Correspondence lllINDIUM-LABELED NEUTROPHILS FOR DETECTING LUNG INJURY
To the Editor: A recent article in the REVIEW by Haslett and colleagues (1) that reported the use of l11indium-Iabeledneutrophils (In-PMN) for detection of intratracheal bleomycin (ITB) injury in rabbits supports previous work from our laboratory in which we examined In-PMN in ITB-injured rats (2). The relatively noninvasive nature of this approach has obvious clinical appeal for following a variety of human diseases in which PMN are thought to playa role; however, the differences in specific methods used in these and related studies vary and raise some important questions of relevance to the potential utility of this approach in clinical medicine. One question of primary importance is how to define a region within images that is representative of lung activity. Haslett and colleagues counted activity in a region-of-interest (ROI) that overlaid the upper lungs (and body) and normalized activity to that of the injectate; however, the methods of defining the specific size of the region for counting were not precisely defined. Although strong correlation with lung lavage and histologic PMN accumulation was demonstrated, the subjectivity of this approach leavesroom for potential error or bias. In this regard, we found it very difficult to identify reproducibly a similar ROI in normal rats, which is consistent with the asymmetric shape of the lung and the accelerating (logarithmic) dissipation of activity measured at different distances from the lung. In contrast, in bleomycin-injured rats, we outlined the whole lung images visualized above the diaphragm, as well as liver and spleen images, to estimate the whole organ activity of each. Unlike Haslett and colleagues, we found that normalizing imaged lung activity to injectate activity did not enhance discrimination of ITB-injured animals, whereas normalizing lung activity for differences in spleen activity did. Further, lung/spleen ratios correlated strongly with the increase in PMN in lavage fluid. We suggested that lung/spleen ratios may normalize for differences not only in distance of the gamma counter from the imaged tissue but also in the percent of normally functioning PMN in the cell preparations. Preparations that accumulate in the spleen more than in the liver have been reported to be associated with less perturbation (3). However, this argument would not be plausible until abnormal cells had been cleared from the lung and, therefore, may not be applied until a late phase of imaging. We also noted a variable rate of clearance of various InPMN preparations from the lungs of normal rats, with discrimination from abnormal rats distinctly enhanced when each cell preparation was split for imaging of a normal and abnormal animal, with paired comparison. Although this approach has proven useful in some animal studies, a paired approach is not feasible for human studies. Therefore, alternatives must be considered. The rate of initial clearance of In-PMN from the lung may also be used to examine pulmonary vascular sequestration of PMN. We used this approach to study the kinetics of labeled cells in normal rats, standardizing for variable distance and size of the ROI by normalizing values to the initial activity in the ROI (4, 5). MacNee and coworkers (6) used similar methods recently to describe differences in the rate of clearance of In-PMN from the lungs of patients with chronic obstructive pulmonary disease after tobacco inhalation. They additionally corrected for lung blood volume by simultaneous use of 99mtechnetium-Iabeled red cells (Tc-RBC). However, their clearance curves closely resemble those that we have obtained in normal rats without correction for blood volume (4, 5), which may reflect the nearly IOO-fold increase in the PMN/RBC ratio reported to occur in the alveolar-microvascular bed of some species (7). Therefore, correction for lung blood volume may be unnecessary. Further, the rate of accumulation in downstream beds may
provide a useful second viewof the rate of clearance of lung activity (4, 5). Some controversy has developed regarding the importance of different methods of cell harvesting and labeling, which likely affects subsequent cell performance. Cells maintained in plasma may clear the lung more quickly than those separated and/or labeled in salt solutions (3). Although histologic observations suggest InPMN labeled in salt solutions distribute similarly to unhandled cells (7), it seems implausible to believe that any separation and labeling method could entirely avoid cell perturbation. Further, because autologous cells harvested clinically for labeling may be expected to be perturbed by the underlying inflammatory process (8), this debate may be overstated. Finally, Haslett and colleagues suggest that In-PMN may have clinical utility for tracking PMN involvement in lung disease; however, for a test to be useful clinically, it must distinguish normal subjects from abnormal subjects reliably. Unfortunately, overlap of unpaired normal and abnormal groups has been observed in virtually all studies of In-PMN to date. As noted above, the utility of normalizing lung activity to that of the injectate or other organs (e.g., spleen) requires confirmation by additional investigation. Regardless, until a reproducible and reliable method emerges, the clinical utility of In-PMN for tracking PMN involvement in lung injury will remain elusive. JAMES
H.
WILLIAMS, JR.,
M.D.
Assistant Professor Pulmonary and Critical Care Medicine University of California, Irvine Irvine, CA
1. Haslett C, Shen AS, Feldsien DC, Allen D, Henson PM, Cherniak RM. l11Indium-labeled neutrophil migration into the lungs ofbleomycin-treated rabbits assessed noninvasively by external scintigraphy. Am Rev Respir Dis 1989; 140:756-63. 2. Williams lH lr, Hartman TM, Lichter 1, Moser KM. Do indium 111-labeled polymorphonuclear leukocytes detect neutrophil-prominent, fibrosing lung disease in rats? 1 Lab Clin Med 1987; 110:55-62. 3. Saverymutto SH, Peters AM, Danpure Hl, Reavy Hl, Osman S, Lavender lP. Lung transit of lllIn-labeled granulocytes: relationship to labeling techniques. Scand 1 Haematol 1983; 30:151-60. 4. Williams lH Jr, Moser KM, Cairo MS. Harvesting the noncirculating pool of polymorphonuclear leukocytes in rats by hetastarch exchange transfusion (HET): yield and functional assessment. 1 Leukocyte BioI 1987; 42:455-62. 5. Williams lH, Wilson AF, Moser KM. Is lung sequestration of indium 111-labeled granulocytes organ specific? 1 Nucl Med 1989; 30:1531-7. 6. MacNee W, Wiggs B, Belzberg AS, Hogg lC. The effect of cigarette smoking on neutrophil kinetics in human lungs. N Engl 1 Med 1989; 321: 924-8. 7. Hogg lC, McLean T, Martin BA, Wiggs B. Erythrocyte transit and neutrophil concentration in the dog lung. 1 Appl Physiol 1988; 65:1217-25. 8. Zimmerman GA, Renzetti AD, Hill HR. Functional and metabolic activity of granulocytes from patients with the adult respiratory distress syndrome: evidence for activated neutrophils in the pulmonary circulation. Am Rev Respir Dis 1983; 127:290-300.
From the Author: A more detailed account of the construction of the lung ROI is given in an earlier publication (1)that was quoted in the abovementioned article (2). Reproducibility is aided by drawing organ regions on a clear plastic template and ensuring that the lung ROI excludes the liver dome. The liver outline, which retains a reasonably constant relationship with the lung in the prone anesthetized rabbit, can be used to orientate the template on subsequent scintigrams. 983
To the Editor: A recent article in the REVIEW by Haslett and colleagues (1) that reported the use of l11indium-Iabeledneutrophils (In-PMN) for detection of intratracheal bleomycin (ITB) injury in rabbits supports previous work from our laboratory in which we examined In-PMN in ITB-injured rats (2). The relatively noninvasive nature of this approach has obvious clinical appeal for following a variety of human diseases in which PMN are thought to playa role; however, the differences in specific methods used in these and related studies vary and raise some important questions of relevance to the potential utility of this approach in clinical medicine. One question of primary importance is how to define a region within images that is representative of lung activity. Haslett and colleagues counted activity in a region-of-interest (ROI) that overlaid the upper lungs (and body) and normalized activity to that of the injectate; however, the methods of defining the specific size of the region for counting were not precisely defined. Although strong correlation with lung lavage and histologic PMN accumulation was demonstrated, the subjectivity of this approach leavesroom for potential error or bias. In this regard, we found it very difficult to identify reproducibly a similar ROI in normal rats, which is consistent with the asymmetric shape of the lung and the accelerating (logarithmic) dissipation of activity measured at different distances from the lung. In contrast, in bleomycin-injured rats, we outlined the whole lung images visualized above the diaphragm, as well as liver and spleen images, to estimate the whole organ activity of each. Unlike Haslett and colleagues, we found that normalizing imaged lung activity to injectate activity did not enhance discrimination of ITB-injured animals, whereas normalizing lung activity for differences in spleen activity did. Further, lung/spleen ratios correlated strongly with the increase in PMN in lavage fluid. We suggested that lung/spleen ratios may normalize for differences not only in distance of the gamma counter from the imaged tissue but also in the percent of normally functioning PMN in the cell preparations. Preparations that accumulate in the spleen more than in the liver have been reported to be associated with less perturbation (3). However, this argument would not be plausible until abnormal cells had been cleared from the lung and, therefore, may not be applied until a late phase of imaging. We also noted a variable rate of clearance of various InPMN preparations from the lungs of normal rats, with discrimination from abnormal rats distinctly enhanced when each cell preparation was split for imaging of a normal and abnormal animal, with paired comparison. Although this approach has proven useful in some animal studies, a paired approach is not feasible for human studies. Therefore, alternatives must be considered. The rate of initial clearance of In-PMN from the lung may also be used to examine pulmonary vascular sequestration of PMN. We used this approach to study the kinetics of labeled cells in normal rats, standardizing for variable distance and size of the ROI by normalizing values to the initial activity in the ROI (4, 5). MacNee and coworkers (6) used similar methods recently to describe differences in the rate of clearance of In-PMN from the lungs of patients with chronic obstructive pulmonary disease after tobacco inhalation. They additionally corrected for lung blood volume by simultaneous use of 99mtechnetium-Iabeled red cells (Tc-RBC). However, their clearance curves closely resemble those that we have obtained in normal rats without correction for blood volume (4, 5), which may reflect the nearly IOO-fold increase in the PMN/RBC ratio reported to occur in the alveolar-microvascular bed of some species (7). Therefore, correction for lung blood volume may be unnecessary. Further, the rate of accumulation in downstream beds may
provide a useful second viewof the rate of clearance of lung activity (4, 5). Some controversy has developed regarding the importance of different methods of cell harvesting and labeling, which likely affects subsequent cell performance. Cells maintained in plasma may clear the lung more quickly than those separated and/or labeled in salt solutions (3). Although histologic observations suggest InPMN labeled in salt solutions distribute similarly to unhandled cells (7), it seems implausible to believe that any separation and labeling method could entirely avoid cell perturbation. Further, because autologous cells harvested clinically for labeling may be expected to be perturbed by the underlying inflammatory process (8), this debate may be overstated. Finally, Haslett and colleagues suggest that In-PMN may have clinical utility for tracking PMN involvement in lung disease; however, for a test to be useful clinically, it must distinguish normal subjects from abnormal subjects reliably. Unfortunately, overlap of unpaired normal and abnormal groups has been observed in virtually all studies of In-PMN to date. As noted above, the utility of normalizing lung activity to that of the injectate or other organs (e.g., spleen) requires confirmation by additional investigation. Regardless, until a reproducible and reliable method emerges, the clinical utility of In-PMN for tracking PMN involvement in lung injury will remain elusive. JAMES
H.
WILLIAMS, JR.,
M.D.
Assistant Professor Pulmonary and Critical Care Medicine University of California, Irvine Irvine, CA
1. Haslett C, Shen AS, Feldsien DC, Allen D, Henson PM, Cherniak RM. l11Indium-labeled neutrophil migration into the lungs ofbleomycin-treated rabbits assessed noninvasively by external scintigraphy. Am Rev Respir Dis 1989; 140:756-63. 2. Williams lH lr, Hartman TM, Lichter 1, Moser KM. Do indium 111-labeled polymorphonuclear leukocytes detect neutrophil-prominent, fibrosing lung disease in rats? 1 Lab Clin Med 1987; 110:55-62. 3. Saverymutto SH, Peters AM, Danpure Hl, Reavy Hl, Osman S, Lavender lP. Lung transit of lllIn-labeled granulocytes: relationship to labeling techniques. Scand 1 Haematol 1983; 30:151-60. 4. Williams lH Jr, Moser KM, Cairo MS. Harvesting the noncirculating pool of polymorphonuclear leukocytes in rats by hetastarch exchange transfusion (HET): yield and functional assessment. 1 Leukocyte BioI 1987; 42:455-62. 5. Williams lH, Wilson AF, Moser KM. Is lung sequestration of indium 111-labeled granulocytes organ specific? 1 Nucl Med 1989; 30:1531-7. 6. MacNee W, Wiggs B, Belzberg AS, Hogg lC. The effect of cigarette smoking on neutrophil kinetics in human lungs. N Engl 1 Med 1989; 321: 924-8. 7. Hogg lC, McLean T, Martin BA, Wiggs B. Erythrocyte transit and neutrophil concentration in the dog lung. 1 Appl Physiol 1988; 65:1217-25. 8. Zimmerman GA, Renzetti AD, Hill HR. Functional and metabolic activity of granulocytes from patients with the adult respiratory distress syndrome: evidence for activated neutrophils in the pulmonary circulation. Am Rev Respir Dis 1983; 127:290-300.
From the Author: A more detailed account of the construction of the lung ROI is given in an earlier publication (1)that was quoted in the abovementioned article (2). Reproducibility is aided by drawing organ regions on a clear plastic template and ensuring that the lung ROI excludes the liver dome. The liver outline, which retains a reasonably constant relationship with the lung in the prone anesthetized rabbit, can be used to orientate the template on subsequent scintigrams. 983
