Integrated F-Fluorodeoxyglucose–Positron Emission Tomography/Dynamic Contrast-Enhanced Computed Tomography to Phenotype Non–Small Cell Lung Carcinoma

Abstract

We applied modern molecular and functional imaging to the pretreatment assessment of lung cancer using combined dynamic contrast-enhanced computed tomography (DCE-CT) and 18 F-fluorodeoxyglucose–positron emission tomography ( 18 F-FDG-PET) to phenotype tumors. Seventy-four lung cancer patients were prospectively recruited for 18 F-FDG-PET/DCE-CT using PET/64-detector CT. After technical failures, there were 64 patients (35 males, 29 females; mean age [± SD] 67.5 ± 7.9 years). DCE-CT yielded tumor peak enhancement (PE) and standardized perfusion value (SPV). The uptake of 18 F-FDG quantified on PET as the standardized uptake value (SUV max ) assessed tumor metabolism. The median values for SUV max and SPV were used to define four vascular-metabolic phenotypes. There were associations (Spearman rank correlation [ rs ]) between tumor size and vascular-metabolic parameters: SUV max versus size ( rs = .40, p = .001) and SUV/PE versus size ( r = .43, p < .001). Patients with earlier-stage (1-HA, n = 30) disease had mean (± SD) SUV/PE 0.36 ± 0.28 versus 0.56 ± 0.32 in later-stage (stage IIB-IV, n = 34) disease ( p = .007). The low metabolism with high vascularity phenotype was significantly more common among adenocarcinomas ( p = .018), whereas the high metabolism with high vascularity phenotype was more common among squamous cell carcinomas ( p = .024). Other non-small cell lung carcinoma tumor types demonstrated a high prevalence of the high metabolism with low vascularity phenotype ( p = .028). We show that tumor subtypes have different vascular-metabolic associations, which can be helpful clinically in managing lung cancer patients to hone targeted therapy.