Prospective Evaluation of Lung Cancer Screening Eligibility Criteria and Lung Cancer Detection in the Yorkshire Lung Screening Trial

Abstract

Introduction

Low-dose computed tomography screening for lung cancer reduces lung cancer mortality, but there is a lack of international consensus regarding the optimal eligibility criteria for screening. The Yorkshire Lung Screening Trial was designed to evaluate lung cancer screening (LCS) implementation, and a primary objective was prospective evaluation of three predefined eligibility criteria.

Methods

Individuals who had ever smoked, aged 55 to 80 years, who responded to written invitation, underwent telephone risk assessment and if eligible by at least one criterion (PLCO_M2012 ≥ 1.51%, LLP_v2 ≥ 5%, USPSTF₂₀₁₃) were offered biennial low-dose computed tomography screening.

Results

Of 44,957 individuals invited, 22,814 responded and underwent eligibility assessment, of whom a total of 7826 were eligible according to any of the three LCS criteria. Comparing PLCO_M2012 ≥ 1.51%, LLP_v2 ≥ 5%, and USPSTF₂₀₁₃, the proportions of responders eligible for screening were 28.0%, 20.5%, and 18.9%, respectively (p < 0.0001 for each comparison), and the proportion of all cancers detected 91.1%, 77.0%, and 62.8%, respectively (p ≤ 0.0002 for each comparison). When risk thresholds were selected to result in equivalent numbers of people eligible for screening, cancer detection proportions were higher for PLCO_M2012 (74.5%) and LLP_v2 (71.3%) than USPSTF₂₀₁₃ (62.8%) (p = 0.0002 and p = 0.032, respectively), but there was no significant difference between the two risk models. Reducing the LLPv2 risk threshold from 5% to 2.5% (as currently used in the English LCS program) and reducing the pack-year requirement for the USPSTF₂₀₂₁ versus the USPSTF₂₀₁₃ criteria increased the numbers eligible for screening, but subsequent cancer yield was not measured in this study.

Conclusion

The PLCO_M2012 ≥ 1.51% criteria identified more people eligible for screening in Yorkshire Lung Screening Trial and resulted in more screen-detected lung cancers than LLP_v2 ≥ 5% or USPSTF₂₀₁₃. When compared in equivalent populations, there was no significant difference between risk models in terms of lung cancer detection and each appeared more efficient at screening population selection than USPSTF₂₀₁₃.