Temozolomide selectively eliminates cancer stem cells in glioblastoma

3208 Glioblastomas (GBM) are among the most aggressive human cancers with a median survival of 14.6 months despite multimodal therapy. Although today’s standard of care - temozolomide - significantly increases the proportion of patients surviving for more than two years, long-term survivors are still a rare exception. The biological properties of cancer stem cells (CSC) maintaining GBM provide a possible explanation for the failure of chemotherapy in the long term: CSC from several tumor entities overexpress multidrug resistance proteins which may protect them against cytotoxic drugs that kill progenitor and differentiated cells. CSC would then give rise to recurrent tumors. However, although multiple authors have postulated this model, it has not yet been convincingly confirmed experimentally for GBM.
 We therefore investigated the effect of temozolomide on CD133+ and CD133- CSC lines (n=5) using an established in vitro model of GBM stem cell lines. Surprisingly, temozolomide induced a dose- and time-dependent decline of the rapidly proliferating stem cell subpopulation while differentiated tumor cells constituting the bulk of all tumor cells were resistant to the cytotoxic effects of the substance. Incubation with sublethal concentrations of temozolomide for 2 days eliminated clonogenic tumor cell growth in vitro and reduced the tumorigenicity in vivo.
 The susceptibility of CSC to temozolomide was meditated by O6-methylguanine-DNA-methyltransferase (MGMT) but not by CD133 expression, the MGMT promoter methylation status did not differ between CSC and differentiated tumor cells. MGMT-expressing CSC showed a 10 fold higher resistance to temozolomide which was reversible by the MGMT inhibitor 6-benzylguanidin.
 Despite the selective elimination of CSC in GBM, we found that CSC in all stem cell lines investigated were completely resistant towards the cytotoxic effects of 5 µM temozolomide which corresponds to the peak concentration measured in the cerebrospinal fluid of patients. While concentrations required to eliminate MGMT positive CSC (500 µM) can not be reached in the patient, concentrations up to 50 µM have been measured in the plasma of patient after standard treatment and sufficed to eliminate CSC in MGMT negative tumors.
 While these findings explain the poor outcome of MGMT positive tumors, it is unclear why patients with MGMT negative tumors die despite the susceptibility of CSC to temozolomide. Considering the impaired blood brain barrier in the tumor and infiltration of CSC in otherwise healthy brain tissue with much lower temozolomide concentrations, we postulate that recurrent GBM in MGMT negative tumor derive from invading CSC protected against temozolomide by the intact blood-brain barrier.
 Together, this study will fundamentally change the understanding of CSC and bears important implications for the conception and interpretation of future and past clinical trials.