DNA Methyltransferase and Histone Deacetylase Inhibitors in the Treatment of Myelodysplastic Syndromes

Original Article

Six (or More) Drugs in Search of a Mechanism: DNA Methyltransferase and Histone Deacetylase Inhibitors in the Treatment of Myelodysplastic Syndromes

Steven D. Gore, MD, Baltimore, Maryland

Key Words negatively impact on a particular cellular pathway or reg- Epigenetics, chromatin, gene silencing, methyltransferase, histone ulatory mechanism critical for the survival, proliferation, deacetylase, myelodysplastic syndromes or metastatic potential of the neoplasm. To date, targeted therapies include drugs that appear truly target-specific Abstract but were developed empirically (e.g., all-trans retinoic The clinical activity of the DNA methyltransferase inhibitors 5-azac- acid [ATRA] in acute promyelocytic leukemia); anti- itidine and 2 -deoxy-5-azacytidine in myelodysplastic syndromes body-based approaches that target specific cancer cells (MDS) suggests that epigenetic modulation of gene transcription may play an important pathogenetic role in the development and expres- and normal counterparts (e.g., rituximab, denileukin); sion of these diseases. Approximately 50% of patients treated with active drugs designed to target a particular pathway but these compounds experience hematologic improvement, making whose clinical activity may not derive from the original these the most active single agents for unselected patients with target (e.g., farnesyl transferase inhibitors in acute myeloid MDS. Responses include complete and partial hematologic responses. leukemia without ras mutations); and active drugs de- Two randomized trials have shown that the use of these drugs significantly alters the natural history of MDS compared with support- signed to target a pathway whose activity does derive ive care. Histone deacetylase inhibitors, which may also impact the from the original target (e.g., imatinib mesylate). expression of genes through epigenetic mechanisms, seem to have The explosive growth of knowledge concerning epi- measurable activity in MDS in preliminary studies. Histone deacety- genetic regulation of gene transcription has led to con- lase inhibitors are most likely used in combination with other agents, siderable interest in the development of therapies that including DNA methyltransferase inhibitors. Despite the clinical ac- modulate gene transcription in cancer cells pharmacolog- tivity of these classes of drugs, there is no conclusive evidence that their clinical activity is attributable to their impact on the epigenome. ically. Epigenetic changes refer to heritable but poten- Such information will be critical in the development of more effec- tially reversible modifications of DNA and chromatin tive congeners and drug combinations in ongoing attempts to im- that regulate gene transcription but that are not caused prove the outcome of patients with MDS. (JNCCN 2006;4:83–90) by irreversible changes such as mutations or deletions. The best-studied epigenetic changes are cytosine methy- The notion of targeted therapies in cancer implies the ra- lation, particularly in CpG dinucleotides in gene pro- tional design or identification of drugs that positively or moter regions, associated with transcriptional silencing,1 and a complex variety of posttranslational modifications to lysine-rich tails of histones, associated with specific From the Sidney Kimmel Comprehensive Cancer Center at Johns transcriptional output. These changes are collectively Hopkins, Baltimore, Maryland. 2 Submitted September 2, 2005; accepted for publication October 10, referred to as the histone code. 2005. Transcriptional silencing associated with methyla- This work was supported in part by grant K24CA111717-01 from the National Cancer Institute. tion of CpG islands in gene promoters is widespread in Correspondence: Steven D. Gore, MD, 1650 Orleans Street, Baltimore, cancer.1 Although these and other epigenetic changes Maryland 21231. E-mail: [email protected] seem to be attractive targets for cancer therapeutics, such

Gore

approaches are currently limited by the complexity of The mechanisms underlying these events are not epigenetic transcriptional control and the lack of speci- well understood. ficity of available pharmacologic agents. Nonetheless, Researchers investigated 5AC for treating MDS reversal of aberrant promoter methylation can be ef- after results showed it could be effective for treating fected through inhibition of the enzymes responsible hemoglobinopathies through induction of hemoglo- for maintaining the methylation pattern, DNA bin F.11–13 The investigation was based on the concept methyltransferases (DNMT). In addition, various com- of effecting more normal differentiation in hematopoi- pounds that lead to histone lysine acetylation through etic cells. In the initial trial of 5AC in MDS, it was inhibition of histone deacetylases (HDAC) are un- administered as a continuous intravenous infusion, der clinical investigation. 75 mg/m2/d for 7 days every 28 days to patients with The historical development of DNMT inhibitors high-grade MDS (refractory anemia with an excess of for the treatment of myelodysplastic syndromes (MDS) blasts [RAEB] and RAEB in transformation [RAEB- most resembles the development of ATRA for acute t] according to French-American-British [FAB] cri- promyelocytic leukemia. Development of these drugs teria). This dose schedule was taken directly from was based on empiric in vitro observations unrelated studies in hemoglobinopathies, with no formal dose- to the knowledge of specific methylated genes in- finding undertaken. In this trial, 21 of 43 patients volved with the pathogenesis or expression of these dis- showed a response, including 5 complete remissions, orders. The identification of the DNMT inhibitors 11 partial responses (PR; a minimum 50% restora- 5-azacitidine (5AC) and 2-deoxy-5-azacitidine tion of the deficit from normal of all 3 peripheral (decitabine [DAC]) as the two most active single blood cell lines, elimination of transfusion require- agents for the treatment of MDS (with the probable ments, and a decrease in percentage bone marrow exception of lenalidomide in patients with abnormal- blasts by at least 50% from prestudy values), and 5 ities of chromosome 5q3) demonstrates the clinical lineage responses. Median response duration was 15 feasibility and activity of drugs that target epigenetic months.14 In a successor phase II study, the drug was changes. However, the mechanism underlying the administered subcutaneously at the same dose sched- clinical activity of these drugs remains uncertain. This ule. This study treated 67 patients with high-risk review details the development of DNMT and HDAC MDS; comparable results were presented in abstract inhibitors in MDS and the current understanding of form.15 the relationship between these drugs’ putative mol- In Cancer and Leukemia Group B (CALGB), ecular targets and their clinical activity. Silverman et al.16 conducted a confirmatory phase III trial in which patients with all FAB subtypes of MDS and chronic myelomonocytic leukemia (CMML) were Clinical Activity of DNA randomly assigned to the subcutaneous administra- Methyltransferase Inhibitors in MDS tion of 5AC or supportive care. Patients with low-risk Both 5AC and DAC were initially studied as classic MDS were required to have significant hematopoi- cytotoxic agents for the treatment of acute myeloid etic failure. After 4 months of supportive care, pa- leukemia (AML).4,5 Jones and Taylor6 noted that tients in the control arm who met specific exit criteria treatment of mouse embryo cells with 5AC inhib- could cross over to receive 5AC. Patients in the sup- ited the methylation of newly synthesized DNA in portive care arm who developed AML as defined by a dose-dependent fashion. This treatment was asso- more than 30% blasts could receive 5AC if blast per- ciated with induction of cell differentiation to form centage did not exceed 40%. Patients in the 5AC arm muscle cells. Both drugs can reverse the methylation had dosage increased by 33% if no response was noted of a wide variety of gene promoters, often associated after 57 days (2 cycles of treatment). Patients experi- with re-expression of the transcriptionally silenced encing complete response (CR) received 3 additional gene products.1 Exposure to azacytosine analogues is cycles of drug; patients with less than CR continued associated with induction of p21WAF1/CIP1 (an unmethy- to receive drug until CR or relapse. lated gene) and concomitant cell cycle arrest and This phase III trial of 191 patients established this can induce apoptosis.7,8 In addition to p21, 5AC and regimen as the most active single-agent therapy for DAC induce expression of p53 and GADD45.9,10 MDS and led to the approval of 5AC by the Food and

Treatment of Myelodysplastic Syndromes

Drug Administration for the treatment of all subsets A phase III study of 170 patients with MDS or of MDS patients. The hematologic response rate in the CMML categorized as intermediate-1, intermediate- 5AC arm was 60%, including CR (7%), PR (16%), 2, or high-risk by the International Prognostic Scoring and hematologic improvement (37%), compared with System compared the same dose schedule of DAC 5% hematologic improvement in the observation arm with supportive care.21 Patients were to receive 2 cy- (in each case, improved neutrophils or platelet counts cles after CR occurred; patients with less than CR in the context of progression to AML). were to continue treatment until disease progression. In patients initially assigned to supportive care In the treatment arm, 9% of patients developed CR, who crossed over, the response to 5AC was 49% (10% 8% developed PR, and 13% experienced hematologic CR, 4% PR, 33% hematologic improvement). Median improvement. In contrast, only 7% of patients expe- time to initial response was 64 days; median time to rienced hematologic improvement in the supportive best response was 93 days. Median duration of response care arm. Median time to response was 3.2 months, was 15 months. Toxicities included mild nausea and with median duration of response 9.5 months. Median vomiting and worsening cytopenias before develop- time to AML or death was 12.1 months in the DAC ment of hematologic responses. Patients assigned to the arm compared with 7.8 months in the supportive care 5AC arm had a median time to AML progression or arm.This difference was not statistically significant.22 death of 21 months, compared with 12 months among A randomized phase II study in which patients re- patients in the supportive care arm (including those ceived doses according to 1 of 3 schedules, each de- who crossed over to receive 5AC). For 15% of the livering 100 mg/m2 of DAC per course, attempted to patients, MDS transformed to AML as a first event, establish optimum dose scheduling of DAC. These compared with 38% in the supportive care arm. A schedules consisted of 10 mg/m2 intravenously over 1 companion quality-of-life study showed improvement hour daily for 10 doses, 20 mg/m2 intravenously over in physical functioning, dyspnea, psychosocial dis- 1 hour daily for 5 doses, or 20 mg/m2 subcutaneously tress, and positive affect in patients on 5AC. Patients daily for 5 doses. All dose schedules resulted in CRs.23 in the supportive care arm had stable or worsening The parallel experience with 5AC and DAC in quality-of-life parameters. The researchers noted im- MDS provides compelling evidence that DNMT in- provements in fatigue, physical functioning, dyspnea, hibitors as a class have important clinical activity for and well-being after crossover.17 the treatment of MDS. Both drugs induce hemato- DAC underwent parallel development for the logic responses in approximately 50% to 60% of pa- treatment of MDS. As with 5AC, formal dose finding tients, including complete clinical and cytogenetic was not performed. DAC was initially administered as responses. Although pharmacodynamic differences a 3-day continuous infusion to 29 elderly patients with may exist between the drugs because 5AC is incorpo- high-risk MDS either at 40 or 50 mg/m2/d. Of these rated into RNA and DNA24 whereas DAC is incorpo- patients, 15 showed a response, including 8 CRs.18 In rated into DNA alone, current differences in efficacy subsequent studies, DAC was administered as a 4-hour and toxicity probably relate to differences in relative intravenous infusion, 3 times daily, for 3 days (total dose schedules and routes of administration. daily dose, 45 mg/m2). Subjects included patients with A critical difference in study design is in duration high-risk MDS or CMML and transfusion-dependent of therapy. The CALGB phase III trial of 5AC spec- patients with refractory anemia and refractory ane- ified ongoing treatment of patients who experienced mia with ringed sideroblasts. Courses were repeated less than CR; patients who experienced CR received every 6 weeks for 6 cycles.19 3 additional cycles after remission. In contrast, the In these studies, 66 patients were treated and 32 European studies of DAC administered only 6 cycles patients showed response, including 9 trilineage and of treatment. The greater remission duration in the 11 bilineage responses. Median response duration was 5AC study may be because of the prolonged adminis- 31 weeks. Cytogenetic responses were analyzed in 61 tration of drug to most patients with response, sug- patients with clonal cytogenetic abnormalities treated gesting a potential role for maintenance therapy. on either of the previously mentioned trials or on a Although the North American phase III trial called subsequent compassionate-use trial. Sixteen patients for ongoing treatment of patients with less than PR, developed complete cytogenetic responses.20 the median number of cycles delivered was only 2.

Gore

The current empiric dosing schedules are probably determine incorporation of the labeled cytosine into not optimal for either drug for the treatment of MDS. DNA-associated cytosine and 5-methylcytosine. Practice patterns in the community have not yet Although total incorporation of the labeled cytosine been studied. Anecdotally, practitioners have been significantly decreased, the ratio of 5-methylcytosine challenged with the administration of the drug in of- to total cytosine also decreased by 70% to 80%, sug- fice settings for 7 sequential days. The expected hema- gesting that methylation was specifically diminished. tologic toxicity of 5AC during the first cycles of The dose of DAC administered was high by current administration and the requirement of 4 cycles to as- standards; in addition, the impact of in vivo adminis- sess response have not been appreciated by all practi- tration of other antimetabolites on the relative 5- tioners and may have led to some patients receiving methylcytosine:total cytosine ratio was not studied. inadequate trials of this drug. Similar problems may Recent studies directly investigating reversal of apply to DAC if it becomes commercially available. promoter methylation in cells from patients treated with DNMT inhibitors have been inconclusive. Lübbert et al.27 used a polymerase chain reaction Relationship Between Clinical Activity (PCR)–based technique (Ms-SNuPE) to detect and Epigenetic Changes changes in methylation of the p15INK4B promoter in The marked activity of DNMT inhibitors for the treat- bone marrow mononuclear cells of MDS patients ment of MDS affords the opportunity to establish treated with DAC. p15 is currently the gene known whether the clinical application of these drugs can be to be most frequently methylated in primary samples 28–30 attributed to re-expression of cell regulatory genes of MDS bone marrow cells ; methylation of p15 is through reversal of promoter methylation and pre- correlated with bone marrow blast percentage in these 31 sumed downstream changes in chromatin conforma- patients. p15 methylation was quantitatively de- 27 tion and gene expression. These remain important creased in 9 of 12 evaluable DAC-treated patients. unresolved questions. Three evaluable patients who experienced CR all Analytic techniques for measuring plasma con- showed decreased p15 methylation; however, methy- centrations of 5AC have only recently been devel- lation reversal was not required for response in other oped. Administration of 5AC subcutaneously at doses patients with lesser responses. Bisulfite sequencing of from 25 to 75 mg/m2 in patients with hematologic ma- the p15 promoter in 4 patients suggested that decitabine lignancies and solid tumors led to dose-dependent in- treatment induced gradual changes in promoter methylation, rather than elimination of clones containing creases in maximal plasma concentration (Cmax), with a half-life of 0.75 to 1.5 hours. At 50 mg/m2, the methylated p15 promoters through cytotoxicity. Issa et al.32 examined p15 methylation in patients mean Cmax was 706 ng/mL, or 2.9 mcmol/L. This concentration should be adequate for DNMT inhibi- treated in a dose-finding study of DAC administered tion.25 In a separate study, mean peak plasma concen- at significantly lower doses than in the European stud- tration after subcutaneous administration of 75 mg/m2 ies and U.S. phase III study. The investigators studied was 687 ng/mL, and mean half-life was 0.69 hours.26 methylation in peripheral blood mononuclear cells Pharmacokinetics of DAC have not been determined using a PCR-based technique known as COBRA using current analytic techniques, nor have they been (combined bisulphite restriction analysis). No signif- performed for doses currently under investigation. icant differences were found in the mean percentage Momparler et al.5 initially addressed the issue of of p15 promoter methylation before treatment be- whether clinical administration of azacytosine nucle- tween patients who experienced response and those osides is associated with changes in DNA methyla- who did not nor in the change in mean or median p15 tion. The investigators isolated leukemic cells from 2 methylation. However, by the investigators’ admis- patients (acute lymphoblastic leukemia and chronic sion, the cells studied (peripheral blood mononuclear myelogenous leukemia in blast phase) receiving cells) may not have been optimally informative and decitabine as a continuous infusion (1 mg/kg/h). The the assay used may not have been sufficiently robust cells were incubated with radiolabeled deoxycytidine to detect biologically important changes. Additionally, before isolation and hydrolysis of the DNA. High- focusing on mean percentage of methylation may have pressure liquid chromatography was then used to obscured important changes in individual patients.

Treatment of Myelodysplastic Syndromes

This group is now focusing on the methylation of non- application of such approaches to MDS comes from the coding repetitive DNA elements, such as Alu and activity of DNMT inhibitors in MDS, suggesting long interspersed nucleotide elements, as a potential important epigenetic mechanisms affecting MDS clin- surrogate for gene-specific methylation studies.33 ical behavior. Compounds with HDAC inhibitory Using micellular electrokinetic chromatography activity typically induce p21WAF1/CIP1 expression con- to assess genomic methylation, Lyko et al.34 demon- comitant with cell cycle arrest and induction of strated decreased methylation in 3 of 4 MDS patients terminal differentiation.35 This effect led to the early treated with DAC. Methylation was studied follow- application of these drugs in myeloid malignancies ing the first cycle of DAC in only 2 patients. as potential differentiation inducers. Significantly decreased methylation could be detected The earliest reported use of a compound with after cycle 2 in one patient, cycle 4 in a second pa- HDAC inhibitory activity in a myeloid malignancy tient, and cycle 5 in the third. However, when evalu- was a case report of a pediatric patient with relapsed able, karyotype normalization preceded changes in AML induced into remission using sodium butyrate.36 genomic methylation; thus, the methylation changes Results showing in vitro activity of the butyrate ana- measured likely occurred in normal cells and were logue sodium phenylbutyrate (NaPB) in the growth likely irrelevant to proximal molecular changes in- arrest and differentiation of primary AML samples37,38 duced by DAC in the early treatment cycles. The led to phase I studies of continuous infusion of NaPB small number of patients studied did not allow eval- in patients with MDS and AML. At a maximum tol- uation of the relationship between demethylation erated dose of 375 mg/kg/d administered as an intra- and clinical response. venous continuous infusion for 7 days, sustained The current data do not allow determination of plasma concentrations of 0.3 mM were achieved, the relevance of methylation changes to clinical re- within the range of concentrations with HDAC in- sponse to DNMT inhibitors. Studies are hampered hibitory activity.39,40 This dose was tolerated by most by the lack of a comprehensive panel of genes that are patients when administered for 7 of 28 days, 7 of 14 methylated in MDS and whose methylation status is days, and 21 of 28 days.39,41 Dose-limiting toxicity was known to impact the phenotype and biologic behav- a reversible encephalopathy apparently caused by ior of the disease. Studies have not examined methy- accumulation of the metabolite phenylacetate.39 lation changes in different cellular compartments. Hematologic improvement was seen in a small num- Most studies have not examined changes early after ber of patients. drug treatment. Treatment with DNMT inhibitors Valproic acid (VPA) has HDAC inhibitory ac- may lead to activation of molecular pathways en- tivity similar to other small chain fatty acids.42,43 The abling terminal differentiation, apoptosis, or senes- clinical availability of oral formulations of VPA has led cence. Through combining genomic methylation to various trials evaluating VPA as a putative HDAC assays with genomic gene expression profiling in inhibitor. A small study of patients with MDS or MDS- treated patients, researchers may identify important related AML showed promising activity, with responses recurrent patterns that explain clinical response. in 8 of 18 patients. This activity included a PR; how- Alternatively, the clinical activity of this class of drugs ever, response duration was brief (median 4 months).44 may be secondary to activation of pathways in re- As with the DNMT inhibitors, the relationship be- sponse to DNA damage.7–10 tween clinical activity of these small chain fatty acids and their ability to inhibit HDAC and impact gene transcription has not yet been demonstrated. Clinical Development of Histone Several more potent and potentially more Deacetylase Inhibitors in MDS specific HDAC inhibitors are under clinical investi- Increasing understanding of the important roles that gation. These include the hydroxamic acid sub- post-translational modifications of histones play in eroylanilide hydroxamic acid (SAHA), the cyclic establishing chromatin confirmation and gene tran- peptide FK228 (depsipeptide), and the benzamide scription has led to interest in pharmacologic target- MS-275.45 The results of recent and current studies of ing of these changes to effect normal transcription these compounds in myeloid malignancies are not yet patterns in malignant cells. Particular interest in the published.

Gore

Combination Therapies Targeting occur in 50% to 60% of patients in low- and high-risk Epigenetic Mechanisms subsets. The mechanisms underlying the clinical ac- Although monotherapy trials may help clarify the tivity of DNMT inhibitors in this group of diseases mechanism of drugs that putatively target epigenetic are unclear, and dissection of these mechanisms is crit- control of transcription, greatest clinical benefit may ical to develop of better congeners and effective com- ultimately derive from biologically rational drug combination therapies. Similarly, first generation HDAC binations. Binding of ligands to nuclear hormone re- inhibitors were studied in MDS as putative differen- ceptors, such as the retinoic acid receptor, leads to the tiating agents and have mild-to-moderate hemato- release of transcriptional corepressors, including logic activity. Based on promising preclinical models, HDACs.46 Butyrate, phenylbutyrate, and VPA inter- combinations of DNMT and HDAC inhibitors form act synergistically with ATRA on myeloid leukemia the basis of important ongoing clinical trials. cells.40,42,47,48 NaPB and ATRA successfully induced Randomized trials that show that DNMT inhibitors remission in ATRA-resistant acute promyelocytic prolong time to progression to AML or death suggest leukemia.49 Preliminary reports of VPA plus ATRA a very important class activity in the natural history suggest that this combination may have activity in of MDS. Unraveling the mechanism underlying this MDS. Three of 11 elderly patients with AML devel- delay in disease progression might provide important oped complete marrow responses to this combina- insight into the biology of disease progression. tion.50 In another study, 6 of 20 elderly patients with AML or MDS developed hematologic improve- References ment.51 1. Jones PA, Baylin SB. The fundamental role of epigenetic events in Several current trials focus on the combination cancer. Nat Rev Genet 2002;3:415–428. of DNMT and HDAC inhibitors. Specific methyl- 2. Rice JC, Allis CD. Code of silence. Nature 2001;414:258–261. binding proteins bind to methylated CpG dinu- 3. List A, Kurtin S, Roe DJ, et al. Efficacy of lenalidomide in myelodys- cleotides and recruit transcriptional corepressors, plastic syndromes. N Engl J Med 2005;352:549–557. including HDAC.52 The sequential addition of HDAC 4. Von Hoff DD, Slavik M, Muggia FM. 5-Azacytidine. A new anti- cancer drug with effectiveness in acute myelogenous leukemia. Ann inhibitors after DNMT inhibitors leads to additive or Intern Med 1976;85:237–245. synergistic re-expression of methylated genes; HDAC 5. Momparler RL, Bouchard J, Onetto N, et al. 5-aza-2-deoxycytidine inhibitors alone do not induce significant re-expres- therapy in patients with acute leukemia inhibits DNA methylation. sion of heavily methylated genes.53 Trials combining Leuk Res 1984;8:181–185. 5AC with NaPB and DAC with VPA in MDS and 6. Jones PA, Taylor SM. Cellular differentiation, cytidine analogs and DNA methylation. Cell 1980;20:85–93. AML have been completed but are not yet published.54 7. Schmelz K, Sattler N, Wagner M, et al. Induction of gene expres- Current and upcoming trials combine 5AC with MS- sion by 5-Aza-2-deoxycytidine in acute myeloid leukemia (AML) 275, DAC with SAHA, and 5AC with SAHA. and myelodysplastic syndrome (MDS) but not epithelial cells by Confirmation of any clinical benefit suggested by these DNA-methylation-dependent and -independent mechanisms. Leukemia 2005;19:103–111. combination trials will require randomized studies 8. Schmelz K, Wagner M, Dorken B, Tamm I. 5-Aza-2-deoxycytidine comparing the combination to the DNMT inhibitor induces p21WAF expression by demethylation of p73 leading to alone. p53-independent apoptosis in myeloid leukemia. Int J Cancer 2005;114:683–695. 9. Schneider-Stock R, Diab-Assef M, Rohrbeck A, et al. 5-Aza-cytidine is a potent inhibitor of DNA methyltransferase 3a and induces Conclusions apoptosis in HCT-116 colon cancer cells via Gadd45- and p53- As with all cancers studied to date, epigenetic changes dependent mechanisms. J Pharmacol Exp Ther 2005;312:525–536. contribute to abnormal gene transcription in MDS. 10. Zhu WG, Hileman T, Ke Y, et al. 5-aza-2-deoxycytidine activates the p53/p21Waf1/Cip1 pathway to inhibit cell proliferation. J Biol Because of long-standing interest in effecting more Chem 2004;279:15161–15166. normal differentiation in MDS bone marrow through 11. Charache S, Dover G, Smith K, et al. Treatment of sickle cell the use of small molecules, prototypes for classes of anemia with 5-azacytidine results in increased fetal hemoglobin drugs that target epigenetic changes were tested in production and is associated with nonrandom hypomethylation of DNA around the gamma-delta-beta globin gene complex. Proc Natl MDS before the term epigenetic was coined. DNMT in- Acad Sci USA 1983;80:4842–4846. hibitors represent the most active class of single agents 12. Dover G, Charache S, Boyer SH, et al. 5-azacytidine increased HbF for unselected MDS patients; hematologic responses production and reduces anemia in sickle cell disease. Dose-response

Treatment of Myelodysplastic Syndromes

analysis of subcutaneous and oral dosing regimens. Blood 1985; 30. Herman JG, Jen J, Merlo A, et al. Hypermethylation-associated in- 66:532. activation indicates a tumor suppressor role for p15(INK4B). Cancer 13. Ley T, DeSimone J, Anagnou NP, et al. 5-azacytidine selectively in- Res 1996;56:722–727. creases gamma-globin synthesis in a patient with beta+ thalassemia. 31. Quesnel B, Guillerm G, Vereecque R, et al. Methylation of the N Engl J Med 1982;307:1469–1475. p15(INK4b) gene in myelodysplastic syndromes is frequent and 14. Silverman LR, Holland JF, Weinberg RS, et al. Effects of treatment acquired during disease progression. Blood 1998;91:2985–2990. with 5-azacytidine on the in vivo and in vitro hematopoiesis 32. Issa JP, Garcia-Manero G, Giles FJ, et al. Phase 1 study of low-dose in patients with myelodysplastic syndromes. Leukemia 1993; prolonged exposure schedules of the hypomethylating agent 5- 7[suppl 1]:21–29. aza-2-deoxycytidine (decitabine) in hematopoietic malignancies. 15. Silverman L, Holland JF, Demakos EP, et al. Azacitidine in myelodys- Blood 2004;103:1635–1640. plastic syndromes: CALGB studies 8421 and 8921. Ann Hematol 33. Yang AS, Estecio MR, Doshi K, et al. A simple method for estimat- 1994;68. ing global DNA methylation using bisulfite PCR of repetitive DNA 16. Silverman LR, Demakos EP, Peterson BL, et al. Randomized con- elements. Nucleic Acids Res 2004;32:e38. trolled trial of azacitidine in patients with the myelodysplastic 34. Mund C, Hackanson B, Stresemann C, et al. Characterization of DNA syndrome: a study of the cancer and leukemia group B. J Clin demethylation effects induced by 5-Aza-2-deoxycytidine in patients Oncol 2002;20:2429–2440. with myelodysplastic syndrome. Cancer Res 2005;65:7086–7090. 17. Kornblith AB, Herndon JE, Silverman LR, et al. Impact of azacyti- 35. Rosato RR, Grant S. Histone deacetylase inhibitors: insights into dine on the quality of life of patients with myelodysplastic syndrome mechanisms of lethality. Expert Opin Ther Targets 2005;9:809–824. treated in a randomized phase III trial: a Cancer and Leukemia Group 36. Novogrodsky A, Dvir A, Ravid A, et al. Effect of polar organic com- B study. J Clin Oncol 2002;20:2441–2452. pounds on leukemic cells. Butyrate-induced partial remission of acute 18. Wijermans PW, Krulder JW, Huijgens PC, et al. Continuous infu- myelogenous leukemia in a child. Cancer 1983;51:9–14. sion of low-dose 5-Aza-2-deoxycytidine in elderly patients with 37. Gore SD, Samid D, Weng LJ. Impact of the putative differentiating high-risk myelodysplastic syndrome. Leukemia 1997;11[suppl 1]: agents sodium phenylbutyrate and sodium phenylacetate on prolif- 19–23. eration, differentiation, and apoptosis of primary neoplastic myeloid 19. Wijermans P, Lubbert M, Verhoef G, et al. Low-dose 5-aza-2-deoxy- cells. Clin Cancer Res 1997;3:1755–1762. cytidine, a DNA hypomethylating agent, for the treatment of high- 38. DiGiuseppe JA, Weng LJ, Yu KH, et al. Phenylbutyrate-induced G1 risk myelodysplastic syndrome: a multicenter phase II study in elderly arrest and apoptosis in myeloid leukemia cells: Structure-Function patients. J Clin Oncol 2000;18:956–962. Analysis. Leukemia 1999;13:1243–1253. 20. Lubbert M, Wijermans P, Kunzmann R, et al. Cytogenetic responses 39. Gore SD, Weng LJ, Zhai S, et al. Impact of the putative differenti- in high-risk myelodysplastic syndrome following low-dose treatment ating agent sodium phenylbutyrate on myelodysplastic syndromes and with the DNA methylation inhibitor 5-aza-2-deoxycytidine. acute myeloid leukemia. Clin Cancer Res 2001;7:2330–2339. Br J Haematol 2001;114:349–357. 40. Yu KH, Weng LJ, Fu S, Gore SD. Augmentation of phenylbutyrate- 21. Greenberg P, Cox C, LeBeau MM, et al. International scoring sys- induced differentiation of myeloid leukemia cells using all trans- tem for evaluating prognosis in myelodysplastic syndromes. Blood retinoic acid. Leukemia 2000;13:1258–1265. 1997;89:2079–2088. 41. Gore SD, Weng LJ, Figg WD, et al. Impact of prolonged infusions 22. Saba HI, Wijermans PW. Decitabine in myelodysplastic syndromes. of the putative differentiating agent sodium phenylbutyrate on Semin Hematol 2005;42[2 suppl 2]:23–31. myelodysplastic syndromes and acute myeloid leukemia. Clin Cancer 23. Kantarjian HM, Issa JP. Decitabine dosing schedules. Semin Hematol Res 2002;8:963–970. 2005;42[suppl 2]:17–22. 42. Gottlicher M, Minucci S, Zhu P, et al. Valproic acid defines a novel 24. Plagemann PG, Behrens M, Abraham D. Metabolism and cytotox- class of HDAC inhibitors inducing differentiation of transformed icity of 5-azacytidine in cultured Novikoff rat hepatoma and P388 cells. EMBO J 2001;20:6969–6978. mouse leukemia cells and their enhancement by preincubation with 43. Gurvich N, Tsygankova OM, Meinkoth JL, Klein PS. Histone pyrazofurin. Cancer Res 1978;38:2458–2466. deacetylase is a target of valproic acid-mediated cellular differenti- 25. Rudek MA, Zhao M, He P, et al. Pharmacokinetics of 5-azacitidine ation. Cancer Res 2004;64:1079–1086. administered with phenylbutyrate in patients with refractory 44. Kuendgen A, Strupp C, Aivado M, et al. Treatment of myelodysplas- solid tumors or hematologic malignancies. J Clin Oncol 2005;23: tic syndromes with valproic acid alone or in combination with all- 3906–3911. trans retinoic acid. Blood 2004;104:1266–1269. 26. Marcucci G, Silverman L, Eller M, et al. Bioavailability of azaciti- 45. Gore SD, Carducci MA. Modifying histones to tame cancer: clini- dine subcutaneous versus intravenous in patients with the myelodys- cal development of sodium phenylbutyrate and other histone deacety- plastic syndromes. J Clin Pharmacol 2005;45:597–602. lase inhibitors. Expert Opin Invest Drugs 2000;9:2923–2934. 27. Daskalakis M, Nguyen TT, Nguyen C, et al. Demethylation of a hy- 46. Minucci S, Pelicci PG. Retinoid receptors in health and disease: permethylated P15/INK4B gene in patients with myelodysplastic co-regulators and the chromatin connection. Semin Cell Dev syndrome by 5-Aza-2-deoxycytidine (decitabine) treatment. Blood Biol 1999;10:215–225. 2002;100:2957–2964. 47. Breitman TR, He R. Combinations of retinoic acid with either 28. Cameron EE, Baylin SB, Herman JG. P15INK4b CpG island methy- sodium butyrate, dimethyl sulfoxide, or hexamethylene bisacetamide lation is heterogeneous in primary acute leukemia and suggests den- synergistically induce differentiation of the human myeloid leukemia sity as a critical factor in transcriptional silencing. Blood (in press). cell line HL60. Cancer Res 1990;60:6268–6273. 29. Herman JG, Civin CI, Issa JP, et al. Distinct patterns of inactivation 48. Trus MR, Yang L, Suarez SF, et al. The histone deacetylase inhibitor of p15INK4B and p16INK4A characterize the major types of hema- valproic acid alters sensitivity towards all trans retinoic acid in acute tological malignancies. Canc Res 1997;57:837–841. myeloblastic leukemia cells. Leukemia 2005;19:1161–1168.

Gore

49. Warrell RP Jr, He LZ, Richon V, et al. Therapeutic targeting of tran- 52. Nan X, Ng HH, Johnson CA, et al. Transcriptional repression by the scription in acute promyelocytic leukemia by use of an inhibitor of methyl-CpG-binding protein MeCP2 involves a histone deacetylase histone deacetylase. JNCI 1998;90:1621–1625. complex. Nature 1998;393:386–389. 50. Raffoux E, Chaibi P, Dombret H, Degos L. Valproic acid and all- 53. Cameron EE, Bachman KE, Myohanen S, et al. Synergy of trans retinoic acid for the treatment of elderly patients with acute demethylation and histone deacetylase inhibition in the re-ex- myeloid leukemia. Haematologica 2005;90:986–988. pression of genes silenced in cancer. Nature Genet 1999;21: 51. Pilatrino C, Cilloni D, Messa E, et al. Increase in platelet count in 103–107. older, poor-risk patients with acute myeloid leukemia or myelodys- 54. Garcia-Manero G, Gore SD. Future directions for the use of plastic syndrome treated with valproic acid and all-trans retinoic hypomethylating agents. Semin Hematol 2005;42(suppl 2): acid. Cancer 2005;104:101–109. 50–59.