RESUMEN
BACKGROUND: In lung cancer screening the ability to distinguish malignant from benign nodules is a key issue. This study evaluates the ability of positron emission tomography (PET) and volume doubling time (VDT) to discriminate between benign and malignant nodules. METHODS: From the Danish Lung Cancer Screening Trial, participants with indeterminate nodules who were referred for a 3-month rescan were investigated. Resected nodules and indolent nodules (ie, stable for at least 2 years) were included. Between the initial scan and the 3-month rescan, participants were referred for PET. Uptake on PET was categorised as most likely benign to malignant (grades I-IV). VDT was calculated from volume measurements on repeated CT scans using semiautomated pulmonary nodule evaluation software. Receiver operating characteristic (ROC) analyses were used to determine the sensitivity and specificity of PET and VDT. RESULTS: A total of 54 nodules were included. The prevalence of lung cancer was 37%. In the multivariate model both PET (OR 2.63, p<0.01) and VDT (OR 2.69, p<0.01) were associated with lung cancer. The sensitivities and specificities of both PET and VDT were 71% and 91%, respectively. Cut-off points for malignancy were PET>II and VDT<1 year, respectively. Combining PET and VDT resulted in a sensitivity of 90% and a specificity of 82%; ROC cut-off point was either PET or VDT indicating malignancy. CONCLUSION: PET and VDT predict lung cancer independently of each other. The use of both PET and VDT in combination is recommended when screening for lung cancer with low-dose CT.