TO THE EDITOR: We read with great interest the article by Harris et al. (1) demonstrating that 18F-FDG uptake rate (Ki, min−1) precisely predicts the degree of eosinophilic inflammatory response in the lungs of asthmatic patients undergoing an allergen challenge. Any result in this field is of tremendous interest because it may provide complementary quantitative insight into the severity of the inflammatory response in asthma and thus may be used for testing innovative therapies. The study of Harris et al. used the gold standard graphical method of Patlak et al. involving dynamic acquisition until about 60 min after injection and venous blood sampling (2,3), allowing determination of the ratio of Ki in allergen-challenged and diluent-control lobe, which was compared with that of eosinophil counts obtained with bronchoalveolar lavage.
Because we have proposed an adapted estimation of 18F-FDG metabolism that can be used for lung examination in clinical practice (4), we suggest that the standardized uptake value (SUV) assessed at late imaging may be used to estimate eosinophilic inflammation in asthma. Let us consider SUV at a late imaging time of t > 60 min:
As a consequence, the ratio of SUV and Ki in allergen-challenged and diluent-control lobe (subscripts 1 and 2, respectively) is the same because f(t) and W are the same:
Harris et al. (1) found that the latter ratio was strongly correlated with eosinophil count ratio (P < 0.001). Furthermore, it should be emphasized, first, that correction for 18F physical decay usually applied to SUV is cancelled out in the former ratio; second, that the use of mean SUV over lung lobe area (or volume) reduces variability in comparison with that of maximal SUV; and importantly, third, that assessing SUV ratio instead of Ki ratio avoids a time-consuming dynamic acquisition and invasive blood sampling.
In conclusion, the method of Harris et al. (1) clearly demonstrates that PET of 18F-FDG uptake rate Ki may serve as a reliable biomarker of eosinophilic inflammation and its functional effects in asthma. In that framework, the current model analysis further suggests that simple SUV may play a similar role when assessed at late imaging. Analysis of SUV data obtained from the last dynamic frames in the study of Harris et al. would give some indication as to whether this suggestion may be tested in a clinical study.
Footnotes
Published online Jun. 1, 2012.
- © 2012 by the Society of Nuclear Medicine and Molecular Imaging, Inc.