Statistical Methods for Cancer Clinical Trials

描述(由申请人提供)：这一雄心勃勃的计划项目的首要目标仍然是为癌症临床试验开发创新的、变革性的统计方法，这些方法有可能加速成功地将新疗法和治疗策略引入实践。生物学、基因组学、统计学和计算科学的进步为个性化癌症治疗的发展带来了巨大的希望。我们的多机构、跨学科的研究团队将利用这些进步来创建新的临床试验设计和数据分析方法，以解决当前统计方法的许多关键限制，并最大限度地提高个性化癌症药物临床试验的有效性。此外，我们将促进将这些方法转化为实践，包括基于新方法开展动物和人体试验研究。该计划将通过五个相互关联的研究项目实现这些目标，这些项目由来自三个机构的研究人员进行，他们在统计、计算和临床科学方面具有互补的专业知识。前四个项目集中于开发新的试验设计和分析方法，将生物标记物纳入有效发现新的个性化治疗；创建分析生物标记物和患者报告结果的现有数据的方法，以便为未来研究的设计提供信息并改进设计；开发最大限度地利用药物基因组学确定生物标记物和候选个体化治疗方法的方法；以及创建发现和验证癌症治疗的连续、个性化决策策略的新方法。第五个项目将把这些方法整合到胰腺癌的新的临床前和临床研究中。我们的全面方法涉及一个积极和协调的实施、沟通和传播成果的过程，包括开发专业的公共使用软件和相关的教程；讲习班和其他外联机制；以及由方案赞助的研讨会和活动，以加速在实践中采用这些方法。拟议的临床试验设计和分析创新有可能改变癌症临床试验的方式，以发现和验证个性化医学。这一全面的、多机构的努力将导致癌症临床试验实践的重大创新，从而改善癌症患者的健康。

DESCRIPTION (provided by applicant): The overarching goal of this ambitious program project continues to be to develop innovative, transformative statistical methods for cancer clinical trials that have the potential to hasten successful introduction of new therapies and treatment strategies into practice. Advances in the biologic, genomic, statistical, and computational sciences hold great promise for the development of personalized cancer treatments. Our multi-institutional, interdisciplinary team of investigators will leverage these advances to create new clinical trial designs and data analysis approaches that resolve many of the key limitations of current statistical methods and that maximize the effectiveness of clinical trials for personalized cancer medicine. In addition, we will foster translation of these methods into practice, including carrying out pilot animal and human studies based on the new methodology. The program will achieve these objectives through five, interrelated research projects carried out by investigators with complementary expertise in the statistical, computational, and clinical sciences from three institutions. The first four projects focus on developing new trial designs and analysis methods that integrate biomarkers for efficient discovery of new, personalized treatments; on creating methods for analysis of existing data on biomarkers and patient reported outcomes to inform and improve the design of future studies; developing methods for maximizing the power of pharmacogenomics for identifying biomarkers and candidate individualized therapies; and creating new methods for discovering and validating sequential, personalized decision-making strategies for cancer treatment. The fifth project will integrate the methods into novel preclinical and clinical studies of pancreatic cancer. Our comprehensive approach involves an energetic and coordinated process for implementation, communication, and dissemination of results, including development of professional, public-use software and associated tutorials; workshops and other outreach mechanisms; and program-sponsored symposia and events, to accelerate the adoption of the methods in practice. The proposed clinical trial design and analysis innovations have the potential to effect a paradigm shift in the way cancer clinical trials are conducted for discovery and validation of personalized medicine. This comprehensive, multi-institutional effort will lead to significant innovations in cancer clinical trial practice that will result in improved health of cancer patients.