Purpose: Anthracyclines have been one of the standard therapies for breast cancer (BC), and dose-related cardiotoxicity is one serious side effect. Exercise is an effective strategy for the prevention and management of BC, endorsed by experts in both exercise and oncology. However, there is a great deal of confusion about the effectiveness of exercise on anthracycline-induced cardiotoxicity and the exercise prescription (i.e., timing, type, and intensity) for cardiotoxicity, which limits its application in clinical settings. The aim of this article is to review the safety of exercise in BC patients receiving anthracyclines and its effectiveness in preventing cardiotoxicity. Methods: Six electronic databases were searched using terms related to exercise, BC, anthracyclines, and cardiotoxicity for retrieving clinical randomized controlled trials in either Chinese or English. A summary of the included literature was also provided. Results: Of 202 records screened, 10 were eligible. A total of 434 BC patients (stage I–IIIC, mean age ranged from 43.5 to 52.4 years) were included. The main findings were that: (1) Acute (a single bout) moderate-to-vigorous aerobic exercise could prevent NT-proBNP elevation beyond the threshold of acute myocardial injury; (2) Long-term (> 8 weeks) moderate-to-high intensity aerobic exercise (continuous or interval) could improve or maintain left ventricular ejection fraction and cardiorespiratory fitness in BC patients. However, the optimal timing, type, and intensity of exercise for people with BC to prevent cardiotoxicity remain unclear. Conclusion: Moderate-to-vigorous intensity exercise may be an effective non-pharmacological approach to mitigate cardiotoxicities induced by anthracyclines in women with BC. However, the optimal exercise prescription for preventing cardiotoxicity remains unclear.

Background/Objectives: Anthracycline-based chemotherapy is associated with systemic toxicities in patients with breast cancer. Aerobic exercise may help mitigate treatment-related effects, but its associated metabolic responses during chemotherapy remain poorly understood. This study assessed biochemical and circulating metabolomic responses to aerobic exercise during anthracycline-based chemotherapy. Methods: This exploratory secondary analysis included 40 participants from a randomized controlled trial (trained n = 21; control, n = 19). Participants in the trained group completed a 12-week aerobic exercise intervention during chemotherapy. Biochemical markers and untargeted plasma LC–MS metabolomic profiles were assessed before and after intervention. Exploratory metabolic candidates were evaluated using linear mixed-effects models with Benjamini–Hochberg correction. Results: Significant Group × Time interactions were observed for 10 of 12 physiological and biochemical markers after BH-FDR correction in the primary analysis. However, baseline-adjusted ANCOVA sensitivity analyses showed that only IL-6 retained statistical significance after BH-FDR correction (adjusted between-group difference: −15.58 pg/mL, 95% CI: −20.72 to −10.45; FDR < 0.001). Despite limited global metabolomic separation, sequential screening identified 26 candidate features, seven of which showed significant Group × Time interactions after correction across candidates. Five had putative annotations, including lysophosphatidylcholine (LPC) 14:1, lysophosphatidylethanolamine (LPE) 17:1, cis-7-hexadecenoic acid, L-ascorbate, and vitamin E acetate. In a full-feature sensitivity analysis, only Com_3317_pos remained significant after BH-FDR correction across all 1150 metabolic features (FDR = 0.0228). Conclusions: Aerobic exercise during anthracycline-based chemotherapy was associated with changes in circulating biochemical markers, although these findings were less robust after baseline adjustment. Untargeted metabolomics identified exploratory exercise-associated metabolic features, particularly those putatively related to lipid metabolism, but most did not remain significant after full-feature multiple-testing correction. Further validation is warranted.