Auranofin Induces a Reversible In-Vivo Stress Response That Correlates With a Transient Clinical Effect In Patients With Chronic Lymphocytic Leukemia

Auranofin (AF) is an oral disease-modifying anti-rheumatic agent. Using a high throughput screening assay of 2,816 FDA-approved drugs against primary tumor cells in vitro, we have previously identified AF as one of five drugs with selective anti-Chronic Lymphocytic Leukemia (CLL) activity. We have shown that AF induces oxidative stress and apoptosis in CLL cells in vitro independent of classic prognostic markers (ASH Meeting Abstract 865, 2012). Here, we evaluated the in vivo effects of AF on CLL using blood samples collected from six patients treated with single agent AF at the NIH as part of a multi-center clinical trial led by the University of Kansas Cancer Center (NCT01419691).

Five CLL patients and one patient with Small Lymphocytic Lymphoma (SLL) were enrolled in this open-label phase II study of AF in relapsed/refractory patients. Patients were started on AF 6 mg daily administered orally on 28-day cycles, with a dose escalation to 9 mg after the first cycle if no grade ≥2 toxicity occurred. AF was generally well tolerated. The best response was stable disease. Sequential blood samples were obtained prior to, and during the first cycle on drug. CLL cells were isolated by density gradient centrifugation. To study the in vivo effect of AF on redox balance we used dihydroethidium (DHE) and concomitantly measured cell viability using 3,3′-dihexyloxacarbocyanine iodide (DiOC6) in CD19 gated fresh cells by flow cytometry. Within 24 hours of the first dose AF induced an average 1.8 fold increase in DHE+ CLL cells indicating increased levels of ROS. Concomitantly there was a similar increase in apoptosis as shown by Annexin V staining. The increase in ROS production and apoptosis was transient; by day 7 all these changes had reverted to baseline or were even below baseline in three patients. A concomitant and equally transient decrease in the absolute lymphocyte count (ALC) and lactate dehydrogenase (LDH) level was also observed.

To investigate the in vivo effect of AF on tumor biology, total RNA (2.5 μg) from CLL cells of three patients treated with AF was profiled on Human Genome U133 Plus 2.0 arrays (Affymetrix). Three time points were analyzed: baseline (D0), after one dose of AF (D1), and a week later (D7). Compared to baseline, there were 182 genes (29-up, 153-down) whose expression changed >1.5-fold at P<.01 on D1, but only 14 genes at D7, suggesting a transient in vivo effect of auranofin on the CLL cells. Next we used a previously characterized gene signature that is regulated by the transcription factor NRF2, which is a key regulator of the cellular response to oxidative stress. We have previously shown that this NRF2 gene signature is upregulated in CLL cells treated with AF in vitro. Surprisingly, we observed that after 24 hours of in vivo therapy, AF induced a statistically significant decrease in the expression of the NRF2 signature in CLL cell and that these changes resolved by D7. To search for biologic processes leading to the gene expression changes, we used Gene Set Enrichment Analysis (GSEA). After 24 hours of in vivo treatment with AF, CLL cells displayed upregulation of genes in pathways involved in drug resistance (such as excretion and transport channels). In contrast, there was a decrease in the expression of genes involved in oxygen metabolism, DNA repair, and protein folding. With the exception of changes in genes regulating redox homeostasis, all these changes reverted back to baseline by D7.

Taken together, in vivo AF appears to transiently decrease anti-oxidant defenses regulated by NRF2 concomitant with an increase in cellular ROS levels and induction of some degree of cellular apoptosis. However, in vivo an adaptive response emerged that mobilized potentially compensatory mechanisms, including induction of pathways that may prevent intracellular accumulation of AF and a decrease in cellular processes that generate ROS, such as protein folding. The transient nature of the cellular response in vivo is consistent with the limited clinical activity seen in our patients. Plans are underway to determine the AF maximum tolerated dose in CLL patients. We will continue to evaluate the in vivo effects of AF on CLL cells at higher doses.

Supported by the Intramural Research Program of NHLBI and NCATS, NIH; a grant from The Leukemia and Lymphoma Society Therapy Acceleration Program to The Learning Collaborative™, as well as philanthropic support.

We thank our patients for participating in these research studies.