MRI Assessment of Baseline White Matter Microstructure and Its Relationship With the Onset and Trajectory of Cancer‐Related Cognitive Impairment in Breast Cancer Patients

Background White matter microstructure may be affected by cancer and its treatment, yet its association with subsequent cancer‐related cognitive impairment (CRCI) remains unclear. Purpose To investigate whether baseline white matter diffusion metrics are associated with the time to CRCI onset and longitudinal cognitive trajectories. Study Type Prospective longitudinal study. Population A total of 230 female patients (49.4 ± 8.6 years) with histopathologically confirmed breast cancer (50 Stage II, 180 Stage III) with planned neoadjuvant chemotherapy before surgery. Field Strength/Sequence A 3 T, multi‐shell SE ‐ EPI DWI . Assessment Diffusion MRI was performed before chemotherapy (baseline). Neuropsychological assessments were carried out at baseline, after the first cycle of chemotherapy, after completion of chemotherapy but before surgery, and at 2–4 years after surgery. Diffusion tensor metrics (mean, radial, and axial diffusivity [MD, RD, and AD]), diffusion kurtosis metrics (axial and radial kurtosis [AK and RK]), and neurite orientation dispersion and density imaging metrics (orientation dispersion index [ODI], neurite density index [NDI]) were extracted from 25 white matter tracts, capturing both mean values and peak width. Statistical Tests Cox regression and linear mixed‐effects models assessed associations with time to CRCI onset and cognitive trajectories, respectively. A p value < 0.05 was considered significant. Results Higher ODI (anterior limb of the internal capsule [ALIC]: HR = 1.28), lower AD (ALIC: HR = 0.72, external capsule [EC]: HR = 0.73, superior fronto‐occipital fasciculus [SFO]: HR = 0.70), and greater peak width of AK (ALIC: HR = 1.29, EC: HR = 1.24, genu of the corpus callosum [GCC]: HR = 1.32) were associated with earlier CRCI onset. Lower FA, RK, and MK ( β = 0.13 to 0.27 × 10 −3 ), higher MD, RD, ODI, and ISOVF ( β = −0.23 to −0.13 × 10 −3 ), and higher peak width metrics ( β = −0.34 to −0.13 × 10 −3 ) were associated with poorer longitudinal cognitive trajectories. Conclusions Pretreatment white matter microstructure may help identify patients at greater risk of CRCI and poorer cognitive trajectories. Level of Evidence 2. Technical Efficacy Stage 3.