Phase I Clinical Trial of Mafosfamide in Infants and Children Aged 3 Years or Younger With Newly Diagnosed Embryonal Tumors: A Pediatric Bra
http://www.100md.com
《临床肿瘤学》
the Texas Children's Cancer Center/Baylor College of Medicine, Houston, TX
Operations and Biostatistical Center, Pediatric Brain Tumor Consortium
St Jude Children's Research Hospital, Memphis, TN
Duke University Medical Center, Durham, NC
Children's Hospital and Medical Center, Seattle, WA
Dana-Farber Cancer Institute, Boston, MA
Children's National Medical Center, Washington, DC
Children's Hospital of Philadelphia, Philadelphia
Children's Hospital of Pittsburgh, PA
University of California, San Francisco, CA. R.H. is currently at the Department of Pediatrics, University of New Mexico, Albuquerque, NM
ABSTRACT
PATIENTS AND METHODS: Twenty-five assessable patients received IT mafosfamide at one of six dose levels ranging from 5 mg to 17 mg. Patients were premedicated with dexamethasone (0.15 mg/kg) and morphine (0.1 mg/kg) before receiving IT mafosfamide. Serial samples of CSF for pharmacokinetic studies were obtained in a subset of patients with Ommaya reservoirs.
RESULTS: Irritability, presumably secondary to pain or headache during mafosfamide administration, was dose limiting in two of three patients at the 17-mg dose level. The maximum-tolerated dose of IT mafosfamide following premedication with dexamethasone and morphine was 14 mg.
CONCLUSION: The maximum tolerated dose and recommended phase II dose of IT mafosfamide in patients younger than 3 years with newly diagnosed embryonal CNS tumors is 14 mg. A trial to assess the efficacy of regional therapy with IT mafosfamide administered with intensive systemic chemotherapy in children younger than 3 years with primary intracranial embryonal tumors is now in progress.
INTRODUCTION
The propensity for widespread neuraxis dissemination in children with embyronal CNS tumors has mandated the use of craniospinal irradiation (CSI) as a component of therapy in older patients. While CSI is an effective part of therapy in older children, it is associated with an unacceptably high incidence of neuropsychological sequelae in younger patients. Because of this, recent treatment strategies for infants and young children have primarily relied on the use of postoperative combination chemotherapy to delay or eliminate the need for radiation therapy (XRT).2-4 Although numerous studies have demonstrated the feasibility of postsurgical chemotherapy, meaningful delay of XRT has been achieved in only a minority of patients.2-4
In response to these treatment issues, the Pediatric Brain Tumor Consortium (PBTC) developed a new therapeutic strategy that utilizes intensive systemic chemotherapy along with intrathecal (IT) chemotherapy for neuraxis prophylaxis and/or treatment in infants and young children with embryonal CNS tumors. Past attempts to incorporate IT therapy, with antimetabolites such as methotrexate or cytarabine, into front-line systemic chemotherapy have not met with success.5,6 We hypothesized that IT therapy with an alkylating agent, similar to those that show substantial activity against embryonal tumors when given systemically, was more likely to be successful.
Cyclophosphamide, a widely used alkylating agent, has demonstrated substantial antitumor activity against embryonal CNS tumors and is a mainstay in virtually all current treatment regimens for these tumors. However, cyclophosphamide is a prodrug, requiring oxidation by hepatic microsomal enzymes to produce the active species, 4-hydroxycyclophosphamide. Thus, cyclophosphamide has no utility if administered intrathecally. However, mafosfamide, a preactivated chemically stable thioethane sulfonic acid derivative of cyclophosphamide, does not require hepatic activation and undergoes spontaneous conversion to the active species in aqueous media (Fig 1). In vitro and in vivo studies have demonstrated that mafosfamide exhibits cytotoxic activity comparable to or exceeding that of activated cyclophosphamide (4-hydroperoxy-cyclophosphamide).7 Thus, IT mafosfamide has offered an opportunity to evaluate the feasibility and utility of neuraxis prophylaxis without the use of CSI.
The starting dose and schedule for this phase I trial were based on the results of a phase I trial of intrathecal mafosfamide in children and adults with neoplastic meningitis, and other published experience with this agent.8-10 In the initial phase I mafosfamide trial, the drug was administered twice weekly for 4 to 6 weeks, and less frequently thereafter at doses ranging from 1.0 to 6.5 mg. CSF pharmacokinetic studies of intraventricular mafosfamide showed that the terminal half-life was approximately 1.8 hours, and that ventricular CSF exposures greatly exceeded those required to achieve a cytotoxic effect in vitro. Unfortunately, lumbar CSF mafosfamide concentrations were generally 1 log lower than simultaneous ventricular levels, and approached, but did not exceed, the 5-μmol/L target exposure required for in vitro cytotoxicity in breast and rhabdomyosarcoma cell lines. To achieve mafosfamide concentrations in excess of the target concentration throughout the neuraxis, a schedule that alternated the site of drug administration between the lumbar and ventricular sites was used. The feasibility of this approach was demonstrated both in the initial phase I trial and in a subsequent pilot study of 10 infants with newly diagnosed embryonal tumors.11 Based on these results, plus cumulative experience by investigators in Europe who routinely used intrathecal mafosfamide doses as high as 20 mg with concomitant analgesia, we felt that further mafosfamide dose escalation beyond 6.5 mg, but not above 20 mg, was warranted for this high-risk patient population. The need for further dose escalation in this population was also supported by results of subsequent studies in medulloblastoma cell lines, which suggested that a target CSF target exposure of 10 μmol/L was required for optimal cytotoxicity in CNS embryonal tumors (Blaney, unpublished data).
In this report, we summarize the initial phase I, dose-escalation component of a phase I-II trial designed to assess the efficacy of the combined systemic/IT chemotherapy approach in infants and young children with embryonal CNS tumors.
PATIENTS AND METHODS
An additional criterion for entry was the completion of a CSF flow study to evaluate for abnormal flow or obstruction of the subarachnoid space. In patients not having a ventriculoperitoneal (VP) or ventriculoatrial (VA) shunt, an Ommaya reservoir was required if their CSF flow study did not show evidence of obstruction or abnormal flow.
Informed consent was obtained from a parent or legal guardian of all patients in accordance with federal and local institutional review board policies before study entry.
Studies Before and During Treatment
A complete history; a physical examination, including a detailed neurological examination; and laboratory studies were obtained before treatment and periodically thereafter. Pretreatment laboratory evaluation included: CBCs and electrolyte, calcium, phosphorus, blood urea nitrogen, creatinine, and liver function tests. Pretreatment CSF studies from lumbar CSF, as well as the ventricular CSF in patients with Ommaya reservoirs, included cytology, cell count, differential, protein, and glucose. Preoperative and postoperative cranial magnetic resonance imaging (MRI) scans (with and without gadolinium) were obtained. The postoperative MRI scan was repeated if it had been obtained more than 3 weeks before the initiation of therapy. A baseline MRI scan of the spine was obtained before surgery or at least 10 days after diagnostic surgery.
As noted earlier, a postoperative radionuclide CSF flow study (111I-DTPA or 99Tc-DTPA) was required for all patients. Those with subarachnoid block were ineligible for IT mafosfamide, but could enter the trial and receive systemic chemotherapy. If a repeat flow study at 10 weeks showed resolution of the block and normal flow, mafosfamide was introduced during the second 10 weeks of systemic chemotherapy.
Complete physical and neurological evaluations were performed weekly for 6 weeks, during the first week of subsequent cycles of chemotherapy, and as clinically indicated. Other regularly scheduled evaluations included monitoring of CBCs, liver and renal function tests, serum chemistries, urinalysis, audiologic evaluations (brainstem auditory evoked response), and CSF studies. Imaging studies were performed during weeks 10 and 20 of preradiation and postradiation chemotherapy and approximately 4 weeks after the completion of radiotherapy. Other laboratory or imaging studies were obtained as clinically indicated.
Dosage and Drug Administration
Mafosfamide (4-sulfoethylthio-cyclophosphamide L-lysine) was supplied by Asta Medica (Frankfurt, Germany) and distributed to PBTC-001 investigators by Texas Children's Cancer Center (Houston, TX). The drug was supplied in 50-mg vials of freeze-dried compound.
The contents of the vial were completely dissolved in 5 mL of preservative-free saline. The final total dose was further diluted with preservative-free saline such that the final total volume of mafosfamide for IT administration was 5 mL. Mafosfamide was administered at a rate of 0.5 mL/min in an isovolumetric fashion (ie, an amount of CSF equivalent to the volume of drug to be administered was removed before injection). Patients were placed prone, in a flat or Trendelenburg position for 1 hour after intralumbar mafosfamide administration. If the drug was administered via Ommaya reservoir, the reservoir was flushed slowly for 1 to 2 minutes with 2 mL of CSF or normal saline after drug administration, and then pumped four to six times.
All patients were premedicated with oral or intravenous dexamethasone (0.15 mg/kg, 6 to 12 hours and 1.5 hours before IT mafosfamide) and morphine (0.1 mg/kg, intravenous, 5 minutes before IT mafosfamide). Patients who required general anesthesia for intralumbar drug administration did not receive morphine until after they had adequately recovered from their anesthetic. All patients received concomitant multiagent chemotherapy with vincristine, cyclophosphamide, cisplatin, and etoposide. The overall treatment schema for the first 20 weeks of therapy is shown in Figure 2.
Mafosfamide was administered twice weekly during weeks 1 to 6 (12 doses), weekly during weeks 7 to 9 (three doses), and then every 3 weeks during weeks 11 to 17 (three doses). Patients without VP or VA shunts and normal CSF flow had an Ommaya reservoir placed before the initiation of IT mafosfamide therapy. IT mafosfamide doses were alternately administered via the intralumbar and inraventricular routes in patients with Ommaya reservoirs. Patients with VP or VA shunts received all doses of IT mafosfamide via lumbar puncture.
Dose Escalation, Dose-Limiting Toxicity, and Maximum Tolerated Dose
A standard phase I design was used. Cohorts of three to six patients were treated at each dose level. Dose escalation was determined by toxicity observed during the first 2 weeks of therapy. Any patient who received even one dose of IT mafosfamide and who experienced DLT was considered in the dose-escalation phase of this trial. Toxicities observed during the last 18 weeks of IT therapy (before irradiation), together with the acute toxicity profiles observed during the initial 2 weeks of IT therapy, were used to derive the mafosfamide dose to be evaluated in the subsequent feasibility component of this trial. Thus, a total of six assessable patients who completed 20 weeks of protocol therapy were enrolled at the dose recommended for subsequent patients enrolled onto this study.
Toxicities associated with mafosfamide and the other systemic anticancer agents administered in this study were evaluated according to the Cancer Therapy Evaluation Program Common Toxicity Criteria version 2. Patients were removed from study if they experienced any grade 3 or greater significant neurotoxicity or other organ toxicity that was judged to be mafosfamide-related.
Neurological toxicity was considered dose limiting, and no further dose escalations were made if one of three to six patients in a cohort experienced a grade 4 mafosfamide-related neurological toxicity; if two of three to six patients experienced a grade 3 mafosfamide-related neurological toxicity; or if two of three to six patients experienced ≥ a grade 3 drug-related non-neurological toxicity. If only one of three patients experienced a grade 3 mafosfamide-related neurological toxicity or a ≥ grade 3 mafosfamide-related non-neurological toxicity, the cohort was expanded to six patients.
The following toxicities were not considered dose-limiting: grade 3 nausea and vomiting; grade 3 hepatic toxicity that resolved to at least grade 1 before the next treatment course; grade 3 fever or infection; or isolated elevated ALT or AST values that returned to baseline before the next treatment course. Grade 3 laboratory/metabolic abnormalities were to be discussed with the Study Chair, taking the patient's clinical condition into account to determine if they were considered to be dose limiting.
Pharmacokinetic Studies
Pharmacokinetic studies of mafosfamide were optional and were performed only in consenting patients with indwelling ventricular access devices. Serial samples of ventricular CSF were obtained on two separate occasions, once after an intraventricular dose, and once after an intralumbar dose. CSF samples (1 mL) were collected before drug administration and at 10 minutes, 2 hours, and 4 hours after mafosfamide administration. CSF was immediately placed in a polypropylene tube containing 4.25 mL of a derivatizing solution (2 mL acetonitrile, 1 mL methanol, 1 mL ammonium phosphate solution [1 mol/L; pH 7], and 0.25 mL methanolic solution of the internal standard d4-aldophosphamidepentafluorobenzyloxim [{2H4} PBOX]). The tube was vortexed for at least 30 seconds and then stored at room temperature for 24 hours to allow for complete derivatization. The tubes were subsequently stored at –20°C or lower until shipment on dry ice to Covance Laboratories (Harrogate, UK) where the samples assayed using a sensitive, validated HPLC/MS/MS assay for the active cytotoxic principle (mafosfamide, and the 4-hydroxycyclophosphamide and aldophosphamide byproducts, which are linked together by a chemical equilibrium) as an aldophosphamide-oxime. The lower limit of quantitation in CSF was 0.03 μmol/L.
Compartmental and noncompartmental analyses were not performed because of the limited number of samples obtained in each patient. Therefore, the results of the pharmacokinetic studies are descriptive.
RESULTS
Toxicity
Intrathecal mafosfamide was generally well tolerated. The dose-limiting toxicity was irritability, presumably secondary to pain, during and immediately after drug administration. This occurred in two of four patients at the 17-mg dose level, and in only one of the initial six patients at the 14-mg dose level.
Mild (grade 1 or 2) drug-related adverse events observed across all mafosfamide dose levels (Table 2) included irritability, and headache or pain during or immediately after drug administration. Occasional or rare adverse events included nausea, vomiting, flushing, arachnoiditis, stridor, anorexia, or fever. A transient mild to moderate diffuse erythematous discoloration involving the face, with or without associated discoloration of the body, was noted immediately after drug infusion in approximately one third of the patients. The erythematous discoloration after intralumbar drug was noted to start on the buttocks and then spread to the lower extremities. In some patients, the area of IT administration that was infiltrated with lidocaine was spared from the discoloration, suggesting that this may have been a dysautonomic reaction rather than a hypersensitivity reaction. The discoloration resolved spontaneously or with administration of diphenydramine.
Two patients at the 14-mg dose level were noted to have stridor after recovery from anesthesia used to facilitate intralumbar drug administration. Although the stridor was initially attributed to the anesthesic, the treating physician subsequently felt that it was related to the mafosfamide, as one of the two patients with stridor also had the discoloration noted above. One patient experienced a seizure of uncertain relationship to mafosfamide. Another patient was noted to experience hearing loss, which again is of uncertain relationship to mafosfamide since all patients enrolled onto this trial received systemic cisplatin.
There were no reported CSF infections in the patients with Ommaya reservoirs. Two patients with VP shunts, who therefore received intralumbar mafosfamide, were reported to have meningitis. In one of these patients the meningitis was attributed to an underlying diarrhea. This same patient also had spine and subdural fluid collection with enhancement with nerve root enhancement in the cauda equina that was attributed by the treating physician to a dural tear. Both patients had an uneventful recovery.
Pharmacokinetics
Evaluation of mafosfamide concentrations in ventricular CSF samples was performed in a subset of patients with indwelling Ommaya reservoirs. Ventricular CSF mafosfamide concentrations after intraventricular dosing were well above the target concentration in all patients at all dose levels. As shown in Figure 3, at the 14-mg dose level ventricular CSF concentrations of mafosfamide exceeded 10 μmol/L (the target concentration based on in vitro studies in medulloblastoma cell lines [Blaney, unpublished data]). Ventricular CSF mafosfamide concentrations after intralumbar dosing approached or exceeded the target mafosfamide level in all but one patient (Fig 4).
DISCUSSION
Alkylators, which possess a steep and linear cytotoxic effect throughout a wide dose range, are more likely to be effective in solid tumors. Until the availability of mafosfamide, thiotepa was the only alkylator available for intrathecal use. However, the utility of thiotepa is limited by its high liposolubility and associated rapid clearance from CSF, which result in poor drug distribution at sites distant from its administration. Mafosfamide, a water-soluble alkylator, has better neuraxis distribution (Figs. 3 and 4). Therefore, at the maximum tolerated dose of 14 mg, mafosfamide concentrations in both the ventricular and lumbar CSF approach or exceed those associated with in vitro cytotoxicity in many tumors.
In an early phase I trial of IT mafosfamide in children older than 3 years, IT mafosfamide dose escalation beyond 5 mg was not possible due to dose-limiting headache and nuchal pain. While this dose resulted in excellent ventricular CSF drug exposures (> the 10 μmol/L target) when given by Ommaya reservoir, lumbar levels only approached but did not meet or exceed the target.8 A later report by European investigators subsequently established that IT mafosfamide could be safely administered to older children at doses as high as 20 mg using concomitant analgesia.9 These experiences prompted a re-evaluation of the maximum tolerated dose of IT mafosfamide as outlined in the current report.
In this limited dosage escalation trial, we determined that the maximum tolerated dose of IT mafosfamide in association with systemic chemotherapy was 14 mg. The administration of concomitant analgesia and decadron before mafosfamide allowed us to markedly increase the dose above that of our prior phase I study. Limited pharmacokinetic data obtained in a subset of patients enrolled onto this trial demonstrate that this dose seems to produce adequate drug exposure throughout the neuraxis, and may thus be useful for the prophylaxis and/or treatment of leptomeningeal disease.
Based on these results, the PBTC is continuing accrual to the phase II component of PBTC-001 study using an IT mafosfamide dose of 14 mg in patients without impaired CSF flow. The primary goal of this continuing trial is to evaluate the overall feasibility and efficacy of administering IT mafosfamide, along with 20 weeks of intensive postoperative systemic chemotherapy. Thereafter, infants who remain or achieve M0 status receive limited-field conformal radiotherapy followed by an additional 20 weeks of systemic chemotherapy. The remainder will go on to receive more intensive systemic therapy with age-adjusted craniospinal irradiation. The results of this ongoing efficacy trial will be updated after completion of patient accrual and follow-up.
Authors' Disclosures of Potential Conflicts of Interest
NOTES
Supported in part by National Institutes of Health (Bethesda, MD) grant CA-98-007, and Asta Medica, Frankfurt, Germany.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Heideman R, Packer R, Albright L, et al: Tumors of the central nervous system. Philadelphia, PA, Lippincott-Raven, 1997
2. Shalet S, Beardwell C, Pearson D, et al: The effect of varying doses of cerebral irradiation on growth hormone production in childhood. Clin Endocrinol (Oxf) 5:287-290, 1976
3. Spunberg J, Chang C, Goldman M, et al: Quality of long-term survival following irradiation for intracranial tumors under the age of two. Int J Radiat Oncol Biol Phys 7:727-736, 1981
4. Mulhern R, Hancock J, Fairclough D, et al: Neuropsychological status of children treated for brain tumors: A critical review and integrative anaysis. Med Pediatr Oncol 20:181-191, 1992
5. van Eys J, Chen T, Moore T, et al: Adjuvant chemotherapy for medulloblastoma and ependymoma using iv vincristine, intrathecal methotrexate, and intrathecal hydrocortisone: A Southwest Oncology Group Study. Cancer Treat Rep 65:681-684, 1981
6. Edwards M, Levin V, Seager M, et al: Intrathecal chemotherapy for leptomeningeal dissemination of medulloblastoma. Childs Brain 8:444-451, 1981
7. Alberts D, Einsphar J, Struck R, et al: Comparative in vitro cytotoxicity of cyclophosphamide, its major active metabolites and the new oxazaphosphorine ASTA Z 7557 (INN mafosfamide). Invest New Drugs 2:141-148, 1984
8. Blaney SM, Balis FM, Arndt CA, et al: Intrathecal mafosfamide: Preclinical and clinical pharmacokinetics from a phase 1 trial. J Clin Oncol (in press)
9. Slavc I, Schuller E, Czech T, et al: Intrathecal mafosfamide therapy for pediatric brain tumors with meningeal dissemination. J Neurooncol 38:213-218, 1998
10. Slavc I, Schuller E, Falger J, et al: Feasibility of long-term intraventricular therapy with mafosfamide (n = 26) and etoposide (n = 11): Experience in 26 children with disseminated malignant brain tumors. J Neurooncol 64:239-247, 2003
11. Blaney S, Horowitz M, Kun L, et al: A new approach to the treatment of infants with CNS tumors. Neuro-Oncology 2:251, 2000 (abstr 24)(Susan M. Blaney, James Bo)
Operations and Biostatistical Center, Pediatric Brain Tumor Consortium
St Jude Children's Research Hospital, Memphis, TN
Duke University Medical Center, Durham, NC
Children's Hospital and Medical Center, Seattle, WA
Dana-Farber Cancer Institute, Boston, MA
Children's National Medical Center, Washington, DC
Children's Hospital of Philadelphia, Philadelphia
Children's Hospital of Pittsburgh, PA
University of California, San Francisco, CA. R.H. is currently at the Department of Pediatrics, University of New Mexico, Albuquerque, NM
ABSTRACT
PATIENTS AND METHODS: Twenty-five assessable patients received IT mafosfamide at one of six dose levels ranging from 5 mg to 17 mg. Patients were premedicated with dexamethasone (0.15 mg/kg) and morphine (0.1 mg/kg) before receiving IT mafosfamide. Serial samples of CSF for pharmacokinetic studies were obtained in a subset of patients with Ommaya reservoirs.
RESULTS: Irritability, presumably secondary to pain or headache during mafosfamide administration, was dose limiting in two of three patients at the 17-mg dose level. The maximum-tolerated dose of IT mafosfamide following premedication with dexamethasone and morphine was 14 mg.
CONCLUSION: The maximum tolerated dose and recommended phase II dose of IT mafosfamide in patients younger than 3 years with newly diagnosed embryonal CNS tumors is 14 mg. A trial to assess the efficacy of regional therapy with IT mafosfamide administered with intensive systemic chemotherapy in children younger than 3 years with primary intracranial embryonal tumors is now in progress.
INTRODUCTION
The propensity for widespread neuraxis dissemination in children with embyronal CNS tumors has mandated the use of craniospinal irradiation (CSI) as a component of therapy in older patients. While CSI is an effective part of therapy in older children, it is associated with an unacceptably high incidence of neuropsychological sequelae in younger patients. Because of this, recent treatment strategies for infants and young children have primarily relied on the use of postoperative combination chemotherapy to delay or eliminate the need for radiation therapy (XRT).2-4 Although numerous studies have demonstrated the feasibility of postsurgical chemotherapy, meaningful delay of XRT has been achieved in only a minority of patients.2-4
In response to these treatment issues, the Pediatric Brain Tumor Consortium (PBTC) developed a new therapeutic strategy that utilizes intensive systemic chemotherapy along with intrathecal (IT) chemotherapy for neuraxis prophylaxis and/or treatment in infants and young children with embryonal CNS tumors. Past attempts to incorporate IT therapy, with antimetabolites such as methotrexate or cytarabine, into front-line systemic chemotherapy have not met with success.5,6 We hypothesized that IT therapy with an alkylating agent, similar to those that show substantial activity against embryonal tumors when given systemically, was more likely to be successful.
Cyclophosphamide, a widely used alkylating agent, has demonstrated substantial antitumor activity against embryonal CNS tumors and is a mainstay in virtually all current treatment regimens for these tumors. However, cyclophosphamide is a prodrug, requiring oxidation by hepatic microsomal enzymes to produce the active species, 4-hydroxycyclophosphamide. Thus, cyclophosphamide has no utility if administered intrathecally. However, mafosfamide, a preactivated chemically stable thioethane sulfonic acid derivative of cyclophosphamide, does not require hepatic activation and undergoes spontaneous conversion to the active species in aqueous media (Fig 1). In vitro and in vivo studies have demonstrated that mafosfamide exhibits cytotoxic activity comparable to or exceeding that of activated cyclophosphamide (4-hydroperoxy-cyclophosphamide).7 Thus, IT mafosfamide has offered an opportunity to evaluate the feasibility and utility of neuraxis prophylaxis without the use of CSI.
The starting dose and schedule for this phase I trial were based on the results of a phase I trial of intrathecal mafosfamide in children and adults with neoplastic meningitis, and other published experience with this agent.8-10 In the initial phase I mafosfamide trial, the drug was administered twice weekly for 4 to 6 weeks, and less frequently thereafter at doses ranging from 1.0 to 6.5 mg. CSF pharmacokinetic studies of intraventricular mafosfamide showed that the terminal half-life was approximately 1.8 hours, and that ventricular CSF exposures greatly exceeded those required to achieve a cytotoxic effect in vitro. Unfortunately, lumbar CSF mafosfamide concentrations were generally 1 log lower than simultaneous ventricular levels, and approached, but did not exceed, the 5-μmol/L target exposure required for in vitro cytotoxicity in breast and rhabdomyosarcoma cell lines. To achieve mafosfamide concentrations in excess of the target concentration throughout the neuraxis, a schedule that alternated the site of drug administration between the lumbar and ventricular sites was used. The feasibility of this approach was demonstrated both in the initial phase I trial and in a subsequent pilot study of 10 infants with newly diagnosed embryonal tumors.11 Based on these results, plus cumulative experience by investigators in Europe who routinely used intrathecal mafosfamide doses as high as 20 mg with concomitant analgesia, we felt that further mafosfamide dose escalation beyond 6.5 mg, but not above 20 mg, was warranted for this high-risk patient population. The need for further dose escalation in this population was also supported by results of subsequent studies in medulloblastoma cell lines, which suggested that a target CSF target exposure of 10 μmol/L was required for optimal cytotoxicity in CNS embryonal tumors (Blaney, unpublished data).
In this report, we summarize the initial phase I, dose-escalation component of a phase I-II trial designed to assess the efficacy of the combined systemic/IT chemotherapy approach in infants and young children with embryonal CNS tumors.
PATIENTS AND METHODS
An additional criterion for entry was the completion of a CSF flow study to evaluate for abnormal flow or obstruction of the subarachnoid space. In patients not having a ventriculoperitoneal (VP) or ventriculoatrial (VA) shunt, an Ommaya reservoir was required if their CSF flow study did not show evidence of obstruction or abnormal flow.
Informed consent was obtained from a parent or legal guardian of all patients in accordance with federal and local institutional review board policies before study entry.
Studies Before and During Treatment
A complete history; a physical examination, including a detailed neurological examination; and laboratory studies were obtained before treatment and periodically thereafter. Pretreatment laboratory evaluation included: CBCs and electrolyte, calcium, phosphorus, blood urea nitrogen, creatinine, and liver function tests. Pretreatment CSF studies from lumbar CSF, as well as the ventricular CSF in patients with Ommaya reservoirs, included cytology, cell count, differential, protein, and glucose. Preoperative and postoperative cranial magnetic resonance imaging (MRI) scans (with and without gadolinium) were obtained. The postoperative MRI scan was repeated if it had been obtained more than 3 weeks before the initiation of therapy. A baseline MRI scan of the spine was obtained before surgery or at least 10 days after diagnostic surgery.
As noted earlier, a postoperative radionuclide CSF flow study (111I-DTPA or 99Tc-DTPA) was required for all patients. Those with subarachnoid block were ineligible for IT mafosfamide, but could enter the trial and receive systemic chemotherapy. If a repeat flow study at 10 weeks showed resolution of the block and normal flow, mafosfamide was introduced during the second 10 weeks of systemic chemotherapy.
Complete physical and neurological evaluations were performed weekly for 6 weeks, during the first week of subsequent cycles of chemotherapy, and as clinically indicated. Other regularly scheduled evaluations included monitoring of CBCs, liver and renal function tests, serum chemistries, urinalysis, audiologic evaluations (brainstem auditory evoked response), and CSF studies. Imaging studies were performed during weeks 10 and 20 of preradiation and postradiation chemotherapy and approximately 4 weeks after the completion of radiotherapy. Other laboratory or imaging studies were obtained as clinically indicated.
Dosage and Drug Administration
Mafosfamide (4-sulfoethylthio-cyclophosphamide L-lysine) was supplied by Asta Medica (Frankfurt, Germany) and distributed to PBTC-001 investigators by Texas Children's Cancer Center (Houston, TX). The drug was supplied in 50-mg vials of freeze-dried compound.
The contents of the vial were completely dissolved in 5 mL of preservative-free saline. The final total dose was further diluted with preservative-free saline such that the final total volume of mafosfamide for IT administration was 5 mL. Mafosfamide was administered at a rate of 0.5 mL/min in an isovolumetric fashion (ie, an amount of CSF equivalent to the volume of drug to be administered was removed before injection). Patients were placed prone, in a flat or Trendelenburg position for 1 hour after intralumbar mafosfamide administration. If the drug was administered via Ommaya reservoir, the reservoir was flushed slowly for 1 to 2 minutes with 2 mL of CSF or normal saline after drug administration, and then pumped four to six times.
All patients were premedicated with oral or intravenous dexamethasone (0.15 mg/kg, 6 to 12 hours and 1.5 hours before IT mafosfamide) and morphine (0.1 mg/kg, intravenous, 5 minutes before IT mafosfamide). Patients who required general anesthesia for intralumbar drug administration did not receive morphine until after they had adequately recovered from their anesthetic. All patients received concomitant multiagent chemotherapy with vincristine, cyclophosphamide, cisplatin, and etoposide. The overall treatment schema for the first 20 weeks of therapy is shown in Figure 2.
Mafosfamide was administered twice weekly during weeks 1 to 6 (12 doses), weekly during weeks 7 to 9 (three doses), and then every 3 weeks during weeks 11 to 17 (three doses). Patients without VP or VA shunts and normal CSF flow had an Ommaya reservoir placed before the initiation of IT mafosfamide therapy. IT mafosfamide doses were alternately administered via the intralumbar and inraventricular routes in patients with Ommaya reservoirs. Patients with VP or VA shunts received all doses of IT mafosfamide via lumbar puncture.
Dose Escalation, Dose-Limiting Toxicity, and Maximum Tolerated Dose
A standard phase I design was used. Cohorts of three to six patients were treated at each dose level. Dose escalation was determined by toxicity observed during the first 2 weeks of therapy. Any patient who received even one dose of IT mafosfamide and who experienced DLT was considered in the dose-escalation phase of this trial. Toxicities observed during the last 18 weeks of IT therapy (before irradiation), together with the acute toxicity profiles observed during the initial 2 weeks of IT therapy, were used to derive the mafosfamide dose to be evaluated in the subsequent feasibility component of this trial. Thus, a total of six assessable patients who completed 20 weeks of protocol therapy were enrolled at the dose recommended for subsequent patients enrolled onto this study.
Toxicities associated with mafosfamide and the other systemic anticancer agents administered in this study were evaluated according to the Cancer Therapy Evaluation Program Common Toxicity Criteria version 2. Patients were removed from study if they experienced any grade 3 or greater significant neurotoxicity or other organ toxicity that was judged to be mafosfamide-related.
Neurological toxicity was considered dose limiting, and no further dose escalations were made if one of three to six patients in a cohort experienced a grade 4 mafosfamide-related neurological toxicity; if two of three to six patients experienced a grade 3 mafosfamide-related neurological toxicity; or if two of three to six patients experienced ≥ a grade 3 drug-related non-neurological toxicity. If only one of three patients experienced a grade 3 mafosfamide-related neurological toxicity or a ≥ grade 3 mafosfamide-related non-neurological toxicity, the cohort was expanded to six patients.
The following toxicities were not considered dose-limiting: grade 3 nausea and vomiting; grade 3 hepatic toxicity that resolved to at least grade 1 before the next treatment course; grade 3 fever or infection; or isolated elevated ALT or AST values that returned to baseline before the next treatment course. Grade 3 laboratory/metabolic abnormalities were to be discussed with the Study Chair, taking the patient's clinical condition into account to determine if they were considered to be dose limiting.
Pharmacokinetic Studies
Pharmacokinetic studies of mafosfamide were optional and were performed only in consenting patients with indwelling ventricular access devices. Serial samples of ventricular CSF were obtained on two separate occasions, once after an intraventricular dose, and once after an intralumbar dose. CSF samples (1 mL) were collected before drug administration and at 10 minutes, 2 hours, and 4 hours after mafosfamide administration. CSF was immediately placed in a polypropylene tube containing 4.25 mL of a derivatizing solution (2 mL acetonitrile, 1 mL methanol, 1 mL ammonium phosphate solution [1 mol/L; pH 7], and 0.25 mL methanolic solution of the internal standard d4-aldophosphamidepentafluorobenzyloxim [{2H4} PBOX]). The tube was vortexed for at least 30 seconds and then stored at room temperature for 24 hours to allow for complete derivatization. The tubes were subsequently stored at –20°C or lower until shipment on dry ice to Covance Laboratories (Harrogate, UK) where the samples assayed using a sensitive, validated HPLC/MS/MS assay for the active cytotoxic principle (mafosfamide, and the 4-hydroxycyclophosphamide and aldophosphamide byproducts, which are linked together by a chemical equilibrium) as an aldophosphamide-oxime. The lower limit of quantitation in CSF was 0.03 μmol/L.
Compartmental and noncompartmental analyses were not performed because of the limited number of samples obtained in each patient. Therefore, the results of the pharmacokinetic studies are descriptive.
RESULTS
Toxicity
Intrathecal mafosfamide was generally well tolerated. The dose-limiting toxicity was irritability, presumably secondary to pain, during and immediately after drug administration. This occurred in two of four patients at the 17-mg dose level, and in only one of the initial six patients at the 14-mg dose level.
Mild (grade 1 or 2) drug-related adverse events observed across all mafosfamide dose levels (Table 2) included irritability, and headache or pain during or immediately after drug administration. Occasional or rare adverse events included nausea, vomiting, flushing, arachnoiditis, stridor, anorexia, or fever. A transient mild to moderate diffuse erythematous discoloration involving the face, with or without associated discoloration of the body, was noted immediately after drug infusion in approximately one third of the patients. The erythematous discoloration after intralumbar drug was noted to start on the buttocks and then spread to the lower extremities. In some patients, the area of IT administration that was infiltrated with lidocaine was spared from the discoloration, suggesting that this may have been a dysautonomic reaction rather than a hypersensitivity reaction. The discoloration resolved spontaneously or with administration of diphenydramine.
Two patients at the 14-mg dose level were noted to have stridor after recovery from anesthesia used to facilitate intralumbar drug administration. Although the stridor was initially attributed to the anesthesic, the treating physician subsequently felt that it was related to the mafosfamide, as one of the two patients with stridor also had the discoloration noted above. One patient experienced a seizure of uncertain relationship to mafosfamide. Another patient was noted to experience hearing loss, which again is of uncertain relationship to mafosfamide since all patients enrolled onto this trial received systemic cisplatin.
There were no reported CSF infections in the patients with Ommaya reservoirs. Two patients with VP shunts, who therefore received intralumbar mafosfamide, were reported to have meningitis. In one of these patients the meningitis was attributed to an underlying diarrhea. This same patient also had spine and subdural fluid collection with enhancement with nerve root enhancement in the cauda equina that was attributed by the treating physician to a dural tear. Both patients had an uneventful recovery.
Pharmacokinetics
Evaluation of mafosfamide concentrations in ventricular CSF samples was performed in a subset of patients with indwelling Ommaya reservoirs. Ventricular CSF mafosfamide concentrations after intraventricular dosing were well above the target concentration in all patients at all dose levels. As shown in Figure 3, at the 14-mg dose level ventricular CSF concentrations of mafosfamide exceeded 10 μmol/L (the target concentration based on in vitro studies in medulloblastoma cell lines [Blaney, unpublished data]). Ventricular CSF mafosfamide concentrations after intralumbar dosing approached or exceeded the target mafosfamide level in all but one patient (Fig 4).
DISCUSSION
Alkylators, which possess a steep and linear cytotoxic effect throughout a wide dose range, are more likely to be effective in solid tumors. Until the availability of mafosfamide, thiotepa was the only alkylator available for intrathecal use. However, the utility of thiotepa is limited by its high liposolubility and associated rapid clearance from CSF, which result in poor drug distribution at sites distant from its administration. Mafosfamide, a water-soluble alkylator, has better neuraxis distribution (Figs. 3 and 4). Therefore, at the maximum tolerated dose of 14 mg, mafosfamide concentrations in both the ventricular and lumbar CSF approach or exceed those associated with in vitro cytotoxicity in many tumors.
In an early phase I trial of IT mafosfamide in children older than 3 years, IT mafosfamide dose escalation beyond 5 mg was not possible due to dose-limiting headache and nuchal pain. While this dose resulted in excellent ventricular CSF drug exposures (> the 10 μmol/L target) when given by Ommaya reservoir, lumbar levels only approached but did not meet or exceed the target.8 A later report by European investigators subsequently established that IT mafosfamide could be safely administered to older children at doses as high as 20 mg using concomitant analgesia.9 These experiences prompted a re-evaluation of the maximum tolerated dose of IT mafosfamide as outlined in the current report.
In this limited dosage escalation trial, we determined that the maximum tolerated dose of IT mafosfamide in association with systemic chemotherapy was 14 mg. The administration of concomitant analgesia and decadron before mafosfamide allowed us to markedly increase the dose above that of our prior phase I study. Limited pharmacokinetic data obtained in a subset of patients enrolled onto this trial demonstrate that this dose seems to produce adequate drug exposure throughout the neuraxis, and may thus be useful for the prophylaxis and/or treatment of leptomeningeal disease.
Based on these results, the PBTC is continuing accrual to the phase II component of PBTC-001 study using an IT mafosfamide dose of 14 mg in patients without impaired CSF flow. The primary goal of this continuing trial is to evaluate the overall feasibility and efficacy of administering IT mafosfamide, along with 20 weeks of intensive postoperative systemic chemotherapy. Thereafter, infants who remain or achieve M0 status receive limited-field conformal radiotherapy followed by an additional 20 weeks of systemic chemotherapy. The remainder will go on to receive more intensive systemic therapy with age-adjusted craniospinal irradiation. The results of this ongoing efficacy trial will be updated after completion of patient accrual and follow-up.
Authors' Disclosures of Potential Conflicts of Interest
NOTES
Supported in part by National Institutes of Health (Bethesda, MD) grant CA-98-007, and Asta Medica, Frankfurt, Germany.
Authors' disclosures of potential conflicts of interest are found at the end of this article.
REFERENCES
1. Heideman R, Packer R, Albright L, et al: Tumors of the central nervous system. Philadelphia, PA, Lippincott-Raven, 1997
2. Shalet S, Beardwell C, Pearson D, et al: The effect of varying doses of cerebral irradiation on growth hormone production in childhood. Clin Endocrinol (Oxf) 5:287-290, 1976
3. Spunberg J, Chang C, Goldman M, et al: Quality of long-term survival following irradiation for intracranial tumors under the age of two. Int J Radiat Oncol Biol Phys 7:727-736, 1981
4. Mulhern R, Hancock J, Fairclough D, et al: Neuropsychological status of children treated for brain tumors: A critical review and integrative anaysis. Med Pediatr Oncol 20:181-191, 1992
5. van Eys J, Chen T, Moore T, et al: Adjuvant chemotherapy for medulloblastoma and ependymoma using iv vincristine, intrathecal methotrexate, and intrathecal hydrocortisone: A Southwest Oncology Group Study. Cancer Treat Rep 65:681-684, 1981
6. Edwards M, Levin V, Seager M, et al: Intrathecal chemotherapy for leptomeningeal dissemination of medulloblastoma. Childs Brain 8:444-451, 1981
7. Alberts D, Einsphar J, Struck R, et al: Comparative in vitro cytotoxicity of cyclophosphamide, its major active metabolites and the new oxazaphosphorine ASTA Z 7557 (INN mafosfamide). Invest New Drugs 2:141-148, 1984
8. Blaney SM, Balis FM, Arndt CA, et al: Intrathecal mafosfamide: Preclinical and clinical pharmacokinetics from a phase 1 trial. J Clin Oncol (in press)
9. Slavc I, Schuller E, Czech T, et al: Intrathecal mafosfamide therapy for pediatric brain tumors with meningeal dissemination. J Neurooncol 38:213-218, 1998
10. Slavc I, Schuller E, Falger J, et al: Feasibility of long-term intraventricular therapy with mafosfamide (n = 26) and etoposide (n = 11): Experience in 26 children with disseminated malignant brain tumors. J Neurooncol 64:239-247, 2003
11. Blaney S, Horowitz M, Kun L, et al: A new approach to the treatment of infants with CNS tumors. Neuro-Oncology 2:251, 2000 (abstr 24)(Susan M. Blaney, James Bo)