The correlation between quantitative parameters in mutiparametric MRI and PI-RADS 2.0.


Year:

Session type:

Theme:

Bowen Jin1,Cheng Wei1,Magdalena Szewczyk-Bieda1,Ghulam Nabi1
1University of Dundee

Abstract

Background

The metabolic and hemodynamic information is evaluated by both qualitative and quantitative parameters in DCE-MRI. The aim of our study was to answer whether these parameters correlate with PIRADS scoring on MRI diagnosis of prostate cancer.

Method

63 men who had an abnormal DRE, elevated PSA and received 3T mp-MRI were recruited into this peer reviewed, protocol based, prospective study, between February 2015 and March 2016. 7 participants with incomplete DCE sequences were withdrawn. 41 of 56 participants with positive MR were analysed and suspicious lesions which have PIRADS score of 3 or above were marked by experienced radiologist. The DCE-MRI semi-quantitative values (wash – in (PE%) and – out curves (SER%)) and qualitative parameters (Ktrans, kep) of the lesions in two kinetic models (Tofts & Kermode, extended Tofts) were recorded in an advanced MRI post-processing software (Olea Medical, La Ciotat, France). The correlation analysis between these quantitative parameters and the previous MRI diagnosis were calculated.

Results

71 lesions were marked in the 41 patients with positive MRI. 21 of the lesions were PIRADS 3. 17 of the lesions were PIRADS 4 and 33 were PIRADS 5. In the extended Tofts model, the mean Ktrans value of PIRADS 3, 4 and 5 lesions were 0.35min-1, 0.55min-1and 0.57 min-1, respectively (p value=0.041). However, in the Tofts & Kermode model, the mean Ktrans value of PIRADS 3, 4 and 5 are 1.0329, 1.0776 and 1.065 min-1 (p value=0.380), respectively. In addition, kep of both models, PE% and SER% were not correlated with PIRADS score, when assessed using a quantitative method rather than shape of the curves.

Conclusion

Ktrans using extended Tofts model, positively correlates with the PIRADS scores but other quantitative methods, assessing wash in and out did not correlate with the PIRADS scoring. Further analysis of histopathology and Gleason score is needed to confirm these findings.