Abstract
The assessment of B-cell clonality is a critical component of the evaluation of suspected lymphoproliferative disorders, but analysis from formalin-fixed, paraffin-embedded tissues can be challenging if fresh tissue is not available for flow cytometry. Immunohistochemical and conventional bright field in situ hybridization stains for kappa and lambda are effective for evaluation of plasma cells but are often insufficiently sensitive to detect the much lower abundance of light chains present in B-cells. We describe an ultrasensitive RNA in situ hybridization assay that has been adapted for use on an automated immunohistochemistry platform and compare results with flow cytometry in 203 consecutive tissues and 104 consecutive bone marrows. Overall, in 203 tissue biopsies, RNA in situ hybridization identified light chain-restricted B-cells in 85 (42%) vs 58 (29%) by flow cytometry. Within 83 B-cell non-Hodgkin lymphomas, RNA in situ hybridization identified restricted B-cells in 74 (89%) vs 56 (67%) by flow cytometry. B-cell clonality could be evaluated in only 23/104 (22%) bone marrow cases owing to poor RNA preservation, but evaluable cases showed 91% concordance with flow cytometry. RNA in situ hybridization allowed for recognition of biclonal/composite lymphomas not identified by flow cytometry and highlighted unexpected findings, such as coexpression of kappa and lambda RNA in 2 cases and the presence of lambda light chain RNA in a T lymphoblastic lymphoma. Automated RNA in situ hybridization showed excellent interobserver reproducibility for manual evaluation (average K=0.92), and an automated image analysis system showed high concordance (97%) with manual evaluation. Automated RNA in situ hybridization staining, which can be adopted on commonly utilized immunohistochemistry instruments, allows for the interpretation of clonality in the context of the morphological features in formalin-fixed, paraffin-embedded tissues with a clinical sensitivity similar or superior to flow cytometry.