Hematopathology / MARROW INVOLVEMENT IN NASAL NK CELL LYMPHOMA

To look for subtle evidence of marrow involvement in nasal NK cell lymphoma at diagnosis, we retrospectively studied trephine biopsy specimens from 25 consecutive patients by 2 sensitive techniques: CD56 immunohistochemistry and Epstein-Barr virus–encoded RNA in situ hybridization (EBER ISH). Only 2 patients had marrow involvement by NK cell lymphoma at diagnosis. In 3 additional patients, marrow involvement developed during or after systemic recurrence. All 5 positive cases were revealed by EBER ISH, but only 3 cases showed CD56 immunoreactivity. Among the 5 cases, only 2 were recognized by morphologic assessment. All 5 patients died, often within a short period, compared with a mortality of 50% for patients without demonstrable marrow involvement. Marrow involvement is distinctly uncommon in nasal NK cell lymphoma at diagnosis, and EBER ISH is the most sensitive technique for the demonstration of occult NK cell lymphoma. Despite the low frequency of marrow involvement in nasal NK cell lymphoma, EBER ISH is worthwhile to identify the minor subgroup of patients with a high likelihood of early death due to disease and when autologous bone marrow or peripheral blood stem cell transplantation is contemplated. NK cell lymphomas are a group of recently characterized hematolymphoid malignant neoplasms comprising at least 3 overlapping categories: nasal NK cell lymphoma, nasal-type (extranasal) NK cell lymphoma, and aggressive NK cell lymphoma/leukemia.1,2 Despite differences in topographic predilection, these categories share many similarities in morphologic features, immunophenotype (CD2+, surface CD3–, cytoplasmic CD3+, CD56+), and genotype (germline T-cell receptor genes, strong association with Epstein-Barr virus [EBV]).3 Nevertheless, the majority of patients with nasal NK cell lymphomas have localized disease at diagnosis (stage I or II) while only approximately one fifth of patients with nasal-type NK cell lymphoma have stage I disease.1,2 Marrow involvement is distinctly uncommon in the former (0%-2%), but more common in the latter (15%-25%).1,2,4 It is, however, well known that in malignant lymphoma, marrow biopsy specimens interpreted on morphologic grounds as “negative for lymphoma” may harbor occult lymphoma cells demonstrable only by more sensitive immunologic or molecular techniques.5-8 Since the outcome of patients with nasal NK cell lymphoma is poor,2-4 with a high proportion of patients eventually developing additional sites of involvement, we question whether the disease might be more disseminated at diagnosis than that revealed by conventional staging (including morphologic evaluation of marrow). Since the bone marrow is a frequent site of dissemination in malignant lymphoma, analysis of marrow involvement provides a means to address this issue. We therefore used CD56 immunohistochemical and EBV-encoded RNA (EBER) in situ hybridization (ISH) techniques to determine the frequency of marrow involvement by nasal NK cell lymphoma at diagnosis. These markers were chosen because they are expressed commonly in nasal NK cell lymphoma but almost never in normal hematopoietic cells.9,10 Hematopathology / ORIGINAL ARTICLE Am J Clin Pathol 2001;115:266-270 267 © American Society of Clinical Pathologists Materials and Methods