Family history of cancer and risk of lung cancer: a retrospective cohort study in a Chinese population

Abstract

Background Family history of cancer is an established risk factor for lung cancer, but its association with the probability of malignancy among patients evaluated for pulmonary nodules—rather than with population-level lung cancer incidence—remains poorly characterized in Chinese populations. Methods We conducted a retrospective cohort study of 3,850 consecutive patients evaluated for pulmonary nodules at a tertiary teaching hospital between January 2018 and December 2021, of whom 1,257 (32.6%) underwent surgical resection or biopsy. Family history of lung and non-lung cancer in first- and second-degree relatives was self-reported at admission. Lung cancer was confirmed histopathologically. Sequential multivariable logistic regression was performed: Model 1 included clinical variables only, and Model 2 additionally included radiological features. Results Among 3,850 patients, 1,158 (30.1%) had pathology-confirmed lung cancer (95.3% adenocarcinoma; 92.2% stage IA). The 2,692 controls comprised 99 patients with pathology-confirmed benign lesions and 2,593 patients managed with imaging surveillance in whom no malignant pathology was recorded. First-degree family history of lung cancer (present in 32.3% of the cohort) was not associated with malignancy in univariate analysis (OR 0.94, 95% CI 0.81–1.09, p = 0.432) but showed an inverse association after adjustment (Model 2: aOR 0.72, 95% CI 0.57–0.90, p = 0.005); family history of non-lung cancer showed a similar pattern (univariate OR 0.79, 95% CI 0.69–0.91; Model 2: aOR 0.59, 95% CI 0.48–0.74, p < 0.001). Radiological features dominated risk prediction (Model 2 AUC 0.807 versus 0.566 for clinical variables alone): nodule size (aOR 1.21 per mm, 95% CI 1.18–1.24), lobulation (aOR 2.32), spiculation (aOR 2.08), and vacuole sign (aOR 1.80) were the strongest predictors. Female sex was independently associated with malignancy (aOR 1.62, 95% CI 1.31–2.01), whereas smoking was not (aOR 1.16, 95% CI 0.90–1.49). Patients with family history were not diagnosed earlier or at an earlier stage than those without. Conclusion In this selected cohort of patients evaluated for pulmonary nodules, family history was not positively associated with malignancy, and the inverse adjusted associations are most plausibly explained by referral- and surveillance-related selection rather than biological protection. Radiological assessment should remain the primary basis for pulmonary nodule management. Family history may be considered during nodule evaluation, but its independent contribution should be validated in prospective cohorts with standardized screening and referral procedures.