Exploration of Interstitial Fibrosis in Chronic Kidney Disease by Diffusion‐Relaxation Correlation Spectrum MR Imaging: A Preliminary Study

Abstract

BackgroundRenal interstitial fibrosis is one of the most common pathways in the progression of chronic kidney disease (CKD). Noninvasive evaluation of interstitial fibrosis would help monitoring CKD progression and prognosis prediction.PurposeTo evaluate the severity of renal interstitial fibrosis by diffusion‐relaxation correlation spectrum imaging (DR‐CSI).Study TypeProspective.SubjectsForty patients with CKD and 10 healthy controls (average age 49.2 ± 14.8 years, 18 females).Field Strength/Sequence3‐T, DR‐CSI with 36 axial spin‐echo echo‐planar diffusion‐weighted images (6 b‐values, 6 echo times).AssessmentInterstitial fibrosis level (IFL) was assessed from biopsy results (IFL = 1, fibrosis percentage <25%, defined as mild; IFL = 2, 25%–50%, moderate; IFL = 3, >50%, severe). Estimated glomerular filtration rate (eGFR) was calculated using serum creatinine. The regions of interest included cortex for both kidneys. The diffusivity‐T2 spectrum was assessed considering three compartments (threshold: T2 30–40 msec, diffusivity 5–9 μm2/msec, according to visible peaks): A (low diffusivity, short T2), B (low diffusivity, long T2), and C (high diffusivity). Volume fractions Vi (i = A, B, C) were calculated.Statistical TestsIntra‐class coefficient (ICC, >0.6 as good) to assess inter‐reader agreement of DR‐CSI Vi. Spearman's correlation to assess relationship of Vi to IFL and eGFR. Receiver operating characteristic analyses with the area under the curve (AUC) to discriminate patients with moderate‐severe fibrosis from mild ones. Statistical significance criteria: P‐value <0.05.ResultsICCs were good for all Vi. Correlations were found between IFL and VB (r = 0.424, significant) and VC (r = −0.400, significant), and between eGFR and VB (r = −0.303, P = 0.058) and VC (r = 0.487, significant). Regarding VB and VC, the AUCs were 0.903 and 0.824.Data ConclusionDR‐CSI help distinguish patients with moderate or severe renal interstitial fibrosis from mild ones.Evidence Level: 2Technical Efficacy: Stage 2

Grading Clear Cell Renal Cell Carcinoma Grade Using Diffusion Relaxation Correlated MR Spectroscopic Imaging

“…The spectral characteristics of normal cortex were found to be consistent with those observed in a previous study focused on renal interstitial fibrosis. 27 The results of this study suggested that sub-regions B and D were associated with the histological grades of ccRCC. The utility of these sub-regions can be reasonably explained.…”

“…In contrast to the ex vivo studies that utilized MDCM methods to determine the specific spectral peaks of intra-voxel tissue components based on prior knowledge of tissue microenvironments, 24,25 T2, ADC, and DR-CSI face challenges in providing biological interpretation for in vivo studies due to their lack of specificity in tissue microenvironments. 18,19,27 This low specificity is not unique to MDCM methods but is a challenge for all MRI methods. The dynamic changes of in vivo tissue microenvironments in tumor, e.g., the intracellular or extracellular water, barriers or membranes, or microvascular structures would interfere with each other, making it difficult for MRI methods to measure and resolve.…”

“…As one of the MDCM methods, DR-CSI has been applied under a two-dimensional (T E and b-value) data acquisition framework. 24,27 As such, the MR signal could be expressed as:…”

“…[24][25][26] Additionally, a recent study has attempted to use DR-CSI for in vivo renal interstitial fibrosis, and has provided better insight into tissue microenvironment than conventional quantitative metrics. 27 We hypothesized that DR-CSI could be used to understand tissue microenvironments in vivo, despite limited MR spatial resolution, by leveraging its capability of powerful spectral analysis based on the correlation of diffusion and relaxation time. Thus, the objective of this study was to compare DR-CSI with quantitative T2 and ADC for grading of ccRCC, using histopathologic results as a reference standard.…”

“…DR‐CSI has been successfully tested on phantoms and subsequently applied to three ex‐vivo cases, two animal spine cord specimens and a malignant prostate cancer, where the distinct spectral peaks and their volume fractions derived from the DR‐CSI results corresponded to different tissue components 24–26 . Additionally, a recent study has attempted to use DR‐CSI for in vivo renal interstitial fibrosis, and has provided better insight into tissue microenvironment than conventional quantitative metrics 27 …”