Conflicts of interest: none declared. Sir, In a recent issue of this Journal, Petrella et al.1 reported a ‘surprising’ case of blastic natural killer (NK)‐cell lymphoma that does not express CD56. Although they state that theirs is the first reported case of CD56– blastic NK‐cell lymphoma (agranular CD4+/CD56+ haematodermic neoplasm), we have previously reported two similar cases as cutaneous lymphoblastic lymphoma of putative plasmacytoid dendritic cell precursor origin.2 Both of our patients were elderly Japanese men who initially presented with cutaneous tumours that showed lymphoblastic/plasmacytoid morphology.2 Patient 1 later developed bone marrow involvement with leukaemia, and both patients died within 1 year.2 Although in our earlier report2 we did not show the detailed immunophenotype of the tumour cells (including CD123 expression), and stated that CD4 expression was not detected by flow cytometry of the minced cutaneous tumour cells in patient 2, we have now confirmed by immunohistochemistry that in both cases the cutaneous tumours exhibited a CD4+, CD56–, CD123+, CD45+, CD43+, HLA‐DR+ phenotype. In contrast to the case reported by Petrella et al.,1 our cases did not express CD2 or CD68.2 However, our cases were also ‘lineage negative’: thus, they lacked conventional B (CD19, CD20, CD79a), T (CD3, CD5), NK (CD57) and myeloid (CD13, CD33, myeloperoxidase) markers.2 Both cases were terminal deoxynucleotidyl transferase‐positive, CD34– and, interestingly, CD10+, but showed no clonal rearrangement in immunoglobulin and T‐cell receptor genes.2
No takes yet. Share an insight, caveat, or question.
Kazuhiro Kawai (2005) studied this question.