Abstract
1384
Objectives To determine the impact of a high-resolution (HR) voxel reconstruction on a state-of-the-art Positron Emission Tomography/Computed Tomography (PET/CT) system for staging of intra-thoracic lymph node metastasis in patients with non-small cell lung cancer (NSCLC).
Methods We included 60 patients with clinically suspected NSCLC, who underwent FDG PET/CT on a novel time-of-flight system (Ingenuity TF, Philips Healthcare). In each patient, cytological/ histological information of at least one suspected intra-thoracic lymph node was obtained. PET data were reconstructed using two reconstruction types: standard 4x4x4 mm3 voxels and high-resolution (HR) 2x2x2 mm3 voxels. We performed quantitative analysis on 103 lymph nodes by measuring the maximum standardized-uptake-value (SUVmax) for both PET reconstruction types. The optimal SUVmax cut-off values to distinguish benign from malignant lymph nodes were considered to be the values with the highest accuracy. Results were compared with literature, where a SUVmax cut-off value of 2.5 is commonly used.
Results On standard reconstructions, SUVmax values were 2.6±1.0 (mean±SD) for benign (n=54) and 7.1±3.1 for malignant (n=49) lymph nodes, respectively (p<0.01). Using the HR reconstruction, these values increased with 20% to 3.1±1.2 and 8.3±3.0 respectively. The optimal SUVmax cut-off value was 3.9 (accuracy 88%) on the standard reconstruction. This cut-off value increased up to 5.0 (accuracy 92%) when using the HR reconstruction.
Conclusions For accurate lymph node staging on a state-of-the-art PET/CT system, it is required to use a higher SUVmax cut-off value than values reported in the literature. The use of a HR reconstruction further increases the measured FDG-uptake. PET readers and referring physicians should update their cut-off values when implementing new PET technology.