Optimizing Lung Cancer Screening Nodule Evaluation

总结 该项目的目标是优化屏幕检测肺结节的管理，从而最大限度地提高 肺癌筛查的好处肺癌是美国癌症死亡的最常见原因。到 为了控制这种疾病的负担，多个国家组织建议用低- 剂量计算机断层扫描（LDCT）。然而，多达三分之一的筛查LDCT识别肺结节 但其中只有1 - 3%是癌症然后对筛查出的肺结节进行额外的随访， 影像学检查，以及在某些情况下，侵入性和潜在有害的程序。后续行动和后续行动 检查程序占了与筛查相关的不必要的伤害和成本的很大一部分。一个 最佳的结节管理算法应该大大减少这些危害，并提供早期癌症 检测优势。然而，肺癌期间检测到的肺结节的最佳管理 筛查目前尚不清楚。LDCT筛查发现肺结节有不同的主要指南 管理大多数得到广泛执行的准则侧重于结核的特性，以决定是否需要 后续的类型。这些指南未能纳入其他关键的患者因素，如年龄，性别，吸烟 病史和合并症此外，其他因素可能严重影响诊断准确性， 结节管理策略的危害，以及最终肺癌筛查的好处。其中包括： 1)基于参与者和结节特征的肺癌风险; 2）癌症侵袭性; 3）类型， 结节随访的顺序和时间; 4）随访和活检相关并发症; 5） 死亡（非肺癌死亡率）; 6）评价对生活质量的影响。此外， 不同种族和民族之间的吸烟模式、肺癌风险和合并症不是 已纳入现行结核管理准则。在这个项目中，我们将使用仿真建模， 有效地确定考虑上述所有问题的最佳算法。我们将建立一个模拟 模型，多种族和民族肺癌模型（MELCAM），基于以前的建模框架 我们的团队使用它来广泛研究肺癌控制的各个方面。该项目的具体目标是： 1)推导并验证MELCAM，以模拟筛查的管理和后续结局 来自不同种族和民族背景的参与者; 2）使用MELCAM比较现有结节 管理协议的整体和质量调整的生命年收益和危害; 3）使用MELCAM， 生成考虑结节和患者因素对结节的影响的结节管理算法 癌症风险、筛查危害和预期寿命，以优化随访程序的类型和时间; 以及4）确定现有的和新的后续算法的成本效益。我们的研究是创新的， 应用最先进的建模技术和个性化方法来优化肺 结核管理最大限度地提高肺癌筛查在不同人群中的效益。

SUMMARY The goal of this project is to optimize the management of screen-detected pulmonary nodules thus maximizing the benefits of lung cancer screening. Lung cancer is the most common cause of cancer death in the US. To curb the burden of this disease, multiple national organizations recommend lung cancer screening with low- dose computed tomography (LDCT). However, up to one third of screening LDCTs identify pulmonary nodules but only 1-3% of these are cancers. Screen-detected pulmonary nodules are then followed-up with additional imaging tests and, in some cases, invasive and potentially harmful procedures. Follow-up and subsequent work-up procedures account for a large portion of screening-associated unnecessary harms and costs. An optimal nodule management algorithm should substantially reduce these harms and provide early cancer detection benefits. However, the optimal management of pulmonary nodules detected during lung cancer screening is currently unknown. There are differing major guidelines for LDCT screen-detected lung nodule management. Most widely implemented guidelines focus on nodule characteristics to decide the need for and type of follow-up. These guidelines fail to incorporate other key patient factors such as age, sex, smoking history, and comorbidities. Furthermore, additional factors can heavily impact the diagnostic accuracy and harms of nodule management strategies and ultimately, the benefits of lung cancer screening. These include: 1) risk of lung cancer based on participant and nodule characteristics; 2) cancer aggressiveness; 3) type, sequence and timing of nodule follow-up; 4) follow-up and biopsy related complications; 5) competing risks of death (non-lung cancer mortality); and 6) impact of evaluation on quality of life. Furthermore, differences in smoking patterns, lung cancer risk, and comorbidities among diverse race and ethnic groups are not incorporated in current nodule management guidelines. In this project, we will use simulation modeling to efficiently determine optimal algorithms that consider all the issues listed above. We will build a simulation model, the Multi-Racial and Ethnic Lung Cancer Model (MELCAM), based on a previous modeling framework used by our team to extensively study various aspects of lung cancer control. The project Specific Aims are to: 1) Derive and validate MELCAM to simulate the management and subsequent outcomes of screening participants from diverse racial and ethnic backgrounds; 2) Use MELCAM to compare existing nodule management protocols in terms of overall and quality-adjusted life-year gains and harms; 3) Use MELCAM to generate nodule management algorithm(s) that consider the impact of both nodule and patient factors on cancer risk, screening harms, and life expectancy to optimize the types and timing of follow-up procedures; and 4) Determine the cost-effectiveness of existing and novel follow-up algorithms. Our study is innovative in applying state-of-the-art modeling techniques and personalized approaches to the optimization of pulmonary nodule management maximizing the benefits of lung cancer screening in diverse populations.