Pattern and predictors of false positive lymph node involvement on positron emission tomography in patients with non-small cell lung cancer.

Objectives The aim of this study was to elucidate the optimal parameters for diagnosing false positive (FP) lymph nodes (LNs) in patients with non-small cell lung cancer.Methods We reviewed 292 patients with non-small cell lung cancer (NSCLC). Fluorodeoxyglucose positron emission tomography (FDG-PET) imaging was performed at 1 hour (early) post-FDG injection and repeated 2 hours (delayed) after injection. We analyzed the relationship between the pathology of LNs and the results of PET, and the percent change in the standardized uptake value (%ΔSUV) between the two time-points.Results Eighteen of 46 cases (39.1%) in the FP group showed higher SUVs for their LNs compared with those for primary tumor, whereas 13.2% in the true positive group (p = 0.032) had higher SUVs for their LNs. Thirty-four of 36 cases in the true positive group had %ΔSUV ranging from 0% to 61.5% compared with only 13 of 33 in the FP group. Twenty out of 22 cases (90.9%) where %ΔSUV was over 61.5% or under 0% were considered as FP.Conclusions Patients with higher SUVs for LNs than for primary tumors and patients with extremely high or low %ΔSUVs tended to have FP LNs.