Abstract
The name “basophil” derives from the staining characteristics of the cell’s granules with Romanowsky stains such as Diff Quik. Basophils are routinely identified in blood smears but seldom in histologic sections from companion or farm animals, so how do the granules of basophils stain with hematoxylin and eosin (HE)? In a report of basophilic enterocolitis in a horse, the granules of the basophils were described as “slightly amphiphilic.” 2 Basophils were detected in the skin of dogs in experimental flea allergy dermatitis; 1 however, the authors used Karnovsky’s fixative and a special processing technique to demonstrate the basophils because they knew that basophils could not be distinguished from other granulocytes in HE sections of formalin-fixed canine skin. Here, we detail what we believe are the features of basophils in HE sections of skin from a cat.
The 10-year-old Persian cat was presented with a 10-mm-diameter, ulcerated, pruritic lesion on the left pinna and a 20-mm-diameter similar lesion on the neck. An EDTA blood sample was drawn just before biopsy of both lesions. Imprints of the biopsy specimens were prepared before the biopsies were immersed in 10% neutral buffered formalin. The cat had peripheral blood basophilia (0.9 × 109/liter; reference range, 0–0.2 × 109/liter) without eosinophilia (0.5 × 109/liter; reference range, 0–1.5 × 109/liter). Imprints from both lesions, stained with Diff Quik (Fig. 1) had numerous basophils and only a few eosinophils (3–12 eosinophils per 100 basophils in five 400× fields). The histologic sections, stained with HE, had a marked dermal infiltration of granulocytes with eosinophilic granules that were assumed to be eosinophils.
The remarkable disparity between the cytologic findings and the histologic interpretation suggested that most of the eosinophilic granulocytes in the histologic sections must be basophils. Upon closer examination of the histologic sections, the granulocytes with eosinophilic granules had 2 distinct appearances (Fig. 2). The more common cell type had colorless cytoplasm with distinct eosinophilic granules and a multilobulated, or less commonly, bilobed nucleus. The less common cell type had mainly a bilobed nucleus and bright eosinophilic cytoplasm crowded with indistinct eosinophilic granules (Fig. 2). Although some cells were difficult to categorize, the ratio of granulocytes with colorless cytoplasm and distinct granules to those with bright eosinophilic cytoplasm and indistinct granules was never lower than 2:1 and generally ranged from 25:1 to 30:1. Our interpretation, therefore, was that the granulocytes with colorless cytoplasm, distinct eosinophilic granules, and more lobulated nuclei were basophils, whereas those with the bilobed nuclei, brightly eosinophilic cytoplasm, and crowded eosinophilic granules were eosinophils. Consistent with these observations, the basophils in the imprints generally had more lobulated nuclei than the eosinophils.
This case offered a unique opportunity for study because the basophil was the main leukocyte in the lesions. We believe that other veterinary anatomic or clinical pathologists as well as dermatologists might be as surprised and intrigued by the case as we were. It will require more study to determine whether basophils can be reliably differentiated from eosinophils in HE sections of cat skin, but, at the very least, this case suggests that granulocytes with eosinophilic granules in histologic sections of feline skin cannot always be assumed to be eosinophils. We hope our letter will inspire others with access to materials and techniques to study this further to refute or refine our observations.
Footnotes
Acknowledgement
Dr Keith Thompson, IVABS, Massey University took the photomicrographs.