Xie Haoling, Zhang Rong, Wei Chunmei, Xu Jinsong, Chu Jie, Wang Xuexing
Department of Oncology, Anning First People's Hospital Affiliated to Kunming University of Science and Technology, Anning, China.
Cadre Medical Department, Third Affiliated Hospital of Kunming Medical University, Kunming, China.
Transl Cancer Res. 2025 May 30;14(5):2885-2899. doi: 10.21037/tcr-24-1513. Epub 2025 May 27.
Triple-negative breast cancer (TNBC) has a poor prognosis due to limited targeted treatments. Chemotherapy often causes chemotherapy-induced myelosuppression (CIM), complicating treatment and raising costs, yet predictive tools for this risk are scarce. This study examined the prevalence and risk factors of CIM in TNBC patients after chemotherapy and created nomograms to predict this risk.
Nomograms were developed from a retrospective study of 316 TNBC patients treated at the Anning First People's Hospital Affiliated to Kunming University of Science and Technology between 1 July 2021 and 31 May 2024. The patients were split into development and validation cohorts in an 8:2 ratio. Least absolute shrinkage and selection operator (LASSO) identified risk factors for CIM, which were used to create the nomograms. The models' accuracy, calibration, and clinical utility were evaluated using the area under the curve (AUC), calibration curves, and decision curve analysis (DCA), with validation through bootstrapping.
In this study of 316 TNBC patients, 102 experienced CIM, an incidence rate of 32.28%. Patient characteristics were similar across cohorts. The development cohort had a mean age of 52.05 years, with a median hospital stay of 5 days. Myelosuppression of degree I was the most common CIM event. LASSO and logistic regression analyses linked CIM to factors like bone metastasis, platinum regimens, chemotherapy cycles, pre-chemotherapy neutrophil count, and drug combinations. The nomograms showed strong predictive accuracy with AUCs of 0.886 [95% confidence interval (CI): 0.836-0.937] and 0.905 (95% CI: 0.834-0.976) in the development and validation cohorts, respectively, and high agreement in calibration curves. DCA confirmed their clinical utility.
This study developed a validated nomogram that accurately predicts the risk of CIM in TNBC patients, helping healthcare providers create personalized treatment plans.
三阴性乳腺癌（TNBC）因靶向治疗有限，预后较差。化疗常导致化疗引起的骨髓抑制（CIM），使治疗复杂化并增加成本，但针对这种风险的预测工具却很稀缺。本研究调查了TNBC患者化疗后CIM的发生率和危险因素，并创建了列线图来预测这种风险。
通过对2021年7月1日至2024年5月31日在昆明理工大学附属安宁市第一人民医院接受治疗的316例TNBC患者进行回顾性研究来制定列线图。患者按8:2的比例分为开发队列和验证队列。最小绝对收缩和选择算子（LASSO）确定了CIM的危险因素，并用于创建列线图。使用曲线下面积（AUC）、校准曲线和决策曲线分析（DCA）评估模型的准确性、校准和临床实用性，并通过自举法进行验证。
在这项对316例TNBC患者的研究中，102例发生了CIM，发生率为32.28%。各队列的患者特征相似。开发队列的平均年龄为52.05岁，中位住院时间为5天。I度骨髓抑制是最常见的CIM事件。LASSO和逻辑回归分析将CIM与骨转移、铂类方案、化疗周期、化疗前中性粒细胞计数和药物组合等因素联系起来。列线图在开发队列和验证队列中的预测准确性很强，AUC分别为0.886 [95%置信区间（CI）：0.836 - 0.937]和0.905（95% CI：0.834 - 0.976），校准曲线的一致性很高。DCA证实了它们的临床实用性。
本研究开发了一种经过验证的列线图，可准确预测TNBC患者发生CIM的风险，有助于医疗保健提供者制定个性化治疗方案。
