The LI in patients with IP was significantly higher than in the control group, while no significant difference in gallium index was found between the control and the IP group (Fig 3). No significant correlation was observed between gallium index and LI among all study participants (Fig 4). In the single patient with acute interstitial pneumonia, gallium index and LI in all lung fields on admission to the hospital (Fig 5, top) were above the mean + 2 SD measured in control patients (55.7% and 13.9 X 10 /min, respectively). Two weeks after the initiation of steroid treatment, the gallium index had decreased markedly, whereas LI and Pao2/fraction of inspired oxygen had remained nearly unchanged (Fig 5, middle). Eleven weeks later, LI was near normal (Fig 5, bottom). LI in patients with active IP was significantly higher than in the group with stable IP, whereas there was no significant difference in gallium index between the two groups (Fig 6). Discussion While the presence of alveolar-capillary membrane injury in interstitial lung disease has been examined by measurements of epithelial permeability to small, water-soluble molecules and of albumin concentration in BAL fluid, an increase in vascular leakiness has not been previously established. In this study of patients with IP, we were able to demonstrate an increase in pulmonary vascular leakiness by measuring LI. However, no correlation between gallium index and LI was found. This suggests that pulmonary Ga uptake is independent of increased pulmonary vascular leakiness and may not merely reflect lung injury in IP. Line et al have suggested that the quantitative evaluation of Ga scans may be useful to stage the activity of idiopathic pulmonary fibrosis and to follow responses to therapy. However, Gelb et al have suggested that positive gallium scan results are associated with the cellular phase of idiopathic pulmonary fibrosis, and do not reliably identify a corticosteroid-responsive group. It has been suggested that Ga first binds to transferrin, and that the gallium-transferrin complex then binds to transferrin receptors distributed on the cell surface. Therefore, these reports suggest that Ga can bind to several pulmonary cell types, which have transferrin receptors, and that a yet unknown pathophysiologic factor causes an increase in pulmonary Ga uptake in interstitial lung disease. Serial measurements of gallium index and LI were performed in the patient with acute interstitial pneumonia. As shown in Figure 5, center, the abnormal Ga uptake had markedly decreased after 2 weeks of steroid therapy, whereas the increased LI and the gas exchange abnormalities had remained unchanged. It is likely that this remarkable decrease in gallium uptake reflected the rapid disappearance of accumulated and movable inflammatory cells. Thus, the rapid change in Ga uptake in response to steroid therapy did not seem to be paralleled by as rapid a resolution of pulmonary septal injury in this case of acute interstitial pneumonia. These results suggest that LI could be useful to assess disease activity in IP. Figure 3. Gallium index and LI in index patients vs control subjects. The average of six ROIs for each individual is shown. Figure 4. Correlation between gallium index and LI. Each circle represents the average of six ROIs per patient. Closed circles indicate patients with IP. Figure 5. Serial gallium scan images with LI, gallium index, and Pao2/fraction of inspired oxygen (FiO2) in a single patient with acute interstitial pneumonia. Top: Scan before treatment. Center: Scan obtained 2 weeks after the initiation of steroid therapy. Bottom: Scan obtained 11 weeks after the initiation of steroid therapy.