PET imaging of pulmonary perfusion and ventilation using ¹³N-Nitrogen

Transported by perfusion and ventilation, we use the dynamics of ¹³NN activity to determine two key variables of respiratory function.

by Tilo Winkler

Key Topic6 June, 2021

Ventilation (V) and perfusion (Q) transport oxygen and carbon in the lungs, and V/Q matching is essential for the efficiency of the gas exchange in the alveoli. However, most pulmonary diseases lead to heterogeneity in V/Q matching reducing the efficiency of the gas exchange. Accurate regional assessments of both ventilation and perfusion are needed to gain deeper insights into the patterns and causes of regional V/Q heterogeneity in different lung diseases.

Our group has developed and refined the ¹³NN-bolus injection method for ventilation-perfusion imaging using Positron Emission Tomography (PET). This method yields accurate values even under conditions where the distributions of ventilation and perfusion in the lungs are highly heterogeneous. Additionally, it allows the assessment of perfusion heterogeneity not biased by imaging noise.

Perfusion

For ¹³NN imaging, we monitor the tidal breathing of the subject. Perfusion imaging starts with a breath-hold at the mean lung volume between end-expiration and end-inspiration. Simultaneously with the beginning of the breath-hold, a bolus of 13-NN saline is injected intravenously and the acquisition of the dynamic PET scan is started.

When the ¹³NN reaches the pulmonary capillaries, it diffuses quickly into the alveoli reaching at the end of the breath-hold a plateau. PET images of the plateau phase are proportional to the relative perfusion showing the heterogeneity in regional perfusion in the lungs (Fig. 1).

13NN tracer kinetics, perfusion, and PET example — Fig. 1 Transport and kinetics of the ¹³NN tracer in the lungs showing a plateau at the end of a breath-hold where the activity is proportional to the regional perfusion illustrated in an example PET image.

Ventilation

After the breath-hold, the tracer kinetics of ¹³NN shows the washout by alveolar ventilation (Fig. 2). The time constant of the washout function is equal to the specific ventilation defines as alveolar ventilation divided by alveolar volume.

13NN tracer kinetics, ventiation, and PET example — Fig. 2 Transport and kinetics of the ¹³NN tracer during the washout by alveolar ventilation. Tracer retention at the end of the washout indicates ventilation defects (VDefs) due to low alveolar ventilation or gas trapping.

Very low ventilation and complete airway closure result in tracer retention at the end of the washout phase, e.g., a minor degree of retention in the example PET image (Fig. 2). During bronchoconstriction, severe airway narrowing and closure result in ventilation defects (VDefs) with very high tracer retention (Video 1).

Video 1 Tracer retention in VDefs inside the lungs during bronchoconstriction in asthma (3D rendering of the tracer retention data from a PET image).

Modeling and parameter estimation

The tracer-kinetics functions of the voxels of the PET image of different regions of interest (ROIs) can be dramatically different from each other if the ventilation is very heterogeneous, e.g., during bronchoconstriction in asthma. We developed for identifying the ventilation and perfusion parameters using four different models for curve fitting and a model selection based on the Akaike Information Criterion (AIC) for selecting the model with the best fit (Fig. 3). Remarkably, modeling of tracer kinetic that are the sum of two different functional compartments allows for the assessment of functional heterogeneity beyond the spatial resolution of the PET scanner.

Parameter estimation of 13NN tracer kinetics — Fig. 3 Schematic of our parameter identification method for ¹³NN kinetics providing up to four parameters per voxel.

Validation of V/Q imaging

We validated the estimated parameters in large animal studies with major V/Q heterogeneities. Our results show the blood gases predicted blood gases based on V/Q imaging were in excellent agreement with the measured blood gases.

Selected papers using this method

Nature 2005;434(7034):777-82. doi: 10.1038/nature03490.
Self-organized patchiness in asthma as a prelude to catastrophic shifts. Jose G Venegas, Tilo Winkler, Guido Musch, Marcos F Vidal Melo, Dominick Layfield, Nora Tgavalekos, Alan J Fischman, Ronald J Callahan, Giacomo Bellani, R Scott Harris.
Am J Respir Crit Care Med. 2017;196(7):834-844. doi: 10.1164/rccm.201612-2438OC.
Hypoxic Pulmonary Vasoconstriction Does Not Explain All Regional Perfusion Redistribution in Asthma. Vanessa J Kelly, Kathryn A Hibbert, Puja Kohli, Mamary Kone, Elliot E Greenblatt, Jose G Venegas, Tilo Winkler, R Scott Harris.
J Nucl Med. 2021 Mar;62(3):405-411. doi: 10.2967/jnumed.120.245977.
PET Imaging Reveals Early Pulmonary Perfusion Abnormalities in HIV Infection Similar to Smoking. Puja Kohli, Vanessa J Kelly , Kathryn A Hibbert , Björn Corleis, Mamary Kone, Josalyn L Cho, Doreen DeFaria-Yeh, Douglas S Kwon, Benjamin D Medoff, R Scott Harris, Tilo Winkler.
J Nucl Med. 2003 Dec;44(12):1982-91.
Quantification of regional ventilation-perfusion ratios with PET. Marcos F Vidal Melo, Dominick Layfield, R Scott Harris, Kevin O'Neill, Guido Musch, Torsten Richter, Tilo Winkler, Alan J Fischman, Jose G Venegas.