Non-obese diabetic-severe combined immunodeficiency (NOD/SCID) mice bearing patient-derived xenografts (PDX) represent highly relevant models for evaluating novel therapeutic approaches1,2. However the expense and time frame of these experiments requires prioritisation of agents for testing. An in vitro system that mimics in vivo response would facilitate prioritisation of agents and may also provide an in vitro tool to predict individual patient responses in a timely fashion.
The curaxin CBL0137, a novel compound that activates p53 and suppresses NFκB protein activity in cancer cells, has entered clinical trials for adult cancers. Recently, the Paediatric Pre-Clinical Testing Program observed potency of CBL0137 in the NOD/SCID mouse model amongst a panel of PDX derived from children with acute lymphoblastic leukaemia (ALL), with individual PDX showing differing degrees of response. However subsequent analyses of the same panel by in vitro cytotoxicity assays failed to show a correlation with in vivo response. A potential source of discrepancy is the influence of bone marrow (BM) microenvironment on the proliferation, maintenance and survival of leukaemia cells, and protection against cytotoxic drugs3.
To develop an in vitro model for co-culture of childhood ALL cells and BM stromal cells, several PDX cell lines were cultured in the presence of GFP-tagged human mesenchymal stromal cells, MSC-hTERT, and viability was tested over 72 hours by measuring 7AAD stain through flow cytometry. Compared to PDX cells cultured alone, the proportion of viable cells increased by up to 5-fold when co-cultured with MSC. Notably, PDX cells co-cultured with MSC were less susceptible to CBL0137 than PDX in mono-culture, with over 3-fold difference in IC50. These data suggest that CBL0137 sensitivity might be influenced by the BM microenvironment. Further investigations across a panel of PDX will determine whether a BM co-culture model more accurately reflects in vivo CBL0137 response.