Dentate nucleus signal changes after multiple GBCAs MRI in breast cancer patients: A generalized additive model...

While research has explored gadolinium retention differences between linear and macrocyclic GBCAs, comparative studies of structurally distinct linear subtypes (ionic vs. nonionic) in high-dose, large cohorts remain limited.

To investigate factors influencing dentate nucleus (DN) signal increase on non-contrast T1WI after multiple GBCA administrations in breast cancer patients, focusing on contrast type (nonionic gadodiamide vs. ionic gadopentetate), interdose intervals, and adjuvant therapies.

Cohort study.

197 female breast cancer patients (excluded from 332) with ≥3 GBCA-enhanced scans and ≥ 2 non-enhanced cranial MRIs (2016-2023), mean age 51.5 years (26-78).

1.5 T and 3.0 T, turbo spin-echo inversion recovery sequence ASSESSMENT: Blinded DN/pons signal ratio measurements; relative change (R_change) reflected gadolinium retention. Associations between R_change and covariates (contrast type, intervals, therapies, etc.) were analyzed.

Generalized additive models (GAMs) and intraclass correlation coefficients (ICC).

107 patients received gadodiamide (mean 7.17 administrations), 60 gadopentetate (mean 7.15), 30 Mixed GBCA (mean 6.80). Excellent interobserver agreement (ICC: 0.860-0.862). GAM showed higher R_change in gadodiamide vs. gadopentetate (0.080 ± 0.041 vs 0.056 ± 0.034; 1.43-fold, P < 0.001). R_change inversely correlated with interdose intervals (P < 0.001) and hormonal therapy (P = 0.021), with no associations with chemotherapy, targeted therapy, calcium supplementation, or age (all P > 0.05).

Nonionic linear gadodiamide has higher intracranial gadolinium retention propensity than ionic linear gadopentetate. Interdose intervals and hormonal therapy are significant modulators.