Peri-treatment Creatinine-Based eGFR Trajectories in Patients Receiving Breast Radiotherapy: A Retrospective Cohort Study
Yunan Qing, Jing Guo, Li Fan, Bin Ouyang, Zhenzhao Pan, Zhaojun Wu, Ruiwan Chen, Xiao Yang, Mengzhen MinAbstract
Background
Longitudinal estimated glomerular filtration rate (eGFR) changes during breast cancer treatment remain incompletely characterized. We examined peri-treatment eGFR trajectories, factors associated with eGFR decline, whether 1-month eGFR change was associated with later decline, and differences by radiotherapy fractionation.
Materials and methods
We retrospectively included 741 patients receiving breast radiotherapy from 2023 to 2025. eGFR was assessed at baseline and 1, 3, 6, 9, and 12 months. The primary endpoint was a ≥ 20% decrease from baseline confirmed at 2 consecutive scheduled visits; ≥30% decrease was the sensitivity endpoint. Trajectories were analyzed with a five-segment linear mixed-effects model. Cox regression evaluated associated factors, a temporally separated 1-month landmark analysis assessed later decline, and fractionation comparisons used multivariable adjustment and overlap weighting.
Results
58 patients (7.8%) met the confirmed ≥20% endpoint and 24 (3.2%) met the ≥30% endpoint. Trajectories diverged most clearly within 3 months. N1–3 disease (HR, 2.60; 95% CI, 1.11–6.12) and recorded CDK4/6 inhibitor use (HR, 8.73; 95% CI, 4.68–16.27) were associated with the endpoint. One-month eGFR decline was not associated with subsequent confirmed ≥20% decline (HR, 1.04 per 10% decline; 95% CI, 0.77–1.41; P = 0.786) but was associated with subsequent confirmed ≥30% decline (HR, 1.61; 95% CI, 1.06–2.44; P = 0.025). No significant renal-risk difference was observed by fractionation scheme.
Conclusion
Peri-treatment creatinine-based eGFR trajectories diverged early in patients who later met the confirmed-decline endpoint. Nodal stage and recorded CDK4/6 inhibitor use were associated with the endpoint; 1-month eGFR change was associated with the ≥30% sensitivity endpoint.