Detection of Lymphoid Markers (CD3 and PAX5) for Immunophenotyping in Dogs and Cats: Comparison of Stained Cytology Slides and Matched Cell Blocks

SIMPLE SUMMARY: The B- or T-cell phenotype of a canine or feline lymphoma is relevant for treatment selection and prognosis. This phenotype is determined by complementary techniques and there is some evidence that the previously stained cytology (RSC) slides used for routine diagnose of lymphoma are suitable for assessing B and T-cell phenotype by immunocytochemistry. Still, immunodetection of the lymphoid markers can also be achieved with cell blocks (CB) prepared from effusions fluids or needle rinses. A comparison between these two RSC and CB is missing in the literature. In this study, lymphoid markers (CD3 and PAX5) were simultaneously studied in RSC and matched CB from 53 lymphomas and 4 chylous (lymphocyte-rich) effusions from dogs and cats. The influence of pre-analytical variables (species, time of archive, type of specimen, and coverslipping) and the interobserver agreement were assessed. Fewer CD3+ lymphocytes were identified in RSC, but the PAX5 positivity in RSC and CB had a substantial agreement. Immunodetection of CD3 and diagnosing a T-cell population on RSC were more difficult. Immunophenotyping was inconclusive in 54% RPSC and in 19% CB. The interobserver reproducibility of immunophenotyping on CB was substantial and higher than in RSC. The performance of RSC from effusions and feline samples was unsatisfactory. The detection of lymphoid markers on RSC was affected by pre-analytical variables, whilst CB was more consistent for assigning a lymphoma phenotype. ABSTRACT: Immunolabeling on Romanowsky-stained cytology (RSC) slides can be used, although there is limited evidence of its suitability for phenotyping canine and feline lymphomas. A comparison with matched cell blocks (CB) is missing. Immunolabeling on RSC and CB was compared for lymphoid markers (CD3 and PAX5) in 53 lymphomas and 4 chylous effusions from dogs and cats. The influence of pre-analytical variables (species, time of archive, type of specimens and coverslipping) and the interobserver agreement among the 2 observers was assessed. Fewer CD3+ lymphocytes were identified in RSC, while the PAX5 positivity by RSC and CB had a substantial agreement. Immunodetection of CD3 and the diagnosis of a T-cell population on RSC was more difficult. Lower intensity and higher background were noted in RSC. Immunophenotyping was inconclusive in 54% RSC and 19% CB. The interobserver reproducibility of immunophenotyping on CB was substantial, being higher than in RSC. The immunolabeling performance on the RSC of effusion and feline samples was unsatisfactory. The detection of lymphoid markers, especially membranous antigens in retrospective RSC, is affected by the pre-analytical variables: species, time of the archive, and type of specimens. CB are a more consistent type of sample for immunophenotyping purposes.