Baseline systemic inflammatory indices predict late radiation-induced fibrosis and toxicity in patients with early breast cancer.

Radiation therapy (RT) is a standard component of treatment for early breast cancer and ductal carcinoma in situ. Although survival is excellent, RT-related side effects can impair long-term quality of life. We aimed to develop models to predict the risk of Grade ≥2 fibrosis and other late toxicities by combining clinical risk factors with baseline systemic inflammatory indices. In this prospective study, 324 women underwent breast-conserving surgery followed by whole-breast RT and were followed for 24 months. Pre-surgical blood counts were used to calculate inflammatory indices, including NLR, dNLR, PLR, MLR, NLPR, SII, SIRI and AISI. Optimal cut-offs for each index was determined using ROC curves. Predictive models integrated clinical factors (surgical complications, post-menopausal status, chemotherapy, smoking, RT dose) with individual inflammatory indices to predict Grade ≥2 fibrosis and overall late side effects (fibrosis, skin side effects, chronic pain, lymphoedema). At 24 months, 13.0 % of patients had developed Grade ≥2 fibrosis and 17.9 % had at least one Grade ≥2 late side effect. Patients who developed fibrosis had higher baseline NLR, dNLR, MLR, NLPR and SIRI. Models incorporating MLR or NLPR achieved area under the curve values > 0.70 and negative predictive values ≥ 93 %, outperforming models based on clinical risk factors alone. These findings suggest that routinely measured systemic inflammatory indices can identify patients at low risk of late RT side effects. If validated in external cohorts, these models may help tailor RT regimens to minimize long-term complications while maintaining therapeutic benefit.