Quantitative volumetric CT-histogram analysis in N-staging of 18F-FDG-equivocal patients with lung cancer

Lung cancer often coexists with acute and chronic lung diseases such as chronic obstructive pulmonary disease. Therefore, mediastinal lymph nodes may be false-positive on (18)F-FDG PET because of the inflammatory disease alone. Nevertheless, (18)F-FDG PET/CT is the primary imaging modality used for staging patients with lung cancer, including nodal status. The purpose of this study was to evaluate whether volumetric CT histogram analysis can improve the characterization of lymph nodes on PET/CT staging of patients with lung cancer.

Sixty histologically proven lymph nodes of 45 patients aged 43-76 y diagnosed with lung cancer were investigated. (18)F-FDG PET/CT, contrast-enhanced CT, and nonenhanced CT were performed before surgery or biopsy as part of the clinical staging procedure. Lymph nodes were analyzed on the basis of the (18)F-FDG standardized uptake value and volumetric CT histogram analysis. These findings were correlated to the gold standard of histopathology.

Histologic examination revealed 36 positive and 24 negative lymph nodes, which were also successfully analyzed by volumetric CT histogram. Median CT density was significantly higher for histologically positive lymph nodes (33.2 Hounsfield units [HU]; range, -29.8 to 59.1) than for histologically negative lymph nodes (10.1 HU; range, -21.0 to 87.4; P = 0.002). The incidence of malignancy was 88% above a cutoff value of 20 HU in the ten (18)F-FDG-equivocal lymph nodes; the incidence of benign findings was 100% in the interval between -20 and +20 HU. Visual- and density-based analysis on contrast-enhanced CT failed to differentiate affected from nonaffected lymph nodes.

Three-dimensional histogram analysis is a promising and potentially valuable imaging surrogate for N-stage stratification in patients with lung cancer with unclear glucose uptake during (18)F-FDG PET imaging. In cases of equivocal (18)F-FDG PET status, this technique might potentially bridge the diagnostic gap between noninvasive techniques and invasive lymph node sampling and could help improve the yield of core biopsies.