Single-Cell Pharmacodynamic Monitoring of S6 Ribosomal Protein in AML Blasts During a Clinical Trial Combining the mTOR Inhibitor Sirolimus with Mitoxantrone, Etoposide, and Cytarabine Chemotherapy.

Abstract 413

The mammalian target of rapamycin (mTOR) is an emerging molecular target in cancer therapy, however the relationship between target activation in individual tumors as well as inhibition of target and clinical response is poorly defined. This is in large part due to the difficulty of real-time monitoring of mTOR activation in patient samples. mTOR is activated in acute myelogneous leukemia (AML) cells and the mTOR inhibitor rapamycin enhances cytotoxic effects of chemotherapy in vitro and in vivo. We therefore developed a new application of a recently-developed, whole blood fixation/permeabilization technique for intracellular flow cytometry (Chow, et al. Cytometry A 2005). Using this approach, we sought to serially monitor S6 ribosomal protein (S6) phosphorylation in peripheral blood leukemic blasts during clinical trials of mTOR inhibitors. S6 is a known target of mTOR and its phosphorylation is a surrogate marker for mTOR kinase activation. We applied this methodology during a recent pilot trial in which an oral mTOR inhibitor, sirolimus (rapamycin), was administered in sequence with intensive combination chemotherapy (mitoxantrone, etoposide, and cytarabine, or MEC) in patients with relapsed, refractory, or secondary AML. The whole blood fixation process sufficiently preserved surface epitopes and light scatter properties for immunophenotyping, allowing specific signaling analysis of leukemic blasts as well as non-malignant cell populations. Importantly, even leukopenic trial samples containing as few as 20% blasts provided robust signaling data in malignant cells. S6 phosphorylation was readily apparent in leukemic blasts prior to therapy and, consistent with prior reports, occurred only in a subset of blasts. Exposing aliquots of pre-treatment whole blood samples to increasing concentrations of rapamycin ex vivo determined that leukemic blasts from most samples showed inhibition of S6 phosphorylation at clinically achievable concentrations (between 10-20 nM). Notably, some subjects' leukemic blasts showed no inhibition to >50 nM rapamycin, which far exceeded trough concentrations measured on our studies. To examine rapamycin's in vivo biochemical effects, we performed a paired analysis of clinical samples drawn at study entry and after 72 hours of oral sirolimus. 10 subjects provided paired samples, of which 2 did not show baseline S6 phosphorylation, 6 showed baseline S6 phosphorylation that inhibited during therapy, and 2 showed baseline S6 phosphorylation but no inhibition. Trough rapamycin levels were similar among rapamycin responsive and resistant subjects. Considering the 6 subjects with in vivo mTOR inhibition, 3 subjects achieved complete or partial remissions from the regimen. Neither subject with in vivo rapamycin resistance had a clinical response. Overall, we conclude that effective inhibition of mTOR signaling in AML blasts occurs in the majority of subjects during sirolimus treatment at the dose studied. However, cell-intrinsic rapamycin resistance occurs in a minority of patients and requires further study to clarify its mechanism and effects upon concurrent chemotherapy response. These data demonstrate the feasibility of real-time, intra-tumoral pharmacodynamic monitoring of S6 phosphorylation by flow cytometry during clinical trials combining intensive chemotherapy and signal transduction inhibitors for leukemia. Our approach greatly clarifies pharmacokinetic/pharmacodynamic relationships and has broad application to pre-clinical and clinical testing of drugs whose direct or downstream effects disrupt PI3K/AKT/mTOR signaling. Such compounds include inhibitors of FLT3, c-KIT, BCR-ABL, JAK2, and ras/raf/MAPK. Multicenter/cooperative group phase II testing of sirolimus plus MEC in AML has been initiated to establish the regimen's response rate and test the extent to which our pharmacodynamic studies predict clinical response.