RATIONALE: Indeterminate pulmonary nodules are a common radiographic finding and require further evaluation because of the concern for lung cancer. OBJECTIVES: We developed an algorithm to assign patients to a low- or high-risk category for lung cancer, based on a combination of serum biomarker levels and nodule size. METHODS: For the serum biomarker assay, we determined levels of carcinoembryonic antigen, α1-antitrypsin, and squamous cell carcinoma antigen. Serum data and nodule size from a training set of 509 patients with (n = 298) and without (n = 211) lung cancer were subjected to classification and regression tree and logistic regression analyses. Multiple models were developed and tested in an independent, masked validation set for their ability to categorize patients with (n = 203) or without (n = 196) lung cancer as being low- or high-risk for lung cancer. MEASUREMENTS AND MAIN RESULTS: In all models, a large percentage of individuals in the validation study with small nodules (<1 cm) were assigned to the low-risk group, and a large percentage of individuals with large nodules (≥3 cm) were assigned to the high-risk group. In the validation study, the classification and regression tree algorithm had overall sensitivity, specificity, and positive and negative predictive values for determining lung cancer of 88%, 82%, 84%, and 87%, respectively. The logistic regression model had overall sensitivity, specificity, and positive and negative predictive values of 80%, 89%, 89%, and 81%, respectively. CONCLUSION: Integration of biomarkers with lung nodule size has the potential to help guide the management of patients with indeterminate pulmonary nodules.
RATIONALE: Indeterminate pulmonary nodules are a common radiographic finding and require further evaluation because of the concern for lung cancer. OBJECTIVES: We developed an algorithm to assign patients to a low- or high-risk category for lung cancer, based on a combination of serum biomarker levels and nodule size. METHODS: For the serum biomarker assay, we determined levels of carcinoembryonic antigen, α1-antitrypsin, and squamous cell carcinoma antigen. Serum data and nodule size from a training set of 509 patients with (n = 298) and without (n = 211) lung cancer were subjected to classification and regression tree and logistic regression analyses. Multiple models were developed and tested in an independent, masked validation set for their ability to categorize patients with (n = 203) or without (n = 196) lung cancer as being low- or high-risk for lung cancer. MEASUREMENTS AND MAIN RESULTS: In all models, a large percentage of individuals in the validation study with small nodules (<1 cm) were assigned to the low-risk group, and a large percentage of individuals with large nodules (≥3 cm) were assigned to the high-risk group. In the validation study, the classification and regression tree algorithm had overall sensitivity, specificity, and positive and negative predictive values for determining lung cancer of 88%, 82%, 84%, and 87%, respectively. The logistic regression model had overall sensitivity, specificity, and positive and negative predictive values of 80%, 89%, 89%, and 81%, respectively. CONCLUSION: Integration of biomarkers with lung nodule size has the potential to help guide the management of patients with indeterminate pulmonary nodules.
Authors: Michael N Kammer; Dhairya A Lakhani; Aneri B Balar; Sanja L Antic; Amanda K Kussrow; Rebekah L Webster; Shayan Mahapatra; Udaykamal Barad; Chirayu Shah; Thomas Atwater; Brenda Diergaarde; Jun Qian; Alexander Kaizer; Melissa New; Erin Hirsch; William J Feser; Jolene Strong; Matthew Rioth; York E Miller; Yoganand Balagurunathan; Dianna J Rowe; Sherif Helmey; Sheau-Chiann Chen; Joseph Bauza; Stephen A Deppen; Kim Sandler; Fabien Maldonado; Avrum Spira; Ehab Billatos; Matthew B Schabath; Robert J Gillies; David O Wilson; Ronald C Walker; Bennett Landman; Heidi Chen; Eric L Grogan; Anna E Barón; Darryl J Bornhop; Pierre P Massion Journal: Am J Respir Crit Care Med Date: 2021-12-01 Impact factor: 30.528
Authors: Michael R Mehan; Stephen A Williams; Jill M Siegfried; William L Bigbee; Joel L Weissfeld; David O Wilson; Harvey I Pass; William N Rom; Thomas Muley; Michael Meister; Wilbur Franklin; York E Miller; Edward N Brody; Rachel M Ostroff Journal: Clin Proteomics Date: 2014-08-01 Impact factor: 3.988
Authors: Sonia Blanco-Prieto; Loretta De Chiara; Mar Rodríguez-Girondo; Lorena Vázquez-Iglesias; Francisco Javier Rodríguez-Berrocal; Alberto Fernández-Villar; María Isabel Botana-Rial; María Páez de la Cadena Journal: Sci Rep Date: 2017-01-24 Impact factor: 4.379
Authors: Charles E Birse; Jennifer L Tomic; Harvey I Pass; William N Rom; Robert J Lagier Journal: Clin Proteomics Date: 2017-07-05 Impact factor: 3.988
Authors: Edward F Patz; Erin Greco; Constantine Gatsonis; Paul Pinsky; Barnett S Kramer; Denise R Aberle Journal: Lancet Oncol Date: 2016-03-18 Impact factor: 41.316