External validation of a published nomogram for prediction of brain metastasis in patients with extra-cerebral metastatic breast cancer and risk regression analysis
External validation of a published nomogram for prediction of brain metastasis in patients with extra-cerebral metastatic breast cancer and risk regression analysis
复制标题
DOI：
10.1016/j.ejca.2016.10.019
复制
发表时间：
2017-02-01
影响因子：
8.4
通讯作者：
Dalenc, Florence
中科院分区：
文献类型：
--
作者：
Genre, Ludivine;Roche, Henri;Dalenc, Florence
背景：转移性乳腺癌(MBC)合并脑转移(BM)患者的存活率有限，通常是致命的。2010年，Graesslin的诺模图被发表，以预测乳腺癌(BC)合并脑外转移疾病患者随后的BM。此模型旨在选择BM的高危人群，从而有助于在前瞻性临床研究中设计预防策略和/或早期治疗BM的影响。患者和方法：对2005年1月至2012年12月在我所治疗的BC及后来的BM患者进行了诺模图外部验证。结果：在492例MBC患者中，116例发生了随后的BM。其中70人被纳入诺模图验证。辨别能力较好(曲线下面积=0.695.95%可信区间0.61~0.77)。BM出现的危险因素有：人表皮生长因子受体2(HER2)过表达/扩增，BC三阴性和脑外转移部位数目(>；1)。使用竞争风险模型，我们排他性地强调了HER2+肿瘤亚组的诺模图兴趣。结论：Graesslin的诺模图外部验证显示了可输出性和重复性。重要的是，竞争风险模型分析为有关脑外转移BC的高危患者的BM早期诊断和/或预防性治疗的前瞻性试验设计提供了额外的信息。(C)2016爱思唯尔有限公司。保留所有权利。
Background: Survival of patients with metastatic breast cancer (MBC) suffering from brain metastasis (BM) is limited and this event is usually fatal. In 2010, the Graesslin's nomogram was published in order to predict subsequent BM in patients with breast cancer (BC) with extra-cerebral metastatic disease. This model aims to select a patient population at high risk for BM and thus will facilitate the design of prevention strategies and/or the impact of early treatment of BM in prospective clinical studies.Patients and methods: Nomogram external validation was retrospectively applied to patients with BC and later BM between January 2005 and December 2012, treated in our institution. Moreover, risk factors of BM appearance were studied by Fine and Gray's competing risk analysis.Results: Among 492 patients with MBC, 116 developed subsequent BM. Seventy of them were included for the nomogram validation. The discrimination is good (area under curve = 0.695 [95% confidence interval, 0.61-0.77]). Risk factors of BM appearance are: human epidermal growth factor receptor 2 (HER2) overexpression/amplification, triple-negative BC and number of extra-cerebral metastatic sites (>1). With a competing risk model, we highlight the nomogram interest for HER2+ tumour subgroup exclusively.Conclusion: Graesslin's nomogram external validation demonstrates exportability and reproducibility. Importantly, the competing risk model analysis provides additional information for the design of prospective trials concerning the early diagnosis of BM and/or preventive treatment on high risk patients with extra-cerebral metastatic BC. (C) 2016 Elsevier Ltd. All rights reserved.