Modeling 18F-FDG kinetics during acute lung injury: experimental data and estimation errors.

Area of Science:

• Nuclear medicine
• Pulmonary imaging
• Pharmacokinetics

Background:

• Positron Emission Tomography (PET) with 2-deoxy-2-[18F]flouro-D-glucose ((18)F-FDG) is increasingly used to assess pulmonary inflammation in acute lung injury (ALI).
• Estimating (18)F-FDG kinetics is crucial for understanding lung pathophysiology during ALI.
• The accuracy of kinetic parameter estimation can be affected by physiological changes like increased extra-vascular lung water.

Purpose of the Study:

• To evaluate the impact of extra-vascular lung water on (18)F-FDG kinetic parameter estimates.
• To compare the accuracy of traditional methods (Patlak and Sokoloff) with a novel four-compartment model in estimating (18)F-FDG kinetics under conditions of lung injury.

Main Methods:

• Experimental study in eleven sheep with induced ALI (lung lavage and mechanical ventilation) and five sheep with endotoxemia.
• Dynamic (18)F-FDG PET scans were performed.
• Simulations of normal and ALI (18)F-FDG kinetics were conducted to assess parameter dependence on transport rates.
• Kinetic parameters (Ki, k(3), F(e)) were estimated using Patlak, Sokoloff, and a four-compartment model.

Main Results:

• The four-compartment model explained 85.7% of the studied (18)F-FDG kinetics better than the Sokoloff model.
• The Sokoloff model showed a significant positive bias in net uptake rate (Ki) and inaccurate estimates of its components (k(3), F(e)) compared to the four-compartment model.
• Simulations revealed that extra-vascular fluid significantly impacted Ki estimates, with larger errors for Patlak and Sokoloff methods in ALI compared to normal lungs.

Conclusions:

• Accumulation of (18)F-FDG in extra-vascular fluid during ALI leads to substantial errors in kinetic parameter estimation using Patlak and Sokoloff methods.
• These errors are influenced by the volume and transport rates of the extra-vascular compartment.
• The four-compartment model provides more accurate quantification of (18)F-FDG kinetics in the presence of increased extra-vascular fluid during lung injury.