Serum Biomarkers May Prognosticate Recurrence in Node-Negative, Non-Small Cell Lung Cancers Less Than 4 Centimeters.

BACKGROUND:A significant proportion of patients who undergo lung resection for less than 4 cm non-small cell lung cancer (NSCLC) will die of disease recurrence within 5 years. The ability to identify patients at greatest risk for recurrence may help individualize treatment and surveillance regimens and improve outcomes. We hypothesized that a serum-based biomarker panel could help risk stratify patients with node-negative NSCLC less than 4 cm for recurrence after lung resection. METHODS:An institutional biorepository of more than 1,800 cases was used to identify patients with resected, node-negative NSCLC less than 4 cm in size. Clinical and radiographic data were collected. Preoperative serum specimens were evaluated in a blinded manner for 47 biomarkers that sampled biological processes associated with metastatic progression, including angiogenesis, energy metabolism, apoptosis, and inflammation. Receiver-operating characteristics curves and log rank tests were used to evaluate individual biomarkers with respect to recurrence, followed by random forest analysis to generate and cross validate a multiple-analyte panel to risk stratify patients for recurrence. RESULTS:The cohort included 123 patients with a median follow-up of 58.2 months; 23 patients had recurrences. A seven-analyte panel consisting of human epididymis protein 4, insulinlike growth factor-binding protein 1, beta-human chorionic gonadotropin, follistatin, prolactin, angiopoietin-2, and hepatocyte growth factor optimally identified patients with disease recurrence with a cross-validated specificity of 91%, sensitivity of 22%, negative predictive value of 83%, positive predictive value of 36%, and accuracy of 78%, providing an area under the receiver-operating characteristics curve of 0.70. CONCLUSIONS:Serum-based biomarkers may be useful for risk stratifying patients with node-negative NSCLC less than 4 cm for recurrence after lung resection.