Blood Test–Informed Model Beats PLCOm2012 for Lung Cancer Risk

A risk prediction model that incorporates routine blood tests and a DNA methylation marker showed better discrimination for 6-year lung cancer risk than the PLCOm2012 model among ever-smokers in 2 Danish cohorts, according to a study published in the Journal of Clinical Oncology.

The investigators noted that lung cancer screening reduces lung cancer mortality, but identifying high-risk individuals remains difficult, and that selection criteria based on smoking history and age include large fractions of the adult population. They developed the Copenhagen Lung Cancer Risk Model (CPH-Lung) to improve risk stratification before image-based screening.

CPH-Lung was trained in a case-control study of ever-smokers (689 cases and 1,924 controls) nested within the Copenhagen General Population Study. It was validated in 3,195 ever-smokers aged 50 years or older in the Copenhagen City Heart Study, 68 of whom developed lung cancer within 6 years. In validation, the area under the curve (AUC) was 0.783 (95% CI, 0.731-0.835) for CPH-Lung vs 0.720 (95% CI, 0.669-0.770) for PLCOm2012 (P =0.006). At a fixed specificity of 77.7%, the level NELSON or NLST criteria would have achieved in the training cohort, sensitivity was 64.7% vs 48.5%.

The final model combines clinical variables, including pack-years, years since smoking cessation, education, prior cancer, and pneumonia within the past 10 years, with aryl hydrocarbon receptor repressor gene (AHRR) methylation, C-reactive protein, monocyte count, hemoglobin, platelet count, and ionized calcium. Low vs high AHRR methylation was associated with 4 times higher lung cancer risk. In the training cohort, sensitivity differences between the models were most pronounced in women, individuals younger than 70 years, and those with lower education, and individuals whose cancers were detected only by CPH-Lung were younger at diagnosis (mean, 68.2 years) than those detected only by PLCOm2012.

The authors noted that the model was developed and validated in populations that were more than 98% Caucasian, which may limit generalizability, and that the validation study was relatively small. In validation, the full model’s AUC was not significantly better than that of a version without blood tests. They stated that validation in other countries and more ethnically diverse populations is needed.

Source: Skov-Jeppesen SM, Kobylecki CJ, Egstrand S, et al. Development and validation of a blood test informed risk prediction model for lung cancer—the Copenhagen Lung Cancer Risk Model (CPH-Lung). J Clin Oncol. Published online October 12, 2026. doi:10.1200/JCO-25-02326