Mantle Cell and Follicular Lymphoma Samples Demonstrate Differing Sensitivity to Bortezomib in a Primary Culture System.

The activity of the proteasome inhibitor bortezomib has been evaluated in follicular lymphoma (FL) and mantle cell lymphoma (MCL) patient samples using a modification of a primary lymphoma culture system based on the CD40-CD40L interaction (Johnson et al, Blood. 1993 5;82(6):1848).

Methods: Single B-cell suspensions were isolated from lymph nodes, ascites or peripheral blood from 12 patients with histologically confirmed FL or MCL. After disaggregation and/or density gradient separation, cells were plated at a density of 5x10⁵ cells/well into 96-well plates containing CHO cells transfected with the CDw32 Fc receptor to express the CD40 ligand, which, when irradiated to prevent further proliferation, provide a stromal layer supporting B-cell proliferation. Malignant or normal B-cells were cultured in IMDM medium supplemented with 10% human serum and 2ng/ml IL-4, and could be maintained for up to 8 days whilst retaining close phenotypic resemblance to the original sample, confirmed by flow cytometric quantitation of cell surface markers. After 24hrs drug was added at varying concentrations in triplicate and cell number and viability were assessed after a further 48hrs using the trypan blue exclusion assay. EC₅₀ values for % viability were derived for each sample using Graphpad Prism.

Results: MCL cells were significantly more sensitive to bortezomib than FL cells (median EC₅₀ 209nM range (122–989nM) n=5, vs 1311nM (153–2211nM) n=8 respectively, Mann-Whitney U-test P<0.05). In contrast, there was no difference in sensitivity to doxorubicin between MCL and FL cells (median EC₅₀ 4.1μM range (2.0–6.3μM) vs 2.3μM(1.5–5.5μM)). Three of the MCL samples were from patients subsequently treated with bortezomib in a phase II clinical trial. In vitro sensitivity of these samples correlated with clinical response. Patient 1, with an in vitro EC₅₀ of 122nM, achieved a complete response (unconfirmed), following 6 cycles of treatment. Patient 2, with an in vitro EC₅₀ of 179nM, initially responded to bortezomib but following two dose reductions, progressed at the end of therapy. The in vitro EC₅₀ had increased to 325nM. A third patient with an in vitro EC₅₀ of 377nM, did not achieve a clinical response to bortezomib. Two patients with FL, both with an EC₅₀ value of 2.2μM, progressed on treatment. CD40 expression in 8 MCL patient samples was measured using flow cytometry. This showed a significant correlation with bortezomib sensitivity (r=−0.92, p=0.002, Spearman’s correlation coefficient vs bortezomib EC₅₀) that was independent of cell proliferation (cell count after 3 days vs CD40 receptor r=−0.34, p=0.43). This correlation between CD40 expression and bortezomib sensitivity was not observed in 5 FL samples (r=−0.5, p=0.45).

Conclusion: These data demonstrate a significant difference in bortezomib sensitivity between MCL and FL, which may be related to differences in CD40 signalling. The difference in in vitro sensitivity correlates with clinical response in 5 patients and suggests the possible use of this system as an in vitro predictor of response to bortezomib treatment.