Abstract 1699

Poster Board I-725

Many malignancies such as myeloid leukemia, breast, colon, and CNS tumors appear to harbor phenotypically and functionally distinct populations of cells with self-renewal, differentiative, and tumorigenic capacity. These unique populations, influencing tumor growth and persistence, have been termed “cancer stem cells,” or “tumor-initiating cells.” No such tumor-initiating capacity has been prospectively identified in non-Hodgkin lymphoma (NHL). Identification of these novel sub-populations in B-NHL is particularly challenging given the unique proliferative capacity, differentiation potential, and population diversity of the peripheral B cell subsets from which lymphomas arise. Our studies have focused on the identification of sub-populations that meet criteria for “lymphoma stem cells” (LySC) in B-NHL samples. Two experimental sources of NHL cells were used: i) primary human DLBCL (TXN190, TXN131, C10, C12) derived at our institution from malignant effusions, and, ii) continuously cultured, widely used DLBCL lines (OCI-Ly-7, -10, -19).

Using multiparameter FACS, we initially performed immunophenotypic studies to investigate a broad range of cell surface markers. Notably, we observed significant heterogeneity in the surface antigen density for CD22, CD27, CD45R/B220, and CXCR5. These data suggest that subpopulations residing within NHL populations may possess functional heterogeneity. To test this concept, we first performed colony-forming unit (CFU) assays in methycellulose medium. Analysis of 7 specimens shows a CFU frequency ranging from .0003% to 10% of the overall population, demonstrating functional heterogeneity with regard to in vitro colony-forming ability. Surprisingly, we found that generation and successful re-plating of B-NHL colonies did not require the addition of exogenous factors such as PHA-conditioned medium, cytokine, or interleukin supplementation. Next, we performed flow cytometric separation of candidate subpopulations as a means of segregating functional potential. These studies showed that CD45R/B220 expression levels appear to correlate with colony-forming ability. B-NHL cell fractions with lower levels of expression of the murine pan-B cell marker CD45R/B220 appear to be enriched for CFU activity in methylcellulose culture compared to B220-high cells. In addition, CD45R/B220-low B-NHL has higher tumorigenic potential (assessed by kinetics of growth rate) in assays of subcutaneous tumor nodule generation. These studies indicate differential potential of NHL subpopulations, and may begin to establish a phenotype of LySC. Finally, we measured in vitro clonogenic potential using two assays – i) limiting dilution suspension culture, and - ii) CFU assays. Interestingly, the results from these two independent assays do not correlate well in B-NHL; clonogenic potential appears to be substantially higher in limiting-dilution suspension cultures when compared to methylcellulose assays. These findings indicate that the relationship between CFU frequency and growth potential in B-NHL populations may differ from those found in normal hematopoietic cell types, and suggest that some caution should be exercised in the interpretation of B-NHL CFU assays.

The adaptation of assays useful for identification of tumor-initiating sub-populations has also been useful in defining subsets of clonally related B-NHL cells that appear immunophenotypically and functionally distinct in vitro and in vivo. Our studies support the notion that B-NHL tumor-initiating cells may comprise a unique population amenable to prospective immunophenotypic and/or functional identification. Implicit in the concept of these putative LySC is that the control, and ultimately the eradication, of such unique NHL populations may result in improved clinical outcomes.

[SAW and RMR contributed equally to this work. Funded by the Lymphoma SPORE, P50 CA130805501A1, PI: Richard Fisher, MD.]


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.