Using mathematical model to analyzing dynamic contrast enhanced MRI to distinguish between stromal benign prostatic hyperplasia and prostate cancer
Xiaobing Fan1, Shiyang Wang1, Milica Medved 1, Tatjana Antic2, Serkan Guneyli 1, Gregory S Karczmar1, and Aytekin Oto1

1Radiology, University of Chicago, Chicago, IL, United States, 2Pathology, University of Chicago, Chicago, IL, United States

Synopsis

Previous dynamic contrast enhanced (DCE) MRI studies demonstrated that stromal benign prostatic hyperplasia (BPH) nodules are difficult to differentiate from transition zone prostate cancer (PCa). Therefore, it is important to improve the accuracy of DCE-MRI to distinguish BPH from PCa. A total of 24 patients with biopsy confirmed PCa were enrolled in this study. DCE-MRI data were acquired at 3 T for a total of ~8.3 minutes. The relative signal enhancement curves for cancer (n=24) and BPH (n=19) were calculated and analyzed using an empirical mathematical model. The ratio of washout-rate/uptake-rate (p<0.01) was significantly smaller in BPH than in cancers.

Introduction

Multi-parametric prostate MRI, including diffusion-weighted imaging, dynamic contrast enhanced (DCE) imaging and/or MR spectroscopy, has become the reference standard for prostate imaging (1). However, the diagnostic accuracy of DCE-MRI is limited in large part because of significant overlap between enhancements produced by contrast media in prostatic carcinoma and benign prostatic hyperplasia (BPH) (2, 3). BPH is extremely common in the transition zone (TZ) where most cancer is found, and can be nodular. Nodular BPH is often categorized into three main subtypes: glandular, stromal and mixed. Among them, glandular BPH can be easily differentiated from TZ PCa. In contrast, stromal BPH nodules can be difficult to differentiate from TZ PCa (3, 4). Therefore, it is important to improve the accuracy of DCE-MRI to distinguish BPH from PCa. Here we demonstrate that the use of an empirical mathematical model (EMM) to fit the relative signal enhancement DCE-MRI curve as a function of time has potential to differentiate BPH from cancer.

Methods

The study was compliant with the HIPAA and approved by our IRB, and all participants provided written informed consent. Twenty-four patients (n = 24) with biopsy confirmed prostate cancer were enrolled in this study (mean age = 58.9 years, range 40 to 70 years). A glomerular filtration rate value greater than or equal to 60 mL/min/1.73 m2 was required. All images were acquired with a Philips Achieva 3 T TX scanner using the combination of a phased array and endorectal coil. After other clinically required scans, axial dual-echo 3D T1-weighted (T1W) DCE-MRI data (TR/TE1/TE2 = 4.8/1.69/3.3 ms, flip angle = 10°, field of view = 250×380×84 mm3, resolution = 1.25×1.75×3.5 mm3, reconstruction resolution = 1.0×1.0×3.5 mm3, SENSE factor = 1.67, partial Fourier factor = 0.675, temporal resolution = 8.3 s) were acquired by using a two-point modified Dixon (mDixon) method to produce the ‘water-only’ images. Gadobenate dimeglumine (Multihance) was injected at the standard dose of 0.1 mmol/kg. The DCE data were acquired for a total of ~8.3 minutes.

Peripheral zone (PZ) and transition zone (TZ) tumor nodules in the prostate were confirmed with biopsy. All patients in the study underwent prostatectomy afterwards. Regions-of-interest (ROIs) were drawn by an experienced radiologist on ADC maps and T2-weighted (T2W) images, and transferred to DCE data for each patient. A total of 24 cancer ROIs on PZ/TZ and 19 BPH ROIs were drawn and used in this study.

The average signal intensity S(t) over ROI was calculated first, then the relative signal enhancement curve (Sr(t)) as function of time (t) was calculated as: $$$S_{r}(t)=(S(t)-S_{0})/S_{0}$$$, where S0 is the baseline signal. The modified three parameters empirical mathematical model (EMM) (5): $$$S_{r}(t)=A\cdot(1-\exp(-\alpha\cdot t))\cdot\exp(-\beta\cdot t)$$$ was used to fit the Sr(t), where A is the scaling constant, α is the rate of contrast agent uptake, and β is the rate of contrast agent washout. The non-parametric Mann-Whitney U-Test was performed to determine whether there was a statistically significant difference between cancer and BPH for fitted EMM parameters. A p-value less than 0.05 were considered significant.

Results

Figure 1 shows T2W image from a patient with cancer (red circle) and BPH (green circle) outlined. The corresponding relative signal enhancement curve Sr(t) from cancer and BPH is shown in Figure 2. For all Sr(t) from 24 cancers and 19 BPH ROIs, the average uptake and washout rates obtained from the EMM are given in Table 1. On average, uptake rate (α) was slightly faster in BPH than cancer (but not statistically significant), and washout rate (β) was significantly slower (p <0.05) in BPH than cancer. Figure 3 shows the box-plot for the ratios of β/α for cancer and BPH. The statistical analysis showed that the BPH had significantly small ratio of β/α (p <0.01) than the cancers.

Discussion

Prostate DCE-MRI data from 24 patients were analyzed by using an EMM based on three parameters. On average, the washout rate for BPH was only about half of the washout rate in cancer. The ratio of contrast agent uptake and washout rates was significantly different in BPH vs. cancer. The results suggest that use of a simple mathematical model to analyze data can improve the diagnostic accuracy of DCE-MRI. Although we followed the contrast agent washout phase up to ~8 min, which was longer than most clinical scans, it is important to follow the washout phase even longer in order to obtain accurate washout rate. The ratio of contrast agent uptake and washout could be a bio-maker to distinguish BPH and PCa.

Acknowledgements

This research is supported by NIH R01 CA172801-01.

References

1. Eberhardt SC, Carter S, Casalino DD, Merrick G, Frank SJ, Gottschalk AR, Leyendecker JR, Nguyen PL, Oto A, Porter C, Remer EM, Rosenthal SA. ACR Appropriateness Criteria prostate cancer--pretreatment detection, staging, and surveillance. J Am Coll Radiol. 2013; 10(2):83-92.

2. Hoeks CM, Hambrock T, Yakar D, Hulsbergen-van de Kaa CA, Feuth T, Witjes JA, Fütterer JJ, Barentsz JO. Transition zone prostate cancer: detection and localization with 3-T multiparametric MR imaging. Radiology. 2013; 266(1):207-17.

3. Oto A, Kayhan A, Jiang Y, Tretiakova M, Yang C, Antic T, Dahi F, Shalhav AL, Karczmar G, Stadler WM. Prostate cancer: differentiation of central gland cancer from benign prostatic hyperplasia by using diffusion-weighted and dynamic contrast-enhanced MR imaging. Radiology. 2010; 257(3):715-23.

4. Rosenkrantz AB, Taneja SS. Radiologist, be aware: ten pitfalls that confound the interpretation of multiparametric prostate MRI. AJR Am J Roentgenol. 2014; 202(1):109-20.

5. Fan X, Abe H, Medved M, Foxley S, Arkani S, Zamora MA, Olopade OI, Newstead GM, Karczmar GS. Fat suppression with spectrally selective inversion vs. high spectral and spatial resolution MRI of breast lesions: qualitative and quantitative comparisons. J Magn Reson Imaging. 2006; 24(6):1311-5.

Figures

Figure 1. T2W image from a 63 years old patient with cancer (red circle) and BPH (green circle) outlined.

Figure 2. The corresponding plots of Sr(t) from cancer and BPH shown in Figure 1.

Figure 3. The box-plot for the ratios of β/α for cancer and BPH. The plus signs (+) indicate mean, and the asterisks (*) indicate the upper and lower limits of the data.

Table 1. Average value (± standard deviation) of the EMM parameters obtained from fitting the relative signal enhancement curves of cancers and BPHs.



Proc. Intl. Soc. Mag. Reson. Med. 24 (2016)
2504