Serum Soluble CD86, Still a Prognostic Factor in the Novel Agent Era in Multiple Myeloma Patients, Is Produced By Myeloma Cells with High CD86 Variant 3 Expression

Introduction: We previously reported that CD86 expressed on tumor cells from multiple myeloma (MM) patients is associated with a proliferative advantage of tumor cells and suppresses the antitumor immune response by inducing the immunosuppressive cytokine IL-10 from CD4⁺ T cells via CD86-CD28 interaction [Yamashita T, Clin Cancer Res 2009]. Furthermore, CD28 expressed on MM cells can mediate pro-survival signaling through CD86-CD28 interaction, resulting in chemotherapeutic resistance [Murray ME, Blood 2014]. Hock et al. reported that serum soluble CD86 levels (sCD86) were a significant independent prognostic marker in MM patients treated with conventional chemotherapy [Br J Haematol 2006]. However, it is unknown whether serum soluble CD86 is still a prognostic marker in the novel agent era and how sCD86 is produced in serum. In this study, we investigated the association of clinical characteristics and prognosis with serum sCD86 and elucidated the mechanism by which sCD86 is produced in MM patients.

Materials and Methods: 1) Peripheral blood and bone marrow (BM) samples were obtained from 294 newly diagnosed (42 asymptomatic and 252 symptomatic) MM patients, 16 patients with monoclonal gammopathy of undetermined significance (MGUS), and 16 healthy controls. 2) sCD86 levels were measured using ELISA. 3) The expression of cell-surface antigen on plasma cells identified as CD138-positive and CD38-strong positive cells were analyzed using flow cytometry. Expression of the CD86 full-length (variant 1) and alternatively spliced variant deleting exon 6 (variant 3), which encodes for the transmembrane domain, in MM cell lines and CD138⁺ plasma cells isolated from BM samples from MM patients were analyzed using real-time PCR.

Results: 1) sCD86 levels were significantly increased in MM patients compared with MGUS patients and healthy controls. Among MM patients, the levels were significantly higher in symptomatic than in asymptomatic patients and markedly increased in advanced stages (International Staging System [ISS] stage II/III and revised ISS II/III). We next investigated the differences in clinical characteristics between two groups according to serum sCD86 levels: high (≥2.16 ng/mL); and low (<2.16 ng/mL). In the high group, plasma cell percentages and corrected calcium, creatinine, and β2-microglobulin levels were markedly higher, and hemoglobin, hematocrit, and estimated glomerular filtration rate significantly lower than those in the low group. Furthermore, MM patients in the high group had significantly shorter overall survival (OS) times than those in the low group (hazard ratio, 0.385; 95% confidence interval, 0.1572-0.9425; P=0.032). 2) CD86 variant 1 and variant 3 were found to be expressed in plasma cells from MM patients. The mRNA levels of CD86 variant 3 in the sCD86-high group were significantly increased in comparison with the low group. These levels were significantly correlated with the concentration of sCD86 (r=0.635, P=0.0003). High levels of CD86 variant 3 mRNA were detected in primary MM cells, but not in hematopoietic BM cells, although the expression of CD86 variant 1 was the same in both MM and hematopoietic cells. On the other hand, sCD86 in cell cultures and mRNA of CD86 variant 3 were not detected in CD86-expressing MM cell lines, suggesting that these phenomena may be seen only in primary MM cells from patients. 3) There was a weak positive correlation between serum sCD86 concentration and CD86 expression levels on MM cells. However, progression-free survival (PFS) and OS were not significantly different between MM patients with high and low expression of CD86. Further, we confirmed that recombinant human CD86 could bind to MM cell lines through CD86-CD28 interaction, which might induce aggressive disease with chemotherapeutic resistance in MM cells. Consistent with this hypothesis, patients with low expression of sCD86/its receptor CD28 had longer OS times than others. In addition, receptor CD28 expression in patients with high CD86 expression was significantly higher than in those with low expression levels.

Conclusions: We demonstrated that sCD86 is produced via expression of CD86 variant 3 in primary MM cells and that sCD86 is a prognostic marker in the novel agent era. These results suggest that serum sCD86 levels may reflect disease progression through CD28 expressed on myeloma cells and be a useful prognostic factor.