Purpose: VNP40101M (Cloretazine), a novel DNA alkylating agent, was evaluated in a phase I study in children with recurrent brain tumors.

Experimental Design: VNP40101M was given i.v. daily for 5 consecutive days every 6 weeks for up to eight cycles. Dose escalation was done independently in patients stratified based on intensity of prior therapy (moderately pretreated, stratum I; heavily pretreated, stratum II). Correlative studies included pharmacokinetics and measurement of O6-alkylguanine-DNA alkyl transferase levels in peripheral blood mononuclear cells before and after treatment.

Results: Forty-one eligible patients (stratum I, 19; stratum II, 22) were enrolled on this study. The dose-limiting toxicity in 35 evaluable patients was myelosuppression, which occurred in 4 of 16 patients in stratum I and 3 of 19 patients in stratum II. Pharmacokinetic studies showed a median terminal half-life of 30 min (range, 14-39.5). The maximum tolerated dose in stratum I and II were 45 and 30 mg/m2/d daily for 5 days every 6 weeks, respectively. Peripheral blood mononuclear cells alkylguanine alkyl transferase levels did not decrease significantly after VNP40101M treatment. Central imaging review confirmed that three patients had stable disease for a median of 45 weeks (range, 37-61+) after therapy.

Conclusions: The recommended dose of VNP40101M for phase II studies in children with brain tumors is 45 mg/m2/d in moderately pretreated and 30 mg/m2/d in heavily pretreated patients when administered for 5 consecutive days every 6 weeks.

Despite advances in surgery, chemotherapy, and irradiation, the prognosis for children with recurrent brain tumors remains dismal. One of the causes for treatment failure in these patients is drug resistance and alternative therapeutic strategies are required to surmount this problem. The novel bifunctional DNA alkylating agent VNP40101M (Cloretazine; Vion Pharmaceuticals) belongs to the group of sulfonyl hydrazine prodrugs that spontaneously generate nucleophilic species that can efficiently alkylate DNA, resulting in DNA cross-linking and cell death (1). This agent also generates methylisocyanate that has been shown in vitro to inhibit alkylguanine alkyl transferase (AGT; ref 2). Phase I and II clinical trials with VNP40101M have been conducted in adults with advanced leukemia and solid tumors using either a single dose every 6 weeks or a fractionated (weekly for 4 weeks every 6 weeks) schedule (38). Promising results have been observed with this drug in patients with untreated acute myelogenous leukemia especially in combination with high-dose Ara-C (4, 6, 9). In addition, preclinical testing of this compound using a fractionated schedule showed excellent activity in pediatric brain tumor xenografts with acceptable toxicity (10). Based on these promising preclinical findings and emerging clinical data, the Pediatric Brain Tumor Consortium initiated a phase I study of VNP40101M using a fractionated schedule in children with recurrent malignant brain tumors.

Objectives. The primary objective of this clinical trial was to estimate the maximum tolerated dose (MTD) of VNP40101M in children when given IV daily for 5 consecutive days every 6 weeks. Secondary objectives included estimation of pharmacokinetics of VNP40101M and its active metabolite, VNP4090CE, after the first dose of VNP40101M; estimation of AGT levels in peripheral blood mononuclear cells (PBMC) before and 4 h after the fifth dose of VNP40101M; and preliminary evaluation of activity in children with recurrent primary brain tumors after two treatment courses.

Patients with recurrent previously biopsy confirmed primary brain tumors (except in those with intrinsic brain stem or optic pathway gliomas where initial biopsy confirmation may not be available before study entry) were eligible for this study. Subjects had to be ages ≤21 years and have a Lansky or Karnofsky performance score of ≥50. The following intervals from prior treatment were required before study entry: at least 3 weeks from prior myelosuppressive chemotherapy or biological therapy (6 weeks for nitrosoureas), 1 week for nonmyelosuppressive biological therapy, 3 months for prior craniospinal irradiation and/or myeloablative chemotherapy with autologous stem cell rescue, and 2 weeks for prior focal radiotherapy. Documentation of adequate organ function was required and included the following: a hemoglobin of ≥8 g/dl, absolute neutrophil count of ≥1,000/μL (unsupported), platelet count of ≥100,000/μL (unsupported), blood urea nitrogen of ≤25 mg/dl (9 mmol/L), serum creatinine of ≤ × 1.5 times upper limit of institutional normal, serum bilirubin of ≤ × 1.5 times upper limit of institutional normal, aspartate aminotransferase/ALT of ≤ × 1.5 upper limit of institutional normal, a cardiac ejection fraction of ≥50%, an electrocardiogram without clinically significant arrhythmias, and a corrected carbon monoxide diffusion capacity of lung of ≥60%. For patients who could not adequately cooperate for pulmonary function testing, a chest radiograph without pulmonary infiltrates, pneumonia, pleural effusion, hemorrhage, or fibrosis, and a resting pulse oximetry in room air of ≥94% were required. Patients of childbearing age were required to use at least two methods of contraception while on study.

Excluded were patients with bone marrow involvement, pregnancy, uncontrolled infection, concurrent treatment with another investigational agent, or allergy to polyethylene glycol.

The institutional review boards of each Pediatric Brain Tumor Consortium institution approved the protocol before initial patient enrollment, and continuing approval was maintained throughout the study. Patients or their legal guardians gave written informed consent, and assent was obtained as appropriate at the time of enrollment.

Study design and treatment. The starting dose level of VNP40101M was 45 mg/m2/d for 5 consecutive days. This dosage was ∼75% of the dose intensity at the adult MTD (305 mg/m2) for recurrent solid tumors (8). Because the dose-limiting toxicity (DLT) for this drug was expected to be myelosuppression, patients were divided into two strata based on intensity of prior therapy. Stratum I was for patients who had received less than or equal to two prior chemotherapy regimens and/or focal radiotherapy. Stratum II was for those who had received more than two prior chemotherapy regimens, craniospinal irradiation, and/or myeloablative chemotherapy with autologous stem cell rescue. Dose escalation was done independently in each stratum, and no intrapatient dose escalations were permitted. A modified Continual Reassessment Method (CRM) as described by Piantadosi et al. (11) was used to estimate the MTD. The latter was defined as the dose at which 25% of patients were expected to experience a DLT. Rounding CRM estimated MTDs to the nearest protocol-prescribed dose level provided dose-finding MTDs. The model used prior information at dose levels of 11 and 484 mg/m2 (which were unlikely to be investigated during the course of the study) with assigned weights of five patients each and toxicity probabilities of 1% and 99%, respectively.

Two patients were enrolled to a dose level under study in each stratum. An additional patient could be assigned to the same level without needing toxicity information from the previous two patients. The CRM model was continually updated, and dose escalation decisions were made as toxicity information became available for each patient. Subsequent dose levels were determined to be the prespecified level closest to the CRM estimated MTD without skipping a level that had been assigned to fewer than two evaluable patients. For de-escalation to lower levels, skipping a dose level was allowed. Additional patients were treated at the dose-finding MTDs to better describe the pattern of toxicities associated with VNP40101M.

Toxicity was graded using the National Cancer Institute Common Terminology Criteria for Adverse Events (version 3.0). The DLT observation period began with the first dose of VNP40101M and concluded at the end of course 1 (first 6 weeks of therapy). Hematologic DLTs included grade 4 neutropenia lasting >7 days, grade 4 anemia, grade 4 thrombocytopenia, or any thrombocytopenia requiring more than or equal to two platelet transfusions within 7 days. Nonhematologic DLTs included any ≥ grade 3 toxicity with the specific exclusion of as follows: grade 3 nausea and vomiting adequately controlled by antiemetics or grade 3 fever or infection of <5 days duration. Additionally, any delay of >14 days between course 1 and 2 due to unresolved hematologic or nonhematologic toxicity was considered a DLT. Patients who died within 14 days of completing the DLT observation period without fully recovering from toxicity were considered to have had a DLT. Patients who came off treatment for reasons other than toxicity before conclusion of the DLT observation period were replaced for purposes of estimating the MTD.

Drug supply and administration. VNP40101M was supplied by Vion Pharmaceuticals as a clear colorless, slightly viscous sterile nonaqueous solution for i.v. administration at a concentration of 10 mg/mL. Each 10-mL vial contained 100 mg of VNP40101M, 3 mL of anhydrous ethyl alcohol, 6 mL of citric acid, and 7 mL of polyethylene glycol 300 (12). Undiluted VNP40101M for injection was refrigerated at 2°C to 8°C. The drug was diluted to concentrations between 0.1 and 4 mg/mL using 5% dextrose injection before administration, kept at room temperature, and given within 2 to 4 h of dilution. Glass, polyethylene, or polyurethane containers and i.v. administration sets were used for drug administration.

Diluted VNP40101M was administered to study subjects i.v. >30 min for 5 consecutive days every 6 weeks. One course was defined as 6 weeks of treatment. Patients were treated for up to eight courses in the absence of disease progression or unacceptable toxicity. The starting dose (dose level 1) was 45 mg/m2/d for each stratum. The study had planned de-escalation dose levels (0 and −1, corresponding to 30 mg/m2 and 20 mg/m2/d, respectively) if the first dose level was found to be too toxic. Antiemetics were used prior each dose per institutional guidelines. Routine use of growth factors was not permitted except in the case of serious neutropenic conditions with prior approval from the protocol chair. Antiepileptic drugs were allowed as required for seizure control. Corticosteroids were permitted but used at the lowest dose necessary to control neurologic symptoms and discontinued if possible. Pneumocystis carinii pneumonia prophylaxis was recommended for all patients.

Patients who experienced dose-limiting toxicity during course 1 or unacceptable toxicity in later courses but who showed clinical and/or radiologic improvement were allowed to dose reduce only once to the next lower dose level and continue treatment after recovery from prior toxicity.

Required clinical and laboratory studies before, during, and at the end of therapy. Patients received physical examination before enrollment and before each course of treatment. A complete blood count was obtained at baseline and weekly thereafter. Electrolytes, blood urea nitrogen, and creatinine, serum bilirubin, aspartate aminotransferase, and alanine aminotransferase were obtained at baseline and every 2 weeks during therapy. Echocardiogram, electrocardiogram, and carbon monoxide diffusion capacity of lung (or chest radiograph plus pulse oximetry) were obtained at baseline and before every other course. Females of childbearing age had a serum pregnancy test done at baseline and before each course.

Response assessment. Magnetic resonance imaging of the brain (and spine if indicated) was done at baseline and before every other course to obtain preliminary evidence of activity of this agent in recurrent brain tumors. Magnetic resonance imaging used three-dimensional magnetic resonance imaging techniques as standardized among the Pediatric Brain Tumor Consortium institutions. Standard response criteria were applied as described previously (13). Treatment response was determined based on the first magnetic resonance imaging obtained after two courses of VNP40101M. The images were electronically transferred to the Pediatric Brain Tumor Consortium Neuroimaging Center for central review along with any subsequent scans that showed evidence of objective response.

Pharmacokinetic studies. Blood samples for pharmacokinetics studies of VNP40101M and its metabolite, VNP4090CE, were obtained in all consenting patients on the first day of course 1. Blood (5 mL) was collected from a peripheral vein or a separate line before the start of the VNP40101M infusion and 0.08, 0.25, 0.5, 1, 2, and 4 h after the end of the infusion. Blood was collected in sodium heparin tubes containing 0.125 mL of 2.0 mol/L citric acid solution and immediately centrifuged at 2°C to 8°C at 1,500 × g for 15 min to obtain plasma. Samples were stored at −80°C until analysis using high performance liquid chromatography with tandem mass spectrometry (14). The lower limit of quantitation of VNP40101M was 1 ng/mL, whereas the lower limit of quantitation of VNP4090CE was 0.5 ng/mL.

Initially, VNP40101M and VNP4090CE concentration-time data were modeled maximum likelihood in ADAPTII (15). Only the subgroup of patients with “adequate” data were used in this initial analysis. For the purposes of this study, adequate was defined by visual inspection of the data by the study pharmacologist, the presence of all data points, and the last scheduled time point (i.e., 4 h). A multicompartment pharmacokinetic model (two compartments each for VNP40101M and VNP4090CE) was fit to the VNP40101M and VNP4090CE plasma concentration-time data simultaneously. Estimated variables included volume of the central compartment (Vc), elimination rate constants for VNP40101M and VNP4090CE, and intercompartment rate constants for VNP40101M and VNP4090CE. The mean and SD from the maximum likelihood estimation of these variables were used to establish the population priors for use in the Bayesian analysis. Finally, the variable distribution from this subgroup of patients was used to apply a maximum a posteriori (MAP) Bayesian approach to the entire study group to generate final variable estimates. VNP40101M systemic clearance was calculated using standard equations. For each patient, the area under the concentration-time curve from zero to infinity (AUC0→∞) for VNP40101M and VNP4090CE was calculated by integration of the simulated concentration-time data from model estimates. Pharmacokinetic variables were summarized using descriptive statistics with the median and range reported.

AGT levels in PBMC. In consenting patients, 5 mL of heparinized blood was collected at two time points (pretherapy and 4 h after the fifth dose) during the first course of treatment. PBMC were isolated by density gradient centrifugation over Ficoll-Paque (Amersham). The cell pellet was frozen at −70°C until assayed for AGT levels as previously described (6).

Forty-two patients were initially enrolled on the study between March 2005 and June 2006. One patient was found to be ineligible due to primary tumor site located in the spinal cord. Of the remaining 41 eligible patients, 35 patients were evaluable for estimation of MTD (16 in stratum I and 19 in stratum II). Six patients were not evaluable for DLT due to disease progression before completion of DLT evaluation period (five patients) or withdrawal from study before starting protocol therapy (one patient). Patient characteristics are summarized in Table 1.

Table 1.

Patient characteristics and observed DLTs by stratum

Patient characteristicStratum IStratum II
No. of patients enrolled on study 20 22 
No. of eligible patients 19 22 
No. of patients evaluable for DLT 16 19 
Median age at enrollment (range) 8.3 y (0.9- 21) 9.9 y (1.5- 21.5) 
Male to female 9:11 11:11 
Histologic diagnosis   
    Malignant glioma 
    Brain stem glioma 
    Medulloblastoma 
    PNET 
    Ependymoma 
    Atypical teratoid rhabdoid tumor 
    Low-grade glioma 
Median number of courses (range) 2 (1-6) 2 (1-8) 
Observed DLTs 4/16 3/19 
    Grade IV thrombocytopenia 1/12 (45)*; 1/2 (60); 1/2 (78) 1/4 (45) 
    Grade IV neutropenia >7 d 1/2 (78)  
    Grade IV thrombocytopenia +grade IV neutropenia >7 d  1/4 (45) 
    Delayed platelet recovery >14 d  1/4 (45) 
Patient characteristicStratum IStratum II
No. of patients enrolled on study 20 22 
No. of eligible patients 19 22 
No. of patients evaluable for DLT 16 19 
Median age at enrollment (range) 8.3 y (0.9- 21) 9.9 y (1.5- 21.5) 
Male to female 9:11 11:11 
Histologic diagnosis   
    Malignant glioma 
    Brain stem glioma 
    Medulloblastoma 
    PNET 
    Ependymoma 
    Atypical teratoid rhabdoid tumor 
    Low-grade glioma 
Median number of courses (range) 2 (1-6) 2 (1-8) 
Observed DLTs 4/16 3/19 
    Grade IV thrombocytopenia 1/12 (45)*; 1/2 (60); 1/2 (78) 1/4 (45) 
    Grade IV neutropenia >7 d 1/2 (78)  
    Grade IV thrombocytopenia +grade IV neutropenia >7 d  1/4 (45) 
    Delayed platelet recovery >14 d  1/4 (45) 
*

Numbers within parentheses indicates dose level (mg/m2/d) at which DLT observed.

The DLT observed in both strata was myelosuppression (grade IV neutropenia for >7 days, and any grade IV thrombocytopenia and/or delayed platelet recovery for >14 days between courses), which occurred in 4 of 16 (25%) patients in stratum I and 3 of 19 (19%) in stratum II (Table 1). The CRM-estimated MTD for stratum I was 46 mg/m2, which corresponds to a dose-finding MTD of 45 mg/m2. The CRM-estimated MTD for stratum II was 41 mg/m2/d but because DLT had already been observed in three of four patients previously treated at 45 mg/m2/d, the decision was made to recommend 30 mg/m2/d as the dose to be used in future phase II trials (Table 1).

In all patients, the median duration to nadir of the absolute neutrophil and platelet counts was 33 days (range, 10-46) and 31 days (range, 17-45) from starting therapy, respectively. The corresponding recovery times for recovery to baseline levels for both counts were 37 days (range, 14-53) and 37 days (range, 24-83), respectively.

Significant toxicities (more than or equal to grade III) encountered in this study during and after the DLT observation period and thought to be related to VNP4010M are reported in Table 2. One patient died as a consequence of neutropenic sepsis caused by a central line infection (at dose level 45 mg/m2/d after the second course of treatment). In addition, two patients presented with renal failure and a clinical and laboratory picture consistent with hemolytic uremic syndrome. One of these patients with recurrent medulloblastoma [prior therapy included craniospinal irradiation, cisplatin, cyclophosphamide, and vincristine at diagnosis and high-dose chemotherapy (carboplatin, cyclophosphamide, and etoposide) at relapse] was initially treated at dose level 1 (45 mg/m2/d), suffered DLT, and was dose reduced to dose level 1 (30 mg/m2/d). She presented with mild renal failure (grade II serum creatinine) and grade IV thrombocytopenia 4 weeks after completing the fourth course of VNP40101M. She was taken off treatment after this complication and sent home for hospice care. Another patient with a recurrent brain stem glioma (prior therapy included focal irradiation and concurrent temozolomide) developed grade IV renal failure (at dose level 78 mg/m2/d after the fourth course of treatment) that stabilized with hemodialysis. She was taken off treatment and subsequently died of progressive disease. An additional patient with brain stem glioma (prior therapy included focal irradiation only) developed grade III pulmonary toxicity (pulmonary edema with pleural effusion on day 3 of the fifth course at 45 mg/m2/d) that resolved after stopping protocol therapy. The renal and pulmonary toxicities were reported to the Food and Drug Administration as being related to VNP40101M.

Table 2.

Highest grade of toxicity related to VNP40101M by stratum during all courses

Type of toxicityToxicity gradeDose (mg/m2/d)
Stratum I
Stratum II
45607820304560
Fatigue II 
Gastrointestinal* II or III 
Anemia II or III 
Neutropenia III, IV, or V 7 
Thrombocytopenia III or IV 
Pulmonary Effusion/infiltrate III 
Renal failure II or IV 
Type of toxicityToxicity gradeDose (mg/m2/d)
Stratum I
Stratum II
45607820304560
Fatigue II 
Gastrointestinal* II or III 
Anemia II or III 
Neutropenia III, IV, or V 7 
Thrombocytopenia III or IV 
Pulmonary Effusion/infiltrate III 
Renal failure II or IV 
*

Nausea and vomiting, or diarrhea.

One patient died of neutropenic sepsis.

Response to treatment. Disease evaluation in 17 evaluable patients in stratum I included stable disease in 1, and progressive disease in 15. One additional patient with recurrent diffuse brain stem glioma who received treatment at dose level 1 (45 mg/m2/d) had stable tumor size after two courses (66 days after enrollment) that was not sustained as she subsequently died of neutropenic sepsis. Stable disease lasted for 61+ weeks in a patient with recurrent atypical central neurocytoma. Of 21 evaluable patients in stratum II, stable disease occurred in 2 and progressive disease in 18. One additional patient with medulloblastoma began treatment at 45 mg/m2/d and was noted to have significant tumor shrinkage 155 days after enrollment. She subsequently came off therapy due to mild renal failure and prolonged thrombocytopenia and did not have further neuroimaging to confirm sustained response. The two patients with stable disease had recurrent optic glioma and anaplastic astrocytoma; stable disease lasted 47 and 37 weeks, respectively.

Pharmacokinetics. VNP40101M and VNP4090CE concentration-time data from 23 patients during course 1 of therapy were available for pharmacokinetic modeling. Pharmacokinetics could not be done on the remaining 18 patients; 17 declined participation, and one had poor peripheral venous access. Visual inspection of the data indicated distinct distribution and elimination phases of VNP40101M and VNP4090CE disposition, and the data were best described by a multicompartmental model. A representative VNP4010M and VNP-4909CE plasma concentration-time plot is shown in Fig. 1.

Fig. 1.

Representative VNP40101M and VNP4090CE plasma concentration-time curves after a 30-min infusion of VNP40101M at a dosage of 45 mg/m2. Closed circles, observed VNP40101M concentrations; open circles, observed 40090CE concentrations. Solid line, best-fit curve from the Bayesian model for VNP40101M; dashed lines, the best-fit curve from the Bayesian model for VNP4090CE.

Fig. 1.

Representative VNP40101M and VNP4090CE plasma concentration-time curves after a 30-min infusion of VNP40101M at a dosage of 45 mg/m2. Closed circles, observed VNP40101M concentrations; open circles, observed 40090CE concentrations. Solid line, best-fit curve from the Bayesian model for VNP40101M; dashed lines, the best-fit curve from the Bayesian model for VNP4090CE.

Close modal

Of the 23 patients, we chose 8 who had adequate sampling (as defined in the Materials and Methods) and used the maximum likelihood results from these patients as the prior density for the subsequent MAP analysis. The VNP40101M and VNP4090CE concentration-time data for all patients were then modeled by using a multicompartment MAP Bayesian approach. The maximum likelihood and MAP variable estimates were highly concordant in the eight patients whose data were used as the prior density. Because the MAP-Bayesian method is more robust than maximum likelihood estimation for modeling sparse data, the MAP-Bayesian analysis offered the most reliable variable estimates for this group. The MAP-Bayesian–estimated pharmacokinetic variables at different dosage levels are summarized in Table 3. The median VNP40101M clearance was 0.17 L/min/m2 (range, 0.07-0.63 L/min/m2). VNP40101M AUC values increased with increasing dosage as shown in Fig. 2; however, no differences were observed in VNP401090CE AUC values at different VNP40101M dose levels (Fig. 2). The median elimination half-life (T1/2β) was 30 min (range, 14-39.5).

Table 3.

Comparison of VNP40101M PK variables between current study and adults with solid tumors

Current Study
Adults with solid tumors*
Dose (mg/m2/d)Cmax (mcg/mL)T1/2β (min)AUC0→∞ (mg/L × min)Vc (L/m2)CL (L/min/m2)Dose (mg/m2/d)Cmax (mcg/mL)T1/2β (min)AUC0→∞ (mg/L × min)Vc (L/m2)CL (L/min/m2)
20 (n = 1) 2.08 29.9 166.8 5.0 0.12 24 0.65 19 25.02  1.05 
30 (n = 9) 2.6 (1.4-4.14) 33.6 (16.0-54.7) 200.6 (130.5-443.9) 5.9 (3.2-12.2) 0.15 (0.07-0.23)       
45 (n = 8) 2.96 (1.41-3.96) 24.7 (14.0-38.2) 234.1 (70.4-338.4) 7.4 (5.8-12.2) 0.17 (0.07-0.63) 40 0.98 25 48.8  0.83 
60 (n = 4) 3.83 (3.64-3.99) 31.8 (16.2-40.7) 284.5 (191.6-353.9) 8.0 (5.7-10.2) 0.18 (0.17-0.32) 60 2.49 12 67.4  0.93 
78 (n = 1) 8.08 39.5 547.5 7.7 0.13 80 3.09 18 102.5  0.79 
Current Study
Adults with solid tumors*
Dose (mg/m2/d)Cmax (mcg/mL)T1/2β (min)AUC0→∞ (mg/L × min)Vc (L/m2)CL (L/min/m2)Dose (mg/m2/d)Cmax (mcg/mL)T1/2β (min)AUC0→∞ (mg/L × min)Vc (L/m2)CL (L/min/m2)
20 (n = 1) 2.08 29.9 166.8 5.0 0.12 24 0.65 19 25.02  1.05 
30 (n = 9) 2.6 (1.4-4.14) 33.6 (16.0-54.7) 200.6 (130.5-443.9) 5.9 (3.2-12.2) 0.15 (0.07-0.23)       
45 (n = 8) 2.96 (1.41-3.96) 24.7 (14.0-38.2) 234.1 (70.4-338.4) 7.4 (5.8-12.2) 0.17 (0.07-0.63) 40 0.98 25 48.8  0.83 
60 (n = 4) 3.83 (3.64-3.99) 31.8 (16.2-40.7) 284.5 (191.6-353.9) 8.0 (5.7-10.2) 0.18 (0.17-0.32) 60 2.49 12 67.4  0.93 
78 (n = 1) 8.08 39.5 547.5 7.7 0.13 80 3.09 18 102.5  0.79 

Abbreviations: Cmax, peak plasma concentration; T1/2β, terminal half life; AUC0→∞, area under the concentration time curve; Vc, volume of central compartment; CL, clearance.

*

Adapted from reference #5.

Fig. 2.

VNP40101M and VNP4090CE AUC0→∞values at different VNP40101M dosage levels. Closed circles, calculated VNP40101M AUC values for individual patients; thin horizontal lines, median AUC at dosage level. Open triangles, calculated AUC VNP4090CE values for individual patients; thick horizontal lines, median AUC at each dosage level.

Fig. 2.

VNP40101M and VNP4090CE AUC0→∞values at different VNP40101M dosage levels. Closed circles, calculated VNP40101M AUC values for individual patients; thin horizontal lines, median AUC at dosage level. Open triangles, calculated AUC VNP4090CE values for individual patients; thick horizontal lines, median AUC at each dosage level.

Close modal

AGT levels in PBMC. Nine patients had blood collected for measurement of pretreatment and posttreatment AGT levels in PBMC (Table 4). Six patients had a median decrease in AGT levels of 30% (range, 8-50%) compared with baseline levels that was not statistically significant. However, there was no significant correlation between dose level, Cmax, or AUC and decrease in AGT levels after treatment (data not shown).

Table 4.

PBMC AGT levels before and after treatment with VNP40101M

Dose LevelPBMC AGT level (fmol/mg protein)
Pair-wise difference (%)
Before first dose4 h after fifth dose
60 1,785 979 −48% 
60 1,395 877 −37% 
60 255 126 −50% 
45 1,815 1,399 −23% 
45 1,948 1,739 −11% 
45 1,697 1,564 −8% 
45 1,329 2,126 15% 
78 876 926 6% 
20 860 1,366 59% 
Dose LevelPBMC AGT level (fmol/mg protein)
Pair-wise difference (%)
Before first dose4 h after fifth dose
60 1,785 979 −48% 
60 1,395 877 −37% 
60 255 126 −50% 
45 1,815 1,399 −23% 
45 1,948 1,739 −11% 
45 1,697 1,564 −8% 
45 1,329 2,126 15% 
78 876 926 6% 
20 860 1,366 59% 

The sulfonyl hydrazines prodrugs, including VNP40101M, are a new class of bifunctional DNA alkylating agents that spontaneously generate nucleophilic species with selective chlorethylating and carbamoylating activities. VNP40101M spontaneously decomposes into VNP4090CE (chlorethylating species) and methylisocyanate (carbamoylating species; ref. 2). Penketh et al. (16) have shown that VNP1010M produces more stable cross-linking of DNA in vitro compared with 1,3-bis(2-chloroethyl)-1-nitrosourea by causing alkylation predominantly in the O6 position of guanine and less N7 alkylation. The latter reaction causes depurination and strand nicking and has less therapeutic value (16). The drug has also shown broad antitumor activity in murine human tumor xenograft models including L1210 leukemia, B16F10 melanoma, M109 lung carcinoma, C26 colon carcinoma, and U251 human glioma (1). The fractionated schedule of drug administration >6 days was superior to a single dose of drug in producing long-term cures (1). The drug seemed to be effective in animals bearing 1,3-bis(2-chloroethyl)-1-nitrosourea or cyclophosphamide resistant L1210 leukemia and also when the leukemia cells were inoculated in the brain, suggesting that the drug had good penetration of the blood-brain barrier (1). In addition, preclinical studies conducted at Duke University Medical Center using a fractionated schedule of VNP40101M in athymic mice bearing subcutaneous tumors of D425 medulloblastoma, D612 ependymoma, and D245 and 456 (1,3-bis(2-chloroethyl)-1-nitrosourea resistant) malignant glioma cell lines showed excellent tumor regressions with minimal toxicity (10). VNP40101M has been extensively tested in adults patients with newly diagnosed and recurrent acute myelogenous leukemia, both as a single agent and in combination with high dose Ara-C with objective responses observed in both phase I and II studies (4, 6, 9). The agent has also been tested in adults with recurrent solid tumors in a phase I setting and, more recently, in a phase II trial in adults with recurrent glioblastoma multiforme (3, 8). The phase I study described in this report is the first time the drug has been used in children.

Our clinical trial confirms that the main DLT associated with VNP40101M in children with recurrent brain tumors is myelosuppression, as has been observed in adults with recurrent solid tumors and acute myeloid leukemia(6, 8). Bone marrow suppression was substantial both in its intensity and duration, resulting in delays in restarting the next course. As expected, the proportion of patients with severe myelosuppression was directly proportional to intensity of prior therapy and administered dose. Thus, patients in stratum II had a higher incidence of myelosuppression for a given dose level (grade IV neutropenia/thrombocytopenia in 1 of 12 patients in stratum I versus 3 of 4 patients in stratum II at the starting dose level of 45 mg/m2/d; Table 1). Similarly, a greater proportion of patients treated at the higher dose levels of 60 and 78 mg/m2/d experienced DLT (Table 1). It is likely that such hematologic toxicity from this drug might limit its use in heavily pretreated patients, especially those who have had craniospinal irradiation or high-dose chemotherapy. Also, due to its potential for causing prolonged myelosuppression, future clinical trials exploring combination therapy with VNP40101M should be designed to either avoid or cautiously combine this drug with agents that are myelosuppressive.

In addition to hematologic toxicity, two patients had hemolytic uremic syndrome and renal failure after four cycles of VNP40101M. The mechanism of renal failure due to VNP40101M is unclear but is likely to be similar to what has been observed with other alkylating agents (17, 18). This complication has not been observed in adult patients treated with this agent thus far (35, 79). One patient with no history of prior pulmonary disease had grade III pulmonary toxicity (pleural effusion and pulmonary edema) that occurred during treatment with VNP40101M. In preclinical toxicity studies, 2 of 10 rats given 18 to 30 mg/m2daily for 5 days developed pleural effusion and pulmonary consolidation ∼4 weeks after therapy (19). Microscopic findings in the lung included alveolar edema, congestion, histiocytic infiltration, and vascular thrombi. Higher doses produced an increased incidence of pulmonary pathology. Similar to renal dysfunction, pulmonary toxicity has only rarely been observed in adult trials of this agent thus far (39).

Pharmacokinetic studies of VNP40101M in this clinical trial showed that the peak plasma concentrations (Cmax) of VNP40101M were linear with dose. As expected, the Cmax and AUC of VNP4090CE was only a fraction of that of the parent compound, consistent with its high reactivity and short half-life in vitro (20). When compared with the published pharmacokinetic variables in adults treated for recurrent solid tumors using the single-dose schedule (8), there were considerable differences in the drug disposition between the two populations for comparable dose levels (Table 3). In particular, children on this study had higher peak plasma concentrations and AUC0→8, longer terminal half-life, and a slower clearance. It is possible that the discordant data could be attributed to differences in techniques for processing the blood sample after collection, pharmacokinetic modeling (one versus four compartmental models), and possibly age-related differences in drug metabolism. Although the role of cytochrome P-450 enzyme system inVNP40101M metabolism is presently unknown (8), it is unlikely that the use of enzyme inducing anticonvulsants (that induce the cytochrome P-450 enzyme system) had any influence on the disposition of the drug in this group of children with brain tumors because the estimated AUC and T1/2 in this study is clearly well above those reported in adult patients who were not receiving any enzyme-inducing anticonvulsant drugs (8).

VNP40101M is known to have a better carbamoylating activity than nitrosoureas (2). Methylisocyanate derived from spontaneous decomposition of VNP40101M mediates the process of carbamylation of sulfhydryl and amine groups of cellular proteins and has been shown to effectively inhibit free AGT in vitro (2). Additional preclinical studies have also confirmed the efficient carbamoylating activity of VNP40101M based on its efficacy in AGT-overexpressing cell lines compared with 90CE that does not possess carbamoylating activity (1, 21). Assessment of AGT depletion in tumor tissue would require brain tumor biopsies before and after therapy and would not be feasible in a clinical setting, particularly in the context of a pediatric phase I study. Therefore, we decided to include evaluation of AGT depletion in PBMCs before and after five doses of VNP40101M as previously described in patients receiving oral temozolomide (22). Although six of nine patients did show varying levels of AGT depletion after therapy, the small sample size precluded meaningful correlation with administered dose or AUC. Also, the three patients who showed stable disease to VNP40101M either did not consent to this portion of the study (two patients) or had samples that could not be assayed properly due to technical difficulties (one patient). Hence, the possible relationship between baseline PBMC AGT levels, or depletion of AGT VNP40101M and the responses observed in these patients remains unknown.

In this heterogeneous group of patients with recurrent brain tumors who had previously failed alkylator therapy (including temozolomide, cyclophosphamide, cisplatin, carboplatin, or nitrosoureas) before VNP40101M exposure, three patients with recurrent optic glioma, atypical neurocytoma, and anaplastic astrocytoma had sustained stable disease that persisted for a median of 47 weeks (range, 37-61+). Two additional patients (one each with medulloblastoma and brain stem glioma) had transient objective responses. It is possible that the other patients did not respond to this treatment due to AGT overexpression in their tumors, as this is the only characterized mechanism of resistance to VNP40101M. Ongoing preclinical and clinical studies will show whether combining VNP40101M with other agents such as temozolomide or O6-benzylguanine might improve tumor AGT depletion and, hence, antitumor activity. It is possible that VNP40101M could be tested in phase II studies of children with brain tumors either as single agent or in combination with alkylators (e.g., temozolomide). However, these trials have to be carefully designed to exclude heavily pretreated patients and those whose tumors overexpress AGT.

Grant support: NIH grant U01 CA81457 for the Pediatric Brain Tumor Consortium, National Center for Research Resources grant M01 RR00188, and the American Lebanese Syrian Associated Charities.

The costs of publication of this article were defrayed in part by the payment of page charges. This article must therefore be hereby marked advertisement in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.

We thank Joyson Pekkattil for the support in protocol development and management.

1
Finch RA, Shyam K, Penketh PG, Sartorelli AC. 1,2-Bis(methylsulfonyl)-1-(2-chloroethyl)-2-(methylamino)carbonylhydrazine (101M): a novel sulfonylhydrazine prodrug with broad-spectrum antineoplastic activity.
Cancer Res
2001
;
61
:
3033
–8.
2
Penketh PG, Shyam K, Baumann RP, Remack JS, Brent TP, Sartorelli AC. 1,2-Bis(methylsulfonyl)-1-(2-chloroethyl)-2-[(methylamino)carbonyl]hydrazine (VNP40101M): I. Direct inhibition of O6-alkylguanine-DNA alkyltransferase (AGT) by electrophilic species generated by decomposition.
Cancer Chemother Pharmacol
2004
;
53
:
279
–87.
3
Badruddoja MA, Penne K, Desjardins A, et al. Phase II study of Cloretazine for the treatment of adults with recurrent glioblastoma multiforme.
Neuro-oncol
2007
;
9
:
70
–4.
4
Giles F, Rizzieri D, Karp J, et al. Cloretazine (VNP40101M), a novel sulfonylhydrazine alkylating agent, in patients age 60 years or older with previously untreated acute myeloid leukemia.
J Clin Oncol
2007
;
25
:
25
–31.
5
Giles F, Verstovsek S, Faderl S, et al. A phase II study of cloretazine (VNP40101M), a novel sulfonylhydrazine alkylating agent, in patients with very high risk relapsed acute myeloid leukemia.
Leuk Res
2006
;
30
:
1591
–5.
6
Giles F, Verstovsek S, Thomas D, et al. Phase I study of cloretazine (VNP40101M), a novel sulfonylhydrazine alkylating agent, combined with cytarabine in patients with refractory leukemia.
Clin Cancer Res
2005
;
11
:
7817
–24.
7
Giles F, Vey N, O'Brien S, et al. Phase II study of VNP40101M(VNP40101M) in patients (pts) with acute myelogenous leukemia (AML) and high risk myelodysplasia (MDS; abstr# 6541). Proc Am Soc Clin Oncol 2005;23:6541.
8
Murren J, Modiano M, Kummar S, et al. A phase I and pharmacokinetic study of VNP40101M, a new alkylating agent, in patients with advanced or metastatic cancer.
Invest New Drugs
2005
;
23
:
123
–35.
9
Giles F, Thomas D, Garcia-Manero G, et al. A phase I and pharmacokinetic study of VNP40101M, a novel sulfonylhydrazine alkylating agent, in patients with refractory leukemia.
Clin Cancer Res
2004
;
10
:
2908
–17.
10
Badruddoja MA, Keir ST, King I, et al. Activity of VNP40101M (Cloretazine) in the treatment of CNS tumor xenografts in athymic mice.
Neuro-oncol
2007
;
9
:
240
–4.
11
Piantadosi S, Fisher JD, Grossman S. Practical implementation of a modified continual reassessment method for dose-finding trials.
Cancer Chemother Pharmacol
1998
;
41
:
429
–36.
12
Krishna G, Hodnick WF, Lang W, et al. Pharmaceutical development and manufacturing of a parenteral formulation of a novel antitumor agent, VNP40101M.
AAPS PharmSciTech
2001
;
2
:
E14
.
13
Macdonald DR, Cascino TL, Schold SC, Jr., Cairncross JG. Response criteria for phase II studies of supratentorial malignant glioma.
J Clin Oncol
1990
;
8
:
1277
–80.
14
Bai F, Minkin P, Fraga CH, O'Shaughnessy MA, Gururangan S, Stewart CF. Determination of Cloretazinetrade mark (VNP40101M) and its active metabolite (VNP4090CE) in human plasma by liquid chromatography electrospray tandem mass spectrometry (LC-ESI-MS/MS).
J Chromatogr B Analyt Technol Biomed Life Sci
2007
;
853
:
97
–104.
15
D'Argenio DZ, Schumitzky A. ADAPT II User's Guide. 1st ed. USC, Los Angeles; 1990.
16
Penketh PG, Shyam K, Sartorelli AC. Comparison of DNA lesions produced by tumor-inhibitory 1,2-bis(sulfonyl)hydrazines and chloroethylnitrosoureas.
Biochem Pharmacol
2000
;
59
:
283
–91.
17
Schacht RG, Feiner HD, Gallo GR, Lieberman A, Baldwin DS. Nephrotoxicity of nitrosoureas.
Cancer
1981
;
48
:
1328
–34.
18
Zakarija A, Bennett C. Drug-induced thrombotic microangiopathy.
Semin Thromb Hemost
2005
;
31
:
681
–90.
19
Lee KC, Almassian B, Noveroske J. Toxicological evaluation of 1,2-bis(methylsulfonyl)-1-(2-chloroethyl)-2-(methylaminocarbonyl)hydrazine (VNP40101M), a novel alkylating agent with potential antitumor activity, with intravenous administration in rats and dogs.
Int J Toxicol
2002
;
21
:
23
–38.
20
Ishiguro K, Shyam K, Penketh PG, Sartorelli AC. Role of O6-alkylguanine-DNA alkyltransferase in the cytotoxic activity of cloretazine.
Mol Cancer Ther
2005
;
4
:
1755
–63.
21
Ishiguro KK, Seow HAHA, Penketh PGPG, Shyam KK, Sartorelli ACAC. Mode of action of the chloroethylating and carbamoylating moieties of the prodrug cloretazine.
Mol Cancer Ther
2006
;
5
:
969
–76.
22
Spiro TP, Liu L, Majka S, Haaga J, Willson JK, Gerson SL. Temozolomide: the effect of once- and twice-a-day dosing on tumor tissue levels of the DNA repair protein O(6)-alkylguanine-DNA-alkyltransferase.
Clin Cancer Res
2001
;
7
:
2309
–17.