Synopsis

We quantified T1, T2, and proton density (PD) of bone metastases in castration-resistant prostate cancer with synthetic MRI (SyMRI). Bone foci of interest were classified into four groups; active disease, red bone marrow (RBM), inactive disease without calcification (Inactive C(-)), and inactive disease with calcification (Inactive C(+)). Active disease group showed very high PD, and Inactive C(+) group showed very low PD. Both Inactive C(-) and RBM showed medium values. Significant differences were noted among these three divisions. SyMRI thus shows clinical potential to differentiate active/inactive lesions, calcifications, and red bone marrow in castration-resistant prostate cancer.

Body of Abstract

Introduction: Bone scintigraphy has traditionally been the most popular method for the evaluation of the bone metastases, although it inherently suffers from the flare phenomenon, as a result of which it cannot be repeatedly used at short time intervals during follow-up. Recently, the European Organization for Research and Treatment of Cancer imaging group (EORTC) positioned whole body diffusion weighted imaging (WB-DWI) as a potential first choice in this setting [1]. With WB–DWI, tumor activity may be evaluated based on signal intensity; high intensity implies active, and low intensity implies inactive disease. However, we still need additional in-phase/out-of-phase images to identify red bone marrow (which may mimic active lesions on WB-DWI), and need computed tomography (CT) to evaluate for calcifications/sclerosis especially in cases with prostate cancer. Recently developed synthetic MRI (SyMRI) allows generating various image contrasts by one single scan (even retrospectively), and it also allows for quantitative analysis of T1, T2, and proton density (PD) [2]. We studied if quantitative values obtained with SyMRI are useful to evaluate tumor activity and presence of calcifications in bone metastases.

Methods: This prospective study includes twenty-one castration-resistant prostate cancer patients. Diffusion-weighted Whole Body Imaging with Background Body Signal Suppression (DWIBS; 2:39 min/station) and SyMRI 3:47 min) were performed in all patients. A total of 68 detected foci of interest were classified into four groups; (1) active disease, (2) red bone marrow (RBM), (3) inactive disease without calcifications (Inactive C(-)), and (4) inactive disease with calcifications (Inactive C(+)), according to signal intensity on DWI,CT, out-of-phase signal loss, prostate specific antigen (PSA), and bone scintigraphy. A 1 cm-diameter region of interest (ROI) was randomly set in each focus of interest by two radiologists independently and corresponding T1, T2, and PD were calculated by dedicated SyMRI software. The differences in T1, T2, and PD values were then assessed among the four groups. Statistical analysis was conducted by Wilcoxon rank sum test and P=0.05 or less was considered significant. The correlations of T1, T2, and PD values between the readers were assessed by Spearman correlation analysis. Additionally, three-dimensional scatter plots were created to visualize the distribution of the values among the 4 groups.

Results: The number of included cases (lesions) in each group (active, RBM, Inactive C(-), and Inactive C(+)) were 12 (25), 7 (10), 10 (19), and 5 (14), respectively. Inter-observer analysis revealed excellent intraclass correlation coefficient for T1, T2, and PD (ρ = 0.72, 0.80, and 0.95, respectively). The difference in PD values among four groups is demonstrated in (Fig.1). The active disease group showed very high PD, and the Inactive C(+) group showed very low PD. Both Inactive C(-) and RBM showed medium values. Significant differences were noted among these three divisions. Regarding T1, a significant difference was noted in both readers only between active and RBM (p=0.0466, and 0.0090, for each reader), and active and Inactive C(-) (p=0.0080, and 0.0060, for each reader). A significant difference in T2 was noted only between active and Inactive C(+) (p=0.0098, and 0.0045, for each reader). Three-dimensional scatter plots (Fig.2 and 3) showed nice visualization of the difference described above. Fig.4 shows an example of a case with an inactive lesion with calcification, characterized by very low PD value. Fig.5 shows an example of a case with red bone marrow, characterized by medium PD value.

Discussion: Based on the results, we can distinguish active lesions from any other inactive or non-tumorous foci. Another impressive thing is that we can know the presence of calcification by SyMRI (without CT). Since the applied TE value was relatively not very low (13ms, shortest echo time among multi echoes), we should lose the signal from lesions with a very short T2. Therefore, the measured PD value may not be the absolute value. Nevertheless, the results encourage us to use visible PD map for the differentiation of active lesions from inactive lesions or red bone marrow, and this may be important for clinical use. Processing the data in a three-dimensional scatter plot may particularly be useful as a powerful visible tool to differentiate the different groups.

Conclusions: SyMRI may differentiate active/inactive lesions, calcifications, and red bone marrow in castration-resistant prostate cancer. The presence of calcifications can be recognized by low PD values.

Acknowledgements

References

1. Lecouvet FE, Talbot JN, Messiou C, et al. Monitoring the response of bone metastases to treatment with Magnetic Resonance Imaging and nuclear medicine techniques: a review and position statement by the European Organisation for Research and Treatment of Cancer imaging group. Eur J Cancer. 2014;50(15):2519-2531.

2. Warntjes JB, Leinhard OD, West J, et al. Rapid magnetic resonance quantification on the brain: Optimization for clinical usage. Magn Reson Med. 2008;60(2):320-329.