4+3. ADC T and ADC ratio were significantly different between the two groups (p<0.001).

AUC, sensitivity and specificity of logistic regression model including ADC T and ADC ratio were 0.98, 0.94 and 0.90 for reader 1, 0.98, 0.95 and 0.86 for reader 2 and 0.96, 0.94 and 0.86 for reader 3.

Intereader agreement was excellent (k=0.9).


ADC parameters could differentiate GP <= 3+4 from >= 4 + 3 PCa with high accuracy and an excellent agreement between the readers.