A Phase I Dose Finding Study of 5-Azacytidine in Combination with Sodium Phenylbutyrate in Patients with Refractory Solid Tumors
. Author manuscript; available in PMC: 2010 Oct 1.
Abstract
Purpose
This was a phase I trial to determine the minimal effective dose and optimal dose schedule of 5-AC in combination with sodium phenylbutyrate in patients with refractory solid tumors. The pharmacokinetics, pharmacodynamics, and antineoplastic effects were also studied.
Experimental design
Three dosing regimens were studied in 27 patients with advanced solid tumors and toxicity was recorded. The pharmacokinetics of the combination of drugs was evaluated. Repeat tumor biopsies and peripheral blood mononuclear cells (PBMC) were analyzed to evaluate epigenetic changes in response to therapy. Epstein Barr Virus (EBV) titers were evaluated as a surrogate measure for gene re-expression of epigenetic modulation in PBMC.
Results
The three dose regimens of 5-AC and PB were generally well tolerated and safe. A total of 48 cycles was administrated to 27 patients. The most common toxicities were bone marrow suppression related neutropenia and anemia, which were minor. The clinical response rate was disappointing for the combination of agents. One patient demonstrated stable disease for 5 months while 26 patients demonstrated progressive disease as best tumor response. The administration of PB and 5-AC did not appear to alter the pharmacokinetics of either drug. Although there were individual cases of targeted DNMT activity and histone H3/4 acetylation changes from paired biopsy or PBMC, no conclusive statement can be made based on these limited correlative studies.
Conclusion
The combination of 5-AC and PB across three dose schedules was generally well tolerated and safe, yet lacked any real evidence for clinical benefit.
Keywords: 5-azacytidine, phenylbutyrate, combination, phase I, correlative study
Introduction
Epigenetic silencing of key genes, such as tumor suppressor genes contribute to carcinogenesis(1–4). Two epigenetic processes, DNA methylation and histone acetylation result in silencing of target genes(5). Methylation of the CpG islands of gene promoter regions induces a conformational change in the chromatin, leading to transcriptional silencing. Methylated DNA then recruits and binds Methyl-CpG binding domain proteins (MBDs). MBD proteins directly repress transcription and also complex with transcriptional co-repressors including HDACs(6). Histone acetylation is associated with an open chromatin conformation, allowing gene transcription. HDACs maintain the chromatin in the closed, nontranscribed state. As individual epigenetic modifications, both DNA methylation and histone deacetylation contribute to a “closed” chromatin conformation, inhibiting transcription. However, epigenetic modulation via HDAC inhibition following demethylation can lead to more robust expression of previously silenced genes than inhibition of either process alone. Thus, drugs targeting DNA methylation and histone deacetylation remain an active and promising area of clinical investigation for cancer therapeutics(7–9).
The cytidine analogue, 5-AC inhibits DNMT, leading to demethylation of DNA in daughter cells with a resultant effect on gene expression and cell differentiation(10). Once incorporated into DNA, 5-AC produces a marked dose-dependent and time-dependent decrease in DNMT activity. Clinical trials of single agent 5-AC in solid tumor malignancies have been disappointing with little clinical activity(11–14). In these early trials of solid tumor malignancies, methylation status and gene expression were not reported. However, 5-AC demonstrated evidence of transcriptional modulation, global hypomethylation and clinical response in trials of patients with hemoglobinopathies and myeloid disorders(10, 15–17). 5-AC is FDA approved for the treatment of myelodysplastic syndrome (MDS) at a dose of 75 mg/m2/day for seven days every four weeks(18). At this dose and schedule, cytopenias, nausea and vomiting proved to be the most common side effects (17).
Phenylbutyrate (PB), a first generation HDAC inhibitor, is FDA approved for the treatment of hyperammonaemia in patients with urea cycle disorder. PB results in the acetylation of histones in vitro, alters gene expression and promotes differentiation(19). Two phase I studies of PB in refractory solid tumors have been reported(20, 21). In the study of intravenous administration, the maximum tolerated dose was 410 mg/kg/day for 5 days. Toxicity was primarily neurocortical and was readily reversible after discontinuation of the drug(21). Evidence of gene expression modulation was demonstrated.
Because demethylation predominantly occurs in S phase, longer periods of exposure to 5-AC may lead to a greater proportion of cells exposed during S phase, potentially leading to a greater effect on methylation status. In contrast, PB leads to G1/G0 growth arrest within 96 hours (22, 23). Maximal re-expression of key genes may require actively dividing cells. Given the epigenetic “layers” of transcriptional silencing and the evidence that demethylation followed by HDAC inhibition results in robust gene expression(7), we hypothesized that maximal gene expression and minimal toxicity would occur with the administration of lower doses of 5-AC for longer periods of time in conjunction with HDAC inhibition. Thus, this study investigated multiple 5-AC dosing schedules in combination with intermittent dosing schedules of PB. The primary objective of this dose finding study was to determine the minimal effective dose (MED) of the combination of agents that results in clinical response and/or target inhibition. Secondarily, we sought the acute and chronic toxicity profile, and pharmacokinetics (PK) of 5-AC and PB when used in combination. Moreover, as methylation silences incorporated viral genomes, such as Epstein Barr Virus (EBV) (24–27), we studied the EBV viral load changes as a surrogate marker of gene re-expression after treatment with 5-AC and sought to correlate changes with the PK.
Patients and methods
Patients age 18 or older with tissue or cytological diagnosis of refractory solid tumor malignancy and no curative options were eligible for the study. Other eligibility criteria included: documentation of evaluable tumor, ECOG PS ≤ 2, life expectancy of 12 weeks or longer, adequate bone marrow, hepatic, and renal function. Patients were required to have a negative serum HIV test and no evidence of CNS metastasis (28). Prior chemotherapy or radiation therapy was acceptable if they were completed ≥ 4 weeks prior to entry, with recovery from any toxicities (to Grade 1 or Grade 0). Patients were required to provide written informed consent prior to study enrollment. The study was approved by the Cancer Review Committee and the Institutional Review Board of The Johns Hopkins University School of Medicine.
Toxicity was classified and graded according to the National Cancer Institute Common Terminology Criteria for Adverse Events (NCI CTCAE) version 2.0. Standard WHO criteria were used to define response or progression. Patients were considered evaluable for toxicity once therapy was initiated. The treatment responses for therapy were evaluated every two cycles. Patients with progressive disease were removed from study. Patients with complete or partial responses or stable disease continued treatment until evidence of progression, unacceptable toxicity, or desire to withdraw from study.
Drug administration
5-AC and PB were supplied by the Cancer Treatment Evaluation Program (CTEP) of the National Cancer Institute (NCI). 5-AC was supplied as 100 mg of white, lyophilized powder with 100 mg mannitol, USP in 30 mL flint vials. The contents of each vial were dissolved in 4 mL sterile water or 0.9% sodium chloride to provide a 25 mg/mL slurry. Doses were divided so that no single injection constituted >2 mL. 5-AC was administered subcutaneously (SC) within 30 minute of reconstitution with daily rotation of the injection site. An administration log that recorded the dates and times of administration was maintained by each patient. When PK studies were performed, the 5-AC dose was administered by trained research staff. PB was supplied as 400 mg/mL (40%) viscous solution in sterile water with a pH of 7.5 to 9.5 in 50 mL vials. Each mL of solution contained 98 mg of sodium. In order to administer this drug as a continuous infusion, the total daily dose was diluted in one liter of sterile water for injection. PB was administered as a continuous infusion (CIV) over 24 hours and bags were changed daily(21).
The study evaluated three schedules of administration of the combination of agents. Please see Table 1. Regimen A evaluated a 14 day low dose regimen of 5-AC with intermittent PB by CIV over 24 hours on days 6 and 13. Regimen B evaluated 5-AC given SC for 7 days at a daily dose of 75 mg/m2/day. 5-AC was followed sequentially by two different doses of CIV PB starting on day 8 and continuing for 7 days. This regimen of 5-AC was chosen due to the fact that the same dose and schedule demonstrated clinical efficacy for the treatment of MDS. Regimen C evaluated the lowest daily doses of 5-AC, given for 21 days with PB given as a CIV infusion over 24 hours once per week while the patient received 5-AC. Cycle length for Regimen A and B was 35 days, and for Regimen C was 42 days.
Table 1.
Schema for Regimens
| Dose levels | AC mg/M2/d (days) | PB mg/m2/d (days) | Number of patients | Days/cycle | |
|---|---|---|---|---|---|
| A | A1 | 25 (D1–14) | 400 (D6, D13) | 3 | 35 |
| A-1 | 18.75 (D1–14) | 400 (D6, D13) | 3 | ||
| A-2 | 15 (D1–14) | 400 (D6, D13) | 3 | ||
| A-3 | 10 (D1–14) | 400 (D6, D13) | 3 | ||
| B | B1 | 75 (D1–7) | 200 (D 8 – 14) | 3 | |
| B2 | 75 (D1–7) | 400 (D 8 – 14) | 4* | ||
| C | C1 | 10 (D1–21) | 400 (D6, D13, D20) | 3 | 42 |
| C2 | 12.5 (D1–21) | 400 (D6, D13, D20) | 6 |
Dose Modification
Non-hematological dose-limiting toxicity (DLT) was defined as NCI/DCT CTC grade 3 or 4 toxicity. Hematological DLT was defined grade 3 or grade 4 neutropenia associated with a fever or lasting 5 days. Anemia was not DLT as red blood cell transfusion was allowed. DLT affecting dose escalation was determined during cycle 1. For patients that developed CTC grade 3 or 4 toxicity that is judged to be clinically significant by the principal investigators, treatment was held until resolution (for no more than two weeks) of toxicity to grade 2 or less. The treatment was then resumed at a 25% dose reduction. If patients experienced a grade 3 or 4 toxicity at the reduced dose, the treatment was interrupted until the toxicity resolved (for no more than two weeks) to grade 2 or less. The treatment resumed at a 50% dose reduction of the initial dose. Patients who required more than two dose-reductions were removed from the study.
If 1 out of 3 patients in a dose level experienced a DLT during cycle 1, then 3 additional patients were enrolled at that dose level. If 0 of 3 or only 1 of the 6 patients in the expanded dose cohort experienced a DLT, then the next dose level was allowed to proceed. If 2 or more out of 6 patients experienced a DLT, then the preceding dose level was considered the MTD. The MTD was defined as the highest dose level of 5-AC in combination with PB that caused 1 or fewer of 6 patients to experience DLTs. Because of the development of neutropenia side effect from dose level A1, cohorts A-1 and A-2 were explored and the protocol amended such that subsequent cohorts were enrolled in the following sequence: A1 -> A-1 -> A-2-> B1 -> A-3 -> B2 -> C1 -> C2.
The MED that elicited a biological or clinical response was also examined. In contrast to a more traditional MTD, the MED was to be determined by evidence of drug target inhibition or clinical response criteria. The MED was not determined on a real-time basis as the results of the correlative studies lagged behind the accrual needs of the trial.
Pharmacokinetic Studies
For 5-AC, samples were obtained prior to treatment and at 0.25, 0.5, 1, 2, 4, 6, 8, and 24 hours after 5-AC. Samples were stored and processed as described (29, 30). PK parameters were determined as previously described(29).
For PB PK, samples were obtained prior to treatment and on the first day of the PB infusion. When PB was administered as a 24-hour infusion, additional samples were obtained at 0.5, 1, and 4 hours during the infusion; immediately before the end of infusion (24 hours); and post infusion at 0.25, 0.5, 1, 2, 4, 8, and 24 hours after the end of the infusion. When PB was administered as a 7-day infusion, additional samples were obtained at 1, 4, 24, and 72 hours during the infusion and immediately before the end of infusion (168 hours). Samples were analyzed for PB and two metabolites, phenylacetylglutamine (PG) and phenylacetic acid (PA), using a high-performance liquid chromatography method using ultraviolet detection(20, 21, 31). The linear calibration curves were generated over the range of 10 to 3000 μM. When PB was administered as a 24-hour infusion, individual concentration-time data were analyzed using noncompartmental methods using WinNonlin Professional version 5.0 as described above for 5-AC with the following exceptions: 1) Maximal plasma concentrations (Cmax) was the concentration obtained prior to the end of the infusion; 2) The AUC value was calculated using the linear trapezoidal rule; and 3) No weight was utilized when calculating λz. When PB was administered as a 7-day infusion, average steady-state concentrations (Css) were calculated for PB, phenylacetylglutamine (PG), and phenylacetate (PA) based on graphical presentation of concentration-time profiles.
Correlative Studies
Biological endpoints were measured using pre and post treatment tumor biopsy, and PBMC. Biopsy of tumor for research purpose was optional for patient enrollment. Tumor markers such as prostate specific antigen (PSA), CA 19-9 were evaluated and followed where appropriate. The correlative studies included: 1) Determination of DNA methyltransferadse enzyme activity(32). Briefly, this was measured by incubating 10 μg of tissue protein lysate with 3 μCi of S-adenosyl-L-[methyl-3H]methionine and 0.5 μg poly(dI-dC)-(dI-dC) or 2.5 μg oligonucleotide 5′-CCAGCCCGGCC CGACCCGACCGCACCCGGCGC-3′ (methylated cytosines are underlined) for 120 min at 37 °C. Results were expressed as the mean disintegrations per minute (d.p.m.) per microgram of protein. Tumor tissues were obtained prior to therapy (within 4 weeks of starting therapy) and on day 14 or 21 in cycle 1 after therapy. 2) Determination of histone acetylation in PBMC by Western blot using polyclonal antibodies directed against acetylated histones H3 and H4 and nonacetylated H2 (33). PBMCs were collected within 4 weeks before treatment as baseline, on treatment days 6 (for dose schedule A and C) or 7 (schedule B) right before starting PB infusion, and days 14 (schedule A&C) or 15 (schedule B) when PB infusion is finished. 3) Determination of EBV viral load in PBMC by real-time PCR. PBMC samples from baseline and treatment days 1, 7,15 were collected (>3 million cells). Samples were snap frozen and DNA was extracted for quantitative competitive PCR analysis using specific primers as described previously(34).
Statistical Considerations
Pharmacokinetic parameters were summarized using descriptive statistics. Graphical presentation of concentration-time profiles consisted of the average and standard deviation of the concentrations determined at each time point. Dose-independent PK parameters (Tmax, T1/2, Cl/F, and Vd/F) and dose-normalized dose-dependent PK parameters (Cmax/Dose and AUC[INF]/Dose) were compared using one-way ANOVA to compare the differences as a function of dose levels. Statistical analysis was done using JMP Statistical Discovery Software version 4.04 (SAS Institute, Cary, NC). The a priori level of significance was P < 0.05. Since a limited number of samples were available for correlative studies, the analysis was descriptive only.
Results
Patient characteristics
Between March 2000 and July 2005, a total 34 patients with advanced solid tumors were consented into the study. Six patients did not meet eligibility requirements and thus were not treated on the protocol. One patient had rapid clinical progression of disease during 5-AC treatment and did not receive PB. This patient was assessable for safety and toxicity but only 27 patients were assessable for pharmacokinetics and efficacy of the combination. The demographic and clinical characteristics of the subjects are summarized in Table 2.
Table 2.
Patient characteristics
| Characteristics | No. of patients |
|---|---|
| No. of patients | 28 |
| Sex (male/female) | 17/11 |
| Age (years), median (range) | 59(34 82) |
| Races | |
| Caucacian/black/Asian | 25/2/1 |
| ECOG Performance Status | |
| 0 | 10 |
| 1 | 17 |
| 2 | 1 |
| Primary tumor site | |
| Colorectal | 7 |
| Prostate | 5 |
| Pancreas (neuroendocrine) | 1 |
| Kidney | 3 |
| Esophageal | 3 |
| Breast | 2 |
| Sarcoma | 2 |
| Bladder | 2 |
| Liver | 1 |
| Lung | 1 |
| Mesothelia | 1 |
| Average prior chemotherapy regimens | 2.5 |
Toxicity
Treatment was generally well tolerated. A total of 48 cycles of treatments were administrated with an average of 1.7 cycles/patient. The most common Grade 3 and Grade 4 toxicities were neutropenia, which happened mostly in cycle 1. All cohorts had three patients except cohort B2 and C2 where four and six patients were enrolled. One patient was replaced in cohort B2 since he had disease progression while on 5-AC treatment but did not receive PB. During the first cycle of dose level C2, one patient had worsening nausea and vomiting and was found to have acute on chronic renal failure. The etiology was multifactorial including new urinary tract obstruction on the left, secondary to tumor bulk, dehydration and fungal urinary tract infection. She required left percutaneous nephrostomy tube placement and exchange of the right side nephroureteral stent. Her treatment continued after creatinine went down to 1.9 but because of her general poor performance status and persistent chronic renal insufficiency she was taken off the study in the middle of the second cycle. This cycle 1 event of renal failure was considered possibly related to study drug and the cohort was expanded. Five other patients in the same cohort experienced no renal toxicities. Because of the slow accrual and the availability of more potent HDAC inhibitors, no further dose escalation was performed after this event.
Potentially treatment related Grade 3 and 4 non- hematologic toxicities during all cycles of therapy are listed in Table 3. At dose level A1, 1 of 3 patients had transient grade 3 hyponatremia. One of 3 patients demonstrated Grade 3 somnolence related to PB at dose level B2 during the first cycle, while 1 of 3 patients demonstrated Grade 3 confusion at dose level C1 of second cycle, which was considered to be possible from PB. Both toxicities reflect the neurocortical side effects of PB. These symptoms were transient and resolved promptly when PB was held.
Table 3.
Summary of nonhematologic and hematologic toxicity (all cycles) presenting in two or more patients
| Dose levels (patients) | A1 (3) | A-1 (3) | A-2 (3) | A-3 (3) | B1 (3) | B2 (3) | C1 (3) | C2 (6) | |||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| events | Any | G 3 | Any | G 3 | Any | G 3 | Any | G 3 | Any | G 3 | Any | G 3 | Any | G 3 | Any | G 3 | |
| Nonhematologic | |||||||||||||||||
| injection site reaction | 23 | 3 | 3 | 2 | 3 | 3 | 3 | 3 | 3 | ||||||||
| Nausea | 18 | 1 | 3 | 1 | 2 | 3 | 4 | 2 | 2 | ||||||||
| Vomiting | 17 | 1 | 3 | 1 | 2 | 3 | 4 | 1 | 2 | 1* | |||||||
| Fatigue | 17 | 3 | 1 | 1 | 2 | 3 | 3 | 1 | 3 | ||||||||
| ALT elevation | 11 | 1 | 3 | 3 | 2 | 2 | |||||||||||
| edema | 10 | 1 | 1 | 1 | 2 | 3 | 1 | 1 | |||||||||
| hyperglycemia | 7 | 1 | 3 | 1 | 1 | 1 | |||||||||||
| lightheadedness | 6 | 2 | 1 | 1 | 1 | 1 | |||||||||||
| anorexia | 5 | 1 | 1 | 3 | |||||||||||||
| diarrhea | 5 | 1 | 1 | 1 | 2 | ||||||||||||
| hyponatremia | 5 | 1 | 1* | 3 | 1 | ||||||||||||
| abdominal pain | 4 | 1 | 1 | 2 | |||||||||||||
| body odor | 4 | 1 | 1 | 1 | 1 | ||||||||||||
| headache | 3 | 1 | 2 | ||||||||||||||
| somnolence | 3 | 1 | 1 | 1 | 1* | ||||||||||||
| alopecia | 3 | 1 | 1 | 1 | |||||||||||||
| flatulence/indigestion | 3 | 1 | 1 | 1 | |||||||||||||
| constipation | 2 | 1 | 1 | ||||||||||||||
| pruritis | 2 | 1 | 1 | ||||||||||||||
| skin rash | 2 | 1 | 1 | ||||||||||||||
| arthralgia | 2 | 1 | 1 | ||||||||||||||
| hypokalemia | 2 | 1 | 1 | ||||||||||||||
| hypoalbuminemia | 2 | 1 | 1 | ||||||||||||||
| confusion | 2 | 1 | 1 | 1 | |||||||||||||
| Hematologic | |||||||||||||||||
| neutropenia | 11 | 3* | 3* | 1* | 1* | 1* | 2* | 2* | 2*+1 | ||||||||
| anemia | 8 | 2 | 1 | 2 | 1 | 1 | 1 | 2 | 1*+1 | ||||||||
| thrombocytopenia | 6 | 1 | 1 | 2 | 1 | 1 |
Hematologic toxicity was moderate (Table 3). Five patients experienced Grade 3 neutropenia lasting for less than 5 days, 3 of 3 on dose level A1, and 2 of 3 on dose level B2. Patients received a 25% dose reduction without recurrence of neutropenia. No treatment delays occurred due to neutropenia. No episodes of febrile neutropenia were noted. For dose level A1, the protocol was amended to begin dose de-escalation, therefore MTD was underexplored and not determined for the dose schedule A. Two of 6 patients experienced Grade 3 anemia at dose level C2 at cycle 2. Both of the patients required transfusion.
For all grade toxicities, the most common adverse events were non-hematological and included 5AC injection site reaction, low grade nausea, vomiting, fatigue, transaminase elevation, edema, hyperglycemia, hyponatremia, light-headedness, anorexia and diarrhea (Table 3).
Response Evaluation
No clinical responses were noted on any regimen or dose level. One patient with leiomyosarcoma on dose level B1 received 4 cycles of therapy and demonstrated stable disease for 4.5 months. Twenty six patients had progressive disease as best clinical response and were removed from study.
Pharmacokinetics
5-AC PK data obtained from an additional 12 patients (data not shown) were consistent with our previously published data (29). PB PK studies were completed in all patients with 21 patients receiving 24-hour infusions and 6 patients received PB as a 7-day infusion. Pharmacokinetic parameters for PB administered as a 24-hour infusion are listed in Table 4. All patients who received the 24-hour infusion received 400 mg/kg/d and there was no statistically significant difference in PB, PA, or PG pharmacokinetic parameters across the various dose-levels (P > 0.05). Therefore, for all analyses, all 24-hour infusion PK parameters were treated the same. The mean ± standard deviation Cmax, AUC[INF], and Vd for PB were 775 ± 467 μM, 17722 ± 8345 h*μM, and 14.0 ± 8.0 L, respectively. The PB concentrations were sustained above 500 μM for 16.46 hours on average. For PA, the mean ± standard deviation Cmax, AUC[INF], and PA:PB AUC ratio were 1395 ± 594 μM, 26680 ± 11236 h*μM, and 1.71 ± 0.92, respectively. There was more variability in the PG exposure with the mean ± standard deviation Cmax, AUC[INF], and PG:PB AUC ratio being 997 ± 507 μM, 25609 ± 16771 h*μM, and 1.85 ± 1.78, respectively. A total of 4 patients had an unidentified PB metabolite in the plasma samples which was quantified from the PB standard curve. For the unknown metabolite, the average Cmax was 316 μM and occurred at 24.2 hours.
Table 4.
Pharmacokinetic parameters of PB administered as a continuous infusion over 24 hours to patients with cancer
| Dose/Regimen | n | Cmax (μM)a | Tmax (h)b | AUC[INF] (h*μM)a | Cl (L/hr)a | T > 500 μM (h)b | metabolite:PB AUC ratioa | |
|---|---|---|---|---|---|---|---|---|
| PB | A-2 | 9 | 714 ± 462 | 24.09 (23.67–25.00) | 16305 ± 8417 | 15.0 ± 10.8 | 14.35 (0.00–25.66) | NC |
| PB | A-3 | 3 | 1004 ± 489 | 24.38 (24.07–24.82) | 21744 ± 7653 | 7.0 ± 3.6 | 23.15 (18.60–26.07) | NC |
| PB | C1 | 3 | 956 ± 184 | 24.06 (24.00–24.12) | 22080 ± 2598 | 6.7 ± 1.3 | 24.10 (23.17–24.68) | NC |
| PB | C2 | 6 | 662 ± 591 | 24.25 (23.83–25.33) | 15659 ± 10453 | 14.8 ± 8.2 | 12.45 (0.00–25.82) | NC |
| Overall | 775 ± 467 | 24.17 (23.67–25.33) | 17722 ± 8345 | 12.6 ± 8.9 | 16.46 (0.00–26.07) | |||
| PA | A-2 | 9 | 1159 ± 445 | 25.16 (24.00–26.95) | 22611 ± 11613 | 13.2 ± 10.2 | NC | 1.39 ± 0.19 |
| PA | A-3 | 3 | 1932 ± 472 | 24.92 (24.25–26.00) | 33398 ± 11492 | 4.9 ± 1.2 | NC | 1.64 ± 0.58 |
| PA | C1 | 3 | 1417 ± 300 | 26.04 (24.12–27.92) | 30544 ± 13291 | 6.3 ± 2.4 | NC | 1.35 ± 0.45 |
| PA | C2 | 6 | 1471 ± 540 | 25.32 (23.83–27.98) | 27492 ± 9956 | 7.9 ± 3.4 | NC | 2.40 ± 1.50 |
| Overall | 1395 ± 504 | 25.30 (23.83–27.98) | 26680 ± 11236 | 9.5 ± 7.6 | 1.71 ± 0.92 | |||
| PG | A-2 | 9 | 1090 ± 647 | 24.93 (23.75–28.17) | 27472 ± 19725 | 6.6 ± 4.0 | NC | 1.99 ± 1.36 |
| PG | A-3 | 3 | 914 ± 439 | 25.47 (24.25–26.17) | 19989 ± 9574 | 5.7 ± 3.6 | NC | 0.90 ± 0.25 |
| PG | C1 | 3 | 854 ± 51 | 24.31 (24.12–24.58) | 20726 ± 3778 | 5.1 ± 1.0 | NC | 0.93 ± 0.06 |
| PG | C2 | 6 | 969 ± 496 | 24.59 (23.83–25.33) | 28068 ± 20331 | 6.6 ± 4.5 | NC | 2.57 ± 2.82 |
| Overall | 997 ± 507 | 24.82 (23.75–28.17) | 25609 ± 16771 | 6.3 ± 3.7 | 1.85 ± 1.78 |
When PB was administered as a 7-day infusion, the mean ± standard deviation steady state PB plasma concentration were 210 ± 73.8 μM at 200 mg/kg/d and 446 ± 211 μM at 400 mg/kg/d. Only 1 patient had PB concentrations sustained about 500 μM. For PA, the mean ± standard deviation steady state plasma concentration were 184 ± 85.2 μM and 1464 ± 1285 μM at 200 and 400 mg/kg/d, respectively. For PG, the median steady state plasma concentration were 427 ± 103 μM at 200 mg/kg/d and 1217 ± 244 μM at 400 mg/kg/d. No patients displayed an unidentified PB metabolite in the plasma samples as was observed during the 24-hour infusion.
Correlative studies
DNA methyltransferase activity was analyzed in five patients, where paired tumor biopsy tissue was available, at baseline and after 5-AC treatment (day 14 or 21 in cycle 1). Table 5 showed the dose level of 5-AC, tumor type, and DNMT activities. DNMT levels varied amongst solid tumors, and patients with prostate adenocarcinoma had low baseline tumor DNMT activity, resulting in the inability to detect significant enzyme inhibition after treatment with 5-AC. However, at the 5-AC dose of 15mg/m2/d for 14 days, a 75%, inhibition of DNMT activity was observed in a patient with hepatocellular carcinoma where basal DNMT activity was much higher. Additionally, a renal cell carcinoma tumor lacked detectable DNMT activity after 5-AC treatment, while the pretreatment enzyme activity was low but detectable. This data suggests that even at daily doses of 15mg/m2, there may be inhibition of DNMT in some patients.
Table 5.
Changes of tumor tissue methyltransferase activity after 5-AC
| Patient ID | Tumor type | Dose level | Methyltransferase activity | |
|---|---|---|---|---|
| Pretreatment | Posttreatment | |||
| 1 | prostate | A1 | 0 | 0 |
| 3 | prostate | A1 | 442 | 448 |
| 4 | prostate | A-1 | 34 | 98 |
| 8 | liver | A-2 | 2650 | 684 |
| 9 | kidney | A-2 | 20 | 0 |
All patients’ PBMC samples were collected for histone H3 and H4 acetylation study. But only 13 samples were available for analysis based on a number of factors relating to the quality of immunoblot (assay not fully worked out, Lab errors etc). Table 6 showed the dose level of PB, tumor type, and changes of histone H3 or H4 acetylation in 13 evaluable Western blots. Eleven of 13 (85%) patients had detectable acetylated forms at baseline (day 0) for both H3 and H4. Two of 13 (15%) had no detectable acetylation for H3 but both were induced to have acetylation after 6 days treatment by 5-AC only at doses of 75mg/m2/d or 12.5mg/m2/d before PB treatment. None of these patients had tissue biopsies so DNMT activities were not available. Seven patients (54%) showed increased acetylation after either 24 hours PB infusion or CIV PB for 7 days (Table 6). At dose schedule B when PB was infused for 7 days, 3 of 5 (60%) patients achieved increased acetylation. Four of 8 (50%) patients showed acetylation when PB was infused for just 24 hours.
Table 6.
Changes of PBMC histone acetylation after treatments
| Patient ID | Tumor type | PB dose level (mg/m2/d)/D | Histone acetylation* | |
|---|---|---|---|---|
| H3 | H4 | |||
| 11 | Rectal | A-2 (400)/D6, D13) | − | − |
| 12 | Renal | B1 (200)/D 8 to 14) | − | + |
| 13 | colon | − | − | |
| 14 | prostate | A-3 (400)/D6, D13 | − | − |
| 15 | prostate | − | − | |
| 16 | breast | − | − | |
| 17 | methothelia | B2 (400)/D8 to 14 | − | − |
| 19 | bladder | + | − | |
| 21 | renal | + | + | |
| 24 | colon | C1 (400)/D6, D13, D20 | + | + |
| 31 | rectal | C2 (400)/D6, D13, D20 | + | − |
| 32 | colon | + | + | |
| 33 | lung | + | + |
Four of 21 patients (19%) had detectable EBV DNA from PBMC at baseline. One patient with leiomyosarcoma and another patient with bladder cancer demonstrated increase EBV viral DNA copy number after 5AC treatment at dose level B (5-AC 75mg/m2) (Supplemental Table 1). Three patients had reduced viral load after 5-AC treatment. Although exploratory, these mixed results do not allow us to draw any conclusions as to the usefulness of this assay in future studies.
Discussion
This report characterizes the safety and toxicity of 5-AC and PB together with PK analysis when 5-AC was combined with PB on different schedules and wide dose ranges in solid tumors. Generally these different regimens were relatively well-tolerated and delivery of this range of dose regimens is feasible. No patients were taken off study because of toxicity. The toxicity profile is consistent with the known adverse effects of 5-AC in hematological malignancies (17, 18). Grade 3 toxicities were bone marrow suppression (likely 5-AC related) and neuro-cortical toxicity (likely PB related). Neutropenia occurred more often in dose level A1 when 5-AC was given 25 mg/m2/d for 2 weeks comparing dose level B when 5-AC was given 75mg/m2/d for 1 week, the FDA approved dose. The high rate of neutropenia (3 of 3, non DLT neutropenia) on the first cohort of the study gave caution to the investigators raising concerns that longer treatment with 5-AC would increase bone marrow suppression and recovery might be prolonged with the addition of an HDAC inhibitor. Longer treatment of 5-AC at lower doses caused neutropenia and anemia in some patients but did not delay subsequent cycles, although cycle length was lengthened to allow three weeks recovery after completing treatment before the next cycle began. The neuro-cortical toxicity attributable to PB (3 of 27 patients) occurred at PB doses of 400 mg/kg/day on either the intermittent or continuous 7 day infusion schedules. This toxicity was consistent with previous reports(21).
Only one patient demonstrated stable disease for 4.5 months and MED was not able to be defined in this study. Attempts to determine the MED through exploration of target inhibition was complicated by the need for tissue acquisition. At study initiation, over half of the patients consented to pre and post treatment biopsies. When accrual slowed while seeking patients agreeable to biopsies, the study moved ahead with accrual at the loss of tissue sample acquisition. However, evidence of biological effect/target inhibition was noted in some patients. Because of availability of tissue pre and post treatment, only 5 samples were available for DNMT analysis and only half of PBMC samples were analyzed. While the primary endpoint to determine the safety and tolerability of this combination across a range of low 5-AC doses was met, the intensity of the trial in terms of the need for correlates hindered accrual. Many of the correlative studies were not conducted preventing us from making any significant conclusions on target inhibition to determine minimal effective dose from a biologic perspective.
The lack of clinical response by the combination of 5-AC and PB may be secondary to the following reasons: 1) solid tumors may have lower DNMT activity comparing to MDS/AML malignant cells. This is most likely the case for prostate cancer patients, and may also reflect the lower S fraction of many patients with solid tumors. Nucleoside analogues require incorporation into DNA for activity. It is possible that 5-AC may inhibit DNMT only in tumors with high DNMT level. Preclinical studies in breast cancer cell lines showed that gene re-expression occurred when DNMT was inhibited at levels of 90% or more(35). Unfortunately, we did not have epigenetic information from the leiomyosarcoma patient who had stable disease after treatment. 2) PB is a relatively weak HDAC inhibitor that only enhances mild to modest acetylation of histone in solid tumors. 3) Epigenetic therapies may also require longer exposure to these agents than cytotoxic therapies. In the late stage of cancer that patients enrolled on clinical trials typically represent, this requirement for longer exposure of the drugs, i.e., more cycles of treatments, for efficacy, is often not possible. Patients in this study had an average of 4 different chemotherapy regimens prior to this treatment and received only 1.7 cycles on average of this treatment in this study.
The administration of PB and 5-AC did not appear to alter the pharmacokinetics of either drug. 5-AC was eliminated rapidly from patients with a t1/2 of 1.08 hours resulting in no accumulation after single daily dosing. The pharmacokinetics of 5-AC administered SC appeared to be linear with respect to Cmax and AUC[INF] at the dose levels studied. These are consistent with our previous report(29) and in patients with hematologic malignancy (33). It was reported that only the higher doses of 5-AC (75 mg/m2/d) could achieve plasma concentrations of 5 μM that were shown in vitro to cause DNMT inhibition(29). Interestingly, in one patient with hepatocellular carcinoma, 5-AC at the 15 mg/m2/d dose for 14 days suppressed DNMT activity by about 75%. Given the measurable inhibition of DNMT at lower concentrations than laboratory assays, which are typically with higher doses of shorter duration, this potentially suggests that pharmacodynamic effects can occur even at lower concentrations. This is consistent with recently published phase I result of 5-AC at dose of 20 – 25 mg/m2 daily for 10 days achieved 4 – 8 months duration of stable disease in 5 of 14 advanced solid tumor patients (35%) (9). The disconnect between 5-AC plasma concentrations achieved at lower 5-AC doses and concentrations required in vitro to produce pharmacodynamic effects needs to be examined further. Several aspects for further study will involve characterizing the plasma protein binding of 5-AC and measurement of drug concentrations intracellular and in tumor.
The incidence of EBV infection in peripheral lymphocytes is about 29% in this small study. Two patients were found to have increased EBV viral load after 5-AC treatment at the dose level of 75 mg/m2. One mechanism could be the demethylation of viral genome induced by 5-AC. Three patients demonstrated reduced viral load. This is possibly secondary to the induced apoptosis or elimination of EBV-infected cells or B cells (34). Because of the low incidence of EBV infection in solid tumor patient PBMC, EBV viral load may not be an ideal surrogate for DNMT inhibition.
We were not able to analyze all the tumor tissues or PBMCs for the correlative studies because of technical difficulties. From the available data, in one example, it seems that 5-AC could inhibit DNMT when the baseline enzyme activity is high. However, this assay may not be optimal to answer if demethylation occurred after 5-AC treatments, due to these considerations. Tumor tissue or PBMC global gene DNA methylation analysis may be more reliable or direct to answer this specific question as how much epigenetic changes happened(36).
In the past several years, extensive and rapid drug development on HDAC inhibitors has lead to the approval of vorinostat (Zolinza™, Merck Inc.) for the treatment of refractory cutaneous T-cell lymphoma(37). More than 10 HDAC inhibitors are in phase I/II clinical trials. Many of them are orally administered (38, 39). Clinical interest in PB had waned given its cumbersome oral and IV administration and based on this and other studies, we closed the study as there was little rationale to proceed with trials using PB in the treatment of solid malignancies. Despite not meeting our original goals for this study, we felt we had explored a wide range of schedules/doses of the more clinically relevant agent, 5-AC.
In summary, the combinations of 5-AC and PB in three dose levels were generally well tolerated and safe, but no obvious clinical responses were achieved. The administration of PB and 5-AC did not appear to alter the pharmacokinetics of either drug. In contrast to clinical response, tumor tissue DNMT inhibition was achieved at a lower dose of 5-AC in one patient suggesting that further studies at low doses of this drug might be warranted. PBMC histone acetylation induction was observed in some patients, but could not provide conclusive evidence for targeted response based on this limited correlative studies.
Translational Relevance.
Preclinical data demonstrates that epigenetic modulation via histone deacetylase (HDAC) and DNA methyltransferadse (DNMT) inhibition can lead to robust expression of previously silenced genes, holding promise as a therapeutic strategy, especially in tumors with silenced transcription of tumor suppressor genes. 5-azacytidine (5-AC) and phenylbutyrate (PB) can individually upregulate the transcription of epigenetically silenced genes. Multiple 5-AC dosing schedules in combination with PB were explored in this study. Furthermore, the study sought to determine whether 5-AC administered at lower doses for longer periods of time would result in consistent biological or clinical responses than shorter duration of 5-AC administration.
Supplementary Material
Acknowledgments
Supported by NCI UO1-CA70095 (RCD, MAC), NCI P50-CA58236 (JGH, MAC), NCI P30-CA08973 (MAR, MAC, RFA), Prostate Cancer Foundation (MAC), AEGON International Fellowship in Oncology (JG), NCI T32 (JL)
This study was sponsored by the Cancer Therapy Evaluation Program (CTEP) of the NCI and was funded through U01 program CA 70095, NCI CA SPORE, AEGON. We thank the following people from Johns Hopkins University for their support during both clinical trials: Rana Sullivan, Kathleen Burks, and Suzanne Dolan for nursing support; Bettye Carr and Jill Stewart for data management; Susan Davidson for quality assurance of the pharmacokinetic data contained in the manuscript; and Jatandra Birney, Carol Hartke, Ping He, Alex Mnatsakanyan, and Yelena Zabelina for their assistance in the pharmacokinetic quantitation.
References
- 1.Baylin SB, Esteller M, Rountree MR, Bachman KE, Schuebel K, Herman JG. Aberrant patterns of DNA methylation, chromatin formation and gene expression in cancer. Hum Mol Genet. 2001;10:687–92. doi: 10.1093/hmg/10.7.687. [DOI] [PubMed] [Google Scholar]
- 2.Worm J, Guldberg P. DNA methylation: an epigenetic pathway to cancer and a promising target for anticancer therapy. J Oral Pathol Med. 2002;31:443–9. doi: 10.1034/j.1600-0714.2002.00034.x. [DOI] [PubMed] [Google Scholar]
- 3.Glozak MA, Seto E. Histone deacetylases and cancer. Oncogene. 2007;26:5420–32. doi: 10.1038/sj.onc.1210610. [DOI] [PubMed] [Google Scholar]
- 4.Razin A, Riggs AD. DNA methylation and gene function. Science. 1980;210:604–10. doi: 10.1126/science.6254144. [DOI] [PubMed] [Google Scholar]
- 5.Rountree MR, Bachman KE, Herman JG, Baylin SB. DNA methylation, chromatin inheritance, and cancer. Oncogene. 2001;20:3156–65. doi: 10.1038/sj.onc.1204339. [DOI] [PubMed] [Google Scholar]
- 6.Nan X, Ng HH, Johnson CA, et al. Transcriptional repression by the methyl-CpG-binding protein MeCP2 involves a histone deacetylase complex. Nature. 1998;393:386–9. doi: 10.1038/30764. [DOI] [PubMed] [Google Scholar]
- 7.Cameron EE, Bachman KE, Myohanen S, Herman JG, Baylin SB. Synergy of demethylation and histone deacetylase inhibition in the re-expression of genes silenced in cancer. Nat Genet. 1999;21:103–7. doi: 10.1038/5047. [DOI] [PubMed] [Google Scholar]
- 8.Esteller M. DNA methylation and cancer therapy: new developments and expectations. Curr Opin Oncol. 2005;17:55–60. doi: 10.1097/01.cco.0000147383.04709.10. [DOI] [PubMed] [Google Scholar]
- 9.Braiteh F, Soriano AO, Garcia-Manero G, et al. Phase I study of epigenetic modulation with 5-azacytidine and valproic acid in patients with advanced cancers. Clin Cancer Res. 2008;14:6296–301. doi: 10.1158/1078-0432.CCR-08-1247. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Creusot F, Acs G, Christman JK. Inhibition of DNA methyltransferase and induction of Friend erythroleukemia cell differentiation by 5-azacytidine and 5-aza-2′-deoxycytidine. The Journal of biological chemistry. 1982;257:2041–8. [PubMed] [Google Scholar]
- 11.Momparler RL, Bouffard DY, Momparler LF, Dionne J, Belanger K, Ayoub J. Pilot phase I–II study on 5-aza-2′-deoxycytidine (Decitabine) in patients with metastatic lung cancer. Anticancer Drugs. 1997;8:358–68. doi: 10.1097/00001813-199704000-00008. [DOI] [PubMed] [Google Scholar]
- 12.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. American journal of clinical oncology. 2002;25:496–501. doi: 10.1097/00000421-200210000-00015. [DOI] [PubMed] [Google Scholar]
- 13.Quagliana JM, O’Bryan RM, Baker L, et al. Phase II study of 5-azacytidine in solid tumors. Cancer treatment reports. 1977;61:51–4. [PubMed] [Google Scholar]
- 14.Srinivasan U, Reaman GH, Poplack DG, Glaubiger DL, LeVine AS. Phase II study of 5-azacytidine in sarcomas of bone. American journal of clinical oncology. 1982;5:411–5. doi: 10.1097/00000421-198208000-00011. [DOI] [PubMed] [Google Scholar]
- 15.Dover GJ, Charache S, Boyer SH, Vogelsang G, Moyer M. 5-Azacytidine increases HbF production and reduces anemia in sickle cell disease: dose-response analysis of subcutaneous and oral dosage regimens. Blood. 1985;66:527–32. [PubMed] [Google Scholar]
- 16.Christman JK, Mendelsohn N, Herzog D, Schneiderman N. Effect of 5-azacytidine on differentiation and DNA methylation in human promyelocytic leukemia cells (HL-60) Cancer research. 1983;43:763–9. [PubMed] [Google Scholar]
- 17.Silverman LR, Holland JF, Weinberg RS, et al. Effects of treatment with 5-azacytidine on the in vivo and in vitro hematopoiesis in patients with myelodysplastic syndromes. Leukemia. 1993;7 (Suppl 1):21–9. [PubMed] [Google Scholar]
- 18.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. 2002;20:2429–40. doi: 10.1200/JCO.2002.04.117. [DOI] [PubMed] [Google Scholar]
- 19.Gore SD, Carducci MA. Modifying histones to tame cancer: clinical development of sodium phenylbutyrate and other histone deacetylase inhibitors. Expert Opin Investig Drugs. 2000;9:2923–34. doi: 10.1517/13543784.9.12.2923. [DOI] [PubMed] [Google Scholar]
- 20.Gilbert J, Baker SD, Bowling MK, et al. A phase I dose escalation and bioavailability study of oral sodium phenylbutyrate in patients with refractory solid tumor malignancies. Clin Cancer Res. 2001;7:2292–300. [PubMed] [Google Scholar]
- 21.Carducci MA, Gilbert J, Bowling MK, et al. A Phase I clinical and pharmacological evaluation of sodium phenylbutyrate on an 120-h infusion schedule. Clin Cancer Res. 2001;7:3047–55. [PubMed] [Google Scholar]
- 22.DiGiuseppe JA, Weng LJ, Yu KH, et al. Phenylbutyrate-induced G1 arrest and apoptosis in myeloid leukemia cells: structure-function analysis. Leukemia. 1999;13:1243–53. doi: 10.1038/sj.leu.2401471. [DOI] [PubMed] [Google Scholar]
- 23.McGrath-Morrow SA, Stahl JL. G(1) Phase growth arrest and induction of p21(Waf1/Cip1/Sdi1) in IB3-1 cells treated with 4-sodium phenylbutyrate. The Journal of pharmacology and experimental therapeutics. 2000;294:941–7. [PubMed] [Google Scholar]
- 24.Robertson KD, Hayward SD, Ling PD, Samid D, Ambinder RF. Transcriptional activation of the Epstein-Barr virus latency C promoter after 5-azacytidine treatment: evidence that demethylation at a single CpG site is crucial. Mol Cell Biol. 1995;15:6150–9. doi: 10.1128/mcb.15.11.6150. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Ben-Sasson SA, Klein G. Activation of the Epstein-Barr virus genome by 5-aza-cytidine in latently infected human lymphoid lines. Int J Cancer. 1981;28:131–5. doi: 10.1002/ijc.2910280204. [DOI] [PubMed] [Google Scholar]
- 26.Chan AT, Tao Q, Robertson KD, et al. Azacitidine induces demethylation of the Epstein-Barr virus genome in tumors. J Clin Oncol. 2004;22:1373–81. doi: 10.1200/JCO.2004.04.185. [DOI] [PubMed] [Google Scholar]
- 27.Robertson KD, Ambinder RF. Methylation of the Epstein-Barr virus genome in normal lymphocytes. Blood. 1997;90:4480–4. [PubMed] [Google Scholar]
- 28.Shahabuddin M, Volsky B, Kim H, Sakai K, Volsky DJ. Regulated expression of human immunodeficiency virus type 1 in human glial cells: induction of dormant virus. Pathobiology. 1992;60:195–205. doi: 10.1159/000163723. [DOI] [PubMed] [Google Scholar]
- 29.Rudek MA, Zhao M, He P, et al. Pharmacokinetics of 5-azacitidine administered with phenylbutyrate in patients with refractory solid tumors or hematologic malignancies. J Clin Oncol. 2005;23:3906–11. doi: 10.1200/JCO.2005.07.450. [DOI] [PubMed] [Google Scholar]
- 30.Zhao M, Rudek MA, He P, et al. Quantification of 5-azacytidine in plasma by electrospray tandem mass spectrometry coupled with high-performance liquid chromatography. J Chromatogr B Analyt Technol Biomed Life Sci. 2004;813:81–8. doi: 10.1016/j.jchromb.2004.09.012. [DOI] [PubMed] [Google Scholar]
- 31.Phuphanich S, Baker SD, Grossman SA, et al. Oral sodium phenylbutyrate in patients with recurrent malignant gliomas: a dose escalation and pharmacologic study. Neuro Oncol. 2005;7:177–82. doi: 10.1215/S1152851704000183. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Rhee I, Jair KW, Yen RW, et al. CpG methylation is maintained in human cancer cells lacking DNMT1. Nature. 2000;404:1003–7. doi: 10.1038/35010000. [DOI] [PubMed] [Google Scholar]
- 33.Gore SD, Baylin S, Sugar E, et al. Combined DNA methyltransferase and histone deacetylase inhibition in the treatment of myeloid neoplasms. Cancer research. 2006;66:6361–9. doi: 10.1158/0008-5472.CAN-06-0080. [DOI] [PubMed] [Google Scholar]
- 34.Yang J, Tao Q, Flinn IW, et al. Characterization of Epstein-Barr virus-infected B cells in patients with posttransplantation lymphoproliferative disease: disappearance after rituximab therapy does not predict clinical response. Blood. 2000;96:4055–63. [PubMed] [Google Scholar]
- 35.Ferguson AT, Vertino PM, Spitzner JR, Baylin SB, Muller MT, Davidson NE. Role of estrogen receptor gene demethylation and DNA methyltransferase. DNA adduct formation in 5-aza-2′deoxycytidine-induced cytotoxicity in human breast cancer cells. The Journal of biological chemistry. 1997;272:32260–6. doi: 10.1074/jbc.272.51.32260. [DOI] [PubMed] [Google Scholar]
- 36.Yegnasubramanian S, Haffner MC, Zhang Y, et al. DNA hypomethylation arises later in prostate cancer progression than CpG island hypermethylation and contributes to metastatic tumor heterogeneity. Cancer research. 2008;68:8954–67. doi: 10.1158/0008-5472.CAN-07-6088. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.Marchion D, Munster P. Development of histone deacetylase inhibitors for cancer treatment. Expert review of anticancer therapy. 2007;7:583–98. doi: 10.1586/14737140.7.4.583. [DOI] [PubMed] [Google Scholar]
- 38.Dokmanovic M, Clarke C, Marks PA. Histone deacetylase inhibitors: overview and perspectives. Mol Cancer Res. 2007;5:981–9. doi: 10.1158/1541-7786.MCR-07-0324. [DOI] [PubMed] [Google Scholar]
- 39.Carew JS, Giles FJ, Nawrocki ST. Histone deacetylase inhibitors: mechanisms of cell death and promise in combination cancer therapy. Cancer letters. 2008;269:7–17. doi: 10.1016/j.canlet.2008.03.037. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.