A Phase I study of concurrent RMP-7 and carboplatin with radiation therapy for children with newly diagnosed brainstem gliomas

Authors

  • Roger J. Packer M.D.,

    Corresponding author
    1. Division of Neurology, Children's National Medical Center, Washington, DC
    2. Department of Neurology, The George Washington University, Washington, DC
    3. Department of Pediatrics, The George Washington University, Washington, DC
    • Center for Neuroscience and Behavioral Medicine, Department of Neurology, Children's National Medical Center;, 111 Michigan Avenue, NW, Washington, DC 20010
    Search for more papers by this author
    • Fax: (202) 884-5226

  • Mark Krailo Ph.D.,

    1. University of Southern California, Keck School of Medicine, Los Angeles, California
    2. The Children's Oncology Group, Arcadia, California
    Search for more papers by this author
  • Minesh Mehta M.D.,

    1. Department of Radiation Oncology, The University of Wisconsin Medical Center, Madison, Wisconsin
    Search for more papers by this author
  • Katherine Warren M.D.,

    1. The National Cancer Institute, Bethesda, Maryland
    Search for more papers by this author
  • Jeffrey Allen M.D.,

    1. Department of Neurology and Pediatrics, New York University, New York, New York
    Search for more papers by this author
  • Regina Jakacki M.D.,

    1. Department of Pediatrics, Children's Hospital of Pittsburgh, Pittsburgh, Pennsylvania
    Search for more papers by this author
  • Judith G. Villablanca M.D.,

    1. Department of Pediatrics, Children's Hospital of Los Angeles, Los Angeles, California
    Search for more papers by this author
  • Akiko Chiba B.S.,

    1. Division of Hematology/Oncology, Children's National Medical Center, Washington, DC
    Search for more papers by this author
  • Gregory Reaman M.D.

    1. Department of Pediatrics, The George Washington University, Washington, DC
    2. The Children's Oncology Group, Arcadia, California
    3. Division of Hematology/Oncology, Children's National Medical Center, Washington, DC
    Search for more papers by this author

  • Presented in part at the Child Neurology Society, Ottawa, Canada, October 15, 2004.

Abstract

BACKGROUND

Ninety percent of children with diffuse, intrinsic brainstem tumors will die within 18 months of diagnosis. Radiotherapy is of transient benefit to these children, and a potential way to improve its efficacy is to add radiosensitizers. Carboplatin is antineoplastic and radiosensitizing; however, its delivery to the primary tumor site is problematic. RMP-7 is a bradykinin analog that causes selective permeability of the blood-brain-tumor interface. The objective of this Phase I study was to determine the toxicity and feasibility of delivering RMP-7 and carboplatin for 5 successive days during radiotherapy to children with newly diagnosed, diffuse, intrinsic brainstem gliomas.

METHODS

RMP-7 was given prior to the end of carboplatin infusion. Local radiotherapy, in dose fractions of 180 centigrays (cGy) per day (to a total dose of 5940 cGy), was given within 4 hours of completion of drug delivery. Duration of treatment was escalated in a stepwise, weekly fashion in cohorts of 3 patients, until there was treatment-limiting toxicity or until radiotherapy was completed. Thirteen patients were treated, and their median age was 7 years (age range, 3–12 yrs).

RESULTS

One child died early during treatment of progressive disease and was not assessable for toxicity. Treatment for 3 weeks, 4 weeks, and 5 weeks was tolerated well, with mild flushing, tachycardia, nausea, emesis, dizziness, and abdominal pain. One of 3 children treated at the full duration of therapy (33 doses over 7 weeks) developed dose-limiting hepatotoxicity and neutropenia. The estimated median survival was 328 days, and 1 patient remained free of disease progression for > 400 days after the initiation of treatment.

CONCLUSIONS

The results of this study confirmed the feasibility of giving RMP-7 and carboplatin daily during radiotherapy to children with brainstem tumors. Cancer 2005. © 2005 American Cancer Society.

Brainstem gliomas, comprising 10% of all childhood brain tumors, remain some of the most difficult tumors to treat successfully.1 Diffuse intrinsic tumors constitute most brainstem gliomas, commonly involving primarily the pons, often with infiltration into other regions of the brainstem and contiguous nonbrainstem sites. They characteristically present with multiple cranial nerve deficits, ataxia, and long tract dysfunction, thus, with the current means of neuroradiographic diagnosis, surgery usually is not needed for diagnosis. Most patients can be diagnosed reliably based on clinical and neuroradiographic findings.2 Radiation therapy remains the most effective treatment, resulting in at least transient disease control in most patients. However, 90% of children with diffuse intrinsic brainstem gliomas will develop recurrences after radiation therapy and will die of their disease. Alterations in dose and fractionations of radiation have not improved survival.3 To date, the addition of chemotherapy or other forms of therapy, before or after radiation, has not improved disease control.1

Radiation therapy is the only effective (albeit transient) treatment for most children with diffuse, intrinsic brainstem gliomas. One potential means to improve the efficacy of radiation is by coupling it with radiosensitizers.4 Carboplatin has antineoplastic activity against several brain tumors, including high-grade gliomas.5, 6 It also has radiosensitization properties.4, 7, 8 An important possible limitation to using carboplatin or any other forms of chemotherapy to potentiate radiation effects is adequate drug delivery to the primary tumor site.9, 10 The blood-brain barrier, which is a monolayer of specialized capillary endothelial cells, forms a relatively continuous barrier between the brain and circulating blood. Most diffuse, intrinsic brainstem tumors do not enhance with intravenous agents, such as gadolinium, and it is believed that they have a relatively intact blood-brain barrier.2 Bradykinin causes transient relaxation of the blood-brain barrier's tight junctions.11, 12 Infusion of bradykinin into cerebral circulation transiently increases permeability of cerebrovasculature.11–14 Of the 2 types of bradykinin receptors, B1 and B2, B2 receptors functionally appear to be expressed predominantly in neuronal and vascular tissue. RMP-7 (Cereport™) is a synthetic bradykinin analog and B2 receptor agonist.14 For this reason, RMP-7 seems to be a reasonable candidate for use with chemotherapy to deliver more therapy to the primary tumor and surrounding region. This approach also attempts to exploit the difference between normal brain and brain tumor blood vessels.15–19, 24 In preclinical studies, RMP-7 caused rapid, reversible, and relatively selected permeability at the blood-brain barrier:tumor interface. This was rapid, transient, and autoregulated, and tachyphylaxis occurred with continuous administration. In animal central nervous system tumor models, tumor levels of carboplatin were increased with the addition of RMP-7, regardless of intracarotid or intravenous administration.15, 16 There was an average 10-fold uptake increase in the tumor bed; and, in experiments, compared with hyperosmolar-induced permeability, there was a significant therapeutic advantage for RMP-7. RMP-7 did not increase cerebral blood flow or volume in the normal brain, but it increased permeability within 2 mL surrounding the tumor and did not affect drug delivery significantly to more distant brain tissue sites. For these reasons, the current study (COG-ADVL0012) was undertaken by the Children's Oncology Group to evaluate the feasibility and potential efficacy of carboplatin and RMP-7 given concurrently with radiotherapy for children with newly diagnosed, diffuse, intrinsic brainstem gliomas.

MATERIALS AND METHODS

Patient Eligibility

Eligible children were between age 3 years and age 21 years, inclusive, with newly diagnosed, diffuse, intrinsic brainstem gliomas. Patients who had diffuse, intrinsic tumors did not require tissue confirmation for entry but were eligible based on neuroradiographic features. Patients who had focal brainstem tumors, which were defined as tumors that occupied < 50% of a single brainstem structure (such as the medulla, pons, or midbrain), and tumors that were exophytic, which were defined as those with > 50% of the tumor lying outside the body of the brainstem, were eligible only if the tumor was a malignant glioma confirmed by biopsy.

Before entry, patients must have had a magnetic resonance imaging (MRI) scan of the brain and spine, with and without gadolinium. All patients must have had measurable disease at the time of study. Lumbar cerebrospinal fluid analysis was not required; however, in patients who underwent a spinal tap, cerebrospinal fluid cytology must have been negative for tumor cells for eligibility. Patients who had disseminated disease confirmed by MRI or cerebrospinal fluid cytology were ineligible.

Before entry, patients were required to have adequate bone marrow and liver functions (within 1.5 times the upper limit of normal). Patients also were required to have normal renal functions, defined as a creatinine level less than the upper limit of normal for age or a 24-hour creatinine clearance or glomerular filtration rate > 80 mL per minute per 1.73 m2. All female patients of child-bearing age were required to have a negative pregnancy test and had to agree to use medically acceptable contraception if they were active sexually.

Patients who had received any prior treatment other than surgery or corticosteroids were ineligible. Patients with neurofibromatosis type 1 also were ineligible. Patients could not have used the following medications in the 24 hours prior to RMP-7 administration: vasodilating compounds, angiotensin-converting enzyme inhibitors, calcium channel blockers, antihistamines, aspirin, or β blockers. All patients had to sign an Institutional Review Board-approved consent or assent forms (if appropriate) prior to entry, in accordance with the Declaration of Helsinki.

Treatment Overview

Patients were to begin therapy within 31 days of diagnosis (for those undergoing surgery, the diagnosis date was considered the date of surgery). Patients were to receive carboplatin and RMP-7 at the assigned duration, intravenously, on a Monday through Friday schedule, within 4 hours before treatment with radiation therapy. Daily doses of the drugs were identical at all durations of therapy.

The duration and total dose of carboplatin and RMP-7 were assigned at entry. The first dose of both was to be administered on the first day of radiation therapy. The initial dose level was 3 weeks (15 doses) of RMP-7 and carboplatin.

Dose Escalation

Patients were enrolled in cohorts of three. If none of the first three patients enrolled at a dose level experienced dose-limiting toxicity (DLT), then the dose or duration was escalated, as appropriate. If one of the first three patients experienced a DLT, then three more patients were enrolled at the dose level. If any of the three patients in the second cohort experienced DLT, then the dose level was considered not tolerated. The dose was reduced one level, and three patients were enrolled at that level provided only three had been evaluated at the previous level. If two or more of the first three patients enrolled at a dose level experienced DLT, then the dose level was considered not tolerated. The dose was reduced one level, and three patients were enrolled at that level provided only three patients had been evaluated at the previous level.

If no DLT was encountered, then duration of carboplatin administration (initially 3 weeks) was extended in subsequent cohorts, so they would receive carboplatin and RMP-7 during the first 4 weeks, 5 weeks, 6 weeks, and 7 weeks of radiation therapy, respectively. Later in the study, the trial was amended to collapse the Weeks 6 and 7 of radiation into 1 stratum, because Week 7 consisted of only 3 days of radiotherapy.

DLT

For this study, DLTs were defined as an absolute neutrophil count < 500 mm3 for 7 consecutive days; a platelet count ≤ 20,000/mm3 for 7 consecutive days; fever and neutropenia for ≥ 7 days during the 7-day treatment cycle; any Grade 3 or 4 nonhematologic toxicity, with the exception of nausea and emesis, which could not be controlled within 7 days; a delay of 2 weeks in completion of radiation therapy because of association with the carboplatin and RMP-7; or death on therapy that was related possibly, probably, or likely to carboplatin and RMP-7.

RMP-7/Carboplatin Dosing Guidelines

RMP-7 was given at a dose of 300 ng/kg of ideal body weight per day as a 10-minute intravenous infusion beginning 5 minutes before the end of the 15-minute carboplatin infusion. The RMP-7 infusion was to be completed 5 minutes after the end of the 15-minute carboplatin infusion. The carboplatin dose, at all durations of therapy, was 35 mg/m2 per day.

Radiation Therapy

Patients were to be treated with a linear accelerator with nominal photon energy between 4 megavolts (MV) and 10 MV. Electrons could not be employed. Any 2-dimensional (2-D) or 3-dimensional (3-D) treatment technique was allowed. For 2-D planning, the margin was to be 1.5 cm2. For 3-D planning, a margin of 1.0 cm in all directions was allowed. The treatment plan was to involve the planning target volume completely within the 95% isodose surface. The total dose of radiation was to be 5940 centigrays (cGy) in 180-cGy dose fractions (33 fractions).

Required Observations During and After Treatment

Because of potential blood pressure problems associated with RMP-7, blood pressure was measured every 15 minutes for the first hour before each dose of RMP-7 to establish baseline average systolic and diastolic blood pressure. Blood pressure was monitored every 5 minutes during RMP-7 infusion and for 20 minutes after completion. If the patient experienced dose-limiting hypotension, then blood pressure was monitored every 2 minutes until it returned to a level that did not meet DLT.

On-Study Evaluation

Patients underwent physical and neurologic examinations weekly throughout their course of radiation therapy, and blood counts and liver functions were to be evaluated twice weekly during radiotherapy. Blood chemistry evaluation was performed at baseline and at Weeks 3 and 5 of treatment. Toxicity was coded according to the National Cancer Institute Common Toxicity Criteria version 2. The maximum grade of each toxicity type observed during the 6-week period of protocol therapy was reported.

Six weeks after radiotherapy, patients were to undergo repeat neurologic and physical examinations and brain MRI scans with and without gadolinium. Patients also were to be evaluated by complete blood counts, serum chemistries, and an audiogram. Patients were then observed at 3-month intervals for the first 2 years after therapy then at 6-month intervals for the next 5 years.

Statistical Methods

Patients who were enrolled and eligible were considered in the calculation of risk of disease progression and risk of death. Progression-free survival (PFS) was defined as the time from study enrollment until disease progression, death, or last contact, whichever came first. A progression event was patient death or disease progression. Survival was defined as the time from study enrollment until death or last patient contact. A survival event was defined as patient death. PFS and survival as functions of time since enrollment were calculated using the method of Kaplan and Meier.20 Confidence intervals for median survival were calculated by the method of Brookmeyer and Crowley.21

RESULTS

Thirteen patients with a median age of 7 years (age range, 3–12 yrs) were enrolled between February 2001 and October 2003. Twelve patients completed treatment, and 1 child died from tumor progression during radiotherapy; thus, 12 patients were fully evaluable for toxicity.

Toxicity

There were no DLTs at the first 3 strata; thus, patients were able to complete 3 weeks, 4 weeks, and 5 weeks of carboplatin and RMP-7 and concurrent radiation without delays (see Table 1). At the fourth dose level (33 days of concurrent therapy), 1 patient developed Grade 3 hepatotoxicity (elevated alanine aminotransferase and aspartate aminotransferase levels) and Grade 3 neutropenia, which did not resolve over a 2-week period. Because there was no resolution, this was considered dose limiting. Radiation was continued, with a 3-day interruption for the patient who ultimately completed therapy at a 25% reduced dose of carboplatin and RMP-7. One patient at Dose Level 2 (who was to receive 4 weeks of RMP-7 and carboplatin) developed progressive disease within 3 weeks after the initiation of treatment, as stated above: Therapy was discontinued, and the child died. We did not believe that this death was related to carboplatin and RMP-7 infusion. All other stratums had only three patients entered.

Table 1. Nondose-Limiting Toxicities Observed during Protocol Therapy According to Dose Level and Grade
ToxicityGradeDose levela
15 doses over 3 weeks (n = 3)b20 doses over 4 weeks (n = 4)b25 doses over 5 weeks (n = 3)b33 doses over 7 weeks (n = 3)b)
  • a

    The number of patients who demonstrated the noted toxicity.

  • b

    The number of patients treated at each dose level.

Hemoglobin1 1  
Lymphopenia3 1  
Neutrophils/granulocytes2  1 
Platelets1 11 
Sinus tachycardia13 1 
Partial thromboplastin time1 1  
Fatigue (lethargy, malaise, asthenia)1 1  
Alopecia2   1
Flushing13212
Rash/desquamation1 1  
Hot flashes/flushes11   
Constipation11   
Nausea111 1
Emesis11  1
 2  1 
Gastrointestinal, other2 1  
Epistaxis11   
Aspartate aminotransferase1 1  
Alanine aminotransferase111  
 2 1  
Infection without neutropenia2   1
Hypocalcemia11   
Dizziness/lightheadedness111  
Hallucinations3   1
Mood alteration-anxiety, agitation2   1
Mood alteration-depression2   1
Ocular/visual, other (specify)11   
Abdominal pain or cramping1   1
Headache112  
Urinary frequency/urgency12   

Grade 1 tachycardia and flushing were common at all doses. Other toxicities, such as headache, gastrointestinal upset, and nausea, were seen at various levels, but none was dose limiting. Because of drug availability, the last cohort (Weeks 6 and 7 of radiotherapy) could not be expanded to ascertain the maximum tolerated dose of carboplatin at 35 mg/m2 per day and RMP-7, the protocol-defined maximum dose.

Response

Radiographic response was determined 6 weeks after radiotherapy. One child developed progressive disease early during radiotherapy, as noted above. Of the remaining 12 patients, 2 patients had partial responses (reduction > 50% in tumor area determined by the product of the 2 greatest dimensions), 2 patients had minor responses (change > 25%, but < 50% in tumor area), 7 patients had stable disease, and 1 patient had progressive disease. One child with stable disease had a marked increased tumoral enhancement, but no change in tumor size was observed 6 weeks after radiotherapy (a concurrent positron emission tomography study showed no increased uptake). Possible intratumoral necrosis was noted 6 weeks after treatment in 3 patients but was not associated with clinical worsening. One child had what we believed was leptomeningeal dissemination (brain and spine meningeal enhancement) 6 weeks after radiotherapy that was associated with increased tumor size. The increased enhancement remained stable for the ensuing 6 months, and the child ultimately died of progressive local disease, raising the issue of whether this represented leptomeningeal tumor spread or increased enhancement due to alterations in the blood-brain barrier.

Outcome

The estimated median survival was 329 days, with a 95% confidence interval of 258–394 days. One patient remained progression free > 400 days after the initiation of treatment, and 1 patient was alive at the time of this writing after disease progression (see Figs. 1, 2). Three of 10 patients who developed disease recurrences had disseminated disease at the time of recurrence. Two patients had diffuse leptomeningeal recurrences that involved the brain and spine, and one patient developed multiple, discrete lesions throughout the cerebral cortex and cerebellum.

Figure 1.

This chart illustrates progression-free survival from study enrollment for patients who were considered evaluable for dose-limiting toxicity.

Figure 2.

This chart illustrates overall survival from study enrollment for patients who were considered evaluable for dose-limiting toxicity.

DISCUSSION

The blood-brain barrier precisely regulates the central nervous system microenvironment, which is critical for normal neuronal function. Lack of fenestrations, decreased pinocytotic activity, tight junctions, and specific receptor-mediated transport mechanisms preserve this microenvironment.9 Protection is the blood-brain barrier's most critical function; however, by excluding toxic substances, it can prevent active antitumor agents from reaching their targets.9 This is an important deterrent to successful treatment of brain tumors with relatively intact barriers, such as diffuse, intrinsic brainstem gliomas. Although many mechanisms increase blood-brain barrier permeability nonspecifically, such as hypertension, hypercapnia, seizures, and hyperosmolar substances, they tend to be transient.10, 22–24 Osmotic disruption has been attempted to enhance treatment efficacy for children with brain tumors, with equivocal success.10 Osmotic disruption is nonselective; and, because tumors usually have inherently leakier blood-brain barriers, the increase in permeability is not as great as in normal brain and brain that surrounds tumor.

In a gallium-68 ethylenediamine tetraacetic acid (68Ga EDTA) positron emission tomography study of 9 adult patients with cortical recurrent malignant gliomas, it was shown that RMP-7 selectively increased transport of 68Ga EDTA into brain tumors without increasing transport into normal tissue.2568Ga EDTA and carboplatin have a similar molecular size, charge, weight, and water solubility. When this study began, Phase I studies in children had shown that RMP-7 infusion with carboplatin was tolerated well over 10 minutes.26 Phase I and Phase II adult trials and preliminary results from a Phase I pediatric study of recurrent brain tumors suggested that carboplatin may be more effective with RMP-7.26, 27 The toxicity profile of the combination seemed reasonable, primarily including vasodilatation, tachycardia, headaches, and nausea.

The rationale for carboplatin with radiotherapy included intrinsic efficacy in many different tumors, including gliomas, and radiosensitization.4, 6, 7 Cytotoxicity of platinum compounds is partially a function of unbound drug that reaches the tumor. Carboplatin binds more slowly to plasma proteins, with a half-life of 6 hours, and continuous infusion is not necessary to ensure free platinum during radiation. Carboplatin was given daily with radiation to patients with other types of malignancies, including nonsmall cell lung carcinoma, with improved survival.8 It also was delivered at 33 mg/m2 per day for 5 days during Weeks 1 and 4 of involved-field, accelerated radiation for glioblastoma multiforme, with reasonable tolerance.28 In a Phase I study, carboplatin was given twice weekly during hyperfractionated radiation for children with brainstem gliomas, and 280/mg/m2 per week for 7 weeks was the maximum tolerated dose.29 Despite responses, PFS was 8 months, and overall survival was only 12 months.29

Overall, in the current study, carboplatin with RMP-7 was tolerated well during radiotherapy. There were frequent minor side effects, such as flushing and tachycardia that occurred during and at all durations of the therapy, similar to the side effects observed previously for carboplatin and RMP-7 without radiotherapy in a pediatric Phase I study.26 Therapy was somewhat inconvenient for patients and caregivers. Patients must have received therapy within 4 hours of radiotherapy initiation. At most sites, chemotherapy and radiation were given in different places, creating logistic difficulties. In addition, the need to monitor blood pressure carefully to insure safety required significant caregiver time before and after infusion. Despite its drawbacks, the therapy was given successfully according to protocol.

One of the rationales for RMP-7 and carboplatin was its possible increased efficacy, compared with carboplatin alone, in adults with high-grade gliomas. A study by Gregor et al. for the RMP-7 European Study Group demonstrated stable disease or response in 79% of patients with recurrent, high-grade gliomas who had not received chemotherapy.27 A randomized, double-blind, placebo-controlled, Phase II study of RMP-7 with carboplatin versus carboplatin alone in adults with recurrent, high-grade gliomas was completed while our study was underway, and RMP-7 did not improve the efficacy carboplatin.30 A Phase III study of RMP-7 and carboplatin with radiation therapy in patients with newly diagnosed, high-grade gliomas also was underway. When preliminary results of that study demonstrated no increased efficacy for RMP-7/carboplatin concurrent with radiation therapy, the study was closed, and the drug supply subsequently became limited.

Conclusions concerning efficacy cannot be drawn reliably from this study. Disseminated disease in three children early after treatment was worrisome and raised the issue of whether blood-brain barrier disruption could lead to a greater incidence of tumor dissemination through vascular spread of tumor cells. Because of drug availability, the last cohort could not be expanded to ascertain the maximum tolerated dose, but we demonstrated the potential use of blood-brain barrier-disruptive agents with chemotherapy during radiotherapy for a tumor that is highly resistant to treatment and the ability to deliver a bone marrow-toxic chemotherapy daily during radiotherapy.

Ancillary