Lung Cancer Screening Beyond Smoking-Based Eligibility: Evidence, Limits, and Lessons From China

Low-dose computed tomography (LDCT) screening lowers lung cancer mortality in populations with substantial cumulative tobacco exposure, and that conclusion rests on two adequately powered randomized controlled trials. Whether the same benefit extends to populations with different risk profiles remains unresolved, particularly in China, where a substantial share of lung cancers arise in individuals without substantial smoking history. This review develops three linked arguments rather than surveying the field exhaustively. First, stage shift and risk-enrichment are necessary but not sufficient; a concurrent reduction in advanced-stage incidence is the more reliable signal of benefit. Second, risk-prediction models built on tobacco-exposed cohorts are least dependable in individuals without substantial smoking history, precisely where expanding eligibility relies on them most. Third, judging evidentiary sufficiency requires an explicit framework that treats sufficiency as a continuum rather than a binary between a randomized trial and nothing. Chinese programs illustrate all three at scale: LDCT can be delivered broadly and can enrich early-stage detection, yet still not establish reduced lung cancer mortality with acceptable harm. Stage-shift findings remain vulnerable to lead-time bias, length bias, and overdiagnosis, and recent data from China and South Korea suggest that diffusion of LDCT into populations at average risk may enlarge overdiagnosis more than net benefit. For populations without substantial smoking history, eligibility decisions should rest on prospective evidence of reduced advanced-stage incidence rather than stage-shift enrichment alone; expansion belongs within disciplined prospective evaluation.