L-Ascorbic Acid Depletion to Treat Acute Myeloid Leukemia and Myelodysplastic Syndromes

The seminal discovery that the in vitro growth of malignant cells could be absolutely dependent on L-ascorbic acid (LAA) was originally published in Sciene. The cell culture assay used with a mouse plasmacytoma model in this discovery was based on colony formation, and was essentially the same as the one used to grow normal hemopoietic colonies, such as CFU-GM and CFU-G, from normal bone marrow specimens. Subsequent analysis of the growth factors for these CFUs using this same culture system eventually led to the discovery of colony stimulating factors, such as GM-CSF and G-CSF, now widely used clinically. Human leukemia, specifically acute myeloid leukemia (AML), cell colonies also grow well in this culture system, as shown by extensive cell biology studies. In addition, cells from patients with myelodysplastic syndrome (MDS), a significant proportion of them progressing to AML behave similarly to AML cells in this culture system . In particular both AML and MDS are identical in that the growth of colonies is enhanced by addition of LAA to the cell culture media in a high proportion of these patients.

The large volume of in vitro data thus generated, including correlations with direct clinical relevance is increasingly convincing that lowering of LAA levels could potentially be developed and utilized as a treatment for specific hemopoeitic malignancies. This was particularly attractive in view of the fact that the growth of normal hemopoietic colonies, such as CFU-GM and CFU-G, is never enhanced by LAA. Such an absolute selectivity would predict a lack of clinical adverse hemopoetic events from an intervention which lowered LAA levels. We also had seemingly-contradictory data that the growth of colonies from AML and MDS patients could be suppressed by addition of LAA, infrequently but sometimes profoundly. However, detailed dose response analysis later clarified this: low physiologic doses enhance and high pharmacologic doses suppress formation of leukemic colonies. From a therapeutic perspective, we would have greater expectations for a depletion strategy than for supplementation, because 1) leukemic suppression by addition of LAA is often accompanied by some mild suppression of the normal CFU and therefore is not absolutely selective; and 2) LAA supplementation has been clinically tested in a variety of solid tumors, with controversial outcomes.

Therefore, our original protocol was developed primarily to accomplish lowering of LAA levels, with a subsequent oral LAA supplementation used primarily to prevent scurvy and only secondarily for possible benefit. However, with the first patient there was a strong indication of antileukemic effects during both the LAA depletion and supplementation phases. Based on this encouragement, the protocol was amended to formally alternate depletion with supplementation, and to utilize intravenous (IV) administration of LAA to achieve high dose supplementation. With 17 subsequent subjects having been treated, this study of the safety and efficacy of cyclic manipulation of LAA levels has demonstrated beneficial outcome in a high proportion of refractory and terminal patients with AML or MDS. Moreover, growing laboratory evidence being produced provides a molecular basis for these clinical outcomes.