A variability study of regional alveolar oxygen tension measurement in humans using hyperpolarized 3He MRI

PURPOSE: A systematic study of the short-term and long-term variability of regional alveolar partial pressure of oxygen tension (pA O2 ) measurements using (3) He magnetic resonance imaging was presented. Additionally, the repeatability of the average evaluated pA O2 was compared with that of the standard pulmonary function tests. METHODS: Pulmonary function test and pA O2 imaging were performed on 4 nonsmokers (1 M, 3 F, 56 ± 1.7 years) and 4 smokers (3 M, 1 F, 52 ± 7.5 years) during three visits over the course of 2 weeks. Two measurements were performed per visit. Variability of pA O2 was assessed using a mixed-effect model, with an intraclass correlation coefficient calculated for each group. The coefficient of variation of pA O2 over the 3-day period was also compared with the coefficient of variation of pulmonary function test results. RESULTS: Short-term regional variability based on intraclass correlation coefficient was 0.71 for nonsmokers, and 0.63 for smokers, with long-term variability significantly lower at 0.59 and 0.47, respectively. While the coefficient of variation of the average pA O2 was similar to the repeatability of the diffusing capacity of CO, it was significantly higher than that of Forced Vital Capacity (P = 0.02). CONCLUSION: Short-term and long-term pA O2 variability differences were used as an indication of true physiological changes in order to measure technical reproducibility. Smokers show higher physiologic variability and less technical reproducibility. The suggested pA O2 -imaging technique showed a reasonable regional repeatability in nonsmokers as well as the ability to detect differences between the two groups with similar reproducibility and superior discriminatory ability when compared with pulmonary function tests.