Evaluation of Diffusion Kurtosis Imaging Versus Standard Diffusion Imaging for Detection and Grading of Peripheral Zone Prostate Cancer

Abstract

Objective To evaluate diffusion kurtosis imaging (DKI) and magnetisation transfer imaging (MTI) compared to standard MRI for prostate cancer assessment in a re-biopsy population. Methods Thirty-patients were imaged at 3 T including DKI (K app and D app) with b-values 150/450/800/1150/1500 s/mm 2 and MTI performed with and without MT saturation. Patients underwent transperineal biopsy based on prospectively defined MRI targets. Receiver-operating characteristic (ROC) analyses assessed the parameters and Wilcoxon-signed ranked test assessed relationships between metrics. Results Twenty patients had ≥ 1 core positive for cancer in a total of 26 MRI targets (Gleason 3+3 in 8, 3+4 in 12, ≥ 4+3 in 6): 13 peripheral (PZ) and 13 transition zone (TZ). The apparent diffusion coefficient (ADC) and D app were significantly lower and the K app and MT ratio (MTR) significantly higher in tumour versus benign tissue (all p ≤ 0.005); ROC values 0.767-1.000. Normal TZ had: lower ADC and D app and higher K app and MTR compared to normal PZ. MTR showed a moderate correlation to K app (r = 0.570) and D app (r =-0.537) in normal tissue but a poor correlation in tumours. No parameter separated low-grade (Gleason 3+3) from high-grade (≥ 3+4) disease for either PZ (p = 0.414-0.825) or TZ (p = 0.148-0.825). Conclusion ADC, D app , K app and MTR all distinguished benign tissue from tumour, but none reliably differentiated low-from high-grade disease. Key Points • MTR was significantly higher in PZ and TZ tumours versus normal tissue • K app was significantly lower and D app higher for PZ and TZ tumours • There was no incremental value for DKI/MTI over mono-exponential ADC parameters • No parameter could consistently differentiate low-grade (Gleason 3+3) from high-grade (≥ 3+4) disease • Divergent MTR/DKI values in TZ tumours suggests they offer different functional information

A Standardized Parameter-Free Algorithm for Combined Intravoxel Incoherent Motion and Diffusion Kurtosis Analysis of Diffusion Imaging Data

“…It has been previously described that introducing kurtosis in the monoexponential diffusion model systematically increases the observed diffusion coefficient, 29,30 which is in line with the findings in our study when kurtosis was introduced in the IVIM model. For positive kurtosis, such as observed in our study as well as in other dedicated DKI studies, 2,[30][31][32][33] at least a part of this effect can be explained with the way kurtosis is introduced in the model, as D and K are introduced in the equation with opposite signs. 31,34,35 ).…”

Imaging of Local Recurrence in Prostate Cancer

“…Diffusional kurtosis imaging has been applied for the identification of PCa, with overall mixed results to date [78][79][80]. High b-value technique, theoretically attenuating signal from water molecules not truly restricted within cells, demonstrates promise for improved detection and characterization of PCa [81,82].…”

MRT der Prostata: Empfehlungen zur Vorbereitung und Durchführung

“…These complex diffusion models are a current topic of research. The currently available data do not show a significant advantage of these methods compared to classic diffusion weighting and should therefore not be used for diagnosis at this time [39].…”

“…Intravoxel incoherent motion imaging (IVIM) takes into consideration the multiexponential behavior of the diffusion signal at different b-values and the influence of the perfusion components of the signal at low b-values. Diffusion kurtosis imaging (DKI) takes into consideration the kurtosis of the tissue which refers to the deviation from the Gaussian distribution [39]. These complex diffusion models are a current topic of research.…”