5-Azacytidine and Decitabine Monotherapies of Myelodysplastic Disorders

OBJECTIVE: To review and differentiate the pharmacology, toxicology, pharmacokinetics, and results of major clinical trials of 5- azacytidine (5-AzaC) and 5-aza-2-deoxycytidine (decitabine) therapy of myelodysplastic disorders.

DATA SOURCES: A PubMed/MEDLINE search was conducted (1966–October 2004) using the following terms: DNA methylation, myelodysplastic disorders, 5-azacytidine, and 5-aza-2-deoxycytidine (decitabine). Additional data sources included bibliographies from identified articles and manufacturer information.

STUDY SELECTION AND DATA EXTRACTION: Clinical trials for the treatment of various malignancies by hypomethylating agents were selected from data sources. All published, major clinical trials evaluating 5-AzaC or decitabine in myelodysplastic disorders and transformed myeloid leukemia treatment were included.

DATA SYNTHESIS: Myelodysplastic disorders are a group of bone marrow stem cell hyperplasias and dysplasias that result in ineffective hematopoiesis. Myelodysplastic disorders and transformed leukemia have poor prognosis and minimal response to chemotherapy. DNA hypomethylating agents have been shown to improve overall response rates (increased neutrophil, leukocyte, and platelet counts), time to leukemic progression, and quality of life compared with supportive therapy. The incidence of the most common adverse effects (nausea, vomiting, myelosuppression) can be reduced by low-dose, continuous, or extended-interval infusion.

CONCLUSIONS: Since appropriate dosing schedules of decitabine are being investigated, comparison of the clinical effectiveness of 5-AzaC and decitabine would be premature at this time. DNA hypomethylating agents show promise as monotherapies of myelodysplastic disorders and transformed leukemia and may be useful as a component of combination chemotherapy of various malignancies.
KEY WORDS: 5-azacytidine, 5-aza-2-deoxycytidine, decitabine; acute myelogenous leukemia, myelodysplastic syndrome.

The potential role of DNA hypomethylating agents in treatment of myelodysplastic syndrome (MDS), myeloid leukemia, and other forms of neoplasia has been extensively reviewed in recent years.1-16 Early work sug- gested that methylation of CpG sites (cytosine residues co- valently bound to guanine by a phosphodiester bond) lo- cated in promoter regions of certain genes may lead to their transcriptional inactivation.17-20 DNA hypermethyla- tion is thought to be involved in selective silencing of tu- mor suppressor gene expression and may play a major role in the neoplastic process.21,22

MDS is often termed pre-leukemia or smoldering leukemia and is characterized as a failure of bone marrow to produce adequate levels of leukocytes and platelets. A reduction in blast numbers may allow repopulation of bone marrow with multipotent precursor cells. Although hy- pomethylating agents cause an initial leukocytopenia and reduction in blast cell number, clinical responses are seen as a delayed improvement in neutrophil, platelet, and leukocyte (trilineage) counts.2,5,9,11,15 This can reduce the need for transfusions and improve the quality of life for pa- tients with MDS.

DNA hypomethylating drugs recently used to treat MDS include 5-azacytidine (5-AzaC) and 5-aza-2-deoxy- cytidine (decitabine). Although 5-AzaC and decitabine dif- fer structurally by only their respective ribose and deoxyri- bose sugar portions (Figure 1A), there are some differ- ences in their pharmacologic and toxicologic profiles. Pharmion Corporation has received approval by the Food and Drug Administration (FDA) for use of 5-AzaC (Vi- daza) as first-line monotherapy of MDS and chronic monomyleocytic leukemia (CMML).23 Similarly, decitabine (Dacogen), which is marketed by SuperGen, is undergoing 2 Phase II trials in the US for treatment of chronic myel- ogenous leukemia (CML) and also Phase III trials for the treatment of MDS. The goals of this review are to examine the differences between these very similar drugs and relate them to the current and future therapeutic roles in MDS and transformed myeloid leukemia.

Data Sources

All published, randomized Phase II and III clinical trials evaluating either 5-AzaC or decitabine treatment of MDS and acute transformed myleoid leukemia were included in this review. Several Phase I studies evaluating either agent in the treatment of various malignancies were included, as needed, for pharmacokinetics and toxicity profiles.

Pharmacology

Most cancer cells represent clonal lines of functionally immature cells that have lost their ability to respond to nor- mal maturation and proliferative control mechanisms. This allows these clonal lines to evade terminal differentiation and, ultimately, programmed cell death. Early studies of chemically induced differentiation of tumor cell lines re- ported a link between DNA hypomethylation and terminal differentiation.24-30 Recently, it has been established that methylated regions of DNA attract DNA binding proteins with subsequent recruitment of a DNA silencing complex that contains histone deacetylases.31,32 Mechanisms in- volved in DNA methylation and histone acetylation work in concert to affect gene expression and are termed “epige- netic modifications” of chromatin.7,8,12 Epigenetics is an emerging field that deals with biochemical alterations in DNA and chromatin that do not involve changes in the DNA coding sequence itself. Epigenetic changes to chro- matin are thought to be crucial in cell development and differentiation. Exposure to DNA hypomethylating agents could, theoretically, result in alteration of the patterns of gene expression without changing the coding sequence of those genes or their transcriptional control elements. More detailed discussions of the chemistry of DNA methylation reactions are available.1,33,34

The inhibition of nucleic acid synthesis by nucleoside analogs has been a mainstay in cancer chemotherapy for >50 years. Most of these agents inhibit formation of pools of nucleotide triphosphate precursors or directly inhibit DNA and/or RNA polymerase reactions. Both 5-AzaC and decitabine are phosphorylated by deoxycytidine kinase and readily incorporated into nucleic acids in growing cell lines,35 being active pharmacologically in the S-phase.36,37 However, hypomethylating agents appear to work by dif- ferent mechanisms, primarily involving processing of RNA and initiation of transcription.
5-AzaC nucleotide triphosphates can be incorporated di- rectly into RNA, while decitabine incorporates into DNA due to its deoxyribose component. Indirect formation of 5- AzaC deoxynucleotide diphosphates, through ribonu- cleotide reductase–catalyzed conversion of ribonucleotide diphosphates (Figure 1B), allows 10 –20% of 5-AzaC in- corporation into the DNA fraction.35 Stated simply, decita- bine preferentially reduces DNA hypomethylation, while 5-AzaC reduces both RNA and, to a lesser extent, DNA methylation (Figure 2). DNA hypomethylation may affect transcription of messenger RNA and differential gene ex- pression, subsequently triggering terminal differentiation and eventually apoptotic processes. Likewise, RNA hy- pomethylation may interfere with transfer RNA and ribo- somal RNA processing, causing direct cytotoxicity by re- ducing protein synthesis.38-45 Therapeutic effects of DNA hypomethylating agents may actually be due to a combina- tion of cytotoxicity and induced differentiation. Further- more, continuous exposure to lower doses of 5-AzaC was found to be more effective in reduction of leukemia cell vi- ability than brief exposure to higher doses.46

Induction of terminal differentiation in human tumor cell lines was noted at drug concentrations in the low mi- cromole range in cell culture,25-30 and reduction of DNA methylation was suggested to be involved. Decitabine is a more potent DNA hypomethylating agent than 5-AzaC in differentiating leukemia cells, being active at 10- to 30- fold lower culture concentrations.26,47 This is not surprising due to preferential incorporation of decitabine into DNA. Mechanistically, DNA methylase becomes covalently bound at sites of 5-AzaC or decitabine incorporation, which may result in a reduction of functional methylase enzyme levels. In addition, covalent attachment of DNA methylase to the DNA strand is thought to be directly in- volved in the mechanism of cell death by blocking DNA synthesis and replication.48-50

It is important to note that inhibition of DNA methyla- tion has not been shown to be inherently cytotoxic, and maximum effects of hypomethylating agents have been observed at very low doses. Notably, increasing the dose will not result in further hypomethylation, but may in- crease toxicity.51 The high doses of decitabine used in clini- cal trials are likely to be cytotoxic, so the role of DNA methylation remains to be determined.2,51

Molecular endpoints, such as methylation status of con- trol elements of tumor suppressor genes, have not been clearly established. Expression of p15(INK4b), a cyclin- dependent kinase inhibitor, is usually up-regulated during granulocytic and megakaryocytic differentiation, but was shown to be down-regulated by hypermethylation in MDS.52 Reduced p15 expression is thought to result in loss of cell cycle control and may be related to disease progression in MDS.53 Reversal of p15 hypermethylation was re- ported in patients with MDS treated with hypomethylating agents.54 Hypermethylation of CpG islands in promoter re- gions is an effective mechanism of gene silencing and has been described in almost every tumor type. Gene inactiva- tion by hypermethylation is not limited to tumor suppres- sor genes (eg, p53), but includes genes of many other cel- lular pathways (Figure 2) including cell cycle control (p15, p16, p73, VHL), DNA repair (hMLH1, MGMT, BRCA1)apoptosis (DAPK, APAF-1), cell adherence (CDH1, CDH13), and detoxification (GSTP1).55,56 Further investi- gation of the gene-specific effects of DNA hypomethyla- tion is needed to understand the pharmacologic effects of hypomethylating agents in hematologic neoplasia.57

Pharmacokinetics

To be active as antineoplastic agents, cytidine analogs must be phosphorylated by deoxycytidine kinase (Figure 1B),58 which represents the rate-limiting step of drug acti- vation.59 The optimum plasma concentration of cytidine nucleoside analogs was postulated to be near the Michaelis constant of deoxycytidine kinase, which was estimated to be approximately 10 µM for decitabine.60 An infusion rate of 30 – 60 mg/m2/h (~1–2 mg/kg/h) would be needed to achieve these plasma concentrations.60,61 By using a mean cell cycle time of about 60 hours for human myeloid leukemia cells,62 it was estimated that the optimal duration of decitabine therapy would be about 90 hours (~4 days).63 Since the plasma half-life of decitabine is only 15–20 min- utes, constant intravenous infusion would be necessary.64 Metabolic inactivation of cytidine derivatives occurs by enzymatic deamination in leukemia cells,65 and higher lev- els of cytosine deaminase activity can lead to drug resis- tance.59,66 This may be problematic in patients previously treated with cytosine arabinoside, where deaminase activi- ty may be elevated.63

A pharmacokinetic comparison of intravenous bolus versus continuous infusion of 5-AzaC was conducted in 8 patients with a variety of malignant tumor types.67 Follow- ing intravenous bolus dosing, disappearance of [14C]-5- AzaC exhibited a multiphasic pattern, with distribution and elimination half-lives determined as 16 –33 minutes and 3.4–6.2 hours, respectively. Less than 2% of the [14C]- 5-AzaC was present in plasma 30 minutes after dosing, with labeled-drug accumulation in the leukocytes. Urinary excretion accounted for 73–98% of the dose within 3 days, with <1% recovery in the feces.

In addition, the relative plasma concentration remained higher following continuous infusion compared with bolus doses, while urinary elimination was not significantly af- fected by route of administration. Continuous infusion was also associated with fewer adverse effects, such as nausea and vomiting, than bolus. The stability of 5-AzaC in Ringer’s lactate (pH 6.2) was confirmed, with a half-life of 94–100 hours at 20 ˚C. The investigators concluded that 5- AzaC should be given by continuous infusion rather than as a bolus dose.67

During early Phase I trials, pharmacokinetic evaluation of decitabine was conducted in 21 patients with advanced solid tumors.68 Infusion of decitabine at 7 dosing levels (25–100 mg/m2) involved three 1-hour infusions separated by 7-hour intervals repeated every 3– 6 weeks. Myelosup- pression was dose limiting, with a white blood cell nadir at 22 days. More recently, 13 patients with non–small cell lung cancer were treated with a 200 – 600 mg/m2/dose in- fusion every 5–7 days for 1–5 courses. After the first 2 pa- tients, the 6-hour infusion every 2–3 days was changed to single 8-hour infusions to reduce leukopenia. From those studies, a dosing regimen of 75 mg/m2 was recommended in 3 infusions every 5 weeks for Phase II trials in solid tumors.68 The optimal dosing of 5-AzaC and decitabine for treat- ment of MDS and transformed acute leukemia is under in- tense investigation. Several small, nonrandomized clinical trials of both solid tumor therapy and salvage therapy for acute myelogenous leukemia (AML) have been carried out over the past several decades and more recently reviewed.3 Response rates were low in studies evaluating monothera- py of solid tumors (including breast, ovarian, colon, melanoma) with 5-AzaC 300–700 mg/m2 cumulative dose given 1–24 mg/kg daily for 8 –10 days through an intra- venous bolus or subcutaneous injection or decitabine 225 mg/m2 as a single-dose infusion.

Cumulative rates of 20% complete remission and 16% partial remission were noted in a review of salvage therapy of 200 patients with AML in several clinical studies using 5-AzaC 100 – 400 mg/m2 total dose (5 days divided infu- sion doses) or decitabine 45–100 mg/m2 continuous infu- sion over 4 days.3 However, it was observed that longer or continuous infusions were more effective than shorter dos- ing, with the schedule being more important than the cu- mulative dose.

Clinical Studies in Patients with Myelodysplastic Syndrome
MDS arises from an abnormal multipotent progenitor cell and represents a heterogeneous hematopoietic disorder characterized by hyperproliferative bone marrow, dyspla- sia of the cellular elements, and ineffective hematopoiesis. In broad terms, MDS can be divided into indolent (early) or aggressive (advanced) disease, where progression fa- vors selection of cells that are less sensitive to apoptotic signals and possess higher proliferative capacity. It is thought that oncogene activation and tumor suppressor gene inactivation facilitate disease progression and eventually transformation to AML. A decrease in bone marrow produc- tion of colony-stimulating factors and reduced progenitor sensitivity to those factors underlies the associated cytopenia. Subtypes of MDS are based on 2 classification systems.
The French–American–British (FAB) system subdivides
MDS into 5 groups: refractory anemia (RA), refractory anemia with ringed sideroblasts (RARS), refractory ane- mia with excess blasts (RAEB), refractory anemia with ex- cess blasts in transformation (RAEB-T), and CMML.69 The FAB system is useful in the prediction of transformation to AML and survival. The International Prognostic Scoring System (IPSS) uses an additive scoring method to separate patients based on cytogenetics, percent bone mar- row blasts, and number of cytopenias.70 Higher scores de- note poorer prognoses. MDS is further classified by risk of developing AML; intermediate-1 risk patients typically have RA, RARS, or RAEB and patients with intermediate- 2 or high risk usually have RAEB-T or CMML.

It is estimated that 12 000 new cases of MDS are report- ed in the US annually, with highest prevalence in patients >60 years of age.71 Death from MDS can result from bleeding and complications of infection, while transforma- tion to AML occurs in up to 40% of patients. Prognosis for patients transforming to AML is exceptionally poor. In high-risk MDS, which includes all FAB subtypes in which elevated levels of immature blast cells are produced, the av- erage patient survival is 6 –12 months. Conventional anti- leukemic therapy has proven ineffective against MDS.72-74 Supportive therapy consists of antibiotic treatment of op- portunistic infections, blood transfusions, and erythropoi- etin therapy for anemia. Although allogeneic stem cell transplantation has curative potential, advanced age and comorbid disease states make most patients with MDS poor candidates for this procedure.74

Several clinical trials have provided evidence of clinical response in patients with MDS treated with hypomethylat- ing agents75-84 including improved survival, reduced need for transfusions, prolonged time to leukemic transforma- tion, and improved quality of life (Table 1).79-81 The ongo- ing clinical approaches to the treatment of MDS and trans- formed myeloid leukemia using demethylating agents are addressed here.
A randomized, controlled Phase III study involving 191 patients with MDS compared 5-AzaC 75 mg/m2/day, giv- en subcutaneously for 7 days every 28 days with support- ive care.79 Both study groups received transfusions and an- tibiotics as required, and patients receiving only supportive care were permitted to cross over to 5-AzaC when their disease worsened. Clinical responses (increased white blood cell count, absolute neutrophil count, or platelet count) were observed in 60% of the 5-AzaC treatment group, while 5% improvement was found in the supportive care group (p < 0.001). Complete remission included reduction of blast cells in bone marrow to <5% and complete normalization of peripheral blood counts. Partial remission included a 50% reduction of initial bone marrow blasts and a trilineage response (>50% restitution of initial deficit of white blood cells, absolute neutrophils, or platelets). Clinical improvement included monolineage or bilineage response of peripheral blood counts or 50% reduction in transfusion requirements.

Although the mean number of erythrocyte transfusions increased initially in patients treated with 5-AzaC, all pa- tients with complete or partial remission no longer needed transfusions. Of the 65 patients who received transfusions at the start of the study, 45% required no more transfu- sions, while another 6% had a 50% reduction in transfu- sions. Median times for initial and best response were 64 and 93 days, respectively, with median duration of response of 15 months for those achieving clinical improvement or better. Leukemic transformation (to AML) occurred as the first event in 15% of 5-AzaC treated patients and 38% of those receiving supportive care (p = 0.001). Median time to trans- formation or death was 20 months for 5-AzaC compared with 14 months for supportive care (p = 0.1). Myelosuppres- sion was the major adverse effect in 5-AzaC treated patients, reported as grade III/IV leukopenia (59%), granulocytopenia (81%), and thrombocytopenia (70%).79

A modest, but not statistically significant, increase in survival of these patients was reported.79 Major improve- ments in physical function, symptoms, and psychological state were noted in quality-of-life assessment for patients initially randomized to receive 5-AzaC, particularly for fa- tigue and dyspnea.80 Results indicate significantly higher response rates, improved quality of life, reduced risk of leukemic transformation, and improved survival compared with supportive care.79,80 It was suggested that 5-AzaC therapy was superior to supportive care for patients with MDS and provides a new treatment option for these pa- tients.85 Based on the results published by Silverman et al.,79,80 the FDA granted approval of 5-AzaC as monothera- py for MDS and CMML.23
Although still in clinical trials, decitabine has also been found to be effective in therapy of MDS (Table 1). Results from 3 combined Phase II European studies were recently re- ported, including 170 patients with MDS treated with decitabine 40–50 mg/m2 administered daily by continuous intra- venous infusion for 3 days every 4 – 6 weeks (135–150 mg/m2 per course).82 Median age for the study group was 70 years, with similar numbers of patients in the IPSS in- termediate-1 (n = 48), intermediate-2 (n = 50), and high- risk (n = 71) groups. Overall platelet response rate was 49%. Responses in platelet count were not significantly different between any FAB groups or IPSS subgroups. Median survival was 15 months, with median response du- ration of 9 months in this study. Myelosuppression was re- ported to be the only major adverse effect.78 Most (n = 162) of the 170 patients were evaluated for effects of low- dose decitabine on platelet count, with 126 (78%) patients being thrombocytopenic at the start of therapy.82 It was re- ported that 54% of the 126 patients showed a rise in platelet count, most (58%) of which occurred after only a single cycle of decitabine. These results indicate that decitabine has a clinically significant and often long- last- ing effect on platelet count in a substantial number of high- risk patients with MDS. Moreover, an increase in platelet count after one cycle of therapy was reported to be highly predictive of survival duration (p < 0.0001).82 It should be noted that similar increases in platelet count were also ob- served in patients treated with 5-AzaC.79

The effectiveness of decitabine was recently evaluated in a Phase I study of low-dose and prolonged schedules in 48 patients with relapsed or refractory MDS, AML, acute lymphocytic leukemia, or CML.81 Decitabine 5, 10, 15, or 20 mg/m2 was administered as a 1-hour intravenous infu- sion 5 days per week for 10 days (5 days on, 2 days off, 5 days on), 10 consecutive days, or 20 consecutive days. The highest response rate was observed with 15 mg/m2 for 10 days with 65% (11/17) response, while a 45% response rate was seen with the first 3 doses (combined 14/31) and 11% with 20 or 15 mg/m2 for >10 days (combined 2/19). Reductions seen in these response rates at higher doses were apparent only when complete remissions were com- pared (8/31 with 15 mg/m2 vs 1/9 with 20 mg/m2; p = 0.06, Fisher’s exact test).

Objective responses were noted in 16 (32%) patients among all dosing schedules. More specifically, in 7 pa- tients with MDS, 2 achieved complete remission (29%) and 2 achieved partial remission (29%). Of the 37 patients with AML, 5 achieved complete remission (14%) and 3 achieved partial remission (8%), while 2 of the 5 patients with CML achieved complete remission (40%) and 2 pa- tients achieved partial remission (40%). The investigators acknowledged that this was not a controlled study and that response rates were across different doses and therefore may be underestimated.81 In response to these results, a large ongoing Phase II study was initiated by these investi- gators to examine the efficacy of different dosing sched- ules of decitabine in patients with MDS.

Adverse Effects and Drug Interactions

Historically, the most relevant adverse effects of 5- AzaC and decitabine therapy have been myelosuppression, nausea, and vomiting.4,75-84 Reduction of these effects has been achieved by continuous intravenous infusion dosing. The most common adverse effects are listed in Table 2.86 Pro- longed myelosuppression was noted following doses of 75 or 100 mg/m2 during 6-hour infusions of 5-AzaC in patients with MDS twice daily for 10 doses.83 Nausea and vomiting, febrile episodes, infections, diarrhea, and mucositis were the most common adverse effects reported in this group.

Since decitabine is not yet available commercially, ad- verse effect profiles are still under investigation and in- complete. In published studies, most common adverse ef- fects following decitabine infusion (90 –120 mg/m2/day) were grade III–IV neutropenia and thrombocytopenia, with median recovery time of 30 and 35 days, respec- tively.82 In the most recent evaluation of low-dose, pro- longed-exposure dosing schedules, investigators conclud- ed that decitabine treatment was well tolerated by patients with MDS, with non-hematologic adverse effects includ- ing liver dysfunction (36%), creatinine elevation (10%), nausea (6%), diarrhea (2%), and skin rashes (2%).81 A- symptomatic but severe elevations in liver function tests (enzymes or bilirubin levels at grades III–IV) were noted in 6 of 50 patients (possibly related to treatment). Febrile episodes were observed in 26 (52%) patients, including documented infections in 18 (36%) of these patients.81

As of this writing, there are no documented drug inter- actions for either 5-AzaC or decitabine, and neither has been reported to induce or inhibit cytochrome P450 en- zymes. However, both drugs should be used with caution, just as with any other antineoplastic agent. Since both 5- AzaC and decitabine can lower blood cell counts initial- ly,77-81 patients administered either agent should not receive vaccinations without their physician’s knowledge and should avoid contact with persons who have recently been given oral polio vaccine. An initial reduction in platelet count can also occur,82 with increased bleeding tendency and bruising. Aspirin or aspirin-containing medicines should be avoided.

Special Populations and Precautions

A sufficient number of randomized controlled studies has not been conducted to provide necessary information for evaluation of racial, ethnic, gender, or age-related prameters of therapy with hypomethylating agents.86 Howev- er, preliminary evidence suggests that there may be a higher incidence of renal failure, seizures, and atrial fibrillation in very elderly patients with MDS treated with decitabine.78

The clinical evaluation of decitabine and 5-AzaC as po- tential human antineoplastic drugs is further complicated by issues of mutagenicity and carcinogenicity. Early stud- ies determined 5-AzaC to be carcinogenic in rodents,87,88 while decitabine was not carcinogenic87 and only minimal- ly mutagenic or non-mutagenic in mammalian cells.37,89 However, both drugs are known teratogens in rodent mod- els.90-93 Birth control measures should be used when either partner is taking one of these drugs, as birth defects may develop if either the male or female is taking them at the time of conception or during pregnancy. 5-AzaC is consid- ered to be risk category D for use during pregnancy and should not be used while breast-feeding.

Caution is needed in patients with severe preexisting hepatic impairment, as 5-AzaC is potentially hepatotoxic. 5-AzaC is contraindicated in patients with prior sensitivity to 5-AzaC or mannitol and in those with advanced malig- nant hepatic tumors due to rare progression to hepatic coma and death (especially in patients with baseline albu- min <30 g/L).94

Approved Labeling and Other Therapeutic Uses

5-AzaC is available as an injectable suspension, market- ed by Pharmion Corporation as an FDA-approved mono- therapy for treatment of MDS. In 2004, the FDA granted Priority Review classification for Pharmion’s New Drug Application (NDA) for 5-AzaC for the treatment of MDS, with the submission based on a National Cancer Institute (NCI)–sponsored Phase III study for the treatment of MDS81,82 and 2 supportive Phase II studies,86 all of which were conducted by the Cancer and Leukemia Group B. 5- AzaC was previously granted Orphan Product designation by the FDA following its submission by Pharmion in 2003.

Decitabine (NSC 127716) was originally developed by

Pharmachemie, who sold worldwide rights of this drug to SuperGen in 1999. SuperGen entered a Cooperative Re- search and Development Agreement with the NCI in May 2000, which initiated and sponsored clinical trials in pa- tients with solid tumors and hematologic malignancies. SuperGen received a rolling NDA submission for decita- bine for injection, with a focus on treatment of patients with MDS. In addition, decitabine is being investigated in an open-label Phase III study versus standard care therapy of advanced MDS and Phase II trials in non–small cell lung cancer in Canada and prostate cancer in the US.94,95 Phase I/II studies are also underway to evaluate decitabine treatment of sickle cell anemia.96

Dosage and Administration

5-AzaC is manufactured by Ben Venue Laboratories for Pharmion Corporation as an injectable suspension. The recommended starting dose is 75 mg/m2 daily given sub- cutaneously for 7 days every 4 weeks. If no beneficial ef- fect is seen after 2 treatment cycles and no toxicity is noted other than nausea and vomiting, the dose may be increased to 100 mg/m2. The dose given will vary according to body surface area, concurrent administration of other medicines, or blood counts. Dosage adjustments may be necessary in geriatric patients or others with reduced renal function. It may be useful to monitor serum bicarbonate, blood urea nitrogen, and serum creatinine levels in patients with renal dysfunction. Although complete or partial remission may require extended therapy, it is recommended that patients receive a minimum of 4 treatment cycles.Decitabine is under Phase III clinical evaluation.

Cost Analysis

5-AzaC for injection is supplied in 100-mg vials with an average wholesale price (AWP) of $467 per vial.86 Us- ing a standard treatment schedule of 75 mg/m2 daily for 7 days, the cost for a patient with a body surface area of 1.71 m2 would be approximately $4200 per treatment cycle. Since >90% of responders initially demonstrated improve- ment by the fifth treatment cycle,79 a minimal initial drug cost would be near $21 000 (AWP).

Summary

DNA hypomethylating agents, such as 5-AzaC and decitabine, have shown promise as single-agent therapy in the treatment of both new-onset and refractory MDS. Cur- rently, only 5-AzaC has been approved for these therapies. Although DNA hypomethylating agent therapy is still in early stages of clinical use, it appears that optimizing the dosing schedules of 5-AzaC and decitabine may be a key issue in treatment of MDS, AML, and CML. From avail- able studies, it appears that decitabine may have a milder adverse effect profile than 5-AzaC, which may influence clinical decisions.

The use of hypomethylating agents in combination with other chemotherapeutic agents is being addressed in ongo- ing clinical trials of both solid tumors and hematologic ma- lignancies. The medicinal chemistry of hypomethylating agents is well known, yet alterations in the patterns of gene expression and their implications on clinical response are still under intense investigation. Although the major role of DNA hypermethylation may indeed involve selective down-regulation of tumor suppressor genes, it is notewor- thy that the full ramification of this effect is not merely si- lencing genes responsible for tumor suppression, but also DNA repair and cell proliferation. Understanding of these changes in gene expression may be critical to optimize therapy and prevent resistance to these new agents.