Kinetic Modeling and Parametric Imaging of N-NH in Treatment-Naïve Lung Cancer.

N-NH PET imaging provides insights into tumor perfusion and metabolism, hence potentially valuable in oncological diagnosis, staging, and prognosis. However, its kinetic characteristics and the optimal protocol for kinetic modeling and parametric imaging remain unclear. This study aims to elucidate the kinetic features of N-NH in lung cancer and its clinical value. Nine lung cancer surgical candidates were prospectively incorporated. Using total-body PET/CT scanner, 35 min N-NH acquisitions were conducted immediately postinjection. Subsequently, routine 5 min F-FDG acquisitions were made. N-NH PET data were reconstructed into dynamic image series. Tumor lesions and normal organs were segmented using nonthreshold dependent automatic or semiautomatic tools. Reversible and irreversible 2-tissue compartment models (2TC vs 2TiC) using image-derived input functions (IDIFs) with population-based metabolite correction were adopted for parametric modeling. Akaike Information Criterion (AIC) was calculated for model selection. Parametric images were produced with the optimal model for lesions. A total of 9 patients presented with 9 primary lung tumors and 17 histologically confirmed lymphadenopathy. All primary lung tumors and regional lymph node metastases were detectable using both N-NH and F-FDG imaging. Primary lung tumors, regional lymphadenopathy, and lung backgrounds demonstrated smaller AIC using 2TC models with pulmonary artery as IDIF, while other organs favored either 2TC or 2TiC models with the descending aorta as IDIF. Lesion-to-background lung ratios reached around 2.218-3.407 for primary lung tumors and 1.932-2.537 for regional lymphadenopathy 10-20 min postinjection of N-NH. Vt images derived from 2TC modeling showed better lesion-to-background lung ratios (4.511 ± 2.955 for primary tumor, and 2.991 ± 2.152 for regional lymphadenopathy). For N-NH imaging in lung cancer, a static image can be acquired at 10-20 min postinjection for clinical diagnosis. The reversible 2TC model is preferred over 2TiC, and the Vt image is preferred over other parametric images in terms of lesion contrast.