New CT-Detected Pulmonary Abnormalities During Breast Cancer Chemotherapy: Frequency, Imaging Features, and...

Background: This study aimed to quantify the frequency and imaging features of new CT-detected pulmonary abnormalities during breast cancer chemotherapy and examine their associations with cumulative cyclophosphamide and paclitaxel exposure. Methods: This single-center retrospective study included 223 patients who underwent an on-treatment chest CT under an early-pandemic infection-control policy. New CT-detected pulmonary abnormalities were defined as new findings on this on-treatment CT that were absent on the pretreatment staging CT. Cumulative cyclophosphamide (C) and paclitaxel (T) doses (mg/m²) were calculated from treatment cycles before the on-treatment CT and dichotomized at 1800 and 960 mg/m², respectively. Associations were assessed using Firth-penalized logistic regression. Results: New CT-detected pulmonary abnormalities were identified in 34/223 patients (15.2%). Among 25 CT-positive cases with sufficiently detailed CT reports, patchy opacities (14/25, 56.0%) and ground-glass opacities (11/25, 44.0%) were the most frequent findings. Compared with patients with low cyclophosphamide and paclitaxel-equivalent exposure, combined exposure categories showed higher odds of the CT-detected outcome. In the Firth model, cyclophosphamide > 1800 mg/m² was associated with the CT-detected outcome (adjusted odds ratio (OR) 2.38, 95% CI 1.06-5.71; p = 0.036), whereas paclitaxel-equivalent exposure > 960 mg/m² (OR 2.46, 95% CI 0.86-6.80; p = 0.092) and age ≥ 60 years (OR 2.08, 95% CI 0.91-4.62; p = 0.083) were not statistically significant. Conclusions: New CT-detected pulmonary abnormalities were common during breast cancer chemotherapy (15.2%). Higher cumulative cyclophosphamide exposure was associated with the CT-detected outcome in multivariable analysis. Prospective longitudinal studies are needed to clarify the clinical significance of these abnormalities and their causality.