Department of Urology, Dow Division of Health Services Research, University of Michigan, USA; Department of Urology, Division of Urologic Oncology, University of Michigan, USA.
Department of Urology, Division of Urologic Oncology, University of Michigan, USA.
Contemp Clin Trials. 2021 Dec;111:106600. doi: 10.1016/j.cct.2021.106600. Epub 2021 Oct 19.
BACKGROUND/AIMS: One in five cancer clinical trials fails with another third failing to meet enrollment goals. Prior efforts to improve enrollment focus on patient facing interventions, but geographic factors such as regional cancer incidence may doom trials before they even begin. For these reasons, we examined associations of regional prostate cancer incidence with trial termination, and identified scientifically-underserved areas where future trials might thrive.
We merged US phase 2-3 prostate cancer clinical trial data from ClinicalTrials.gov with prostate cancer incidence data from statecancerprofiles.cancer.gov. We matched trial information from 293 closed and 560 active trials with incidence data for 2947 counties. Using multivariable logistic regression, we identified associations with trial termination. We identified 'scientifically-underserved' counties with the highest cancer incidence quintile (>61 annual cases) but lowest active trials quintile (0 or 1 trial).
Of 293 closed trials, one in three was terminated (n = 96, 32.8%). On multivariable analysis, only lower regional prostate cancer incidence was associated with higher likelihood of premature trial termination (OR 0.98, 95% CI [0.96-0.99] for every 100 cases, p = 0.03). We identified 188 counties with >61 annual prostate cancer cases but 0 or 1 active trials, indicating potential scientifically-underserved areas.
In this novel study, we found prostate cancer trials in areas with low prostate cancer incidence were more likely to fail. We also identified scientifically-underserved areas where trials might thrive. Our findings provide a more nuanced understanding of clinical trial feasibility and upstream opportunities for improvement.
背景/目的：五分之一的癌症临床试验以失败告终，另有三分之一未能达到入组目标。先前为提高入组率而进行的努力主要集中在面向患者的干预措施上，但地理因素（如区域性癌症发病率）可能会使试验在开始之前就失败。出于这些原因，我们研究了区域前列腺癌发病率与试验终止之间的关系，并确定了未来可能蓬勃发展的科学上服务不足的领域。
我们将来自 ClinicalTrials.gov 的美国 2 期-3 期前列腺癌临床试验数据与来自 statecancerprofiles.cancer.gov 的前列腺癌发病率数据合并。我们将 293 项已关闭和 560 项正在进行的试验的试验信息与 2947 个县的发病率数据进行了匹配。使用多变量逻辑回归，我们确定了与试验终止相关的因素。我们确定了“科学上服务不足”的县，这些县的癌症发病率最高（>61 例/年），但活跃试验最低（0 或 1 项试验）。
在 293 项已关闭的试验中，三分之一（n=96，32.8%）提前终止。多变量分析显示，只有较低的区域前列腺癌发病率与更高的试验提前终止可能性相关（每 100 例发病率降低 0.98，95%CI[0.96-0.99]，p=0.03）。我们确定了 188 个县，这些县的前列腺癌发病率>61 例/年，但没有或仅有 1 项活跃的试验，表明这些县可能存在科学上服务不足的情况。
在这项新颖的研究中，我们发现前列腺癌发病率较低的地区的试验更有可能失败。我们还确定了科学上服务不足的领域，这些领域可能会有试验蓬勃发展。我们的研究结果提供了对临床试验可行性的更细致的理解，以及改进的上游机会。
