Biomarker-Integrated Risk Model for Lung Cancer


Key Points

  • Inclusion of a biomarker panel into the smoking-exposure risk model improved predictive ability.
  • The circulating biomarkers consisted of CA125, CEA, cytokeratin-19 fragment, and precursor form of surfactant protein B.

As reported in JAMA Oncology by the Integrative Analysis of Lung Cancer Etiology and Risk (INTEGRAL) Consortium for Early Detection of Lung Cancer, incorporating biomarkers into a lung cancer risk prediction model may improve performance compared with risk based on age and smoking exposure alone.

In the study, prediagnostic blood samples from 108 ever-smoking patients with lung cancer diagnosed within 1 year after sampling and samples from 216 smoking-matched controls from the Carotene and Retinol Efficacy Trial cohort were used to develop a biomarker risk score based on 4 circulating proteins: cancer antigen 125 (CA-125), carcinoembryonic antigen (CEA), cytokeratin-19 fragment (CYFRA 21-1), and precursor form of surfactant protein B (Pro-SFTPB).

The biomarker score was validated using absolute risk estimates for 63 ever-smoking patients with lung cancer diagnosed within 1 year after blood collection and 90 matched controls from 2 large European population-based cohorts (European Prospective Investigation into Cancer and Nutrition and the Northern Sweden Health and Disease Study). Predictive abilities of a traditional smoking history–based risk model (smoking model) and a model combining biomarker score and the smoking model (integrated model) were compared.

Risk Prediction With Integrated Model

In the validation study of 63 ever-smoking patients with lung cancer and 90 matched controls, the predicted risk of a diagnosis of lung cancer within 1 year for a 60-year-old man with a smoking history of 30 pack-years was estimated at 0.37% using the smoking model; on the integrated model, risk was 0.07% and 1.56% for the same man with a biomarker score equal to the average of the first and fourth quartiles.

Compared with the smoking model, median 1-year risk estimates from the integrated model increased for cases from 0.27% to 0.45% and decreased for controls from 0.12% to 0.04%. The area under the receiver-operating characteristics curve (AUC) was 0.83 for the integrated model vs 0.73 for the smoking model alone (P = .003).

At an overall specificity of 0.83 based on the U.S. Preventive Services Task Force (USPSTF) computed tomography screening criteria, the sensitivity of the integrated model was 0.63 vs 0.43 for the smoking model. At an overall sensitivity of 0.42 based on USPSTF screening criteria, the specificity of the integrated model was 0.95 vs 0.86 for the smoking model.

The investigators concluded, “This study provided a proof of principle in showing that a panel of circulating protein biomarkers may improve lung cancer risk assessment and may be used to define eligibility for computed tomography screening.”

The study was supported by grants from the National Cancer Institute and National Cancer Institute Early Detection Research Network, Fondation ARC pour la recherche sur le cancer and INCa, MD Anderson Lung Cancer Moon Shot Program, Lyda Hill Foundation, Canary Foundation, Lungevity Foundation, and S. Rubenstein Family Foundation.

Samir M. Hanash, MD, PhD, of The University of Texas MD Anderson Cancer Center, and Mattias Johansson, PhD, of the International Agency for Research on Cancer, Lyon, are the corresponding authors for the JAMA Oncology article.

The content in this post has not been reviewed by the American Society of Clinical Oncology, Inc. (ASCO®) and does not necessarily reflect the ideas and opinions of ASCO®.