An integrated proteomic classifier accurately identified benign lung nodules when used in patients with a pretest probability of malignancy (pCA) ≤50%, according to the results of a clinical trial published in CHEST.
Gerard A. Silvestri, MD, from the Thoracic Oncology Research Group, Division of Pulmonary and Critical Care Medicine, Medical University of South Carolina in Charleston, and colleagues conducted the prospective multicenter observational trial, Pulmonary Nodule Plasma Proteomic Classifier (PANOPTIC; ClinicalTrials.gov Identifier: NCT01752114) using multiple reaction monitoring mass spectrometry to measure the relative abundance of 2 plasma proteins — LG3BP and C163A. A total of 685 patients with 8- to 30-mm lung nodules were included.
The results were integrated with a clinical risk prediction model to identify lung nodules that were likely to be benign and calculated sensitivity, specificity, and negative predictive value (NPV). The researchers then estimated how decisions regarding invasive testing might have changed had the integrated classifier results been available before those decisions were made.
A subgroup of 178 patients with clinician assessed pCA ≤50% had a 16% prevalence of lung cancer. The integrated classifier demonstrated a sensitivity of 97% (95% CI, 82%-100%), a specificity of 44% (95% CI, 36%-52%), and an NPV of 98% (95% CI, 92%-100%) in distinguishing benign nodules from malignant ones. The integrated proteomic classifier performed better than positron emission tomography, validated lung nodule risk models, and physician cancer probability estimates (P <.001).
The authors determined that if the integrated classifier results were used to direct care, 40% fewer invasive procedures would be performed on benign modules and 3% of malignant nodules would be misclassified. They suggested that the biomarker should be assessed in a clinical utility study, as the implications of a negative result clearly support such assessment.
Limitations of the study included the lack of 2-year nodule stability data, the underrepresentation of community practices, and the retrospective analysis of the estimate of effect on invasive testing decisions.
Disclosures: This study was funded by Integrated Diagnostics. Several authors report financial relationships with pharmaceutical and device companies.
Silverstri GA, Tanner NT, Kearney P, et al. Assessment of plasma proteomics biomarker’s ability to distinguish benign from malignant lung nodules: results of the PANOPTIC (PulmonAry NOdule Plasma proTeomIc Classifier) trial [published online February 26, 2018]. CHEST. doi: 10.1016/j.chest.2018.02.012