Development of a pragmatic LDH–magnesium–HER2 risk score for severe chemotherapy toxicity in breast cancer: a prospective study in a resource-limited oncology setting
Discuss this preprint
Start a discussion What are Sciety discussions?Listed in
This article is not in any list yet, why not save it to one of your lists.Abstract
Background: Chemotherapy toxicity frequently necessitates dose reductions, treatment delays, or early discontinuation, which undermine therapeutic efficacy and worsen patient outcomes. In resource-limited oncology settings, where high patient volumes and constrained supportive care make comprehensive geriatric assessments impractical, there is an urgent need for simple, pragmatic tools to pre-emptively identify patients at high risk for chemotherapy toxicity. We aimed to develop and internally validate a biomarker-based risk score using only routine clinical data to predict severe toxicity in breast cancer patients. Methods: We conducted a longitudinal clinical investigation combining three datasets: (1) a prospective cohort of 88 breast cancer patients with serial biochemical measurements at baseline (T1), mid-therapy (T2, after 3 cycles), end-of-therapy (T3), and 30 days post-treatment (T4); (2) a 226-patient cohort for risk score development and internal validation; and (3) an institutional survival registry (n=2,635; 2020–2024) for contextual survival analysis. Baseline lactate dehydrogenase (LDH), serum magnesium, and HER2 status were evaluated as independent predictors of CTCAE v5.0 grade ≥3 toxicity using logistic regression. The score was internally validated using bootstrap resampling (1,000 iterations). Results: Among 226 patients, 200 were included and 44 (22%) developed grade ≥3 toxicity. The LDH-magnesium-HER2 score demonstrated strong discrimination (AUC 0.82, 95% CI 0.73-0.89) with 76% sensitivity and 75% specificity. Toxicity rates were 8.5% in low-risk (0-1 points), 20.0% in intermediate-risk (2 points), and 57.1% in high-risk (≥3 points) groups (OR 28.6, 95% CI 8.3-98.2 for high vs. low). Bootstrap validation confirmed minimal overfitting (optimism-adjusted AUC 0.81) . Conclusion: The LDH–magnesium–HER2 score is a simple, low-cost, biomarker-based tool requiring no additional testing beyond routine oncology care that effectively stratifies breast cancer patients at risk for severe chemotherapy toxicity. The score enables risk-adapted chemotherapy delivery, intensified supportive care, and personalized dosing strategies, particularly valuable in resource-constrained oncology centers. External multicenter validation in diverse populations is warranted before clinical implementation.