Quantitative Ex Vivo MRI Changes due to Progressive Formalin Fixation in Whole Human Brain Specimens: Longitudinal Characterization of Diffusion, Relaxometry, and Myelin Water Fraction Measurements at 3T

PURPOSE: Postmortem MRI can be used to reveal important pathologies and establish radiology–pathology correlations. However, quantitative MRI values are altered by tissue fixation. Therefore, the purpose of this study was to investigate time-dependent effects of formalin fixation on MRI relaxometry (T(1) and T(2)), diffusion tensor imaging (fractional anisotropy, FA; and mean diffusivity, MD), and myelin water fraction (MWF) measurements throughout intact human brain specimens. METHODS: Two whole, neurologically-healthy human brains were immersed in 10% formalin solution and scanned at 13 time points between 0 and 1,032 h. Whole-brain maps of longitudinal (T(1)) and transverse (T(2)) relaxation times, FA, MD, and MWF were generated at each time point to illustrate spatiotemporal changes, and region-of-interest analyses were then performed in eight brain structures to quantify temporal changes with progressive fixation. RESULTS: Although neither of the diffusion measures (FA nor MD) showed significant changes as a function of formalin fixation time, both T(1) and T(2)-relaxation times significantly decreased, and MWF estimates significantly increased with progressive fixation until (and likely beyond) our final measurements were taken at 1,032 h. CONCLUSION: These results suggest that T(1)-relaxation, T(2)-relaxation and MWF estimates must be performed quite early in the fixation process to avoid formalin-induced changes compared to in vivo values; and furthermore, that different ex vivo scans within an experiment must be acquired at consistent (albeit still early) fixation intervals to avoid fixative-related differences between samples. Conversely, ex vivo diffusion measures (FA and MD) appear to depend more on other factors (e.g., pulse sequence optimization, sample temperature, etc.).