Other site principal investigators: Wayne Furman, MD (St. Jude Children's Research Hospital, Memphis, Tennessee); Rochelle Bagatell, MD (University of Arizona Health Sciences Center, Tucson, Arizona); Lori Luchtman-Jones, MD (Washington University School of Medicine, St. Louis, Missouri); Luis Eduardo Garcia, MD (Hospital CIMA, San Jose, Costa Rica); and Eric Sandler, MD (Nemours Children's Clinic, Jacksonville, Florida).
Preclinical testing suggests the combination of carboplatin and irinotecan has at least additive antitumor activity. The primary objectives of the current study were to determine the maximum tolerated doses (MTDs) and recommended phase 2 doses of carboplatin administered with irinotecan to pediatric patients with refractory solid tumors.
This was a multicenter, open-label, single-arm dose escalation study in which subjects with refractory solid tumors received 21-day treatment cycles of intravenous carboplatin on Day 1 followed by intravenous irinotecan administered daily for 10 days within 2 consecutive weeks. The plasma pharmacokinetics of ultrafiltrable platinum, irinotecan, and 2 irinotecan metabolites were determined during Cycle 1. The interpatient plan for dose escalation at study initiation was to increase irinotecan first followed by increases in carboplatin.
Twenty-eight patients with a median age of 8.5 years (range, 1-21 years) were enrolled with a variety of solid tumors. Two of 6 subjects at the first dose level (carboplatin target area under the curve [AUC], 4.0 mg/mL*min; irinotecan, 18 mg/m2/dose) experienced dose-limiting gastrointestinal toxicities requiring a dose de-escalation scheme (carboplatin AUC, 4.0 mg/mL*min; irinotecan, 15 mg/m2/dose). Three of 6 subjects at the second dose level experienced dose-limiting gastrointestinal complications and bone marrow suppression. A further dose de-escalation to carboplatin AUC of 4.0 mg/mL*min and irinotecan of 12 mg/m2/dose resulted in dose-limiting bone marrow suppression in 1 of 13 patients treated at that dose, and therefore was determined to be the MTD. One complete response (in a patient with medulloblastoma) and 3 partial responses (in patients with neuroblastoma, medulloblastoma, and lymphoendothelial carcinoma, respectively) were observed.
Irinotecan is a water-soluble camptothecan derivative that inhibits topoisomerase I (topo I), an enzyme involved in DNA repair, transcription, and replication.1-3 Single-agent irinotecan has demonstrated antitumor activity in xenograft models of childhood tumors including neuroblastoma, sarcomas, and brain tumors.4-6 Irinotecan is a prodrug that is metabolized by carboxylesterase enzymes to form SN-38, which is 100 to 1000 times more potent as a topo I inhibitor than irinotecan.7-11 Common toxicities include diarrhea, elevated transaminases, and myelosuppression when given in combination with other chemotherapeutic agents. A variety of schedules for irinotecan have been used, although there is preclinical evidence to suggest that a protracted administration schedule enhances efficacy while decreasing hematologic toxicity.12 However, the Children's Oncology Group recently completed a study comparing the response rate in an “upfront phase 2 window” to 2 different irinotecan schedules (daily × 5 days vs daily × 5 days for 2 consecutive weeks) in combination with vincristine in children experiencing their first recurrence/progression of rhabdomyosarcoma or undifferentiated sarcoma. In these patients with recurrent disease, the prolonged schedule appeared to offer no added benefit over the shorter 5-day schedule (unpublished data). Given the importance of irinotecan metabolism on its efficacy and toxicity, it is necessary to understand how the pharmacokinetics of irinotecan are affected by combination chemotherapy regimens.
Carboplatin is a platinum-based compound that causes DNA cross-links and single-strand breaks.13, 14 It has established activity against several pediatric tumors, including neuroblastoma and brain tumors.15-24 Major toxicities include myelosuppression and renal impairment.
The combination of carboplatin and irinotecan has been studied in adults with a variety of cancers and is well tolerated.25-31 Toxicities are mostly nonoverlapping; however, the combination can result in enhanced myelosuppression. Given the single-agent efficacy of carboplatin and irinotecan in a variety of pediatric tumors and the promising experience in adults, this combination is worthy of evaluation in pediatric patients with recurrent solid tumors. Therefore, a phase 1 study to evaluate the safety and feasibility of the combination was conducted.
The primary objective was to determine a maximum tolerated dose (MTD) and recommended phase 2 dose for pediatric patients. Secondary objectives included determining dose-limiting toxicities (DLTs) and plasma pharmacokinetics of carboplatin, irinotecan, and irinotecan metabolites, as well as assessing antitumor activity within the context of a phase 1 trial. At the time of study design, it was expected that combination therapy could be achieved with close to MTDs of each agent. Given particular interest in the use of irinotecan for the target population, irinotecan dosing was initiated at 90% of the single-agent phase 2 dose.6 Carboplatin was dosed at approximately 67% of the recommended dose when used in combination with other chemotherapeutic agents.32 The intent was to dose escalate the irinotecan and then the carboplatin.
MATERIALS AND METHODS
Patients ages 1 to 21 years with a histologically or cytologically confirmed diagnosis of a solid malignancy that had progressed on standard therapy or for which no standard effective therapy was known were eligible to enroll after written and voluntary informed consent from patients or their legal guardians. Patient assent was obtained if appropriate. The study was approved by each institution's institutional review board. Other criteria included: measurable or nonmeasurable disease; adequate bone marrow function (absolute neutrophil count [ANC] >1.0 × 109/L and platelet count >100,000 × 106/L); adequate hepatic function (total bilirubin ≤1.5 mg/dL and alanine transaminase ≤2.5 times the institutional upper limit of normal); adequate renal function (glomerular filtration rate [GFR] >30 mL/min/m2 as measured by creatinine clearance or radionuclide scan); at least a 4-week interval since the last immunotherapy, radiotherapy, or chemotherapy; life expectancy of at least 8 weeks; recovery to baseline or grade 1 from toxicities resulting from prior therapy; Karnofsky score or Lansky play scale ≥50; and a negative pregnancy test. Patients were ineligible or excluded for any of the following reasons: >3 prior chemotherapy regimens; serious uncontrolled medical disorder or infection; symptomatic brain metastases; receiving phenytoin, phenobarbitol, primidone, carbamazepine, or valproic acid; and having previously underwent bone marrow transplantation.
Study Design, Treatment, and Dose Escalation
This was an open-label, single-arm dose escalation study in which subjects received a 21-day treatment cycle of intravenous carboplatin on Day 1 followed by intravenous irinotecan administered daily for 10 days within 2 consecutive weeks (Fig. 1) at the following institutions: Children's National Medical Center, Washington, DC; Memorial Sloan-Kettering Cancer Center, New York, New York; St. Jude Children's Research Hospital, Memphis, Tennessee; Children's Hospital of Pittsburgh, Pittsburgh, Pennsylvania; University of Arizona Health Sciences Center, Tucson, Arizona; Washington University School of Medicine, St. Louis, Missouri; Hospital CIMA Centro Internacional de Medicina Avanzada, San Jose, Costa Rica; and Nemours Children's Clinic, Jacksonville, Florida. Commercial supplies of carboplatin and irinotecan were provided by Bristol-Myers Squibb.
The initial dose level was carboplatin given at a target area under the curve (AUC) of 4.0 mg/mL/min administered over 50 minutes followed 10 minutes later by intravenous irinotecan at a dose of 18 mg/m2 administered over 60 minutes. A total of 10 doses of irinotecan were given over the first 15 days of each 21-day cycle. The carboplatin dose was calculated as previously described32 using the following formula:
The initial dose escalation schema planned to increase irinotecan first followed by increases in carboplatin using a traditional 3-plus-3 phase 1 trial design such that cohorts of 3 subjects would be treated at each dose level. Escalation to the next dose level was permitted if each subject at the current dose level completed the first cycle of treatment without experiencing a DLT. If 1 of the first 3 evaluable patients developed a DLT during Course 1, up to 3 additional subjects were enrolled at that dose level. If ≥2 subjects at a particular dose level experienced a DLT at any time during the first cycle, then the MTD would be exceeded. The MTD was to be expanded to treat at least 9 subjects to determine a recommended phase 2 dose. Dose escalation within individual subjects was not permitted.
Given unanticipated toxicities at the planned initial dose level, the study was amended to include a dose de-escalation scheme (Table 1).
Table 1. Dose Escalation Scheme
Carboplatin (AUC, mg/ mL*min)
Irinotecan (mg/m2/Day ×x 5 for 2 Weeks)
No. of Patients per Cohort
AUC indicates target area under the curve.
1 (starting dose level)
DLTs were defined as any of the following events that could be considered possibly, most likely, or definitely related to study drug therapy: grade 4 neutropenia (ie, ANC <500 cells/mm3) for >7 consecutive days; grade 4 febrile neutropenia (ie, life-threatening infection); thrombocytopenia (<25,000/mm3) for >7 consecutive days; thrombocytopenia requiring >2 platelet transfusions within a period of 7 consecutive days; grade 3 or greater nausea and/or vomiting despite the use of adequate/maximal medical intervention and/or prophylaxis; grade 3 or greater diarrhea despite the use of adequate/maximal medical intervention and/or prophylaxis; any other grade 3 or greater nonhematologic toxicity except alopecia, anorexia, fatigue/asthenia, or transient arthralgia/myalgia (unless unresponsive to maximal medical intervention); retreatment delay of >1 week between consecutive irinotecan doses within a cycle because of delayed recovery from a toxicity related to treatment with carboplatin and irinotecan; and retreatment delay of >2 weeks from the scheduled start date of a subsequent cycle because of delayed recovery from a toxicity related to treatment with carboplatin and irinotecan.
All subjects were followed for at least 30 days after the last dose of therapy and then every 4 weeks until study drug-related toxicities were resolved or were deemed irreversible. Subjects were to receive treatment at the assigned dose level for a maximum of 12 courses provided they did not experience a DLT. Escalation to the next higher dose level was permitted if all 3 patients at the current dose level completed their first cycle of treatment and no patient experienced a DLT. Enrollment to an open cohort occurred simultaneously (ie, data from the initial patients enrolled to a given cohort were not required to continue enrollment to that cohort).
Patients were required to receive the Day 1 infusion and at least 1 irinotecan dose in the second week of each course at the treating institution. Home infusions of other irinotecan doses were permitted.
No other chemotherapy, immunotherapy, hormonal therapy, or radiotherapy was permitted while the subject was in the study. Granulocyte-colony-stimulating factor was administered at a dose of 5 μg/kg/day subcutaneously or intravenously 1 day after the 10th dose of irinotecan in a treatment cycle and continued until the ANC was >1.0 × 109/L. Atropine was prescribed for early diarrhea and loperamide for delayed onset diarrhea. Antiemetic regimens were prescribed at the discretion of the treating physician.
The plasma pharmacokinetics of ultrafiltrable platinum, irinotecan, SN-38 (the active irinotecan metabolite), and APC (RPR-121056; another irinotecan metabolite) were determined during Cycle 1 with scheduled blood collections immediately before chemotherapy, at 30 minutes, 48 minutes, 90 minutes, 2 hours, 4 hours, 6 hours, 8 hours, 24 hours, and then on Days 3, 5, 8, 10, and 12. Approximately 18 mL of blood was obtained from each patient for pharmacokinetic study during the first 24 hours of the study, and an additional 5 mL was collected for the remainder of the study. Samples were collected in tubes containing liquid K3 ethylenediamine tetraacetic acid as the anticoagulant and immediately centrifuged for 5 minutes at 3000 × g to generate plasma. Plasma (0.4 mL) was taken for ultrafiltration, and the remaining plasma was flash frozen and stored at −70°C until further analysis.
The 0.4-mL plasma sample was immediately transferred to an ultrafiltration device, and centrifuged with a fixed angle head at 1000 × g for 20 minutes at 10°C to 15°C. Ultrafiltrate was transferred to a labeled tube and stored at −20°C or colder until shipment for further analysis.
Plasma samples were assayed for the concentration of the lactone form of irinotecan, SN-38, and APC by a validated high-performance liquid chromatographic-mass spectrometric method.33 Plasma ultrafiltrate samples were analyzed for platinum concentrations by a validated flameless atomic absorption method.34 The following pharmacokinetic values for irinotecan, SN-38, APC, and ultrafiltrable platinum were determined: maximum plasma concentration (Cmax), the time to reach Cmax, mean residence time, plasma elimination half-life, and AUC. In addition, total body clearance and volume of distribution at steady state were calculated for irinotecan and ultrafiltrable platinum, and the predose plasma concentrations of irinotecan and its metabolites on Days 3, 5, 8, 10, and 12 were evaluated for evidence of accumulation.
Pre-enrollment evaluations included a complete medical history and physical examination, complete blood count (CBC), comprehensive metabolic panel (CMP), audiogram, creatinine clearance, chest x-ray, and tumor measurements. CBCs were performed weekly. Physical examinations with evaluation of symptoms, CBC, CMP, and creatinine clearance were performed before the start of each cycle. Audiograms were repeated before every other cycle.
Adverse events were assessed using the National Cancer Institute Common Toxicity Criteria (version 2.0).35 Consistent imaging for tumor assessments was performed at enrollment, after every second cycle of therapy, and upon study completion, withdrawal, or discontinuation of therapy. Patients were evaluated for response according to the international Response Evaluation Criteria in Solid Tumors (RECIST) criteria. Complete response (CR) was defined as the disappearance of all clinical and radiologic evidence of tumor, a partial response (PR) was defined as a decrease of ≥30% in the sum of the longest dimension of all lesions, and progressive disease was defined as a >20% increase in the sum of the longest dimension of all lesions or the presence of a new lesion. Stable disease indicated neither sufficient increase nor decrease in tumor burden to categorize otherwise.
Descriptive statistics were used to summarize clinical safety and laboratory observations
Patient Characteristics and Treatment Administration
Twenty-eight patients with a median age of 8.5 years (range, 1-21 years) with the following tumor types were enrolled: osteosarcoma (5 patients), neuroblastoma (4 patients), hepatoblastoma (4 patients), astrocytoma (3 patients), rhabdomyosarcoma (3 patients), medulloblastoma (2 patients), Wilms tumor (2 patients), atypical teratoid (1 patient), synovial sarcoma (1 patient), glioma (1 patient), lymphoepithelial carcinoma (1 patient), and alveolar soft part sarcoma (1 patient) (Table 2).
Table 2. Patient Characteristics
Median age (range), y
No. of prior chemotherapy regimens
Twenty-three of the 28 patients had received prior chemotherapy; 10 had received prior carboplatin, 17 had received prior cisplatin, and 13 patients had received at least 2 prior treatment regimens. Four patients had received prior topotecan, but none had received irinotecan. Eighteen of these 28 patients had received prior radiotherapy.
Six subjects were enrolled on Dose Level 1 (carboplatin AUC of 4 and irinotecan, 18 mg/m2/day). Two subjects experienced a DLT. One patient experienced grade 3 diarrhea. Another patient had grade 3 epistaxis, grade 3 diarrhea, grade 3 ileus, and grade 3 dehydration. Because the initial dose level exceeded the MTD, the protocol was amended, and a de-escalation scheme was designed (Table 1) using lower irinotecan doses given particular concerns regarding gastrointestinal DLT.
Six subjects were enrolled on Dose Level −1 (carboplatin AUC 4 and irinotecan 15 mg/m2/day). Three of 6 subjects at the second dose level experienced DLTs, including 1 patient with grade 4 neutropenia for >7 days, grade 4 thrombocytopenia, and grade 3 hemorrhage. One patient had grade 4 abdominal pain and a grade 3 catheter infection. One patient had delayed hematologic recovery. The subsequent cohort of subjects was treated at Dose Level −2 with further de-escalation of the irinotecan dose.
Four patients were assigned to Dose Level −2 (carboplatin AUC of 4 and irinotecan, 12 mg/m2/day). No DLTs were observed, and 3 subsequent patients were enrolled at Dose Level −2a with escalation of the carboplatin dose to an AUC of 5 with irinotecan, 12 mg/m2/day. All 3 subjects treated on Dose Level −2a experienced a DLT, including at least 1 of the following: prolonged thrombocytopenia, grade 3 diarrhea, grade 3 abdominal pain, and/or excessive platelet transfusion requirement.
Dose Level −2 was identified as the MTD. Nine additional subjects were treated at this dose level. Only 1 of the 13 patients treated at the MTD experienced DLTs (grade 3 bone pain, grade 4 neutropenia for >7 days, and prolonged hematologic recovery). Thus, Dose Level −2 was determined to be the recommended phase 2 dose.
The mean plasma concentration-time profile of irinotecan, SN-38, and APC after iv administration of carboplatin (AUC of 4) and irinotecan (12, 15, and 18 mg/m2) on Day 1 are depicted in Figure 2a. The mean plasma concentration-time profile of platinum after the intravenous administration of carboplatin (AUC of 4) and irinotecan (12, 15, and 18 mg/m2) on Day 1 is depicted in Figure 2b.
Pharmacokinetic parameter values of irinotecan and metabolites are detailed in Table 3, and pharmacokinetic parameter values of platinum from carboplatin are detailed in Table 4. The exposure of patients to SN-38 and APC determined in this study are consistent with values previously reported for this protracted low-dose regimen.36 As such, the pharmacokinetics of irinotecan are apparently not affected by concurrent carboplatin. However, the trough plasma concentrations provide evidence of possible SN-38 and APC accumulation for this protracted regimen.
Table 3. Pharamcokinetic Parameter Values of Irinotecan, SN-38, and APC
Cmax indicates maximum plasma concentration; Tmax, time to reach Cmax; AUC, target area under the curve; MRT, mean residence time; T1/2, half-life; CLT, clearance; Vss, volume of distribution.
Values presented as the mean (standard deviation) and range.
Dose of carboplatin (AUC)/dose of irinotecan (mg/m2).
Values are presented as the median and range.
Tumor evaluations were centrally reviewed for all patients to determine each patient's best overall response.
Of the 28 patients treated in this study, 1 patient achieved a CR, 3 patients acheived a PR, and 10 patients had stable disease. Among the patients who had not previously received chemotherapy, there were no objective radiographic responses to protocol therapy noted.
A 2-year-old with medulloblastoma achieved a CR after the fourth cycle of therapy at Dose Level −2 (MTD). The disease recurred 9 months later. The patient had received prior carboplatin and cisplatin.
One PR occurred in a 15-year-old with lymphoepithelial carcinoma after the second cycle of therapy at Dose Level 1, and the disease recurred 3 months later. The patient had received prior cisplatin. Two other PRs occurred in patients who were treated at Dose Level −2 (MTD). A PR occurred in a 4-year-old with medulloblastoma after the second cycle of therapy, and the disease recurred 3 months later. The patient had received prior cisplatin. Another PR occurred in a 3-year-old with neuroblastoma after the second cycle of therapy, and disease recurrence was documented 2 months later. The patient had received prior carboplatin and cisplatin as well as high-dose chemotherapy with peripheral stem cell transplant.
Ten patients had stable disease while receiving therapy (2 patients with osteosarcoma, 2 with astrocytoma, 1 with glioma, 1 with synovial sarcoma, 1 with rhabdomyosarcoma, 1 with hepatoblastoma, 1 with Wilms tumor, and 1 with alveolar soft part sarcoma).
This phase 1 study was conducted to characterize a novel schedule of the combination of carboplatin and irinotecan, 2 agents that have demonstrated single-agent activity in a variety of pediatric solid malignancies. Irinotecan dosing was initiated at 90% of the single-agent phase 2 dose,6 and carboplatin was dosed at approximately 67% of the recommended dose when used in combination with other chemotherapeutic agents.32 The intent was to dose escalate the irinotecan and then the carboplatin. However, significant gastrointestinal and hematologic toxicities were encountered that required de-escalation of the irinotecan dose. Despite the irinotecan dose de-escalation, the concurrent carboplatin dose could not be further escalated.
The protracted irinotecan regimen used (ie, daily × 5 for 2 weeks) was based on preclinical data suggesting improved efficacy with this schedule.6, 12 Other studies using this schedule reported toxicities similar to those found in this study. The Memorial Sloan-Kettering study with single-agent irinotecan reported diarrhea as the predominant toxicity, with neutropenia and thrombocytopenia observed in fewer patients.37 The Italian Soft Tissue Sarcoma Committee performed a phase 2 study of irinotecan (20 mg/m2/day for 5 days a week, for 2 consecutive weeks) and found the main toxicity to be diarrhea with limited hematologic toxicity.38 In 2 consecutive phase 2 window studies of irinotecan alone and in combination with vincristine for patients with metastatic rhabdomyosarcoma, the Children's Oncology Group concluded that irinotecan (20 mg/m2/day for 5 days a week, for 2 consecutive weeks) was effective and generally well tolerated, with gastrointestinal toxicities being the most common adverse events.39 A Children's Oncology Group phase 2 trial of irinotecan (50 mg/m2/day for 5 days repeated every 3 weeks) in refractory pediatric solid tumors was well tolerated but not effective as a single agent in a variety of solid tumors with the exception of medulloblastoma (16% response rate).40 A recently completed “upfront phase 2 window” study found similar response rates when comparing 2 different irinotecan schedules (daily × 5 days vs daily × 5 days for 2 consecutive weeks) in combination with vincristine in children experiencing their first recurrence/progression of rhabdomyosarcoma or undifferentiated sarcoma (unpublished data). Thus, although there is preclinical evidence to suggest that a prolonged administration schedule is more efficacious, it is unclear precisely which schedule is ideal. Particularly in patients with recurrent disease, a shorter 5-day schedule would be more convenient than a longer schedule and would be a reasonable consideration for future studies using irinotecan.
In contrast to irinotecan, carboplatin's predominant DLT is myelosuppression. Combined with ifosfamide and etoposide, the recommended dose for carboplatin is an AUC 6-7.32 As demonstrated by the finding that it was not possible to escalate the carboplatin dose beyond an AUC of 4, the current study found increased hematologic toxicity in this moderately to heavily pretreated patient population. The hematologic toxicity observed likely demonstrates the potential additive myelosuppressive effects of irinotecan (and/or its metabolites) in previously treated patients.
Multiple dosing schedules of the combination of irinotecan and carboplatin have been studied in adults with a variety of malignancies,25-27 including lung cancer,29-31 ovarian cancer,28 and non-Hodgkins lymphoma.41 As in the current study, the main toxicities were gastrointestinal and hematologic, and efficacy in phase 1 of 2 studies has been encouraging. Studies in adults tended not to use a protracted irinotecan schedule, so comparison between the adult experience and this study is limited.
The pharmacokinetics of single-agent irinotecan and its metabolites has been reported in adults using a variety of dosing schedules9, 42, 43 and in children with solid tumors receiving 20, 24, or 29 mg/m2 of irinotecan daily for 5 days for 2 weeks36, 44 or 50 mg/m2/day for 5 days repeated every 3 weeks.40 To our knowledge, the current study is the first report of irinotecan pharmacokinetics in children when given in combination with carboplatin. Pharmacokinetic data were available from 20 patients for platinum and 22 patients for irinotecan and its metabolites. As seen in single-agent irinotecan studies,7, 36, 44 the exposure levels of SN-38 were very variable. In a previous study in children receiving 20 mg/m2 of irinotecan, the mean AUC of irinotecan lactones, SN-38, and APC were 423, 28.4, and 57.0 ng/mL/hour, respectively.44 These results are consistent with the values observed in the current study. In the current study, irinotecan plasma clearance when given with carboplatin was similar to that reported in the Children's Oncology Group phase 2 trial of irinotecan.40 The current study suggests there is no effect of carboplatin on the pharmacokinetics of irinotecan or its metabolites. Limited patient numbers prevent a significant correlation between individual patient pharmacokinetics and outcome.
A recommended phase 2 dose of combination carboplatin and irinotecan for heavily pretreated patients was determined to be carboplatin (AUC of 4 mg/mL/min × 1 day) and irinotecan (12 mg/m2/day × 10 days). Although the study was not designed to study efficacy, the activity observed was encouraging. Of 13 patients treated at the recommended phase 2 dose, objective responses were documented in 3 patients. All responders had received prior cisplatin, and 2 had also received prior carboplatin.
Although the combination of carboplatin and irinotecan was tolerated only when using doses of each drug that are well below the single-agent MTD, activity was observed for some heavily pretreated patients with refractory tumors. This suggests potential additive antitumor activity through as yet unknown mechanisms. For medulloblastomas, neuroblastomas, and sarcomas in particular, these findings support further investigation of combination carboplatin and irinotecan for pediatric solid malignancies.
Conflict of Interest Disclosures
Ms. Ringuette and Mr. Cohen are both employees of Bristol-Myers Squibb and own stock in the company. Supported by Bristol-Myers Squibb, Lawrenceville, New Jersey.