Serial T₁ and T₂ measurements of metastatic bone lesions in prostate cancer patients: MR fingerprinting vs conventional MRI.

Abstract

This study evaluated serial magnetic resonance fingerprinting (MRF)-derived T₁ and T₂ relaxivities of prostate bone metastasis compared with conventional T₁ and T₂ measurements. This prospective study (July 2020 to July 2022) included MRF and conventional MRI acquisitions (T₁: inversion-recovery turbo spin echo; T₂: dual spin echo) from participants with bone metastasis from primary prostate cancer from two cohorts: pre-treatment (N = 34) and pre/post-treatment (N = 19). Phantom/human data were acquired on a 1.5-T scanner using an MRF sequence outputting T₁ and T₂ maps. Regions of interest (ROIs) of bone metastasis were drawn per visit on both MRF and conventional MRI. Inter-method reproducibility of T₁ and T₂ was assessed using Bland-Altman plots, reproducibility, intraclass correlation, and Spearman correlation coefficients. A delta parameter [post-treatment - pre-treatment] of method-specific T₁ and T₂ was reported. Thirty-four patients with metastatic prostate cancer (mean age, 68 years ± 7 [standard deviation]) were evaluated pre-treatment; 19 participants were further scanned post-treatment. MRF-derived mean T₁ and T₂ in bone metastasis were slightly higher than the conventional MR measurements: 10.8% (T₁) and 15.5% (T₂). The reproducibility coefficient (r%) was 19.3% for T₁ and 32.5% for T₂, whilst the Spearman correlation coefficient was strong for both parameters (0.66, p < 0.001 and 0.70, p < 0.001). The MRF-derived delta T₁ parameter was moderately correlated to the inversion-recovery method (0.59, p = 0.008), whilst the MRF-derived delta T₂ was very strongly correlated to the dual spin echo method (0.80, p < 0.001). A good correlation of MRF-derived T₁ and T₂ measurements with conventional quantitative methods was demonstrated in bone metastasis. Question MR fingerprinting (MRF)-derived T₁ and T₂ values have the potential to characterise bone metastasis and treatment response, but their performance against conventional MRI is unclear. Findings The inter-method reproducibility coefficient was 19.3% for T₁ and 32.5% for T₂, whilst the Spearman correlation coefficient was strong for both parameters. Clinical relevance Serial MRF-derived T₁ and T₂ measurements in bone metastasis in patients with prostate cancer correlated well with conventional MRI measurements, supporting MRF use for faster quantitative measurements in bone lesions.