In Vitro Stromal Co-Culture Model For The Evaluation Of Reovirus Oncolytic Therapy For Refractory Pediatric Leukemia

Acute lymphoblastic leukemia (ALL) is the most common malignant disease in children, constituting more than a quarter of all childhood cancers. The survival rate of standard/good risk ALL now exceeds 80%. In spite of the significant improvements in outcome, the emergence of resistant disease remains the most common cause of death. Escalation of the intensity of combination chemotherapy with the introduction of second-line drugs is accompanied by cumulative toxicity, with marginal incremental benefits. Hence novel therapeutic approaches are urgently needed to improve the outcome in these patients.

In addition to the inherent molecular alterations, recent reports have suggested an active role for the bone marrow stromal microenvironment in promoting both leukemia cell viability and drug resistance. Previous studies have shown a pivotal role for the stromal chemokine CXCL12 (SDF-1) and its receptor CXCR4 in this process. In the multiple myeloma model, inhibition of CXCR4 by the targeted agent AMD3100 (Plerixafor) abrogates the chemotherapy protection conferred by stromal cells.

Recent years have seen evidence for the effectiveness of oncolytic viruses as cytotoxic agents against refractory tumors. The strategy behind this approach is to develop viruses that can replicate and cause cell lysis specifically in cancer cells while leaving non-malignant cells unaffected. In this study we explore the utility of reovirus mediated anti-leukemic therapy in an experimental model of pediatric leukemia in the context of co-culture with bone marrow stromal cells.

A panel of leukemia cell lines (n=4) representing pediatric ALL and acute myeloid leukemia (AML) were evaluated when co-cultured in the presence of stromal conditioned medium, live bone marrow stromal cells, stromal cells fixed with 0.5% glutaraldehyde, or control human skin fibroblasts (1Br3 cells). After three days in culture, cell proliferation kinetics under each condition was measured by WST-1 assay. The influence of nonspecific leukemia cell attachment was evaluated by the culture of leukemia cells in poly-L-lysine coated plates. Next the leukemic cells were treated with oncolytic reovirus, at various concentrations with and without co-cultures conditions to evaluate the potential of leukemia cells to escape reovirus mediated cytolysis. Finally, the ability of stromal cells to modulate reovirus anti-leukemic activity was evaluated in the presence of various concentrations of AMD3100.

An average of a 2-fold increase in leukemia cell proliferation (in the absence of reovirus) was seen with live stromal co-cultures compared to conditioned media or fixed stromal cells, indicating the requirement of live cell-cell contact with stromal cells in this process. Poly-L lysine coated plates did not increase leukemia cell growth. Our data reveals that, as previously seen with chemotherapeutic drug exposure, the presence of stromal cells is able to decrease the cytolytic activity of reovirus on leukemic cells. Leukemia cells alone exhibit a survival of 10% following a three-day exposure to reovirus (50 MOI). However, in the presence of stromal the survival rate was increased to 40% (p < 0.05). Furthermore, the addition of AMD3100 provided partial restoration of the oncolytic activity (mean 15% survival, n=3 experiments). Stromal conditioned media also decreased reovirus infectivity, although to a lower extent (mean 17% cell survival). Co-culture of leukemia cells with control 1br3 human skin fibroblast cells provided no protection from reoviral cytolysis (mean 10% survival).

In this proof-of-concept study, we provide evidence for the first time that the bone marrow niche may provide at least partial protection from oncolytic virotherapy. This indicates the importance of evaluating the efficacy of oncolytic viruses in the context of the tumor microenvironment in future studies. Furthermore, we provide evidence for the consideration of combination therapies with targeted agents such as AMD3100, which may disrupt the stromal protective activities, to increase the effectiveness virotherapy in the treatment of refractory leukemia.

No relevant conflicts of interest to declare.