#2478 Multiparametric magnetic resonance imaging predicts decline of kidney function but does not perform better than diffusion-weighted MRI alone

Abstract

Abstract Background and Aims Renal cortical interstitial fibrosis, typically measured by biopsy, is a critical factor for kidney function prognosis. Diffusion-weighted magnetic resonance imaging (DW-MRI) has emerged as a novel promising non-invasive method to assess the degree of fibrosis. We have previously shown that there is a strong correlation between the cortico-medullary apparent diffusion coefficient difference (ΔADC) and histologically assessed fibrosis. We also developed a model based on ΔADC, eGFR and proteinuria that calculates a risk score for future renal function decline. Multiparametric MRI combines DW-MRI with other sequences and can provide synchronized insights into perfusion, diffusion, oxygenation, and tissue characterization (T1 and T2 mappings). Our current research evaluates whether additional multiparametric MRI sequences increase the power of the DW-MRI based model to predict renal function decline. Method We used data from 197 patients with either chronic kidney disease (42) or allograft kidneys (155). Each participant underwent a biopsy followed by a multiparametric MRI within the following week. The median follow-up period was 2.2 years, during which laboratory parameters were recorded. The primary endpoint was defined as a rapid decline in kidney function, characterized by a reduction in eGFR of over 30%, or the initiation of dialysis. Prognostic factors for the primary outcome were analyzed using both univariable and multivariable Cox regression models, incorporating MRI parameters (ΔADC, cortical and ΔT1, cortical and ΔT2), eGFR, and proteinuria. Results At baseline, mean eGFR was 54.5 ml/min/1.73 m2 (SD 23.8) and mean proteinuria was 0.90 g/24 h (SD 2.4). Rapid decline of renal function occurred in 54 patients after a median time of 1.1 years (interquartile range, 0.9–2.1 years). In univariable survival analysis, cortical T1 measure and cortico-medullar difference ΔT2 have showed the best association to rapid kidney function decline. Compared to our previously established model incorporating ΔADC, eGFR, and proteinuria, adding cortical T1 did not significantly improve the hazard ratio (from 4.62 (95% CI: 1.56–13.67) to 4.36 (95% CI: 1.46–13.02) with cortical T1 inclusion). Harrell's C-index marginally increased with cortical T1, suggesting no significative improvement in prognostic capacity (0.77 versus 0.79 with cortical T1). Additionally, adjusting the regression model for ΔT2 yielded no enhancement in predictive power. Conclusion Our results did not demonstrate an enhancement of the prognostic power of the composite score with the inclusion of additional MRI sequences. Consequently, no clear benefit of multiparametric MRI on top of DW-MRI was observed. Additional research is required to investigate this issue more thoroughly.