Department of Radiation Oncology, Massachusetts General Hospital, Harvard Medical School, Boston.
Dana-Farber Cancer Institute, Boston, Massachusetts.
JAMA Netw Open. 2021 Nov 1;4(11):e2133205. doi: 10.1001/jamanetworkopen.2021.33205.
Precision oncology is revolutionizing cancer care, allowing for personalized treatments to improve outcomes. Cancer research has benefitted from well-designed studies incorporating precision medicine objectives, but it is unclear if these studies are representative of the diverse cancer population.
To evaluate racial and ethnic representation in breast, prostate, lung, and colorectal cancer studies incorporating precision oncology objectives in the Clinicaltrials.gov registry and compare with the incidence of these cancer types in racial and ethnic minority groups in the US population.
DESIGN, SETTING, AND PARTICIPANTS: This cross-sectional study identified US-based breast, prostate, lung, and colorectal cancer studies incorporating precision oncology objectives for reporting of race and ethnicity. The Surveillance, Epidemiology, and End Results and US Census databases were used to determine cancer incidence by race and ethnicity, linked with cancer type and median year of enrollment for each trial. Data were collected and analyzed between December 2020 and April 2021.
The expected number of participants per study by each racial and ethnic group was calculated based on the corresponding US-based proportion. Under- and overrepresentation was defined as the ratio of the actual number of enrolled cases to the expected number of cases for each trial by cancer type. Ratios above 1 indicated overrepresentation while a ratio below 1 indicated underrepresentation. Random-effects meta-analysis of representation ratios of individual trials was performed to weigh each individual study.
Of 93 studies encompassing 5867 enrollees with race and ethnicity data; 4826 participants (82.3%) were non-Hispanic White, 587 (10.0%) were Black, and 238 (4.1%) were Asian. Per observed-to-expected ratios, White participants were overrepresented in all studies, with a ratio of 1.35 (95% CI, 1.30-1.37), as well as Asian participants, with a ratio of 1.46 (95% CI, 1.28-1.66), while Black participants (ratio, 0.49; 95% CI, 0.45-0.54), Hispanic participants (ratio, 0.24; 95% CI, 0.20-0.28), and American Indian and Alaskan Native participants (ratio, 0.43; 95% CI, 0.24-0.78) were underrepresented. By individual cancer site, White participants were consistently overrepresented in all studies, while Black and Hispanic participants were underrepresented.
This analysis found that precision oncology studies for breast, lung, prostate, and colorectal cancers vastly underrepresent racial and ethnic minority populations relative to their cancer incidence in the US population. It is imperative to increase diversity among enrollees so that all individuals may benefit from cancer research breakthroughs and personalized treatments.
精准肿瘤学正在彻底改变癌症治疗方式,使个性化治疗得以改善治疗效果。癌症研究从纳入精准医学目标的精心设计的研究中受益,但这些研究是否能代表美国癌症患者的多样化情况尚不清楚。
评估纳入精准肿瘤学目标的乳腺癌、前列腺癌、肺癌和结直肠癌研究在美国临床试验注册处的种族和民族代表性,并与美国人群中这些癌症类型在种族和少数民族中的发病率进行比较。
设计、地点和参与者:本横断面研究确定了纳入报告种族和民族信息的美国乳腺癌、前列腺癌、肺癌和结直肠癌精准肿瘤学目标的研究。利用监测、流行病学和最终结果数据库和美国人口普查数据库,根据种族和民族确定癌症发病率,并将其与每个试验的癌症类型和中位入组年份相关联。数据于 2020 年 12 月至 2021 年 4 月间收集和分析。
根据美国相应比例,计算了每个种族和民族组中每一项研究预期的参与者人数。未充分代表和过度代表的定义是,每个试验中实际入组病例数与相应癌症类型的预期病例数的比值。比值大于 1 表示过度代表,比值小于 1 表示未充分代表。对个别试验的代表性比率进行随机效应荟萃分析,以权衡每个单独研究的权重。
在纳入种族和民族数据的 93 项研究中,共纳入了 5867 名参与者,其中 4826 名参与者(82.3%)是非西班牙裔白人,587 名参与者(10.0%)为黑人,238 名参与者(4.1%)为亚洲人。根据观察到的与预期的比值,所有研究中白人参与者都存在过度代表的情况,其比值为 1.35(95%CI,1.30-1.37),而亚洲人参与者的比值为 1.46(95%CI,1.28-1.66),黑人参与者(比值为 0.49;95%CI,0.45-0.54)、西班牙裔参与者(比值为 0.24;95%CI,0.20-0.28)和美洲印第安人和阿拉斯加原住民参与者(比值为 0.43;95%CI,0.24-0.78)则存在未充分代表的情况。按个别癌症部位来看,白人参与者在所有研究中均持续存在过度代表的情况,而黑人参与者和西班牙裔参与者则存在未充分代表的情况。
本分析发现,乳腺癌、肺癌、前列腺癌和结直肠癌的精准肿瘤学研究在很大程度上未能充分代表美国癌症患者中的少数族裔人群,与这些人群在美国癌症发病率不成比例。必须增加参与者的多样性,以使所有个人都能从癌症研究突破和个性化治疗中受益。