Risk-based lung cancer screening strategies may improve screening efficiency and reduce differences across racial and ethnic groups compared with current US Preventive Services Task Force (USPSTF) eligibility criteria, according to a large cohort study published in the Annals of Internal Medicine.
The study evaluated 16 lung cancer risk prediction models among more than 641,000 US adults aged 50 to 80 years with a smoking history from the Lung Cancer Cohort Consortium. Participants included Asian, Hispanic, non-Hispanic Black, and non-Hispanic White individuals.
According to the study, the performance of existing prediction models varied substantially across racial and ethnic groups.
“General patterns across the 16 models included substantial underestimation of lung cancer risk in non-Hispanic Black participants, lower discrimination in Asian participants than all other groups, and lower discrimination in non-Hispanic Black than non-Hispanic White participants,” wrote lead study author Xiaoshuang Feng, PhD, of the International Agency for Research on Cancer, and colleagues.
To compare screening approaches, the investigators applied risk-model thresholds that selected approximately the same proportion of participants for screening as the 2021 USPSTF criteria. According to the study, every risk-based strategy produced better average estimated screening efficiency and smaller differences in efficiency across the four racial and ethnic groups overall than the USPSTF criteria. However, some models did not improve estimated screening efficiency among non-Hispanic Black participants compared with USPSTF criteria.
Among the models evaluated, the Prostate, Lung, Colorectal, and Ovarian Cancer Screening Trial Model 2012 and the Life Years Gained From Screening–Computed Tomography model demonstrated the strongest overall performance based on estimated screening efficiency, according to the study.
However, the authors emphasized that no single strategy addressed every performance measure equally well, noting that “no strategy could simultaneously optimize eligibility, sensitivity, and efficiency while also reducing racial and ethnic differences.”
The researchers noted that models incorporating race and ethnicity as predictors generally showed smaller differences in calibration across racial and ethnic groups than models that did not include those variables. They also reported that a model developed in a Chinese population demonstrated stronger discrimination among Asian participants than models developed primarily in Western populations.
According to the authors, expanding screening eligibility alone may not fully address disparities. The study found that although all risk-based approaches improved average screening efficiency compared with USPSTF criteria, performance varied across racial and ethnic groups and across measures of calibration, discrimination, eligibility, and sensitivity.
“To optimize efficiency and minimize its variation across racial and ethnic groups, risk-based strategies were superior to USPSTF criteria. Further optimization of prediction models for the diverse US population is needed,” wrote Dr. Feng and colleagues.
The researchers acknowledged several limitations, including the underrepresentation of racial and ethnic minority groups compared with national data, particularly Asian and Hispanic participants, and the inability to evaluate model performance among Native Hawaiian or other Pacific Islander and American Indian or Alaska Native participants because of small numbers.
Disclosures: The study was funded by the US National Cancer Institute, the Lung Cancer Research Foundation, and Cancer Research UK. According to the authors, the funding organizations had no role in the study design, data analysis, data interpretation, writing of the report, or decision to submit the manuscript for publication.
Source: Annals of Internal Medicine
