White matter microstructure damage measured by automated fiber quantification correlates with pain symptoms in lung cancer patients.

ABSTRACT

To investigative the white matter (WM) alterations in lung cancer patients with cancer pain (CP+), and explore the correlations between damaged WM fiber tracts and clinical indicators. Twenty-six CP+, 26 lung cancer patients without CP (CP-), and 31 healthy controls (HC) were recruited. All participants underwent diffusion tensor imaging (DTI) and clinical assessments. Automated fiber quantification (AFQ) technique was performed to identify the 20 WM fiber bundles, and the fractional anisotropy (FA), mean diffusivity (MD), axial diffusivity (AD), and radial diffusivity (RD) were extracted. Intergroup comparisons of these diffusion metrics were conducted based on the entire fiber bundle level and 100 node levels along each tract. The associations between altered diffusion metrics and the numeric rating scale (NRS) scores, as well as the pain duration, were analyzed. At the entire level, the CP + group showed impaired WM structure in the right cingulum hippocampus (CH_R). At the pointwise level, the CP + group exhibited extensive nodal FA reduction or MD, RD, and AD elevation. In addition, the AD of the posterior portion of the right inferior longitudinal fasciculus (ILF_R, nodes 71-75) in the CP + group was positively correlated with the pain duration, and the FA of CH_R (nodes 22-38) was negatively correlated with NRS score. Extensive WM microstructural damage may be a pattern of brain abnormalities in lung cancer patients with CP, and in particular, specific nodal disruption along pain-related fiber tracts may be a sensitive imaging biomarker to characterize the severity and duration of CP.