An 80-year old woman presented with cervical lymphadenopathy and was referred to a surgeon for excisional biopsy for her enlarged right cervical lymph node. Intraoperative consultation was suggestive of a hematopoietic neoplasm. The pathologist performing the intraoperative consultation sent a portion of the lymph node in Roswell Park Memorial Institute (RPMI) media to his flow cytometry reference lab. The reference lab received the tissue and demographic information, and no prior history of hematopoietic neoplasm was provided.
FLOW CYTOMETRIC STUDIES
A portion of lymph node tissue was received in a 50mL conical tube containing RPMI media. After touch imprints were made, the tissue was placed in a Petri dish containing Phosphate Buffered Saline (PBS-P) and minced. The suspension was filtered through a nylon mesh, and then washed twice with PBS-P. Following the washes, cells were counted with a hemacytometer, and an aliquot of the sample was stained with 50 μL of the antibody cocktail containing polyclonal lambda-Fluorescein Isothiocyanate (FITC), polyclonal kappa-Phycoerytherin (PE), CD14-PerCP, CD5-PE-Cy7, CD10-APC, CD20-APC-H7, CD19-V450, and CD45-V500 for 15 min in the dark at a 37°C. Red cells in the sample were lysed with ammonium chloride, and then washed once with PBS-P. The sample was collected on a Becton Dickinson (BD) Fluorescent Activated Cell Sorting (FACS) Canto II flow cytometer and analyzed using Kaluza v 1.2 software. FCS files and example analysis are available as Supporting Information. The results of immunophenotypic evaluation are presented in Figure 1. Viable events were inclusively gated based on FSC and SSC, and doublets were excluded using height versus width comparisons using both FSC and SSC (Fig. 1A). The key to analysis in this case is to include both B cell markers CD19 and CD20 in the panel and analyze. For example, if the cells are gated with a single B cell marker, such as those cells with bright CD20, a lambda restricted B cell population is identified (Fig. 1B). This is an atypical population, though it only represents part of the case. Similarly if only CD19 is used to gate the B cells, a kappa restricted B cell population is identified (Fig. 1C). It is also an atypical population, though it incompletely represents all the elements of the node. Using CD19 and CD20 in combination allows the interpreter to identify two distinct B cell populations—one that is CD20+CD19dim and another that is CD20dimCD19+ (Fig. 1D). However, a Boolean gate that includes all CD19 and CD20 positive events demonstrate a mix of kappa and lambda expressing cells (Fig. 1D). Color gating is a useful strategy to distinguish these two populations, and in this portion of the analysis, we now incorporate CD5. Color gating of the first B cell population (33% of leukocytes) is performed by selecting the population that has bright CD20 but lacks CD5 (Fig. 1E). Subsequent analysis demonstrates that this population has dim CD19, is lambda monotypic, and has increased FSC (Fig. 1E). Additional studies (data not shown) demonstrate that this population expresses partial CD38 and CD79b, but lacks CD23. Color gating of the second B cell population (49% of leukocytes) is performed by selecting the population that has dim CD20 and aberrantly coexpresses CD5 (Fig. 1F). Subsequent analysis demonstrates that this population expresses CD19 at a level comparable to normal B cells, is kappa monotypic, and does not have appreciably increased forward scatter (Fig. 1F). Additional studies (data not shown) demonstrate that this population expresses CD23, but lacks CD38 (dim to absent) and CD79b. Our final flow cytometry interpretation was that the first population represented a diffuse large B cell lymphoma (DLBCL) and the second population represented a chronic lymphocytic leukemia/small lymphocytic lymphoma (CLL/SLL), though we were unsure of the relationship between the two populations. Fluorescent in situ Hybridization (FISH) was performed on a portion of the cell suspension not used for flow cytometry and was negative for a t(11;14), arguing against a mantle cell lymphoma. About 25.5% of the cells had only one MYB signal (deletion of 6q), 14% had only one D13S319 signal, and 8% had only one TP53. The referring pathologist sent us an H&E recut of one of the tissue blocks. The morphology is consistent with immunophenotypic analysis of the populations—sections of large B cell lymphoma are juxtaposed near sections of small lymphocytic lymphoma (Fig. 2).
This case was selected for presentation and discussion to illustrate key principles of immunophenotypic analysis of B cell neoplasms. While many large flow cytometry labs have extensive multiple tube or multicolor flow cytometry panels to screen lymph node and other material for atypical B cell populations, some of the smaller labs in our area use more limited panels that contain a single four color tube containing CD2/CD19/KAPPA/LAMBDA. We routinely use more than one B cell antigen in our tubes to evaluate B cell neoplasms, as the use of a single lineage associated antigen may not always detect a neoplastic population or all populations present (as in this case). Often this can be because the abnormal population loses surface expression of that antigen.
As with any case in hematopathology, the clues in this case are to distinguish normal from abnormal populations. Things to look for include different expression of lineage-defining antigens, aberrant expression of antigens, and abnormal light scatter. While CD19 is thought to be a very useful lineage defining antigen for B cells, decreased or absent CD19 expression can be seen in B cell malignancies (). Decreased or absent CD19 is more common in follicular lymphoma or DLBCL, but aberrant decrease or loss can occur in a variety of B lineage neoplasms ([1, 2]). CD20 is also typically seen in mature B cells, though decreased CD20 expression is a classic immunophenotypic aberrancy of the neoplastic B cells comprising a CLL/SLL clone (). In this case, simultaneously looking at CD19 and CD20 identified distinct populations—the first CD20+CD19dim and another that is CD20dimCD19+. These populations were further distinguished based on increased FSC in one population, aberrant CD5 expression by the other population, and monotypic light chain expression in each population.
Richter syndrome (RS) describes the development of an aggressive lymphoma in the context of CLL/SLL ([3, 4]). Upon discussion of the case with the pathologist reviewing the lymph node, it was mentioned that the patient did have a remote history of CLL/SLL. Thus, the clinical impression, with the new diagnosis of a DLBCL, was that of RS. Approximately, 2–8% of patients with CLL/SLL develop DLBCL, and less than 1% develop classical Hodgkin lymphoma (). A recent study of 86 patients with pathologically proven RS who developed DLBCL found that TP53 disruption (47.1%) was the most frequent genetic alteration (by mutation of TP53 or deletion of 17p13), and that its disruption was a poor predictor of RS survival (). In addition, this study also looked at the clonal relationship (by molecular analysis of IgH VDJ rearrangements) of the original CLL/SLL and the subsequent DLBCL. Paired data were available for 63 patients. The RS lymphoma was clonally related to the CLL phase in 79.3% of cases and unrelated in the remaining 20.6% of cases (). Interestingly, patients with a clonally unrelated RS neoplasm had significantly longer survival (median, 62.5 months) compared to those with clonally related transformation (median, 14.2 months) (). While the two B cell populations identified in the node of this patient had different light chains inferring that they are not related, their clonal relationship, as determined by B cell receptor gene rearrangement studies, is unclear. The cytogeneticist who interpreted the FISH studies did conclude that the presence of smaller populations of cells with loss of D13S319 and TP53 may reflect two discrete populations of neoplastic cells or may represent cytogenetic evolution that can be associated with disease progression. Ideally, were our lab to have a flow cytometric cell sorter for clinical use, we could sort the two populations and perform parallel molecular studies to identify whether the two populations harbored clonotypic B cell receptor gene rearrangements. Finally, as both B cell neoplasms were found in the same node, one could consider the possibility of a composite lymphoma. We do not favor classifying this patient's neoplasm as a composite lymphoma, however, for while they were detected in the same anatomic site, the history of CLL/SLL preceded the DLBCL. Moreover, literature describing composite lymphoma does not describe this sequence, and a recent flow cytometry study of composite lymphomas described cases involving DLBCL, but not CLL/SLL ().
In summary, this is a case of RS of a patient with a history of CLL/SLL developing DLBCL—both were distinct and detected by multicolor flow cytometry. The essentials of this case are threefold. First, it was helpful to have both CD19 and CD20 in the panel as B lineage antigens to identify both populations. Second, having a multicolor panel that included CD5 in addition to the B cell antigens helped to distinguish the two atypical populations. Finally, cases of RS can be detected immunophenotypically, though the CLL/SLL and more aggressive neoplasm may have distinct light chain restriction, as in this case.
CASE 4 DIAGNOSIS: Richter Syndrome Characterized by Diffuse Large B Cell Lymphoma in the Context of Prior Chronic Lymphocytic Leukemia/Small Lymphocytic Lymphoma.