Acute lymphoblastic leukemia (ALL) is the most common pediatric cancer. While significant progress has been made in the therapy of leukemia, several obstacles still hinder cure. Central nervous system (CNS) relapse is a major cause of treatment failure among patients with ALL and current CNS-directed therapies are also associated with significant morbidities including neurocognitive deficits, endocrinopathies, and secondary malignancies. Consequently, novel CNS-directed leukemia therapies are needed to improve long-term outcomes in ALL while decreasing treatment-related morbidity.

We developed in vitro and in vivo models of CNS leukemia in order to identify the unique characteristics of the CNS microenvironment that create a sanctuary site for leukemia cells. We transplanted multiple human ALL cell lines into immune-compromised mice (NSG). Mice were not irradiated or conditioned prior to transplantation to avoid perturbing leukemia niches. Using this xenotransplant model, we then identified the meninges as the predominant CNS site that harbors leukemia cells both before and after treatment with systemic cytarabine. This anatomic distribution of CNS leukemia agrees with other murine studies and also recapitulates the histopathology of human CNS leukemia.

Having demonstrated that the meninges provide a unique niche for leukemia, we then developed ex vivo co-culture approaches to focus more specifically on the effects of the meninges on leukemia biology. To more accurately model the leukemia-meningeal niche in co-culture we substituted tissue culture media for cerebral spinal fluid (CSF). However, given the unique composition and concentration of many substrates in CSF relative to serum or media we first characterized the effects of CSF on leukemia cells. We found that leukemia cells in CSF have limited survival even when replenished daily with fresh CSF. We hypothesized this decrease in viability may be secondary to elevated reactive oxygen species (ROS) given the lower levels of redox proteins in CSF. Accordingly, we found leukemia cells in CSF showed a higher level of ROS compared to the leukemia cells in regular media. Addition of N-acetyl-cysteine, a ROS scavenger, to CSF decreased ROS levels and cell death in leukemia cells. Moreover, leukemia cells co-cultured with meningeal cells in CSF showed decreased ROS levels compared to leukemia cells growing in suspension in CSF.

We found the effect of the meninges on leukemia biology extends beyond ROS regulation. Leukemia cells co-cultured with meningeal cells were also significantly more resistant to chemotherapy-induced apoptosis through effects on apoptosis balance and enhanced quiescence. These effects were reversed when the leukemia cells were removed from the meninges and placed back into suspension. Moreover, leukemia cells cultured in meningeal conditioned media also exhibited mild chemoresistance, indicating a role for a soluble factor(s) secreted by meningeal cells. Using proteomic approaches we then identified candidate soluble factor(s) secreted by meningeal cells that may contribute to leukemia chemoresistance.

These results show that meningeal cells influence key aspects of leukemia biology, including ROS regulation and chemoresistance, and that the pathophysiology of CNS leukemia is not only related to the ability of leukemia cells and chemotherapy to access the restricted CNS environment. Finally we are leveraging this knowledge of the CNS leukemia niche into the development of novel CNS-directed therapies that modulate ROS levels, target candidate soluble factors, or that disrupt the interactions between leukemia and meningeal cells.


No relevant conflicts of interest to declare.

Author notes


Asterisk with author names denotes non-ASH members.