A Model for Individualized Risk Prediction of Contralateral Breast Cancer

描述(申请人提供)：被诊断为浸润性乳腺癌或导管原位癌的患者越来越多地选择接受对侧预防性乳房切除术(CPM)，以降低他们患对侧乳腺癌(CBC)的风险。这是一个特别令人不安的趋势，因为大量这样的中成药被认为在医学上是不必要的。这是因为在过去的二十年里，由于可获得有效的辅助治疗，对大多数患者来说，CBC的风险已经显著下降。尽管如此，被诊断为第一原发乳腺癌的患者往往会大大高估他们的CBC风险。同时，他们低估了与CPM相关的并发症、风险和负面影响。这些错误的认识部分解释了美国CPM发生率上升的原因。此外，几乎没有证据表明CPM有助于延长生存期。因此，需要适当地权衡CPM的好处和缺点。鉴于CPM的侵袭性和不可逆转性，我们有责任为正在经历情感挑战时期的乳腺癌患者提供健全和有效的教育。医生确实试图教育他们的病人；然而，他们缺乏工具来帮助他们进行这一努力。特别是，他们需要一种CBC风险预测模型，可以为散发性(非遗传)乳腺癌患者提供个性化的风险估计。这个项目旨在通过开发这样的模型，验证它，并 在免费提供的软件包中实施它，以便立即临床使用。为了建立模型，我们将使用监测、流行病学和最终结果(SEER)计划的数据和文献中风险估计的荟萃分析。拟议中的模型将采用盖尔模型的风格--这是一种流行的工具，用于就患乳腺癌的风险向女性提供咨询--但它专门为患有单侧乳腺癌的女性提供关于患CBC风险的咨询。在建立模型后，我们将在从四个机构-德克萨斯大学西南医学中心、达拉斯帕克兰纪念医院、M D安德森癌症中心和达特茅斯医学院-收集的乳腺癌患者的前瞻性数据上对其进行验证。一旦模型被验证，我们将在统计软件R中创建一个用户友好的包来实现该模型。然后我们将把这个包集成到CancerGene中，这是一个广泛使用的免费临床软件，用于就乳腺癌的风险向患者提供咨询。 Cancergene被授权给全球4000多个地点，因此将是向大量从业者提供新模型的完美途径。我们相信，我们建议的模式将极大地促进患者的教育，并将有助于遏制医学上不必要的CPM日益增长的趋势。

DESCRIPTION (provided by applicant): Patients diagnosed with invasive breast cancer or ductal carcinoma in situ are increasingly choosing to undergo contralateral prophylactic mastectomy (CPM) to reduce their risk of contralateral breast cancer (CBC). This is a particularly disturbing trend as a large number of these CPMs are believed to be medically unnecessary. This is because the risk of CBC has dropped markedly for most patients in the last two decades due to availability of effective adjuvant therapies. Despite this fact, patients diagnosed with first primary breast cancer tend to substantially overestimate their CBC risk. At the same time, they underestimate the complications, risks, and negative effects associated with CPM. These incorrect perceptions partly explain the rising CPM rates in the U.S. Moreover, there is little evidence that CPM helps in prolonging survival. Thus, the benefits of CPM need to be weighed properly with its drawbacks. Given the invasive and irreversible nature of CPM, it behooves us to provide sound and effective education to breast cancer patients, who are going through an emotionally challenging period. Physicians do try to educate their patients; however, they lack tools that can help them in this endeavor. In particular, they need a CBC risk prediction model that can provide individualized risk estimates for sporadic (non-genetic) breast cancer patients. This project aims to fill this need by developing such a model, validating it, and implementing it in a freely available software package for immediate clinical use. To build the model, we will use data from Surveillance, Epidemiology, and End Results (SEER) Program and meta-analysis of risk estimates from literature. The proposed model will be in the style of the Gail model - a popular tool for counseling women on the risk of developing breast cancer - but one that is exclusively designed for counseling women with unilateral breast cancer on the risk of developing CBC. After building the model, we will validate it on prospectively collected data on breast cancer patients from four institutions - University of Texas at Southwestern Medical Center, Parkland Memorial Hospital in Dallas, M D Anderson Cancer Center, and Dartmouth Medical School. Once the model is validated, we will create a user-friendly package in statistical software R for implementing the model. Then we will integrate the package into CancerGene, a widely used and freely available clinical software for counseling patients on the risk of breast cancer. CancerGene is licensed to more than 4000 sites worldwide and therefore will be a perfect gateway to make the new model available to a large number of practitioners. We believe our proposed model will greatly facilitate patients' education and will help stem the increasing trend of medically unnecessary CPMs.