A phase I pharmacokinetic and pharmacodynamic study of the DNA methyltransferase 1 inhibitor MG98 administered twice weekly

Original article

Annals of Oncology 14: 766–774, 2003 DOI: 10.1093/annonc/mdg216

A phase I pharmacokinetic and pharmacodynamic study of the DNA methyltransferase 1 inhibitor MG98 administered twice weekly D. J. Stewart1, R. C. Donehower2, E. A. Eisenhauer3*, N. Wainman3, A. K. Shah4, C. Bonfils5, A. R. MacLeod5, J. M. Besterman5 & G. K. Reid5 1

Ottawa Regional Cancer Centre, Ottawa, ON, Canada; 2Johns Hopkins Oncology Center, Baltimore, MD, USA; 3National Cancer Institute of Canada Clinical Trials Group, Kingston, ON, Canada; 4MGI Pharma, Bloomington, MN, USA; 5MethylGene, Inc., Montreal, QC, Canada

Background: Hypermethylation and inactivation of tumor suppressor genes by the enzyme DNA methyltransferase may lead to neoplastic transformation. MG98, a phosphorothioate antisense oligodeoxynucleotide that is a specific inhibitor of mRNA for human DNA methyltransferase 1 (DNMT1), was evaluated in a phase I study. Patients and methods: MG98 was given as a 2 h i.v. infusion twice weekly three weeks out of every four to patients with solid tumors. Pharmacokinetic evaluation was performed on days 1 and 15 of cycle 1 and mRNA expression of DNMT1 was measured in peripheral blood mononuclear cells (PBMCs). Results: Nineteen patients were entered onto the study. A total of 74 cycles (range 1–18 cycles) were administered at dose levels from 40 to 480 mg/m2. Dose limiting toxicity was seen in two of three patients at 480 mg/m2 and consisted of a constellation of fever, chills, fatigue and, in one case, confusion beginning within 6 h after the first infusion. Other toxic effects included fatigue, anorexia, nausea, vomiting and diarrhea, reversible elevations in transaminases and partial thromboplastin time. Pharmacokinetic evaluation showed Cmax and AUC to be dose proportional with low inter- and intra-patient variability. No consistent changes in DNMT1 mRNA expression were noted in PBMCs. One partial response was documented in a patient with renal cell carcinoma treated at 80 mg/m2. Conclusions: The recommended dose of MG98 was 360 mg/m2 given by 2 h infusion twice a week for three weeks out of every four. Phase II trials using this dose and schedule are underway. Key words: anti-sense oligodeoxynucleotide, DNA methyltransferase I, MG98, phase I

Introduction While all somatic cells in the body possess the same genes, the structure and function varies markedly between individual cells. This difference in structure and function between genetically identical cells arises largely as a result of the selective inactivation of different genes in different cells. Evidence suggests that a gene may be selectively inactivated as a result of being hypermethylated by DNA methyltransferase enzymes [1]. As described in a recent review [2], the process of oncogenesis involves numerous genetic and epigenetic changes in the cell. Among the changes that may occur are alterations of tumor suppressor genes such as p53 and the retinoblastoma gene [3]. Since DNA methylation patterns are aberrant in tumor cells [4], and since hypermethylation has been noted in a number of tumor suppressor loci and genes in cancer cells [5], it has been hypothesized that DNA methylation may be important in tumor

*Correspondence to: Dr E. A. Eisenhauer, Investigational New Drug Program, NCIC Clinical Trials Group, 82–84 Barrie Street, Queen’s University, Kingston, ON, Canada K7L 3N6. Tel: +1-613-533-6430; Fax: +1-613-533-2411; E-mail: [email protected] © 2003 European Society for Medical Oncology

progression [4, 6]. This hypothesis is supported by observations that DNA methyltransferase activity is often increased in tumor cell lines [7] and induction of abnormal expression of DNA methyltransferase via gene transfer may result in cellular transformation [8]. DNA methylation patterns are established during development by the de novo methyltransferase enzymes DNMT3a and DNMT3b, and maintained by DNMT1 [9]. DNMT1 is the preferred target to reverse established aberrant methylation in cancer cells as it encodes cellular maintenance methyltransferase activity. However, recent studies suggest that the de novo methyltransferase DNMT3b may also play a role in human cancer [10, 11]. Antisense oligonucleotides are synthetic nucleic acids that are designed to inhibit the translation of specific mRNAs by binding to a target region within the mRNA via Watson–Crick base pairing [12]. MG98, a second-generation phosphorothioate antisense oligodeoxynucleotide, is a specific inhibitor of human DNMT1 mRNA. It produces a dose-dependent reduction in cellular DNMT1 protein levels, resulting in the reactivation of aberrantly silenced tumor suppressor genes in human cancer cells [13]. In cell lines, MG98 has IC50 values from 50–70 nM. In human

Downloaded from http://annonc.oxfordjournals.org/ at College of Saint Rose on June 29, 2015

Received 8 January 2003; revised 9 February 2003; accepted 28 February 2003

767

Patients and methods

one to six patients per dose level. Dose escalations were permitted in individual patients provided they had completed at least one full cycle with no grade >2 MG98-related toxicity, and provided that accrual had been completed for the higher dose level without dose-limiting toxicity. At least one patient was to be entered per dose level. Once any patient experienced grade ≥2 toxicity, a minimum of three patients was required at that and all subsequent dose levels. If one of three patients on a dose level experienced doselimiting toxicity, then a further three patients were to be enrolled, for a total of six. If two or more of three or two or more of six patients experienced doselimiting toxicity, then that dose was considered to be the maximum tolerated dose (MTD). The next lowest dose from the MTD was to be considered the recommended dose for future phase II studies using this schedule. The protocol permitted expansion of this dose level to further assess its safety profile. The definition of dose-limiting toxicity (DLT) was the cycle one occurrences of one or more of: absolute granulocyte count <0.5 × 109/l for ≥5 days, febrile neutropenia or grade ≥3 neutropenic infection, platelets <50 × 109/l or thrombocytopenic bleeding, sustained grade ≥3 PTT, grade ≥3 non-hematological toxicity, or any toxic effect resulting in the patient missing two or more doses per cycle.

Patient evaluation Patients were examined and interviewed for symptoms and toxicity at least once every 4 weeks. Toxicities were evaluated using the NCI Common Toxicity Criteria (NCI CTC), version 2.0. A complete blood count, differential and platelet count was performed twice weekly for cycles 1 and 2, and weekly thereafter. Serum biochemistry was evaluated weekly. Urinalyses were carried out weekly for cycles 1 and 2, then day 1 each cycle. Appropriate radiological exams to assess known disease were repeated every 8 weeks and, if bidimensionally measurable disease was present at baseline, response was assessed according to NCIC CTG (WHO based) criteria. An electrocardiogram was done when patients went off protocol treatment.

Eligibility criteria

Planned dose alterations

This phase I study was conducted at the Ottawa Regional Cancer Centre, Ottawa, Canada and at the Johns Hopkins Oncology Center, Baltimore, MD, USA under the auspices of the NCIC CTG. The Research Ethics Committees of both institutions approved the study. The committee of the US institution has an assurance approved by the Department of Health and Human Services. All patients gave written informed consent. Eligibility criteria included: a malignant solid tumor not curable by standard therapy, age ≥18 years, ECOG performance status ≤2, life expectancy ≥12 weeks, granulocytes ≥1.5 × 109/l, platelets ≥100 × 109/l, creatinine and bilirubin ≤1.25 times the upper limit of normal (ULN), AST or ALT ≤3× ULN, proteinuria on urinalysis <2+ (or ≤500 mg protein/24 h), and normal partial thromboplastin time (PTT). Patients may have received up to three prior chemotherapy regimens. Pregnant or lactating women were ineligible. Patients with brain or leptomeningeal disease were eligible only if they had been adequately treated, had been stable for >4 months, and were asymptomatic. Ineligibility criteria were: active or uncontrolled infection, serious illness or medical condition that would not permit management according to protocol, or concomitant therapy with coumadin or heparin. Patients were not to be enrolled if they were on other anticancer agents or had received an investigational treatment within 3 weeks before registration.

Individual doses of MG98 were to be held if any of the following events were documented at the time of planned treatment: platelets <50 × 109/l, granulocytes <0.5 × 109/l, febrile neutropenia or grade ≥3 neutropenic infection, PTT grade ≥2, grade 4 nausea or vomiting despite antiemetics, or any other major organ toxicity grade ≥3 (except alopecia). Doses were to be held until the toxicity was grade ≤1, and then the patient was to be treated at the next lowest dose level. The patient was to go off treatment if there was no recovery after 2 weeks. Day 1 of a new cycle was delayed by 1 week if platelets were <100 × 109/l or granulocytes were <1.5 × 109/l, or if the PTT was grade ≥2 at the time of retreatment.

Study design MG98 (supplied by MethylGene, Inc., Montreal, Canada) was administered by i.v. infusion over 2 h twice weekly (Monday and Thursday or Tuesday and Friday) for three weeks out of every four. The starting dose of MG98 was 40 mg/m2/injection, one-sixth of the highest dose tested in monkeys [14]. Planned dose levels were 40, 80, 160, 240, 360 and 480 mg/m2/injection, with

Pharmacokinetics All participating patients underwent pharmacokinetic evaluation. On days 1 and 15 of cycle 1, 5 ml of blood were collected into vacutainer tubes containing EDTA 10 min prior to the start of MG98 infusion, then at 1, 2 h (just prior to the end of drug infusion), and at 4, 5, 6 and 8 h after the start of the infusion. Blood samples for pharmacokinetics were also collected 10 min prior to the start of MG98 infusion on day 8, cycle 1, and days 1 and 15 for cycle 2, and every second cycle thereafter. Blood samples were centrifuged at 2500 r.p.m. for 10 min at 4°C, and plasma was separated and stored in two 2 ml tubes at –20°C until assay.

MG98 assay Plasma samples were assayed for full-length MG98 by ion exchange highperformance liquid chromatography (HPLC) with ultraviolet detection, using a method adapted and validated by Anapharm, Inc. (Ste Foy, Quebec,

Downloaded from http://annonc.oxfordjournals.org/ at College of Saint Rose on June 29, 2015

tumor xenografts models it caused dose-dependent growth delay and tumor regression when compared with mismatch control oligodeoxynucleotides [14]. More frequent exposure was associated with greater antitumour effects. MG98 was generally well tolerated in animals. Reversible prolongation of partial thromboplastin time and slightly elevated transaminases, blood–urea–nitrogen and creatinine were seen in rats [14]. In monkeys, histopathological changes were found in the liver and kidneys, but it was unclear whether these were due to MG98 or to exposure to natural parasites and bacteria [14]. MG98 binds extensively to human plasma proteins. When 35 S-labeled MG98 was administered to rats, its major route of excretion was renal, and radioactivity levels in plasma, blood and tissues were measurable up to 30 days post-dose. Pharmacokinetic parameters were linear with dose in both monkeys and rats, and did not change over multiple cycles of drug administration [14]. The National Cancer Institute of Canada Clinical Trials Group (NCIC CTG) undertook two phase I trials of MG98 to determine the recommended dose of MG98 for further evaluation. The first study involved a continuous infusion schedule [15] and the second study, reported here, gave a 2-h intravenous infusion twice per week, three weeks out of every four. It was felt that this latter schedule was of interest to evaluate since the therapeutic index of antisense oligonucleotides in some instances appears better with an intermittent schedule than with a continuous highdose schedule [16].

768 Canada). In brief, MG98 was extracted from an aliquot of human EDTA plasma using a filtration procedure. Quantitation of MG98 was performed by peak area method using a weighted (1/C) linear regression to determine the concentration of MG98 in plasma samples. The method was found to be suitable for the determination of MG98 in human plasma over the range of 0.2–10 µg/ml. The lower limit of detection was 0.050 µg/ml. The precision determined by coefficients of variation (CV%) and accuracy measured of percentage of nominal value at the lower and upper limits of quantitations were 2.81% and 103%, respectively, and 1.29% and 100%, respectively. The between-run precision and accuracy of low, medium and high quality control samples ranged between 1.55% and 2.04%, and 98.1% and 98.8%, respectively. CV% within-run precision and accuracy of lower and higher limits of quantitation and low, medium and high quality control samples ranged between 0.32% and 3.04%, and 95.6% and 98.4%, respectively. MG98 was found to be stable in human EDTA plasma for 624 days at –20°C storage conditions.

Pharmacokinetic analysis was performed using WinNonlin™ 3.2 (Pharsight Corp., CA, USA) by non-compartmental methods. Cmax was the maximal observed plasma concentration. Elimination rate constant (Kel) was determined from the terminal slope of the plasma concentration–time data using ln-linear regression; terminal elimination half-life (T½) was calculated as (ln2)/Kel. Area under the plasma concentration–time curve from time zero to the last non-zero concentration, or Ct (AUC0–t) was calculated by the linear trapezoidal rule and area under the concentration–time curve from time zero to infinity (AUC0–∞) was calculated as AUC0–t + (Ct/Kel). Total body clearance (CLp) was calculated as dose/AUC0–∞ and volume of distribution at steady-state (Vdss) was calculated as MRT × CLp, where mean residence time (MRT) was calculated as (AUMC∞/AUC∞) – τ/2, and τ is time of infusion. The area under the first moment curve (AUMC) was measured using the linear trapezoidal rule and extrapolated to infinity as: AUMClast + t × Ct /Kel + Ct/(Kel)2.

Complement activation On days 1 and 15 of cycle 1, 7 ml of blood were collected 10 min before the start of the MG98 infusion (time 0), then at 2 and 4 h for measurement of complement factor Bb. Blood samples were collected in EDTA-containing tubes, inverted 10 times, and immediately placed on ice until centrifuged. Plasma was separated by centrifugation within 1 h of sample collection at 3000 r.p.m. for 15 min at 4°C. The resulting plasma samples were frozen at –70°C until analysis. All specimens taken for the measurement of complement factor Bb were analyzed by a central laboratory (Specialty Laboratories, Santa Monica, CA, USA) for consistency and standardization.

DNA methyltransferase in peripheral blood mononuclear cells Patients had 20 ml of blood collected into green-topped heparin-coated Vacutainer tubes for harvesting of peripheral blood mononuclear cells (PBMCs) 10 min prior to the start of MG98 i.v. infusion on days 1, 8 and 15 of cycle 1, days 1 and 15 of cycle 2, and every other cycle thereafter. Specimens were kept at controlled ambient temperature until processing (24 h later).

PBMC mRNA analysis PBMCs were isolated on Lymphoprep gradients (Nycomed) and total RNA was isolated from PBMC using RNeasy extraction kit (Qiagen) according to the manufacturers’ instructions. The level of DNMT1 mRNA relative to β-actin was measured from total RNA by RT–PCR according to the following method. For primer annealing of the RT step, 1 µg total RNA was mixed with

PCR was performed using 2 µl of the RT reaction and Expand LT DNA Polymerase (Roche), following the manufacturer’s recommendations. The final reaction contained 500 µM each dNTP, 600 nM each primer, a total of 3 mM MgCl2 and 2.5 U enzyme, in the provided buffer 3. Each cycle consisted of a denaturation step at 94°C for 10 s, followed by annealing at 60°C for 30 s, and by extension at 68°C for 2 min, with a 20-s increment per cycle after the first 10 cycles. A preliminary time-course PCR was performed with DNMT1 and actin primers to determine the number of cycles within which the reaction is in the linear range. For quantitative reactions, DNMT1 and actin were amplified together in a multiplex reaction where the active primers are added for the 17 cycles. The PCR product was resolved on 1% agarose gel, visualized by ethidium bromide, and quantified by alpha-imager. Each sample was measured a minimum of three times, from three independent RT reactions.

Results Patient demographics and treatment The characteristics of the 19 patients enrolled are shown in Table 1. All except one patient were male and all except two had received prior chemotherapy, including four who had received three prior chemotherapy regimens. A total of 74 cycles of treatment were administered at doses ranging from 40 to 480 mg/m2, as shown in Table 2. Table 2 also shows the number of cycles given at reduced or escalated doses.

DLT and dose recommended for phase II studies The DLT of MG98 was a constellation of fever, chills, rigors, weakness or fatigue, and confusion. This was seen in two out of three patients following their first dose of MG98 at 480 mg/m2; it was significant enough to lead to admission to hospital in one case and required a return to the outpatient department in the other. Onset of symptoms occurred within 6 h after completion of the infusion and resolved spontaneously over the ensuing hours. At 480 mg/m2, the first patient with these symptoms experienced a recurrence of the same symptoms when rechallenged at the same dose despite premedication with antihistamine and acetaminophen, but had no recurrence when retreated with a lower dose. Because of the failure of premedication to prevent symptom recurrence, when the second patient who experienced DLT at 480 mg/m2 was retreated, the dose was lowered to 360 mg/m2. No recurrence of symptoms was seen. Thus 480 mg/m2 was determined to be the MTD. With this observation the next lowest dose level, 360 mg/m2, was reopened to enter three additional patients (for a total of six). Although the first three patients entered at 360 mg/m2 had tolerated their first cycle well, one of the three in the expansion cohort encountered the same DLT syndrome after day 11 dose in cycle 1 of therapy. Interestingly, when pharmacokinetic data became available, this patient, who had

Downloaded from http://annonc.oxfordjournals.org/ at College of Saint Rose on June 29, 2015

Pharmacokinetic analysis

500 ng of oligo(dT)12–18 (Gibco) in 0.2 ml thin-walled tubes, heated at 65°C for 10 min, then chilled on ice. Reverse transcription was performed with 50 U Expand RT (Roche) in the provided buffer supplemented with 10 mM DTT, 1 mM each dNTP, and 20 U RNasin and incubated at 42°C for 60 min. The RT product was neither heated nor frozen, but stored at 4°C until used for PCR. A no-RT control, where the enzyme was omitted, was run for each blood sample.

769 received a nephrectomy in the past for renal cell carcinoma, had day 1 Cmax and AUC values which were considerably higher than any other patient at that dose (Cmax 254.6 versus a mean of 91.4 µg/ml; AUC 647.7 versus a mean of 298.8 µg·h/ml for that dose level). This observation, in addition to the timing of the reaction, suggests the DLT syndrome observed may have been related to MG98 exposure. The day 1 Cmax values in the patients at 480 mg/m2 were 115.6, 136.2 and 97.7 µg/ml, and AUC values were 460.9, 557.4 and 384.1 µg·h/ml, respectively. The latter two patients had dose limiting symptoms. Fever and chills were also noted in occasional patients at lower doses, but were less severe and not part of an acute syndrome as described. The recommended dose of MG98 for phase II studies is 360 mg/m2 when MG98 is administered as a 2-h i.v. infusion on days 1, 4, 8, 11, 15 and 18 of a 28 day cycle.

Table 1. Patient characteristics (n = 19) Characteristic

No. of patients

Median age, years (range)

66 (42–82)

Gender Male

18

Female

1

ECOG performance status 0

2

1

15

2

2

No. of disease sites 6

2

3

3

2

≥4

8

Other toxic effects

Prior therapy Chemotherapy

17

Radiotherapy

9

Immunotherapy

3

Other therapy

4

No. of prior chemotherapy regimens 0

2

1

8

2

5

3

4

Tumor type Mesothelioma

6

Lung (non-small-cell)

2

Renal cell

6

Colorectal

2

Pancreas

1

Unknown primary

1

Other

1

Table 2. Patient entry and cycles of treatment Dose level (mg/m2)

No. of pts starting at this dose

No. of cycles given In pts starting at this dose

Escalated from lower dose

Reduced from higher dose

Total

40

1

12

0

0

12

80

2

20

4

6

30

160

3

5

1

0

6

240

4

6

0

2

8

360

6

11

0

1

12

480

3

6

0

0

6

Total

19

pts, patients.

74

Other toxic effects seen with MG98 are shown in Tables 3, 4 and 5. Most patients experienced fatigue, which was grade 3 in severity in nine patients at some point during their treatment. In two cases the severe fatigue was documented during the same cycle that DLT was noted at 480 mg/m2 and represented part of the symptom complex of the DLT. In the remainder it was not considered dose limiting, either because it occurred later than cycle 1 (four patients: fatigue grade 3 was seen in cycles 4, 5, 17 and 18), was considered unrelated to therapy (one patient) or was seen in patients with grade 2 fatigue at baseline and thus the degree of fatigue attributable to drug was uncertain. Gastrointestinal toxicity (nausea, vomiting and diarrhea) seemed to be dose related, and was generally limited to grade 1 or 2. The solitary episode of grade 3 diarrhea occurred at 480 mg/m2 during the patient’s second cycle of therapy. Despite having no dose reduction, the patient did not have diarrhea of grade >1 thereafter. Laboratory evidence of reversible dose-related hepatic toxicity was also common. One patient with grade 3 elevation in bilirubin had a blocked stent, not drug toxicity as the underlying cause. When first versus worst cycle were compared (Table 3A and 3B), there was a suggestion that transaminase increases became more common and severe with repeated dosing, although numbers are small. Creatinine rises were documented in six cases: five were grade 1–2 and one was grade 3. MG98 may have been responsible for the rise in creatinine in this latter patient since no other cause could be found and following discontinuation of treatment, the creatinine level gradually fell. Myelosuppression was minimal. As expected, dose-related prolongation of PTT was common immediately after MG98, but in most patients it had recovered to normal values by the time of the next treatment (Table 5). No bleeding was seen. Two patients had events late in their treatment that were considered to be possible hypersensitivity reactions. The first patient developed rash, dizziness and hypotension after being dose escalated to 160 mg/m2 from a lower dose. His next dose was reduced to 80 mg/m2 given with antihistamines and dexamethasone. He experienced no recurrence of this reaction. A second patient re-escalated to 160 mg/m2 following a dose reduction for PTT experienced chills and fever after day 1, cycle 7. While receiving

Downloaded from http://annonc.oxfordjournals.org/ at College of Saint Rose on June 29, 2015

1

770 Table 3A. Biochemical toxicity: worst toxicity by grade cycle 1 at each dose level Dose (mg/m2/day)

No. of patients

AST

ALT

1

2

1

3

4

1

40

1

80

2

160

3

1

240

4

2

2

360

6

6

5

480

3

2

Total

19

11

Bilirubin 2

3

4

1

2

Creatinine 3

4

1

2

3

4

1 1 1 1

1

8

1

1

1

1

1

2

1

1

1

2

1

Dose (mg/m2/day)

No. of patients

AST 1

ALT 2

3

4

1

Bilirubin 2

40

1

1

80

2

1

160

3

240

4

2

360

6

4

2

2

2

480

3

1

1

1

1

Total

19

9

5

5

3

3

4

1

2

Creatinine 3

4

1

2

3

4

1 2

1 2

day 15, cycle 7, he developed severe bronchospasm, hypoxia, cyanosis, rigors and tachycardia. The reaction resolved with administration of oxygen and diphenhydramine. This patient was not retreated with MG98.

Complement activation There was no clinical evidence of complement activation. Two patients treated with MG98 at 160 mg/m2 had minor elevations of complement factor Bb levels of 1.6–2.4 µg/ml. In all other patients, Bb factor levels were <1.6 µg/ml (normal).

Response Fourteen patients were evaluable for response. One patient with renal cell carcinoma previously treated with retinoid plus interferon and an investigational alkylating agent had a partial response in lung metastases lasting 9 months. One patient with renal cell cancer and one with colorectal carcinoma had stable disease for 3.7 and 2.6 months, respectively. The remainder of patients with measurable disease had a best response of disease progression.

Pharmacokinetics Plasma MG98 concentration-time profiles are shown in Figure 1. Plasma concentrations declined monoexponentially after ter-

2 1

2

1 1

2

1

1 1

1

1

3

1

1

2

1

mination of infusion when plotted as log concentration versus time (not shown). Pharmacokinetic parameters are presented in Table 6 and Figure 2. The AUC0–∞ and Cmax both increased linearly with dose. The Vdss was low, approximating that of blood volume. There was a trend towards slightly lower CLp and longer T½ with higher MG98 doses. This may have been due to patient factors or the small numbers in each group. Mean elimination phase half-life was short and ranged from 0.9 to 2.6 h. Pharmacokinetic data from patients receiving the same dose on days 1 and 15 showed low inter- and intra-patient variability and no evidence of drug accumulation. There was an outlier at 360 mg/m2 (day 1 AUC 647.74 µg·h/ml versus day 1 mean of 342.1 µg·h/ml) (see Figure 2). This patient had higher Cmax and AUC values than others treated at that dose, and also experienced dose-limiting symptoms as described previously. He had undergone a nephrectomy in the past for renal cell carcinoma and this observation led us to explore other dose levels for patients with nephrectomy. There were four such patients identified: one at 80 mg/m2, one at 240 mg/m2 and two at 360 mg/m2, including the one with high Cmax and AUC. The other patient at 360 mg/m2 and the one at 240 mg/m2 both had pharmacokinetic values in the same range as other patients at the same dose. The patient at 80 mg/m2 had higher day 1 Cmax (23.9 µg/ml) and AUC (86.3 µg·h/ml) than the other patient at that dose (values: Cmax 13.3 µg/ml, AUC 46.1 µg·h/ml).

Downloaded from http://annonc.oxfordjournals.org/ at College of Saint Rose on June 29, 2015

Table 3B. Biochemical toxicity: worst toxicity by grade by patient by dose level

771

1 4 3 2 5

1 2

3

1 4 5

1 1 1

2 1

1 1 1

2 4

2

1 4 5

1 1

1 1

2

1

7 4

3

2 1

2

1 1

4

2

9

1

1

8

3

1 2

3

19 Total

6

480

360

4 240

3 160

b

1 4

1 1 1 2 1

1

1 1 1 1 1 1

1 1 1 1

1 1

2 80

40

a

1

1 2

1 1

1 1 1 1

4 3 2 1 4 3 2 1 4 3 2 1

Dose in mg/m2/day. Hypersensitivity reactions: patient entered at 40 mg/m2 had been escalated to 160 mg/m2 and developed rash, nausea and dizziness at end of infusion day 1 cycle 14. Patient was retreated with premedication at 80 mg/m2 without recurrence. A second patient entered at 160 mg/m2 developed bronchospasm, cyanosis and rigors 1 h after treatment day 15 cycle 7. Patient recovered with treatment.

1

1b

4 3 b

2 1 3 2 1 2 1 2

3

4

2 1 1

3

4

1

2

3

4

Diarrhea Vomiting Nausea Chills/rigors Fever Anorexia Fatigue No. of patients Dosea

Table 4. Common non-hematological toxicity: worst toxicity by grade per patient (all cycles) at each dose level

Changes in PBMC DNMT1 mRNA are presented in Figure 3A and B. Eighteen patients had baseline and follow-up measurements and were included in this analysis. No consistent changes in DNMT1 mRNA expression were noted in PBMCs. No doserelated trend was seen.

Discussion This study demonstrated that the DNMT1 antisense oligonucleotide MG98 was generally well tolerated when administered as a 2-h i.v. infusion twice per week for three weeks out of every four. Post-infusion reactions consisting of significant fever, chills, rigors, fatigue/weakness and confusion were dose limiting at an MG98 dose of 480 mg/m2. Attempts to modify this by pretreatment with antihistamines and acetaminophen were not successful in the first patient and so were not pursued further. The dose recommended for phase II studies is therefore 360 mg/m2 when administered on this schedule. Fatigue was common, as were dose-related prolongations of PTT and reversible elevations of hepatic enzymes. The elevations seen in serum creatinine were not clearly drug related except in one instance. Gastrointestinal toxicity and myelosuppression were minimal. Overall, MG98 was better tolerated in this study than in the other NCIC CTG phase I study in which the drug was administered by continuous intravenous infusion [15]. In that study, dose-limiting transaminitis and fatigue were noted at above 80 mg/m2/day. In light of the fact that MG98 was generally well tolerated in the intermittent schedule, was easier to administer, could deliver higher total doses and produced one prolonged partial remission, phase II studies have been initiated using this schedule. Table 5. Partial thromboplastin time alterations with MG98 Dose (mg/m2) a

No. of doses

40

80

160

240

360

480

102

118

64

28

75

41

5

1

1

2

1b

1b

Pre-dose partial thromboplastin time prolongation Grade 1 2 3 Post-dose partial thromboplastin time prolongation Grade 1

10

2

1

3 a

16

45 4

25

56

17

6

20

1

1

Six doses intended per cycle (days 1, 4, 8, 11, 15 and 18 of each 28 day cycle). b Neither of the grade 3 pre-dose PTT values qualified as DLT: one was drawn from a heparinized line and the other occurred in association with and was attributed to a stent-induced septicemia.

Downloaded from http://annonc.oxfordjournals.org/ at College of Saint Rose on June 29, 2015

3

4

Stomatitis

4

Hyper-sensitivity

DNMT1 mRNA expression

772

Table 6. Summary MG98 pharmacokinetics (mean ± SD) after 2-h i.v. infusion in patients with solid tumors Parameter

Day

AUC0–∞ (µg·h/ml) Cmax (µg/ml) 2

Vdss (l/m ) 2

CLp (l/h/m ) T½ (h)

Dose, mg/m2 [no. of pts (day 1/15)] 40 (1/1)

80 (2/3)

160 (3/2)

240 (4/3)

360 (6/6)

480 (3/1)

1

22.7

66.2 ± 28.4

145.5 ± 39.1

169.5 ± 32.4

342.1 ±151.7

467.5 ± 86.9

15

21.3

60.8 ± 20.1

133.2 ± 68.2

179.2 ± 40.4

255.5± 59.5

502.4 116.5 ± 19.3

1

7.0

18.6 ± 7.5

40.0 ± 9.7

46.2 ± 6.7

110.3 ± 71.0

15

6.6

18.3 ± 5.5

36.5 ± 14.2

46.2 ± 6.2

72.4 ± 14.4

141.1

1

4.6

3.6 ± 2.2

2.6 ± 0.6

3.3 ± 0.3

2.8 ± 0.9

3.3 ± 0.9

15

5.0

3.4 ± 2.1

3.0 ± 1.0

3.6 ± 0.6

3.4 ± 0.8

3.2

1

1.8

1.3 ± 0.6

1.2 ± 0.4

1.5 ± 0.3

1.2 ± 0.3

1.1 ± 0.2

15

1.9

1.4 ± 0.5

1.4 ± 0.7

1.4 ± 0.4

1.5 ± 0.4

1.0

1

0.8

1.2 ± 0.1

1.6 ± 0.2

1.8 ± 0.4

2.1 ± 0.2

2.3 ± 0.5

15

0.9

1.2 ± 0.2

1.7 ± 0.6

1.7 ± 0.4

2.0 ± 0.5

2.6

AUC0–∞, area under the concentration–time curve from time zero to infinity; Cmax, mean plasm concentration; Vdss, volume of distrubution at steady state; CLp, total body clearance; T½, half life; pts, patients.

Whether or not MG98 proves active in any tumor type as a single agent, it could be of interest to test in combination with other anticancer agents. In addition to arising from overexpression of various characteristics, chemotherapy resistance may also

be due to underexpression of some factors [17]. Underexpression of a characteristic may arise as a result of hypermethylation of the gene responsible. For example, resistance to cisplatin may be conferred by hypermethylation of the hMLH1 gene [18]. Hence,

Downloaded from http://annonc.oxfordjournals.org/ at College of Saint Rose on June 29, 2015

Figure 1. Mean plasma concentration–time profiles of MG98 on days 1 and 15 following 2-h i.v. infusion in patients with solid tumors.

773

Figure 2. Plot of AUC and Cmax versus dose following 2-h i.v. infusion of MG98 administered on days 1 and 15 in patients with solid tumors.

Downloaded from http://annonc.oxfordjournals.org/ at College of Saint Rose on June 29, 2015

Figure 3. Percentage change from baseline in peripheral blood mononuclear cell DNMT1 mRNA by patient. (A) Dose levels 40–240 mg/m2 and (B) dose levels 360–480 mg/m2. *Values >300% therefore bar truncated.

reversal of gene hypermethylation might reduce resistance to some chemotherapy agents. The mechanism by which MG98 caused acute infusion reactions is unclear. Other antisense oligonucleotides may activate complement [19], but there was little evidence of complement activation in patients in our study. The risk of these reactions appeared to be dose-related and dose reduction prevented their recurrence. The timing of the reactions plus the observation of the increased exposure to the drug in the single patient at 360 mg/m2 who experienced these effects makes it probable that the event is exposure related. No consistent changes in DNMT1 mRNA expression were noted in PBMCs. The degree of variability in DNMT1 expression in PBMCs is unknown, and it is unclear whether the expression in these cells is reflective of changes in tumor tissues. Unfortunately, the validity of using PBMC as a surrogate for tumor tissue levels of DNMT1 could not be assessed in murine models in advance of human studies. In addition, mRNA measurements were done at times of MG98 trough levels in order to determine whether a sustained decrease in message could be detected. Measurements taken 24–48 h after the completion of the infusion may have been better suited to explore the impact of the drug and drug dose on DNMT1 mRNA expression. In some of the ongoing phase II studies, an attempt is being made to evaluate changes in tumor cell DNMT1expression following MG98 treatment. After 2-h i.v. infusion of MG98, plasma concentration declined mono-exponentially. MG98 AUC0–∞ and Cmax increased linearly with dose and the CLp and T½ remained unchanged on days 1 and 15. The CLp of MG98 was low, as was the Vdss, suggesting that MG98 does not distribute extensively into tissues. The impact of this characteristic on its efficacy is unclear, but direct measures of

774

Acknowledgements This study was supported by grants from the National Cancer Institute of Canada and MethylGene, Inc., Montreal, Canada.

References 1. Becker PB, Ruppert S, Schutz G. Genomic footprinting reveals cell typespecific DNA binding of ubiquitous factors. Cell 1987; 51: 435–443. 2. Goffin J, Eisenhauer E. DNA methyltransferase inhibitors – state of the art. Ann Oncol 2002; 13: 1699–1716. 3. Lehman TA, Reddel R, Peiifer AM et al. Oncogenes and tumor suppressor genes. Environ Health Perspect 1991; 93: 133–134. 4. Szyf M. DNA methylation properties: consequences for pharmacology. Trends Pharmacol Sci 1994; 15: 233–238. 5. Merlo A, Herman JG, Mao L et al. 5′ CpG island methylation is associated with transcriptional silencing of the tumor suppressor p16/CDKN2/ MTS1 in human cancers. Nat Med 1995; 1: 686–692.

6. Szyf M. The DNA methylation machinery as a target for anticancer therapy. Pharmacol Ther 1996; 70: 1–37. 7. Kautiainen TL, Jones PA. DNA methyltransferase levels in tumorigenic and nontumorigenic cells in culture. J Biol Chem 1986; 261: 1594–1598. 8. Wu J, Issa JP, Herman J et al. Expression of an exogenous eukaryotic DNA methyltransferase gene induces transformation of NIH 3T3 cells. Proc Natl Acad Sci USA 1993; 90: 8891–8895. 9. Okano M, Bell DW, Haber DA et al. DNA methyltransferases DNMT3a and DNMT3b are essential for de novo methylation and mammalian development. Cell 1999; 99: 247–257. 10. Beaulieu N, Morin S, Chute I et al. An essential role for DNA methyltransferase DNMT3b in cancer cell survival. J Biol Chem 2002; 277: 28176–28181. 11. Rhee I, Bachman KE, Park BH et al. DNMT1 and DNMT3b cooperate to silence genes in human cancer cells. Nature 2002; 416: 552–556. 12. Crooke ST, Bennett CF. Progress in antisense oligonucleotide therapeutics. Ann Rev Phamacol Toxicol 1996; 36: 107–129. 13. Fournel M, Sapieha P, Beaulieu N et al. Down-regulation of human DNA-(cytosine-5) methyltransferase induces cell cycle regulators p16ink4A and p21WAF/Cip1 by distinct mechanisms. J Biol Chem 1999; 274: 24250–24256. 14. MG98 Investigator’s Brochure. MethylGene, Inc., Montreal, Canada, 1998. 15. Siu L, Gelmon K, Moore M et al. A phase I and pharmacokinetic study of the human DNA methyltransferase (MeTase) antisense oligodeoxynucleotide MG98 given as a 21-day continuous infusion every 4 weeks. Proc Am Soc Clin Oncol 2000; 19: 189a (Abstr 733). 16. Leonetti C, D’Agnano I, Lozupone F et al. Antitumor effect of c-myc antisense phosphorothioate oligodeoxynucleotides on human melanoma cells in vitro and in mice. J Natl Cancer Inst 1996; 88: 419–429. 17. Stewart DJ, Raaphorst P, Yau J et al. Active vs passive resistance, dose– response relationships, high dose chemotherapy, and resistance modulation: a hypothesis. Invest New Drugs 1996; 14: 115–130. 18. Plumb JA, Strathdee G, Sludden J et al. Reversal of drug resistance in human tumor xenografts by 2′-deoxy-5-azacytidine-induced demethylation of the hMLH1 gene promoter. Cancer Res 2000; 60: 6039–6044. 19. Monteith DK, Levin AA. Synthetic oligonucleotides: the development of antisense therapeutics. Toxicol Pathol 1999; 27: 8–13.

Downloaded from http://annonc.oxfordjournals.org/ at College of Saint Rose on June 29, 2015

target effect in tumor tissue would be the preferred approach to evaluating this. Since the phase II program of MG98 includes one study with renal cell carcinoma, it will be of interest to continue to explore the relationships between toxicity, nephrectomy and pharmacokinetics. Two of four patients with nephrectomy in this phase I trial had lower clearance and higher Cmax and AUC values than non-nephrectomized patients treated at the same doses. Since many patients with renal cell carcinoma will have had previous nephrectomy, this population will be important to follow closely for toxicity and the effect of drug on creatinine levels. In summary, MG98 was generally well tolerated when given as a 2-h infusion on 2 days per week for three weeks out of four. Phase II studies in head and neck cancer, and renal cell carcinoma have recently been completed, and studies in other tumor types are underway.