Fibrosis imaging with multiparametric proton and sodium MRI in pig injury models

Camilla W. Rasmussen,Nikolaj Bøgh,Sabrina K. Bech,Thomas H. Thorsen,Esben S. S. Hansen,Lotte B. Bertelsen,Christoffer Laustsen
DOI: https://doi.org/10.1002/nbm.4838
2022-09-25
NMR in Biomedicine
Abstract:Chronic kidney disease (CKD) is common and has vast implications for health and mortality. It is aggravated by intrarenal fibrosis, but the assessment of fibrosis is limited to kidney biopsies, which contains risk of complications and sampling errors. This calls for a noninvasive modality for diagnosing and staging intrarenal fibrosis. The present study is an exploratory study that evaluates a multiparametric MRI protocol including sodium (23Na‐MRI) imaging to determine the opportunities within this modality to assess kidney injury as a surrogate endpoint of fibrosis. The study includes 43 pigs exposed to either ischemia reperfusion injury (IRI), unilateral ureteral obstruction (UUO), or serving as healthy controls. Fibrosis was determined using gene expression analysis of collagen. The medulla/cortex ratio of 23Na‐MRI decreased in the injured kidney in the IRI pigs, but not in the UUO pigs (p = 0.0180, p = 0.0754). To assess the combination of MRI parameters in estimating fibrosis, we performed a linear regression model consisting of cortical ADC, ΔR2*, ΔT1, 23Na medulla/cortex ratio, and plasma creatinine (R2 = 0.8009, p = 0.0117). The 23Na medulla/cortex ratio did only slightly improve the fibrosis prediction model, leaving 23Na‐MRI in an ambiguous place for evaluation of intrarenal fibrosis. Use of multiparametric MRI in combination with plasma creatinine show potential for estimation of fibrosis in human kidney disease, but more translational and clinical work is warranted before MRI can contribute to earlier diagnosis and evaluation of treatment of AKI and CKD.
radiology, nuclear medicine & medical imaging,biophysics,spectroscopy
What problem does this paper attempt to address?