FDG-PET in Staging of Lung Cancer
The treatment plan for newly-diagnosed lung cancer is mostly influenced by mediastinal lymph node status. False-negative findings on PET/CT may cause futile thoracotomy. The aim of the study was to identify the practical predictors of occult metastasis on PET/CT defined negative mediastinal node.
We retrospectively reviewed 66 documented lung cancer patients. All had FDG-PET/CT preoperatively and underwent surgical resection of primary tumor with either lymph node dissection or sampling. Gold standard was histopathology. The characteristics of patients in false negative (FN) and true negative (TN) groups defined by PET scans were compared to identify the predictors of occult mediastinal metastasis. The diagnostic efficacy was re-evaluated with the combination of PET/CT findings and these predictors. The classification tree was plotted to sort high-risk and low-risk patients step by step with the predictors.
Right lower lobe of primary tumor was significantly different between PET-defined FN and -TN groups (P = 0.037) and between PET-defined FN and true positive (TP) groups (P = 0.0.15). Marginal significance of a higher PET defined-positive N1 was noted in -FN than –TN groups (P= 0.089). If we combined these risk factors with original PET image findings, the sensitivity reached from 52.94% to 88.24%. In the classification tree using RLL, central tumor, and N1 (PET-defined TP) on PET as input variables in three subsequent steps, all the cases with target variable (PET-defined FN) could be classified in the groups as predicted to be high-risk for mediastinal metastasis.
According to our study, N1 could be sentinel lymph node, which took more FDG when tumor cells spread through. RLL might have specific lymphatic drainage pattern so most false-negative findings occurred in RLL tumor. With these factors in concern, most of the false-negative results could be identified. The underlying mechanism and whether PET-defined FN would have any clinical impact in outcomes are under investigation.