Introduction:

Fractional Ventilation, defined as the ratio of fresh gas reaching each compartment of the lung to the total gas in that compartment at end-inhale, can be estimated using MR imaging and a multi-breath wash-in of hyperpolarized (HP) gas. Mimicking physiological breathing, a multi-breath wash-in of HP gas gradually illuminates regions of the diseased lung that imaging gas cannot reach in the short time-constants of single breath imaging, providing a comprehensive measure of ventilation and gas mixing efficiency, as well as additional information about the damaged regions of the lung that are of paramount interest. In response to the global ³He shortage, last year we presented FV imaging using HP ¹²⁹Xe in large mammals. In this work, we present an initial demonstration of the feasibility of using HP ¹²⁹Xe in conjunction with a multi-breath wash-in MR imaging technique to measure fractional ventilation (FV) in human subjects¹.

Material and Methods:

All subjects underwent an initial dry-run maneuver outside of the MRI room to study the effect of back-to-back inhalations of a diluted xenon/oxygen mixture using the gas delivery system shown in Figure 1. Based on our observations in the first group of subjects (3 healthy nonsmokers and 2 smokers), modifications to both the delivery system and imaging gas mixtures were deemed necessary. Imaging experiments were performed using a 1.5-T MRI Siemens system and an 8-channel ¹²⁹Xe coil (Stark, Germany). A 2D multi-slice GRE sequence (T_R/T_E = 12.0/4.5 ms) was employed, with coronal slices (resolution~8x8x35 mm³), covering the whole-lung (matrix size: 48x36). The FV image-series consisted of 6 wash-in breaths of HP ¹²⁹Xe polarized to ~45% using Xemed’s XeBox polarizer (Durham, USA). Due to the anesthetic effect and high density of xenon, different gas mixtures (He and N_²) were studied while always keeping the xenon ratio below 30%. SPO₂, heart rate and end-tidal gases were monitored throughout the study, and subjects were asked about any adverse events.

Results:

While all subjects experienced minimal tingling and/or vertigo with or without euphoria after the fourth/fifth breath of imaging gas mixture, none experienced a drop in SPO₂ > 2% during the breathing maneuver, except for the first healthy subject, whose SPO₂ momentarily dropped down to 90% due to poor oxygen mixing. Due to the paramagnetic effect of oxygen, the imaging gas cannot be premixed and two separate lines are necessary, as shown in Figure 1. When using two separate bags and delivery lines, it is important to use larger diameter tubes for the oxygen bag to ensure proper gas mixing: for example, when breathing an average of 700 mL of O₂ in 5s through an ID=5mm tube, flow will stay in the laminar regime; meanwhile, flow of the denser xenon will pass the transient Reynolds’ number, becoming turbulent and causing poor gas mixing. At the same time, due to the relative difference in the gas densities, the tube relative IDs should be chosen proportionally to keep the required normoxic gas ratio. Experimenting with different mixtures, we had our best results when diluting xenon with helium in the imaging gas mixture and adding nitrogen to the oxygen bag to balance the relative density of the bags (Xe:He, 0.3VT:0.2VT and O₂:He, 0.2VT:0.3VT, VT: Tidal Volume). Figure 2 shows the six signal build-ups for two slices as well as FV maps (mean±std) for four slices in one healthy and one COPD subject. As previously shown for HP ³He¹, a very large effect-size (Cohen’s d=1.4) can be detected between the FV-heterogeneity (sd of SV distribution) of healthy vs. COPD subjects, making it a suitable marker of lung function deterioration.

Conclusion:

We have demonstrated the feasibility of multi-breath wash-in imaging of regional fractional ventilation using HP 1²⁹Xe MRI for the first time in both healthy and diseased human subjects. While no subject reported any adverse events from back-to-back inhalation of an imaging gas mixture containing 30% ¹²⁹Xe, it is important to guarantee a normoxic gas mixture when using separate bags for oxygen and xenon due to their drastically different densities. It is also interesting to note that, among all subjects, the COPD subject was least affected by xenon—a fact which may be explained by her compromised percent predicted DLCO of 25% (a measure of gas uptake).

Acknowledgements

This work was funded by NIH R01-HL127969 04.

References

1- Hamedani H, et al. Radiology 279 (3), 917-924