RESUMO
The presence of the histone 3 (H3) K27M mutation in diffuse midline glioma has implications for diagnosis, prognosis, and treatment, making rapid and accurate H3 K27M characterization vital for optimal treatment. This study evaluated an immunohistochemical assay using a commercially available monoclonal anti-H3 K27M in human central nervous system tumors. H3 K27M-positive glioma specimens were obtained from clinical sites with prior H3 K27M testing using local methods; negative control glioblastoma tissue was obtained from a tissue library. Specimens were stained with a rabbit anti-H3 K27M monoclonal antibody; slides were evaluated for the proportion of H3 K27M-positive tumor cells and staining intensity by a board-certified pathologist. H-score was calculated for each sample. Sensitivity, specificity, accuracy, repeatability, and reproducibility were evaluated. Fifty-one central nervous system specimens were stained (H3 K27M, n=41; H3 wild type, n=10). All H3 K27M-mutant specimens had positive nuclear staining, and most specimens had an H-score ≥150 (31/40, 77.5%). No nuclear staining occurred in H3 wild-type specimens; all cores in the normal tissue microarray were negative. Results were 100% sensitive, specific, and accurate for H3 K27M detection relative to local methods. Repeatability and reproducibility analyses were 100%, with a high degree of concordance for staining intensity. H3 K27M antigen was stable for at least 12 months at ambient temperature. Immunohistochemistry using a commercially available anti-H3 K27M monoclonal antibody provides a highly sensitive, specific, and stable method of establishing H3 K27M status in human glioma; this method may facilitate diagnosis in cases where sequencing is not feasible or available.