Epigenetics, Vitamin C, and Abnormal Blood Cell Formation - Vitamin C in Patients With Low-Risk Myeloid Malignancies (EVITA)

BACKGROUND Recent investigations have shown that mutations in epigenetic regulators are common, both in the apparently normal hematopoiesis of the elderly and in patients (pts) with myeloid cancers. It was long anticipated that DNA methylation was a permanent silencing mark, but with the discovery of the ten eleven translocation (TET) enzymes it became clear that active demethylation occurs. The initial steps in this process are catalyzed by TET enzymes, which are, however, frequently mutated and methylated in hematological cancers. The Jumonji enzymes, which catalyze histone demethylation, are also aberrantly regulated in hematological cancers.

Vitamin C (VitC) was identified in the 1930'ies as the necessary micronutrient in the prevention of scurvy. Unlike plants and most animals, humans are unable to synthesize vitC from glucose due to lack of the required enzyme, L-gulonolactone oxidase. Therefore, vitC must be provided through the diet. Recent studies recognize vitC as an important cofactor for the Fe(II)- and 2-oxoglutarate dioxygenase family. These include the TET enzymes, which are involved in the conversion of 5-methylcytosine (5-mC) to its oxidized derivatives 5-hydroxymethylcytosine (5-hmC), 5-carboxyl cytosine (5-caC), and 5-formylcytosine (5-fC), and the Jumonji enzymes that are involved in histone demethylation. Accordingly, vitC may potentially play an important role in the regulation of DNA and histone demethylation. However, > 80 percentage of hematological cancer pts were found to be severely vitC deficient. Interestingly, analyses of 20 participants included in the investigators' recently conducted randomized, placebo-controlled pilot study (NCT02877277) show that the level of vitC in MDS and CMML pts undergoing treatment with azacitidine, is easily elevated to the normal range by oral vitC supplement (unpublished data). When pts that were already taking vitC supplements were switched to placebo, the vitC levels quickly dropped below the normal range.

It has also been shown that the formation of 5-hmC and its derivatives may be compromised in healthy individuals and pts with TET mutations. However, since many of these mutations are heterozygous, and since the three TET enzymes (TET1, TET2, and TET3) may have some redundancy, restoration of vitC to physiological levels might have an impact on the level of 5-hmC/5-mC in individuals with TET mutant clonal hematopoiesis or hematological cancer.

Analyses of 5-hmC/5-mC levels in peripheral blood (PB) mononuclear cells (MNCs) from the participants in the pilot study also showed a clear trend toward increased 5-hmC in the vitC arm, however, after designing the trial the investigators realized that 5-hmC/5-mC levels are better measured in hematopoietic stem cells in the bone marrow (BM) where the levels are 10-20 fold higher. Thus, the pilot study will be followed-up with a randomized placebo-controlled trial of oral vitC in individuals with low-risk myeloid malignancies; i.e., CCUS or low-risk MDS/CMML, to investigate if oral vitC can change the biology of these disease entities and ultimately prevent progression.

Hypotheses:

1. The investigators and collaborators have previously shown that cancer pts are vitC deficient, and individuals with CCUS, which represents pre-MDS, might also be vitC deficient. The hypothesis is that this may lead to reduced levels of 5-hmC/5-mC in vivo in both cancer pts and individuals with CCUS

2. Elevating serum vitC levels to the normal range in CCUS and low-risk MDS/CMML pts by oral supplementation with vitC may

   • reduce the malignant clone,
   • increase the 5-hmC/5-mC ratio,
   • change the plasma cytokine profile towards a less inflammatory, less tumorigenic profile,
   • change gene expression

Aims:

To determine if restoring vitC to the normal range in CCUS and low-risk MDS/CMML pts can:

1. reduce the malignant clone,
2. increase the 5-hmC/5-mC ratio in CCUS and low-risk MDS/CMML pts
3. reduce accumulation of 5-mC at promoters/enhancers/long terminal repeats (LTRs), or at other regulatory genomic regions of tumor suppressors/methylated driver genes/genes involved in hematopoietic development,
4. upregulate the expression of these genes,
5. change the plasma cytokine profile,
6. alter intestinal permeability (measured by concentration of bacterial DNA in peripheral blood) and/or composition of gut microbiota,
7. entail any safety risks.

RESEARCH PLAN A total of 100 patients is planned for enrolment. Individuals with CCUS, low-risk MDS, or CMML-0 or -1 will be included from Rigshospitalet, Herlev University Hospital, Odense University Hospital, Aalborg University Hospital or Keck Hospital of University of Southern California between November 2017-December 2021.

The participants will enter block randomization with a ratio of 1:1; vitC 1000 mg/day p.o. versus placebo for one year.

Up to 30 healthy elderly volunteers are planned for enrolment. Amount and type of bacterial DNA in PB will be determined and serve as controls.

MATERIAL AND MEASUREMENTS PB: PB samples (45 mL) will be taken at study entry and every 3 months (or more if required according to physician's choice) during the first year.

Measurements include blood counts including differential count, levels of folic acid, vitamin B12, vitamin D, iron, ferritin, transferrin, transferrin saturation, plasma vitC levels, cytokines, and bacterial DNA content (referenced to bacterial DNA on patient's skin and in feces).

BM: BM samples (18 mL) will be taken at study entry, after 3 months, and after one year.

Levels of vitC and various cytokines will be measured. Mutations in genes frequently involved in myeloid cancer/clonal hematopoiesis, including epigenetic regulators, and variant allele frequencies (VAF) will be investigated by targeted next generation sequencing in sorted BM CD34+ cells (if available) or BM negative fraction of CD34+ sorting at study entry, after 3 months and after one year.

RNA sequencing and total 5-hmC/5-mC assessment will be performed on BM CD34+ hematopoietic stem cells.

Feces: Feces samples (9 mL) will be collected from a subset of patients at study entry, after 3 months, and after one year. Bacterial DNA will be extracted and sequenced.

RESEARCH BIOBANK A research biobank will be established at The Epi-/Genome Laboratory, Rigshospitalet / Biotech Research and Innovation Centre to store the biological samples from the participants. The research biobank is approved by the Regional Science Ethics Committee and the Danish Data Protection Agency in accordance with the Act on Processing of Personal Data (license no. H-16022249 and 04864/RH-2016-259, respectively). Cryopreserved separated MNCs from BM and PB will be stored in addition to granulocyte pellets and plasma. The date for closing the research biobank is 31-12-2030.

For correlative studies, biological samples will be sent to Van Andel Research Institute, Grand Rapids, US, and Imperial College, London, UK (external collaborators), for RNA sequencing and analyses of DNA methylation and hydroxymethylation, respectively. Biological samples will also be sent to Life Science Faculty, University of Copenhagen, and The National Veterinary Institute, Technical University of Denmark (external collaborators) for measurement of serum vitC concentrations and analyses of T cell responses, respectively.

METHODS VitC measurement: Ascorbate and total vitC, i.e., ascorbate + dehydroascorbic acid (the oxidized form of vitC; DHA), are quantified by high-performance liquid chromatography (HPLC) with coulometric detection; DHA is assessed by subtraction of ascorbate from total vitC. Uric acid is used as endogenous internal standard.

Cell sorting: Magnetic-activated cell sorting (MACS, using a EasySep device).

Total 5-hmC/5-mC measurement: Dot blot analysis of 5-hmC. 5-hmC/5-mC measurement by Mass Spectrometry.

Locus specific 5-hmC/5-mC measurement: "EPIC" 850 K BeadChips. 5-hmC/5-mC at selected sites will be measured by pyrosequencing.

Gene expression: Total RNA sequencing and reverse transcriptase-quantitative polymerase chain reaction.

Mutation detection: Targeted next generation sequencing of a panel of genes recurrently mutated in myeloid cancer as described.

STATISTICAL CONSIDERATIONS AND POWER CALCULATION This study is the first study to examine the effects of vitC as monotherapy on 5-hmC/5-mC levels in hematopoietic stem cells in humans in vivo. Therefore, it is not possible to perform a power calculation or a sample size calculation. The number of participants (n=100) is set as an estimate of the number needed to observe a potential significant difference between the groups (vitC vs. placebo) in the primary endpoint; median change in VAF of somatic mutations from baseline to 12 months.

Efficacy analyses are by intention-to-treat. Safety analyses include all participants who receive at least one dose of protocol therapy.

ETHICAL CONSIDERATIONS The study has been approved by the Regional Science Ethics Committee (H-16022249) and the Danish Data Protection Agency (04864/RH-2016-259).

All participants included in the project will be informed orally and in writing. Participation will only be accepted after written consent. Participants will be informed that they can at any time for any reason withdraw from the study without it affecting their treatment in the health care system.

Using the targeted DNA sequencing approach described, there is a small risk of detecting a germline mutation in the participants related to myeloid malignancy. The participants will be asked to state in the informed consent if they do not want to receive any further relevant health-related information that may appear during the project analyses. Unless the participant explicitly states that he or she does not want to be informed of any potential health-related random findings in the study, the participant will be informed and offered further investigations and genetic counseling at the local hospital in case of random findings.

Patient disadvantages, side effects, risks, and complications Blood sampling is associated with brief discomfort and/or pain. No significant risks are associated with the blood sampling. Local bleeding can occur, which in rare cases can cause discomfort and discoloration for a few days. Rarely, a blood sampling can cause vasovagal reaction leading to a brief loss of consciousness.

BM aspiration is associated with brief pain while the local anesthesia is given. Furthermore, many individuals experience an uncomfortable feeling in the nates and leg while the BM is aspirated. This lasts approximately 30 seconds. Finally, some tenderness can occur for a few days after the procedure. Possible complications to undergoing a BM investigation include bleeding and infection. However, the incidence of these complications is extremely low.

According to the Nordic Nutrition Recommendations there is no evidence that intake of vitC above 1000 mg/day are either carcinogenic or teratogenic. However, high intakes (> 1000-2000 mg/day) may cause diarrhea and other gastrointestinal disturbances and susceptible individuals may experience kidney stone formation from increased oxalate formation. Since vitC is only given at physiological doses, it is anticipated that the study is safe and will provide no additional risk for the participants; an assumption that is supported by the experience from the pilot study.