PAX immunoreactivity identifies alveolar rhabdomyosarcoma.

PAX5 is a member of the paired box transcription factors involved in development and its expression has been well characterized among hematopoietic malignancies of B-cell lineage. Its expression has also been reported in a subset of neuroendocrine carcinomas, urothelial tumors, Merkel cell carcinoma, glioblastoma, and neuroblastoma cell lines. As such, we sought to assess it as a diagnostic marker in the evaluation of pediatric small round blue cell tumors. Tumors selected for evaluation included embryonal rhabdomyosarcoma (55 cases), alveolar rhabdomyosarcoma (ARMS) (51 cases), neuroblastoma (22 cases), Wilms tumor (18 cases), Ewing Family of Tumors (11 cases), lymphoblastic lymphoma (8 cases), hepatoblastoma (6 cases), and granulocytic sarcoma (3 cases) as either cores in a tissue microarray or whole mount sections. All cases were immunostained using an antibody directed toward PAX5 and immunoreactivity was scored semiquantitatively according to percentage of nuclear staining. As expected, all B-cell lymphoblastic lymphomas were strongly immunoreactive against PAX5. Additionally, all Wilms tumors showed staining of variable intensity, most intensely in the epithelial component. Of the rhabdomyosarcoma cases, 34 of 51 (67%) ARMS were immunoreactive whereas none of the 55 embryonal rhabdomyosarcoma cases stained. No other tumor type on the array was immunoreactive toward PAX5. Genetic information was available on 7 ARMS, 5 of which had characteristic translocations involving PAX genes, either t(2:13) or t(1;13). Of the translocation-positive cases, all showed nuclear reactivity toward PAX5, and both the translocation-negative cases did not. Possible explanations of PAX5 staining include aberrant expression of the PAX5 transcription factor, PAX5 expression in normal tissue at the time the tumors most closely recapitulates in development or crossreactivity with another member of the PAX family. PAX3 and PAX7 fusion genes characterize the majority of ARMS making crossreactivity with these proteins an attractive theory, and suggest that PAX5 immunoreactivity may be specific for translocation-positive ARMS. Further study in a larger series of rhabdomyosarcomas is warranted to assess the sensitivity and specificity of PAX5 immunoreactivity for the ARMS variant.