Evaluation of a 7-Day Continuous Intravenous Infusion of Decitabine: Inhibition of Promoter-Specific and Global Genomic DNA Methylation
http://www.100md.com
《临床肿瘤学》
the Huntsman Cancer Institute and the Departments of Internal Medicine (Oncology), Dermatology, Oncological Sciences, and Medicinal Chemistry, University of Utah, Salt Lake City, UT
ABSTRACT
PURPOSE: The nucleoside analog 5-aza-2'-deoxycytidine (5-aza-CdR, decitabine) is a potent inhibitor of DNA methylation in vitro. Cellular treatment with this agent induces the re-expression of methylation-silenced genes. It remains unclear to what extent this compound inhibits DNA methylation in vivo. A clinical study was designed to examine the molecular effects and toxicity of a continuous 1-week intravenous infusion of decitabine in solid tumor patients.
METHODS: Ten patients with refractory solid tumors were included in this study. Decitabine was administered at 2 mg/m2/d via continuous infusion for 168 hours. Quantitative polymerase chain reaction and high performance liquid chromatography were utilized to measure promoter-specific and global DNA methylation in peripheral-blood cells before and after treatment.
RESULTS: Transient grade III/IV neutropenia (two patients) and grade II thrombocytopenia (one patient) was observed at the lowest planned dose step (2 mg/m2/d for 7 days). Nonhematologic toxicities were not observed. Quantitative polymerase chain reaction demonstrated significant MAGE-1 promoter hypomethylation by 14 days after the start of treatment in all 13 treatment cycles examined. Significant genomic DNA hypomethylation was also seen by day 14 in 11 of 13 treatment cycles analyzed. Genomic DNA methylation reverted to baseline levels by 28 to 35 days after the start of treatment, demonstrating that inhibition of DNA methylation by decitabine is transient.
CONCLUSION: A 168-hour continuous infusion of decitabine is well tolerated and results in the inhibition of promoter-specific and genomic DNA methylation in vivo. This treatment schedule is suitable for evaluation of decitabine in combination with agents whose activity may be enhanced by the reversal of DNA methylation–mediated gene silencing.
INTRODUCTION
The recognition of aberrant DNA methylation as a contributing factor to tumorigenesis has resulted in the testing of DNA methyltransferase inhibitors as anticancer agents.1-3 The cytosine analogues 5-aza-2'-deoxycytidine (5-aza-2'-CdR or decitabine) and 5-aza-cytidine (5-aza-CR) are potent pharmacologic inhibitors of DNA methylation.2,4 The activity of these drugs requires incorporation into cellular DNA, and the subsequent covalent sequestration of DNA methyltransferase enzymes.5-7 A number of studies have demonstrated that decitabine treatment elicits the reactivation of methylation-silenced tumor suppressor genes and causes phenotypic reversion in cancer cell lines.8-11
In clinical studies, DNA methyltransferase inhibitors have shown activity in the treatment of chronic myelogenous leukemia, sickle cell anemia, and myelodysplastic syndrome (MDS).12-20 Notably, decitabine treatment has induced clinical responses in approximately 50% of patients with MDS, with myelosuppression being the major adverse effect.17 Decitabine has achieved orphan drug status for this malignancy. At the molecular level, cytogenetic responses to decitabine have been noted in approximately one third of MDS patients who presented with clonal chromosomal abnormalities.18 Issa et al21 recently demonstrated that low-dose prolonged exposure to decitabine correlated with clinical responses in hematologic malignancies. Furthermore, Dasalakis et al19 and Issa et al21 have presented evidence showing that decitabine treatment of MDS resulted in the hypomethylation of the p15INK4b gene in vivo.
In contrast to results observed in the treatment of hematologic disorders, DNA methyltransferase inhibitors have, to date, shown little activity as single agents in the treatment of solid tumors.7,17,22,23 Our group has put forward the hypothesis that DNA methyltransferase inhibitors may be best utilized as a treatment for solid tumors as part of a combination chemotherapy approach.3 Successful combination chemotherapy may be expected to require effective inhibition of DNA methylation by decitabine treatment before the use of a secondary agent. To test the potential utility of this approach, we attempted to develop a clinical protocol using decitabine that inhibited DNA methylation without high levels of systemic toxicity. Another goal was to develop assays that could identify methylation changes in human clinical samples.
We hypothesized that long-term (7-day) continuous exposure would allow for greater incorporation of decitabine into DNA, thus enhancing hypomethylation of DNA and increasing the potential for potentially synergistic therapeutic effects when combined with other agents. We therefore designed a phase I clinical trial to evaluate the maximum tolerated dose and dose-limiting toxicities of a 7-day (168-hour) continuous intravenous (IV) infusion of decitabine in solid tumor patients. A dose escalation scheme was designed, starting at approximately 1/20 of the previous maximum tolerated short-term infusion dose. Gene-specific and genomic changes in DNA methylation in peripheral-blood mononuclear cells were monitored during the course of treatment.
METHODS
Eligibility Criteria
This phase I study was reviewed and approved by the University of Utah (Salt Lake City, UT) institutional review board in accordance with assurances filed with the US Department of Health and Human Services. Patients were required to have recurrent or metastatic cancer. All curative or survival-prolonging standard treatment options had to have been exhausted. Prior treatment with chemotherapy, radiotherapy, immunotherapy, and/or surgery was allowed. Timing of the last previous treatment had to be greater than 4 weeks before protocol entry, and patients were required to have recovered from side effects of prior treatment, except hair loss. Patients had to be 18 years of age, with an estimated life expectancy of 12 weeks and a Southwest Oncology Group performance status of 0 to 2. Laboratory-based eligibility testing was required, including an absolute granulocyte count of > 2,000/μL and a platelet count of 100,000/μL; a bilirubin of 2.0 mg/dL; and a calculated or measured creatinine clearance of > 50 mL/min. Patients with significant comorbid medical or psychosocial conditions were excluded, as were pregnant or lactating women. All patients signed acknowledgment of informed consent.
Drug Administration
The initial study design was for five patients to be enrolled in sequential dose escalation steps, starting at 2 mg/m2/d. Subsequent dose-escalation steps of 5, 10, 20, and 40 mg/m2/d were planned. Decitabine was administered as a 7-day (168-hour) continuous IV infusion via indwelling central venous catheter. Decitabine (provided by Supergen, Dublin, CA) was stored as a stable freeze-dried powder (50 mg/20 mL glass vial). Immediately before use, the content of each vial was reconstituted with Sterile Water for Injection, USP (Baxter, Deerfield, IL). The reconstituted solution was immediately diluted to a concentration of 0.1 to 1 mg/mL with ice-cold infusion fluids (either 5% dextrose in water, or 0.9% sodium chloride injection [USP, Baxter]). Decitabine was reconstituted from powder every 12 hours and kept refrigerated (for a maximum of 12 hours). Once warmed to room temperature, each bag of solution was infused over a maximum period of 3 hours before changing to a new infusion bag of drug.
Dose Escalation Schema
Patients were admitted to the Huntsman General Clinical Research Center for the entire 168-hour treatment period. The main clinical objective of the study was to identify a well-tolerated continuous infusion dose that effectively inhibited DNA methylation. Another goal was to identify dose-limiting toxicities (DLTs) of decitabine on this schedule. DLT was defined as any grade III nonhematologic and/or grade III or IV hematologic toxicity. The maximum tolerated dose (MTD) was defined as the dose level below which DLT was observed in two of five patients. If fewer than two patients experienced DLT, then the next five patients were to be treated at the next higher dose. With this design, the probability of stopping the trial due to DLT was calculated to be 8%, 47%, and 81% if the true rate of DLT is 10%, 30%, and 50%, respectively (conditional on having administered the dose level to all five patients).24,25
Treatment cycles were planned every 28 days. Dose escalation was planned for each cohort of five patients if no grade III/IV toxicity was seen. Treatment response was assessed every two cycles by computed tomography scans. Patients who had stable disease or better were continued on treatment until disease progression was observed. If grade III/IV hematologic toxicity developed during a treatment course, treatment was interrupted until toxicity decreased to grade 0 or I. Treatment was resumed once blood counts recovered to > 2,000 absolute neutrophil count and > 100,000 platelets. Treatment could be delayed for a maximum of 3 weeks to allow blood count recovery. Hematopoietic growth factor support was allowed for neutropenia or neutropenic fever. Subsequent cycles could either be administered with hematopoietic growth factor or with a dose reduction to the previous step. Criteria for removal from protocol treatment included: cancer progression, unacceptable toxicity, treatment delay > 3 weeks due to toxicity, development of grade 3 nonhematologic toxicity, or voluntary withdrawal by the patient.
Response Assessment
Complete response (CR) was defined as complete disappearance of all measurable and assessable disease without any new lesions or disease-related symptoms. Partial response (PR): was defined as 50% decrease under baseline in the sum of products of perpendicular diameters of all measurable lesions without progression at any site or any new lesions. Progression was defined as a 50% increase or an increase of 10 cm2 (whichever is smaller) in the sum of products of all measurable lesions over smallest sum observed (over baseline if no decrease) or appearance of any new lesion/site. Stable disease (SD) does not qualify for CR, PR or progression.
DNA Isolation
Patient blood samples (7 mL) were collected into sodium citrate tubes at baseline (day 0) and on days 7 and 14 days after the start of each cycle of decitabine treatment. Samples were immediately placed on ice following a blood draw from a central venous catheter. DNA extraction from whole blood was performed within 1 hour of collection, using the Puregene DNA isolation kit (Gentra Systems, Minneapolis, MN).
MAGE-1 Promoter Methylation Analysis
A number of gene-specific methylation sites were evaluated to identify a marker for the effects of decitabine on normal peripheral blood leukocytes. The promoter of the MAGE-1 gene, which is methylated in somatic tissues, proved informative.26 The methylation status of the MAGE-1 promoter was evaluated using an assay developed by De Smet et al27 (schema, Fig 1A). This assay uses enzymatic digestion of genomic DNA with a methylation-sensitive restriction enzyme (HpaII), coupled with polymerase chain reaction (PCR), to evaluate the methylation status of CpG sites present in the MAGE-1 promoter. We modified the assay by performing real-time PCR amplifications using a Lightcycler instrument (Roche, Indianapolis, IN) to derive a quantitative measure of the extent of methylation of the MAGE-1 promoter in clinical samples. DNA samples obtained from each individual patient were processed and analyzed in parallel. Each PCR reaction contained approximately 1/25 of the DNA recovered from digestion and also contained Taq (Biolase, Bioline, Reno, NV), Taq antibody (Clonetech, Palo Alto, CA), Sybr green (Molecular Probes, Eugene, OR), 1 μM of each oligonucleotide primer,27 200 nM deoxy-nucleotide triphosphates, and 1X PCR buffer (50 mmol/L Tris, 0.25 mg/mL bovine serum albumin, 3 mM MgCl2). Lightcycler PCR parameters were as follows: initial denaturing for 10 seconds at 95°, followed by 35 cycles at 95° for 1 second, 56° for 1 second, and 72° for 2 seconds. Copy numbers and melting curves were calculated using software provided with the Lightcycler instrument. For each sample, the copy number obtained from the test reaction (using CDS20-EPD4 primers) was divided by the copy number derived from a control reaction (CDS21-EPD4 primers) to generate a ratio. This ratio directly reflects the average methylation state of the MAGE-1 promoter HpaII sites. We normalized the ratio to 1.0 for the day-0 sample of each individual patient treatment cycle. All PCR reactions were performed in triplicate and were repeated at least once, on a different day.
Genomic DNA Methylation Analyses
Genomic DNA was treated with nuclease P1 (Sigma, St Louis, MO) and bacterial alkaline phosphatase (Sigma) as described by Gehrke et al28 to evaluate changes in methyl-cytosine levels in leukocyte DNA (schema, Fig 1B). Digested DNA samples (90 μL, derived from approximately 5 μg DNA) were injected onto a Beckman System Gold HPLC system equipped with a Phenomenex Luna C18, 25 x 4.6 mm column, 126 nM solvent module, a 168 nM photodiode array detector and a 507e auto sampler. HPLC was performed using solvent A (2.5% MeOH/50 mmol/L potassium phosphate pH 4.0) and solvent B (30% MeOH/50 mmol/L potassium phosphate pH 4.0). The column was eluted at a constant flow rate of 1 mL/min beginning with 100% A for 10 minutes followed by a linear gradient over a period of 15 minutes to 100% B. After holding at 100% B for 10 minutes, the column was recycled to 100% A with a linear gradient over a period of 10 minutes. Absorbance of the eluent at 200 to 300 nM was monitored, generating chromatograms that plotted absorbance at 284 nM (max of 5-Me-dC at pH 4) versus time. The ratio of 5-Me-dC to dC was determined using a calibration curve resulting from the analysis of a serial dilution of nucleosides produced from digestion of normal human DNA. Samples were assayed in triplicate and results are presented as mean ± SD.
Statistical Analysis
Repeated measures of analysis of variance (ANOVA) were used to analyze the data, using Statistica software 6.0 (Statsoft, Tulsa, OK). Logarithmic transformation was applied to the data shown in Figures 2A and B data to equalize the variance at the three time points. This was not necessary for data shown in Figures 2C and D.
RESULTS
Patient Characteristics and Toxicity Assessment
Eleven patients signed informed consent to participate in the study after eligibility testing was successfully completed (Table 1). One patient (patient No. 7) was deemed ineligible due to changes in serum bilirubin between the prestudy visit (and informed consent) and admission for treatment (< 2 weeks). Another patient (patient No. 8) was withdrawn on day 2 of decitabine infusion due to rapid progression of head and neck cancer and pre-existing edema involving the oropharynx and face, resulting in respiratory difficulty. This patient died 1 week later, without completing the first treatment cycle. Thus, nine of 11 patients were assessable for toxicity.
The initial two patients accrued to the initial dose step (2 mg/m2/d) successfully completed the infusion without apparent early toxicity. One patient subsequently developed grade III neutropenia (day 28), and the second patient developed grade IV neutropenia (141 neutrophils/μL on day 27) and grade II thrombocytopenia. Both were completely asymptomatic and the hematologic alterations reversed within 3 days without sequelae. However, because two of the first five patients demonstrated grade III or IV toxicity, dose escalation was halted at this initial dose step (see Methods). Analysis of these two patients suggested possible contributing factors: one patient had had prior extensive (18 months) treatment with the nucleoside analog gemcitabine. The second patient had completed irradiation to the central pelvis and lumbosacral spine 1 month before protocol entry. After extensive discussions with the institutional review board, and due to the absence of other adverse effects, the protocol was amended to allow an additional five patients to be treated at the 2 mg/m2/d dose (a total of 10 patients) to better assess the incidence of DLT and planned biologic end points. Patients with extensive radiotherapy to marrow-containing bones or extensive nucleoside analog pretreatment were subsequently excluded. Further escalation was not permitted, however. Closer monitoring of blood counts was subsequently performed.
The revised analysis plan was that if nonhematologic grade III or hematologic grade IV toxicity was observed in three patients overall (in 10 patients), the starting dose would be judged too high. If hematologic toxicities resulted in neutropenia < 500 cells for > 5 days, or associated fever > 38.5°C or evidence of infection was observed, this counted as a grade III/IV toxicity toward establishing the MTD. A total of seven additional patients were accrued to the study without any further grade III/IV toxicity. Evaluation of toxicity suggested that a 2.0 mg/m2/d dose of decitabine resulted in a level of toxicity suitable for further evaluation of inhibition of DNA methylation in patients. The main DLT observed was neutropenia (Table 1).
Assessment of Clinical Response
All patients had extensively pretreated solid tumors. As noted above, there were a number of early tumor-related events that occurred that indicated the advanced state of cancer in the study patients. One additional patient (patient No. 3) became symptomatic from tumor-related small intestinal obstruction 1 week following completion of decitabine infusion, which required palliative resection. Following recovery (28 days), the patient subsequently completed two additional cycles of decitabine without sequelae. Eight patients completed at least two cycles of treatment and were assessable for tumor response (Table 1). No objective responses were observed, and seven patients progressed after two cycles of decitabine treatment. Two patients had stable disease. One of these, a woman with metastatic ovarian cancer (patient No. 6), received four cycles and had stable disease for 4 months before progression. The other, a man with renal carcinoma (patient No. 10), had stable disease lasting 6 months before progression.
Low-Dose Continuous Infusion of Decitabine Inhibits Promoter-Specific DNA Methylation
We analyzed the effects of decitabine treatment on promoter-specific DNA methylation in DNA isolated from blood. For this analysis, we exploited the fact that the MAGE-1 promoter CpG island is normally methylated in somatic tissues.26 Quantitative real-time PCR was employed to measure the extent of promoter CpG methylation in clinical samples. We analyzed genomic DNA isolated from patient blood samples before treatment (day 0), immediately after termination of continuous infusion (day 7), and 1 week after the end of continuous infusion (day 14). A total of 13 patient treatment cycles were analyzed, including eight patients receiving the first cycle of decitabine and five patients receiving the second cycle (Figs 2A and B). In seven of 13 treatment cycles, a decline in MAGE-1 promoter methylation was apparent by day 7, while at day 14, all 13 decitabine cycles resulted in MAGE-1 hypomethylation (Figs 2A and B; Figs 3A and B). At day 14, the mean level of MAGE-1 methylation across all patient samples declined by 60% following cycle 1, and by 65% following cycle 2 (Figs 4A and B). The observed decline in MAGE-1 promoter methylation over the entire patient population was significant at day 7 and day 14 of the first treatment cycle (P = .005), as determined by repeated measures ANOVA (Fig 3A). Similar results were observed in the second treatment cycle (Fig 3B; P = .0029).
Low-Dose Continuous Infusion of Decitabine Inhibits Global DNA Methylation
The effect of decitabine infusion on overall methylation of genomic DNA was also evaluated, utilizing a quantitative HPLC-based assay.28 DNA from patient blood was obtained before treatment (day 0), immediately following continuous infusion (day 7), and 1 week after continuous infusion (day 14). Genomic DNA methylation was evaluated in individual patient samples derived from the first (Fig 2C; Fig 3C) and second cycle of decitabine treatment (Fig 2D; Fig 3D), provided that sufficient DNA was available (20 μg). A total of five of nine decitabine cycles analyzed showed hypomethylation at day 7, while at day 14, 11 of 13 patient samples showed hypomethylation (Figs 2C and D; Figs 3C and D). The decline in genomic DNA methylation over the entire patient population at day 14 was highly significant (P < .001) for both the first cycle (P = .0059) and second cycles (P = .0048) of decitabine treatment by repeated measures ANOVA.
Duration of Genomic DNA Hypomethylation in Patient Blood Samples
Next, we analyzed the kinetics of recovery of genomic DNA methylation levels in patient samples by comparing the level of methylation at the end of the first cycle of decitabine treatment to the level at the beginning of the second cycle. Unplanned variations in the time between the first and second cycles of treatment in individual patients were utilized to evaluate the extent of recovery to baseline methylation levels following treatment. In five of six patients, there was a significant decline in the level of genomic DNA methylation 14 days after the start of the first cycle of treatment (Fig 4). If treatment was delayed to day 35 in responding patients, global methylation reverted to baseline levels by the start of the second cycle. In contrast, DNA from the single patient retreated on day 28 (patient No. 10) remained hypomethylated as compared with the pretreatment sample.
DISCUSSION
Cytosine methylation is an epigenetic event in mammalian DNA that correlates with loss of gene expression. Aberrant promoter hypermethylation modifies expression of genes involved in diverse cellular processes that impact tumor cell growth and behavior, including growth factor and cytokine signaling, as well as the expression of drug sensitivity or resistance phenotypes.29 The importance of methylation in regulating cell growth and differentiation has resulted in the testing of DNA methyltransferase inhibitors as anticancer agents and differentiation agents.1-3 The nucleoside analogs 5-aza-2'-deoxycytidine (decitabine) and 5-aza-cytidine are potent pharmacologic inhibitors of DNA methylation in vitro.2,4 Decitabine-induced DNA hypomethylation, gene reactivation, and effects on cell function require incorporation into DNA followed by several cellular division cycles in vitro.4 We performed a phase I clinical trial of decitabine as a 168-hour continuous IV infusion in an attempt to develop a clinical protocol that would result in significant DNA hypomethylation in vivo. As a starting point, a dose estimated to be 1/20 of the MTD of short-term intermittent IV infusion was selected (2 mg/m2/d). This dose resulted in asymptomatic grade III/IV hematologic toxicity in two of 10 patients. The predominant toxicity was neutropenia in both patients. This toxicity was not symptomatic and appeared to resolve rapidly (2 to 3 days). While technically we established a MTD using a rigorous definition of hematologic toxicity, it should be apparent that further dose escalation of decitabine might be possible with platelet and growth factor support. A more extensive evaluation of less heavily pretreated patients is being planned to evaluate the relationship of marrow hypomethylation and the potential for hematologic toxicity.
As the overall goal of the study was to identify a tolerable dose of decitabine that inhibited DNA methylation in vivo, for combination with other antineoplastic agents, we did not pursue further decitabine dose escalation. With the exception of the described hematologic toxicity, the drug adverse effect profile was excellent, with no other toxicities noted. There were no objective responses to this treatment regimen in patients treated in this study. However, two patients showed disease stabilization for 4 to 6 months. The lack of patient responses was not unexpected, as the patients enrolled in the trial had heavily pretreated solid tumors and the trial was designed to establish a biologic end point: a safe dose of decitabine suitable for inhibiting DNA methylation for use in combination with other anticancer agents. While the current schedule of drug preparation is not easily adapted to out-patient use, due to the instability of decitabine, changes in the diluent have resulted in a more stable formulation that may facilitate out-patient administration via a portable belt pump.
Molecular correlation studies were performed to evaluate the effect of decitabine on DNA methylation in vivo. We isolated DNA from peripheral blood to assess gene-specific and genomic changes in DNA methylation. We utilized a quantitative PCR assay that requires as little as 2 μg of DNA to assess methylation of the MAGE-1 promoter.26 Significant hypomethylation of MAGE-1 promoter was observed in patient blood samples both at the end of continuous infusion (day 7) and also 7 days later (day 14). In the current study, it was not possible to assess whether MAGE-1 promoter hypomethylation led to MAGE-1 gene expression, but notably, Sigalotti et al30 recently reported de novo expression of MAGE-1 and other cancer-testis antigens in acute myelogenous leukemia and MDS patients treated with decitabine. In addition to MAGE-1 promoter methylation, we utilized a quantitative HPLC assay to measure the global level of 5-methylcytosine in patient blood samples. One limitation of this assay is that is requires larger amounts of DNA (20 to 25 μg), and thus may not be suitable for smaller specimens, such as tumor biopsies. Using this assay, we observed that genomic methylation decreased progressively from day 7 to 14. The kinetics of decitabine-induced hypomethylation are consistent with the maturation kinetics of peripheral-blood leukocytes, which take approximately 7 days to differentiate from marrow stem cells to mature forms released into the peripheral blood, suggesting that the target of decitabine was hematopoietic cells localized in the bone marrow.31 We also measured the extent of remethylation of blood DNA by analyzing genomic methylation levels at the beginning of the second cycle. Based on this exploratory data analysis, remethylation appears to occur between 28 and 49 days after the start of the 7-day continuous decitabine infusion. Analysis of a larger number of patients will be needed to more precisely establish the kinetics of remethylation in vivo.
Identification of suitable molecular end points to assess DNA methylation changes following decitabine treatment in patients with solid tumors have proven challenging. In studies in sickle cell anemia patients, fetal hemoglobin expression has provided a useful clinical end point.12,13 In hematologic malignancies, p15INK4b methylation in blood or bone marrow leukemia cells has been measured.19,21 Recently, Chan et al32 reported hypomethylation of the Epstein-Barr virus genome in vivo following treatment with 5-azacytidine. Difficulties in routinely obtaining adequate tumor tissue samples prompted us to measure methylation changes in peripheral blood as a surrogate marker in the current study. We analyzed DNA methylation changes both at a specific promoter locus (MAGE-1) and in total genomic DNA. An advantage of the methodology employed in the current study is that the methylation changes we observed are at nonselectable end points (ie, specific genes with oncogenic or cell-growth related properties). In addition, to our knowledge, this is the first study to show an effect of decitabine on global genomic 5mC levels in clinical samples. Future planned clinical studies will measure these DNA methylation end points in tumor biopsies from decitabine-treated patients. We have recently developed a high-sensitivity mass spectrometry–based assay to quantify 5mC that will facilitate these future studies.33
Consistent with the phase I nature of this trial, we did not anticipate decitabine clinical activity as a single agent. It is our expectation that the best potential for clinical success for decitabine will reside in combinations with agents such as cisplatin or interferon-alfa.34,35 Notably, the efficacy of decitabine in combination with cisplatin has recently been tested in a phase II clinical trial.23 While this drug combination showed moderate activity, the regimen, which utilized 50 mg/m2/d for 3 consecutive days of decitabine, was limited by significant hematologic toxicity.23 In the current study, we found that a 7-day continuous IV infusion of 2 mg/m2/d decitabine is well tolerated without significant adverse effects. Furthermore, this treatment schedule induced significant hypomethylation of both genomic DNA and a specific gene promoter. Thus, the clinical strategy outlined here may prove suitable for evaluating the clinical potential of combination treatments involving decitabine and agents whose activity may be enhanced by the reversal of DNA methylation-mediated gene silencing.3
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
ACKNOWLEDGMENTS
This trial was supported in part by a grant from Supergen, Inc. (Dublin, CA) and the Huntsman Cancer Institute Cancer Center support grant (NIH/NCI 5P30 CA420-14). The Huntsman General Clinical Research Center is supported by a grant from the NIH/NCRR (M01 RR00064). The Huntsman Cancer Foundation, Doris Duke Charitable Foundation, and the Howard Hughes Medical Institute provided additional support. Dr Karpf (now in the Department of Pharmacology and Therapeutics, Roswell Park Cancer Institute) was supported by Postdoctoral Fellowship PF-99-151-01-CDD from the American Cancer Society.
NOTES
Terms in blue are defined in the glossary, found at the end of this issue and online at www.jco.org.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Jones PA, Baylin SB: The fundamental role of epigenetic events in cancer. Nat Rev Genet 3:415-428, 2002
Bender CM, Zingg JM, Jones PA: DNA methylation as a target for drug design. Pharm Res 15:175-187, 1998
Karpf AR, Jones DA: Reactivating the expression of methylation silenced genes in human cancer. Oncogene 21:5496-5503, 2002
Jones PA, Taylor SM: Cellular differentiation, cytidine analogs and DNA methylation. Cell 20:85-93, 1980
Momparler RL: Molecular, cellular and animal pharmacology of 5-aza-2'-deoxycytidine. Pharmacol Ther 30:287-299, 1985
Juttermann R, Li E, Jaenisch R: Toxicity of 5-aza-2'-deoxycytidine to mammalian cells is mediated primarily by covalent trapping of DNA methyltransferase rather than DNA demethylation. Proc Natl Acad Sci U S A 91:11797-11801, 1994
Ferguson AT, Vertino PM, Spitzner JR, et al: Role of estrogen receptor gene demethylation and DNA methyltransferase. DNA adduct formation in 5-aza-2'deoxycytidine-induced cytotoxicity in human breast cancer cells. J Biol Chem 272:32260-32266, 1997
Yoshiura K, Kanai Y, Ochiai A, et al: Silencing of the E-cadherin invasion-suppressor gene by CpG methylation in human carcinomas. Proc Natl Acad Sci U S A 92:7416-7419, 1995
Herman JG, Umar A, Polyak K, et al: Incidence and functional consequences of hMLH1 promoter hypermethylation in colorectal carcinoma. Proc Natl Acad Sci U S A 95:6870-6875, 1998
Myohanen SK, Baylin SB, Herman JG: Hypermethylation can selectively silence individual p16ink4A alleles in neoplasia. Cancer Res 58:591-593, 1998
Bender CM, Pao MM, Jones PA: Inhibition of DNA methylation by 5-aza-2'-deoxycytidine suppresses the growth of human tumor cell lines. Cancer Res 58:95-101, 1998
Koshy M, Dorn L, Bressler L, et al: 2-deoxy 5-azacytidine and fetal hemoglobin induction in sickle cell anemia. Blood 96:2379-2384, 2000
DeSimone J, Koshy M, Dorn L, et al: Maintenance of elevated fetal hemoglobin levels by decitabine during dose interval treatment of sickle cell anemia. Blood 99:3905-3908, 2002
Silverman LR: Targeting hypomethylation of DNA to achieve cellular differentiation in myelodysplastic syndromes (MDS). Oncologist 6:8-14, 2001 (5 suppl)
Silverman LR, Demakos EP, Peterson BL, et al: Randomized controlled trial of azacitidine in patients with the myelodysplastic syndrome: A study of the cancer and leukemia group B. J Clin Oncol 20:2429-2440, 2002
Kantarjian HM: Treatment of myelodysplastic syndrome: Questions raised by the azacitidine experience. J Clin Oncol 20:2415-2416, 2002
Wijermans P, Lubbert M, Verhoef G, et al: Low-dose 5-aza-2'-deoxycytidine, a DNA hypomethylating agent, for the treatment of high-risk myelodysplastic syndrome: A multicenter phase II study in elderly patients. J Clin Oncol 18:956-962, 2000
Lubbert M, Wijermans P, Kunzmann R, et al: Cytogenetic responses in high-risk myelodysplastic syndrome following low-dose treatment with the DNA methylation inhibitor 5-aza-2'- deoxycytidine. Br J Haematol 114:349-357, 2001
Daskalakis M, Nguyen TT, Nguyen C, et al: Demethylation of a hypermethylated P15/INK4B gene in patients with myelodysplastic syndrome by 5-aza-2'-deoxycytidine (decitabine) treatment. Blood 100:2957-2964, 2002
Sacchi S, Kantarjian HM, O'Brien S, et al: Chronic myelogenous leukemia in nonlymphoid blastic phase: Analysis of the results of first salvage therapy with three different treatment approaches for 162 patients. Cancer 86:2632-2641, 1999
Issa JP, Garcia-Manero G, Giles FJ, et al: Phase 1 study of low-dose prolonged exposure schedules of the hypomethylating agent 5-aza-2'-deoxycytidine (decitabine) in hematopoietic malignancies. Blood 103:1635-1640, 2004
Bovenzi V, Le NL, Cote S, et al: DNA methylation of retinoic acid receptor beta in breast cancer and possible therapeutic role of 5-aza-2'-deoxycytidine. Anticancer Drugs 10:471-476, 1999
Pohlmann P, DiLeone LP, Cancella AI, et al: Phase II trial of cisplatin plus decitabine, a new DNA hypomethylating agent, in patients with advanced squamous cell carcinoma of the cervix. Am J Clin Oncol 25:496-501, 2002
Storer BE: Design and analysis of phase I clinical trials. Biometrics 45:925-937, 1989
Wetherill GB: Serial estimation of quantal response curves. J Royal Stat Soc Series B 25:1-48, 1963
De Smet C, Lurquin C, Lethe B, et al: DNA methylation is the primary silencing mechanism for a set of germ line- and tumor-specific genes with a CpG-rich promoter. Mol Cell Biol 19:7327-7335, 1999
De Smet C, De Backer O, Faraoni I, et al: The activation of human gene MAGE-1 in tumor cells is correlated with genome-wide demethylation. Proc Natl Acad Sci USA 93:7149-7153, 1996
Gehrke CW, McCune RA, Gama-Sosa MA, et al: Quantitative reversed-phase high-performance liquid chromatography of major and modified nucleosides in DNA. J Chromatogr 301:199-219, 1984
Esteller M, Corn PG, Baylin SB, et al: A gene hypermethylation profile of human cancer. Cancer Res 61:3225-3229, 2001
Sigalotti L, Altomonte M, Colizzi F, et al: 5-Aza-2'-deoxycytidine (decitabine) treatment of hematopoietic malignancies: A multimechanism therapeutic approach Blood 101:4644-4646, 2003
Athens JW: Disorders of neutrophil proliferation and circulations: A pathophysiological view. Clin Haematol 4:553-566, 1975
Chan AT, Tao Q, Robertson KD, et al: Azacitidine induces demethylation of the Epstein-Barr virus genome in tumors. J Clin Oncol 22:1373-1381, 2004
Song L, James SR, Kazim L, et al: Specific method for determinational genomic DNA methylation by liquid chromatography-electrospray ionization tandem mass spectroscopy. Anal Chem 77:504-510, 2005
Strathdee G, MacKean MJ, Illand M, et al: A role for methylation of the hMLH1 promoter in loss of hMLH1 expression and drug resistance in ovarian cancer. Oncogene 18:2335-2341, 1999
Karpf AR, Peterson PW, Rawlins JT, et al: Inhibition of DNA methyltransferase stimulates the expression of signal transducer and activator of transcription 1, 2, and 3 genes in colon tumor cells. Proc Natl Acad Sci U S A 96:14007-14012, 1999(Wolfram E. Samlowski, San)
ABSTRACT
PURPOSE: The nucleoside analog 5-aza-2'-deoxycytidine (5-aza-CdR, decitabine) is a potent inhibitor of DNA methylation in vitro. Cellular treatment with this agent induces the re-expression of methylation-silenced genes. It remains unclear to what extent this compound inhibits DNA methylation in vivo. A clinical study was designed to examine the molecular effects and toxicity of a continuous 1-week intravenous infusion of decitabine in solid tumor patients.
METHODS: Ten patients with refractory solid tumors were included in this study. Decitabine was administered at 2 mg/m2/d via continuous infusion for 168 hours. Quantitative polymerase chain reaction and high performance liquid chromatography were utilized to measure promoter-specific and global DNA methylation in peripheral-blood cells before and after treatment.
RESULTS: Transient grade III/IV neutropenia (two patients) and grade II thrombocytopenia (one patient) was observed at the lowest planned dose step (2 mg/m2/d for 7 days). Nonhematologic toxicities were not observed. Quantitative polymerase chain reaction demonstrated significant MAGE-1 promoter hypomethylation by 14 days after the start of treatment in all 13 treatment cycles examined. Significant genomic DNA hypomethylation was also seen by day 14 in 11 of 13 treatment cycles analyzed. Genomic DNA methylation reverted to baseline levels by 28 to 35 days after the start of treatment, demonstrating that inhibition of DNA methylation by decitabine is transient.
CONCLUSION: A 168-hour continuous infusion of decitabine is well tolerated and results in the inhibition of promoter-specific and genomic DNA methylation in vivo. This treatment schedule is suitable for evaluation of decitabine in combination with agents whose activity may be enhanced by the reversal of DNA methylation–mediated gene silencing.
INTRODUCTION
The recognition of aberrant DNA methylation as a contributing factor to tumorigenesis has resulted in the testing of DNA methyltransferase inhibitors as anticancer agents.1-3 The cytosine analogues 5-aza-2'-deoxycytidine (5-aza-2'-CdR or decitabine) and 5-aza-cytidine (5-aza-CR) are potent pharmacologic inhibitors of DNA methylation.2,4 The activity of these drugs requires incorporation into cellular DNA, and the subsequent covalent sequestration of DNA methyltransferase enzymes.5-7 A number of studies have demonstrated that decitabine treatment elicits the reactivation of methylation-silenced tumor suppressor genes and causes phenotypic reversion in cancer cell lines.8-11
In clinical studies, DNA methyltransferase inhibitors have shown activity in the treatment of chronic myelogenous leukemia, sickle cell anemia, and myelodysplastic syndrome (MDS).12-20 Notably, decitabine treatment has induced clinical responses in approximately 50% of patients with MDS, with myelosuppression being the major adverse effect.17 Decitabine has achieved orphan drug status for this malignancy. At the molecular level, cytogenetic responses to decitabine have been noted in approximately one third of MDS patients who presented with clonal chromosomal abnormalities.18 Issa et al21 recently demonstrated that low-dose prolonged exposure to decitabine correlated with clinical responses in hematologic malignancies. Furthermore, Dasalakis et al19 and Issa et al21 have presented evidence showing that decitabine treatment of MDS resulted in the hypomethylation of the p15INK4b gene in vivo.
In contrast to results observed in the treatment of hematologic disorders, DNA methyltransferase inhibitors have, to date, shown little activity as single agents in the treatment of solid tumors.7,17,22,23 Our group has put forward the hypothesis that DNA methyltransferase inhibitors may be best utilized as a treatment for solid tumors as part of a combination chemotherapy approach.3 Successful combination chemotherapy may be expected to require effective inhibition of DNA methylation by decitabine treatment before the use of a secondary agent. To test the potential utility of this approach, we attempted to develop a clinical protocol using decitabine that inhibited DNA methylation without high levels of systemic toxicity. Another goal was to develop assays that could identify methylation changes in human clinical samples.
We hypothesized that long-term (7-day) continuous exposure would allow for greater incorporation of decitabine into DNA, thus enhancing hypomethylation of DNA and increasing the potential for potentially synergistic therapeutic effects when combined with other agents. We therefore designed a phase I clinical trial to evaluate the maximum tolerated dose and dose-limiting toxicities of a 7-day (168-hour) continuous intravenous (IV) infusion of decitabine in solid tumor patients. A dose escalation scheme was designed, starting at approximately 1/20 of the previous maximum tolerated short-term infusion dose. Gene-specific and genomic changes in DNA methylation in peripheral-blood mononuclear cells were monitored during the course of treatment.
METHODS
Eligibility Criteria
This phase I study was reviewed and approved by the University of Utah (Salt Lake City, UT) institutional review board in accordance with assurances filed with the US Department of Health and Human Services. Patients were required to have recurrent or metastatic cancer. All curative or survival-prolonging standard treatment options had to have been exhausted. Prior treatment with chemotherapy, radiotherapy, immunotherapy, and/or surgery was allowed. Timing of the last previous treatment had to be greater than 4 weeks before protocol entry, and patients were required to have recovered from side effects of prior treatment, except hair loss. Patients had to be 18 years of age, with an estimated life expectancy of 12 weeks and a Southwest Oncology Group performance status of 0 to 2. Laboratory-based eligibility testing was required, including an absolute granulocyte count of > 2,000/μL and a platelet count of 100,000/μL; a bilirubin of 2.0 mg/dL; and a calculated or measured creatinine clearance of > 50 mL/min. Patients with significant comorbid medical or psychosocial conditions were excluded, as were pregnant or lactating women. All patients signed acknowledgment of informed consent.
Drug Administration
The initial study design was for five patients to be enrolled in sequential dose escalation steps, starting at 2 mg/m2/d. Subsequent dose-escalation steps of 5, 10, 20, and 40 mg/m2/d were planned. Decitabine was administered as a 7-day (168-hour) continuous IV infusion via indwelling central venous catheter. Decitabine (provided by Supergen, Dublin, CA) was stored as a stable freeze-dried powder (50 mg/20 mL glass vial). Immediately before use, the content of each vial was reconstituted with Sterile Water for Injection, USP (Baxter, Deerfield, IL). The reconstituted solution was immediately diluted to a concentration of 0.1 to 1 mg/mL with ice-cold infusion fluids (either 5% dextrose in water, or 0.9% sodium chloride injection [USP, Baxter]). Decitabine was reconstituted from powder every 12 hours and kept refrigerated (for a maximum of 12 hours). Once warmed to room temperature, each bag of solution was infused over a maximum period of 3 hours before changing to a new infusion bag of drug.
Dose Escalation Schema
Patients were admitted to the Huntsman General Clinical Research Center for the entire 168-hour treatment period. The main clinical objective of the study was to identify a well-tolerated continuous infusion dose that effectively inhibited DNA methylation. Another goal was to identify dose-limiting toxicities (DLTs) of decitabine on this schedule. DLT was defined as any grade III nonhematologic and/or grade III or IV hematologic toxicity. The maximum tolerated dose (MTD) was defined as the dose level below which DLT was observed in two of five patients. If fewer than two patients experienced DLT, then the next five patients were to be treated at the next higher dose. With this design, the probability of stopping the trial due to DLT was calculated to be 8%, 47%, and 81% if the true rate of DLT is 10%, 30%, and 50%, respectively (conditional on having administered the dose level to all five patients).24,25
Treatment cycles were planned every 28 days. Dose escalation was planned for each cohort of five patients if no grade III/IV toxicity was seen. Treatment response was assessed every two cycles by computed tomography scans. Patients who had stable disease or better were continued on treatment until disease progression was observed. If grade III/IV hematologic toxicity developed during a treatment course, treatment was interrupted until toxicity decreased to grade 0 or I. Treatment was resumed once blood counts recovered to > 2,000 absolute neutrophil count and > 100,000 platelets. Treatment could be delayed for a maximum of 3 weeks to allow blood count recovery. Hematopoietic growth factor support was allowed for neutropenia or neutropenic fever. Subsequent cycles could either be administered with hematopoietic growth factor or with a dose reduction to the previous step. Criteria for removal from protocol treatment included: cancer progression, unacceptable toxicity, treatment delay > 3 weeks due to toxicity, development of grade 3 nonhematologic toxicity, or voluntary withdrawal by the patient.
Response Assessment
Complete response (CR) was defined as complete disappearance of all measurable and assessable disease without any new lesions or disease-related symptoms. Partial response (PR): was defined as 50% decrease under baseline in the sum of products of perpendicular diameters of all measurable lesions without progression at any site or any new lesions. Progression was defined as a 50% increase or an increase of 10 cm2 (whichever is smaller) in the sum of products of all measurable lesions over smallest sum observed (over baseline if no decrease) or appearance of any new lesion/site. Stable disease (SD) does not qualify for CR, PR or progression.
DNA Isolation
Patient blood samples (7 mL) were collected into sodium citrate tubes at baseline (day 0) and on days 7 and 14 days after the start of each cycle of decitabine treatment. Samples were immediately placed on ice following a blood draw from a central venous catheter. DNA extraction from whole blood was performed within 1 hour of collection, using the Puregene DNA isolation kit (Gentra Systems, Minneapolis, MN).
MAGE-1 Promoter Methylation Analysis
A number of gene-specific methylation sites were evaluated to identify a marker for the effects of decitabine on normal peripheral blood leukocytes. The promoter of the MAGE-1 gene, which is methylated in somatic tissues, proved informative.26 The methylation status of the MAGE-1 promoter was evaluated using an assay developed by De Smet et al27 (schema, Fig 1A). This assay uses enzymatic digestion of genomic DNA with a methylation-sensitive restriction enzyme (HpaII), coupled with polymerase chain reaction (PCR), to evaluate the methylation status of CpG sites present in the MAGE-1 promoter. We modified the assay by performing real-time PCR amplifications using a Lightcycler instrument (Roche, Indianapolis, IN) to derive a quantitative measure of the extent of methylation of the MAGE-1 promoter in clinical samples. DNA samples obtained from each individual patient were processed and analyzed in parallel. Each PCR reaction contained approximately 1/25 of the DNA recovered from digestion and also contained Taq (Biolase, Bioline, Reno, NV), Taq antibody (Clonetech, Palo Alto, CA), Sybr green (Molecular Probes, Eugene, OR), 1 μM of each oligonucleotide primer,27 200 nM deoxy-nucleotide triphosphates, and 1X PCR buffer (50 mmol/L Tris, 0.25 mg/mL bovine serum albumin, 3 mM MgCl2). Lightcycler PCR parameters were as follows: initial denaturing for 10 seconds at 95°, followed by 35 cycles at 95° for 1 second, 56° for 1 second, and 72° for 2 seconds. Copy numbers and melting curves were calculated using software provided with the Lightcycler instrument. For each sample, the copy number obtained from the test reaction (using CDS20-EPD4 primers) was divided by the copy number derived from a control reaction (CDS21-EPD4 primers) to generate a ratio. This ratio directly reflects the average methylation state of the MAGE-1 promoter HpaII sites. We normalized the ratio to 1.0 for the day-0 sample of each individual patient treatment cycle. All PCR reactions were performed in triplicate and were repeated at least once, on a different day.
Genomic DNA Methylation Analyses
Genomic DNA was treated with nuclease P1 (Sigma, St Louis, MO) and bacterial alkaline phosphatase (Sigma) as described by Gehrke et al28 to evaluate changes in methyl-cytosine levels in leukocyte DNA (schema, Fig 1B). Digested DNA samples (90 μL, derived from approximately 5 μg DNA) were injected onto a Beckman System Gold HPLC system equipped with a Phenomenex Luna C18, 25 x 4.6 mm column, 126 nM solvent module, a 168 nM photodiode array detector and a 507e auto sampler. HPLC was performed using solvent A (2.5% MeOH/50 mmol/L potassium phosphate pH 4.0) and solvent B (30% MeOH/50 mmol/L potassium phosphate pH 4.0). The column was eluted at a constant flow rate of 1 mL/min beginning with 100% A for 10 minutes followed by a linear gradient over a period of 15 minutes to 100% B. After holding at 100% B for 10 minutes, the column was recycled to 100% A with a linear gradient over a period of 10 minutes. Absorbance of the eluent at 200 to 300 nM was monitored, generating chromatograms that plotted absorbance at 284 nM (max of 5-Me-dC at pH 4) versus time. The ratio of 5-Me-dC to dC was determined using a calibration curve resulting from the analysis of a serial dilution of nucleosides produced from digestion of normal human DNA. Samples were assayed in triplicate and results are presented as mean ± SD.
Statistical Analysis
Repeated measures of analysis of variance (ANOVA) were used to analyze the data, using Statistica software 6.0 (Statsoft, Tulsa, OK). Logarithmic transformation was applied to the data shown in Figures 2A and B data to equalize the variance at the three time points. This was not necessary for data shown in Figures 2C and D.
RESULTS
Patient Characteristics and Toxicity Assessment
Eleven patients signed informed consent to participate in the study after eligibility testing was successfully completed (Table 1). One patient (patient No. 7) was deemed ineligible due to changes in serum bilirubin between the prestudy visit (and informed consent) and admission for treatment (< 2 weeks). Another patient (patient No. 8) was withdrawn on day 2 of decitabine infusion due to rapid progression of head and neck cancer and pre-existing edema involving the oropharynx and face, resulting in respiratory difficulty. This patient died 1 week later, without completing the first treatment cycle. Thus, nine of 11 patients were assessable for toxicity.
The initial two patients accrued to the initial dose step (2 mg/m2/d) successfully completed the infusion without apparent early toxicity. One patient subsequently developed grade III neutropenia (day 28), and the second patient developed grade IV neutropenia (141 neutrophils/μL on day 27) and grade II thrombocytopenia. Both were completely asymptomatic and the hematologic alterations reversed within 3 days without sequelae. However, because two of the first five patients demonstrated grade III or IV toxicity, dose escalation was halted at this initial dose step (see Methods). Analysis of these two patients suggested possible contributing factors: one patient had had prior extensive (18 months) treatment with the nucleoside analog gemcitabine. The second patient had completed irradiation to the central pelvis and lumbosacral spine 1 month before protocol entry. After extensive discussions with the institutional review board, and due to the absence of other adverse effects, the protocol was amended to allow an additional five patients to be treated at the 2 mg/m2/d dose (a total of 10 patients) to better assess the incidence of DLT and planned biologic end points. Patients with extensive radiotherapy to marrow-containing bones or extensive nucleoside analog pretreatment were subsequently excluded. Further escalation was not permitted, however. Closer monitoring of blood counts was subsequently performed.
The revised analysis plan was that if nonhematologic grade III or hematologic grade IV toxicity was observed in three patients overall (in 10 patients), the starting dose would be judged too high. If hematologic toxicities resulted in neutropenia < 500 cells for > 5 days, or associated fever > 38.5°C or evidence of infection was observed, this counted as a grade III/IV toxicity toward establishing the MTD. A total of seven additional patients were accrued to the study without any further grade III/IV toxicity. Evaluation of toxicity suggested that a 2.0 mg/m2/d dose of decitabine resulted in a level of toxicity suitable for further evaluation of inhibition of DNA methylation in patients. The main DLT observed was neutropenia (Table 1).
Assessment of Clinical Response
All patients had extensively pretreated solid tumors. As noted above, there were a number of early tumor-related events that occurred that indicated the advanced state of cancer in the study patients. One additional patient (patient No. 3) became symptomatic from tumor-related small intestinal obstruction 1 week following completion of decitabine infusion, which required palliative resection. Following recovery (28 days), the patient subsequently completed two additional cycles of decitabine without sequelae. Eight patients completed at least two cycles of treatment and were assessable for tumor response (Table 1). No objective responses were observed, and seven patients progressed after two cycles of decitabine treatment. Two patients had stable disease. One of these, a woman with metastatic ovarian cancer (patient No. 6), received four cycles and had stable disease for 4 months before progression. The other, a man with renal carcinoma (patient No. 10), had stable disease lasting 6 months before progression.
Low-Dose Continuous Infusion of Decitabine Inhibits Promoter-Specific DNA Methylation
We analyzed the effects of decitabine treatment on promoter-specific DNA methylation in DNA isolated from blood. For this analysis, we exploited the fact that the MAGE-1 promoter CpG island is normally methylated in somatic tissues.26 Quantitative real-time PCR was employed to measure the extent of promoter CpG methylation in clinical samples. We analyzed genomic DNA isolated from patient blood samples before treatment (day 0), immediately after termination of continuous infusion (day 7), and 1 week after the end of continuous infusion (day 14). A total of 13 patient treatment cycles were analyzed, including eight patients receiving the first cycle of decitabine and five patients receiving the second cycle (Figs 2A and B). In seven of 13 treatment cycles, a decline in MAGE-1 promoter methylation was apparent by day 7, while at day 14, all 13 decitabine cycles resulted in MAGE-1 hypomethylation (Figs 2A and B; Figs 3A and B). At day 14, the mean level of MAGE-1 methylation across all patient samples declined by 60% following cycle 1, and by 65% following cycle 2 (Figs 4A and B). The observed decline in MAGE-1 promoter methylation over the entire patient population was significant at day 7 and day 14 of the first treatment cycle (P = .005), as determined by repeated measures ANOVA (Fig 3A). Similar results were observed in the second treatment cycle (Fig 3B; P = .0029).
Low-Dose Continuous Infusion of Decitabine Inhibits Global DNA Methylation
The effect of decitabine infusion on overall methylation of genomic DNA was also evaluated, utilizing a quantitative HPLC-based assay.28 DNA from patient blood was obtained before treatment (day 0), immediately following continuous infusion (day 7), and 1 week after continuous infusion (day 14). Genomic DNA methylation was evaluated in individual patient samples derived from the first (Fig 2C; Fig 3C) and second cycle of decitabine treatment (Fig 2D; Fig 3D), provided that sufficient DNA was available (20 μg). A total of five of nine decitabine cycles analyzed showed hypomethylation at day 7, while at day 14, 11 of 13 patient samples showed hypomethylation (Figs 2C and D; Figs 3C and D). The decline in genomic DNA methylation over the entire patient population at day 14 was highly significant (P < .001) for both the first cycle (P = .0059) and second cycles (P = .0048) of decitabine treatment by repeated measures ANOVA.
Duration of Genomic DNA Hypomethylation in Patient Blood Samples
Next, we analyzed the kinetics of recovery of genomic DNA methylation levels in patient samples by comparing the level of methylation at the end of the first cycle of decitabine treatment to the level at the beginning of the second cycle. Unplanned variations in the time between the first and second cycles of treatment in individual patients were utilized to evaluate the extent of recovery to baseline methylation levels following treatment. In five of six patients, there was a significant decline in the level of genomic DNA methylation 14 days after the start of the first cycle of treatment (Fig 4). If treatment was delayed to day 35 in responding patients, global methylation reverted to baseline levels by the start of the second cycle. In contrast, DNA from the single patient retreated on day 28 (patient No. 10) remained hypomethylated as compared with the pretreatment sample.
DISCUSSION
Cytosine methylation is an epigenetic event in mammalian DNA that correlates with loss of gene expression. Aberrant promoter hypermethylation modifies expression of genes involved in diverse cellular processes that impact tumor cell growth and behavior, including growth factor and cytokine signaling, as well as the expression of drug sensitivity or resistance phenotypes.29 The importance of methylation in regulating cell growth and differentiation has resulted in the testing of DNA methyltransferase inhibitors as anticancer agents and differentiation agents.1-3 The nucleoside analogs 5-aza-2'-deoxycytidine (decitabine) and 5-aza-cytidine are potent pharmacologic inhibitors of DNA methylation in vitro.2,4 Decitabine-induced DNA hypomethylation, gene reactivation, and effects on cell function require incorporation into DNA followed by several cellular division cycles in vitro.4 We performed a phase I clinical trial of decitabine as a 168-hour continuous IV infusion in an attempt to develop a clinical protocol that would result in significant DNA hypomethylation in vivo. As a starting point, a dose estimated to be 1/20 of the MTD of short-term intermittent IV infusion was selected (2 mg/m2/d). This dose resulted in asymptomatic grade III/IV hematologic toxicity in two of 10 patients. The predominant toxicity was neutropenia in both patients. This toxicity was not symptomatic and appeared to resolve rapidly (2 to 3 days). While technically we established a MTD using a rigorous definition of hematologic toxicity, it should be apparent that further dose escalation of decitabine might be possible with platelet and growth factor support. A more extensive evaluation of less heavily pretreated patients is being planned to evaluate the relationship of marrow hypomethylation and the potential for hematologic toxicity.
As the overall goal of the study was to identify a tolerable dose of decitabine that inhibited DNA methylation in vivo, for combination with other antineoplastic agents, we did not pursue further decitabine dose escalation. With the exception of the described hematologic toxicity, the drug adverse effect profile was excellent, with no other toxicities noted. There were no objective responses to this treatment regimen in patients treated in this study. However, two patients showed disease stabilization for 4 to 6 months. The lack of patient responses was not unexpected, as the patients enrolled in the trial had heavily pretreated solid tumors and the trial was designed to establish a biologic end point: a safe dose of decitabine suitable for inhibiting DNA methylation for use in combination with other anticancer agents. While the current schedule of drug preparation is not easily adapted to out-patient use, due to the instability of decitabine, changes in the diluent have resulted in a more stable formulation that may facilitate out-patient administration via a portable belt pump.
Molecular correlation studies were performed to evaluate the effect of decitabine on DNA methylation in vivo. We isolated DNA from peripheral blood to assess gene-specific and genomic changes in DNA methylation. We utilized a quantitative PCR assay that requires as little as 2 μg of DNA to assess methylation of the MAGE-1 promoter.26 Significant hypomethylation of MAGE-1 promoter was observed in patient blood samples both at the end of continuous infusion (day 7) and also 7 days later (day 14). In the current study, it was not possible to assess whether MAGE-1 promoter hypomethylation led to MAGE-1 gene expression, but notably, Sigalotti et al30 recently reported de novo expression of MAGE-1 and other cancer-testis antigens in acute myelogenous leukemia and MDS patients treated with decitabine. In addition to MAGE-1 promoter methylation, we utilized a quantitative HPLC assay to measure the global level of 5-methylcytosine in patient blood samples. One limitation of this assay is that is requires larger amounts of DNA (20 to 25 μg), and thus may not be suitable for smaller specimens, such as tumor biopsies. Using this assay, we observed that genomic methylation decreased progressively from day 7 to 14. The kinetics of decitabine-induced hypomethylation are consistent with the maturation kinetics of peripheral-blood leukocytes, which take approximately 7 days to differentiate from marrow stem cells to mature forms released into the peripheral blood, suggesting that the target of decitabine was hematopoietic cells localized in the bone marrow.31 We also measured the extent of remethylation of blood DNA by analyzing genomic methylation levels at the beginning of the second cycle. Based on this exploratory data analysis, remethylation appears to occur between 28 and 49 days after the start of the 7-day continuous decitabine infusion. Analysis of a larger number of patients will be needed to more precisely establish the kinetics of remethylation in vivo.
Identification of suitable molecular end points to assess DNA methylation changes following decitabine treatment in patients with solid tumors have proven challenging. In studies in sickle cell anemia patients, fetal hemoglobin expression has provided a useful clinical end point.12,13 In hematologic malignancies, p15INK4b methylation in blood or bone marrow leukemia cells has been measured.19,21 Recently, Chan et al32 reported hypomethylation of the Epstein-Barr virus genome in vivo following treatment with 5-azacytidine. Difficulties in routinely obtaining adequate tumor tissue samples prompted us to measure methylation changes in peripheral blood as a surrogate marker in the current study. We analyzed DNA methylation changes both at a specific promoter locus (MAGE-1) and in total genomic DNA. An advantage of the methodology employed in the current study is that the methylation changes we observed are at nonselectable end points (ie, specific genes with oncogenic or cell-growth related properties). In addition, to our knowledge, this is the first study to show an effect of decitabine on global genomic 5mC levels in clinical samples. Future planned clinical studies will measure these DNA methylation end points in tumor biopsies from decitabine-treated patients. We have recently developed a high-sensitivity mass spectrometry–based assay to quantify 5mC that will facilitate these future studies.33
Consistent with the phase I nature of this trial, we did not anticipate decitabine clinical activity as a single agent. It is our expectation that the best potential for clinical success for decitabine will reside in combinations with agents such as cisplatin or interferon-alfa.34,35 Notably, the efficacy of decitabine in combination with cisplatin has recently been tested in a phase II clinical trial.23 While this drug combination showed moderate activity, the regimen, which utilized 50 mg/m2/d for 3 consecutive days of decitabine, was limited by significant hematologic toxicity.23 In the current study, we found that a 7-day continuous IV infusion of 2 mg/m2/d decitabine is well tolerated without significant adverse effects. Furthermore, this treatment schedule induced significant hypomethylation of both genomic DNA and a specific gene promoter. Thus, the clinical strategy outlined here may prove suitable for evaluating the clinical potential of combination treatments involving decitabine and agents whose activity may be enhanced by the reversal of DNA methylation-mediated gene silencing.3
Authors' Disclosures of Potential Conflicts of Interest
The authors indicated no potential conflicts of interest.
ACKNOWLEDGMENTS
This trial was supported in part by a grant from Supergen, Inc. (Dublin, CA) and the Huntsman Cancer Institute Cancer Center support grant (NIH/NCI 5P30 CA420-14). The Huntsman General Clinical Research Center is supported by a grant from the NIH/NCRR (M01 RR00064). The Huntsman Cancer Foundation, Doris Duke Charitable Foundation, and the Howard Hughes Medical Institute provided additional support. Dr Karpf (now in the Department of Pharmacology and Therapeutics, Roswell Park Cancer Institute) was supported by Postdoctoral Fellowship PF-99-151-01-CDD from the American Cancer Society.
NOTES
Terms in blue are defined in the glossary, found at the end of this issue and online at www.jco.org.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
Jones PA, Baylin SB: The fundamental role of epigenetic events in cancer. Nat Rev Genet 3:415-428, 2002
Bender CM, Zingg JM, Jones PA: DNA methylation as a target for drug design. Pharm Res 15:175-187, 1998
Karpf AR, Jones DA: Reactivating the expression of methylation silenced genes in human cancer. Oncogene 21:5496-5503, 2002
Jones PA, Taylor SM: Cellular differentiation, cytidine analogs and DNA methylation. Cell 20:85-93, 1980
Momparler RL: Molecular, cellular and animal pharmacology of 5-aza-2'-deoxycytidine. Pharmacol Ther 30:287-299, 1985
Juttermann R, Li E, Jaenisch R: Toxicity of 5-aza-2'-deoxycytidine to mammalian cells is mediated primarily by covalent trapping of DNA methyltransferase rather than DNA demethylation. Proc Natl Acad Sci U S A 91:11797-11801, 1994
Ferguson AT, Vertino PM, Spitzner JR, et al: Role of estrogen receptor gene demethylation and DNA methyltransferase. DNA adduct formation in 5-aza-2'deoxycytidine-induced cytotoxicity in human breast cancer cells. J Biol Chem 272:32260-32266, 1997
Yoshiura K, Kanai Y, Ochiai A, et al: Silencing of the E-cadherin invasion-suppressor gene by CpG methylation in human carcinomas. Proc Natl Acad Sci U S A 92:7416-7419, 1995
Herman JG, Umar A, Polyak K, et al: Incidence and functional consequences of hMLH1 promoter hypermethylation in colorectal carcinoma. Proc Natl Acad Sci U S A 95:6870-6875, 1998
Myohanen SK, Baylin SB, Herman JG: Hypermethylation can selectively silence individual p16ink4A alleles in neoplasia. Cancer Res 58:591-593, 1998
Bender CM, Pao MM, Jones PA: Inhibition of DNA methylation by 5-aza-2'-deoxycytidine suppresses the growth of human tumor cell lines. Cancer Res 58:95-101, 1998
Koshy M, Dorn L, Bressler L, et al: 2-deoxy 5-azacytidine and fetal hemoglobin induction in sickle cell anemia. Blood 96:2379-2384, 2000
DeSimone J, Koshy M, Dorn L, et al: Maintenance of elevated fetal hemoglobin levels by decitabine during dose interval treatment of sickle cell anemia. Blood 99:3905-3908, 2002
Silverman LR: Targeting hypomethylation of DNA to achieve cellular differentiation in myelodysplastic syndromes (MDS). Oncologist 6:8-14, 2001 (5 suppl)
Silverman LR, Demakos EP, Peterson BL, et al: Randomized controlled trial of azacitidine in patients with the myelodysplastic syndrome: A study of the cancer and leukemia group B. J Clin Oncol 20:2429-2440, 2002
Kantarjian HM: Treatment of myelodysplastic syndrome: Questions raised by the azacitidine experience. J Clin Oncol 20:2415-2416, 2002
Wijermans P, Lubbert M, Verhoef G, et al: Low-dose 5-aza-2'-deoxycytidine, a DNA hypomethylating agent, for the treatment of high-risk myelodysplastic syndrome: A multicenter phase II study in elderly patients. J Clin Oncol 18:956-962, 2000
Lubbert M, Wijermans P, Kunzmann R, et al: Cytogenetic responses in high-risk myelodysplastic syndrome following low-dose treatment with the DNA methylation inhibitor 5-aza-2'- deoxycytidine. Br J Haematol 114:349-357, 2001
Daskalakis M, Nguyen TT, Nguyen C, et al: Demethylation of a hypermethylated P15/INK4B gene in patients with myelodysplastic syndrome by 5-aza-2'-deoxycytidine (decitabine) treatment. Blood 100:2957-2964, 2002
Sacchi S, Kantarjian HM, O'Brien S, et al: Chronic myelogenous leukemia in nonlymphoid blastic phase: Analysis of the results of first salvage therapy with three different treatment approaches for 162 patients. Cancer 86:2632-2641, 1999
Issa JP, Garcia-Manero G, Giles FJ, et al: Phase 1 study of low-dose prolonged exposure schedules of the hypomethylating agent 5-aza-2'-deoxycytidine (decitabine) in hematopoietic malignancies. Blood 103:1635-1640, 2004
Bovenzi V, Le NL, Cote S, et al: DNA methylation of retinoic acid receptor beta in breast cancer and possible therapeutic role of 5-aza-2'-deoxycytidine. Anticancer Drugs 10:471-476, 1999
Pohlmann P, DiLeone LP, Cancella AI, et al: Phase II trial of cisplatin plus decitabine, a new DNA hypomethylating agent, in patients with advanced squamous cell carcinoma of the cervix. Am J Clin Oncol 25:496-501, 2002
Storer BE: Design and analysis of phase I clinical trials. Biometrics 45:925-937, 1989
Wetherill GB: Serial estimation of quantal response curves. J Royal Stat Soc Series B 25:1-48, 1963
De Smet C, Lurquin C, Lethe B, et al: DNA methylation is the primary silencing mechanism for a set of germ line- and tumor-specific genes with a CpG-rich promoter. Mol Cell Biol 19:7327-7335, 1999
De Smet C, De Backer O, Faraoni I, et al: The activation of human gene MAGE-1 in tumor cells is correlated with genome-wide demethylation. Proc Natl Acad Sci USA 93:7149-7153, 1996
Gehrke CW, McCune RA, Gama-Sosa MA, et al: Quantitative reversed-phase high-performance liquid chromatography of major and modified nucleosides in DNA. J Chromatogr 301:199-219, 1984
Esteller M, Corn PG, Baylin SB, et al: A gene hypermethylation profile of human cancer. Cancer Res 61:3225-3229, 2001
Sigalotti L, Altomonte M, Colizzi F, et al: 5-Aza-2'-deoxycytidine (decitabine) treatment of hematopoietic malignancies: A multimechanism therapeutic approach Blood 101:4644-4646, 2003
Athens JW: Disorders of neutrophil proliferation and circulations: A pathophysiological view. Clin Haematol 4:553-566, 1975
Chan AT, Tao Q, Robertson KD, et al: Azacitidine induces demethylation of the Epstein-Barr virus genome in tumors. J Clin Oncol 22:1373-1381, 2004
Song L, James SR, Kazim L, et al: Specific method for determinational genomic DNA methylation by liquid chromatography-electrospray ionization tandem mass spectroscopy. Anal Chem 77:504-510, 2005
Strathdee G, MacKean MJ, Illand M, et al: A role for methylation of the hMLH1 promoter in loss of hMLH1 expression and drug resistance in ovarian cancer. Oncogene 18:2335-2341, 1999
Karpf AR, Peterson PW, Rawlins JT, et al: Inhibition of DNA methyltransferase stimulates the expression of signal transducer and activator of transcription 1, 2, and 3 genes in colon tumor cells. Proc Natl Acad Sci U S A 96:14007-14012, 1999(Wolfram E. Samlowski, San)