Intact Cell Differential Count on a Single Occasion Underestimates Sputum Eosinophilic Activity

RATIONALE: On Wright-stained sputum cytospins, eosinophil differential of >1.2% is considered abnormal and 3% is generally considered clinically relevant, identifies airway eosinophilia when blood eosinophil counts are within normal limits (particularly in severe asthmatics), and is used to define an "eosinophil phenotype". We hypothesized that failure to take into consideration free eosinophil granules (FEG) identified by light microscopy, and the re-emergence of eosinophilia (3%) during subsequent exacerbations or when neutrophilia resolves or with steroid reduction may underestimate the prevalence of the eosinophilic phenotype. METHODS: This is a retrospective analysis of our Institutional Review Board-approved clinical database of 24,176 sputum examinations of which 17,693 had viable cell counts (the rest being just saliva in the expectorate) represented from 9570 patients (6604 had sputum examined on one occasion, 2966 on more than one occasion) with various airway diseases. 2415 of these samples had excessive cell degeneration and therefore accurate differential count could not be made. But in all samples, FEGs were examined and semi-quantified to estimate the prevalence of granules in those patients who may not have an intact cell differential or an eosinophil <3%. In those patients who had sputum examined on more than one occasion, follow up results were examined to identify re-emergence of eosinophil 3%. RESULTS: Of those with intact cell counts (n=15,278), 11,282 (73.8%) would have been classified as normal or clinically non-relevant eosinophilia of <3%. However, 1298 (11.5%) of these had FEGs indicating clinically relevant eosinophilia. Additionally, 1177 samples with excessive cell degeneration showed FEGs. Further, among those with two or more intact cell counts, 998 samples (33.6%) had re-emergence of eosinophilia either when a previous neutrophilia resolved (n=165 (16.5%) or due to a reduction in the dose of corticosteroids (n=236 (23.6%). CONCLUSIONS: A total of 3473 (1298+1177+998) samples (of the 11,282 identified as non-eosinophilic), ie. a third, had clinically relevant airway eosinophilia that would not have been identified if the phenotypic classification was limited to 3% intact eosinophils on one occasion. If the threshold is lowered to 2%, and other airway eosinophilic activity includes macrophage granular inclusions, Charcot-Leyden crystals, and eosinophil traps, this underestimation is likely to be significantly more. In addition, mast cell activity, that is not currently quantified by sputum cytometry, may also contribute to the underappreciation of T2 inflammatory process in the airway.