Purpose: Vinorelbine (Navelbine) is an orally absorbable Vinca with broad antitumor activity. It differs from other Vinca in that it is structurally modified on the catharanthine nucleus and has differential actions on tubulin that render it less neurotoxic than other compounds in this class. We conducted a phase I study of vinorelbine given the activity of Vinca alkaloids in many pediatric tumors.

Experimental Design: We evaluated the safety and pharmacokinetics of oral and i.v. vinorelbine administered weekly × 6 in children (age, 2-17 years) with different tumors. Patients with disease involvement in the bone marrow were eligible but were stratified and dose-escalated separately. Oral vinorelbine (week 1) was administered as liquid-filled gelatin capsules at thrice the i.v. dose. Intravenous vinorelbine doses of 24 to 37.5 mg/m2 were administered on weeks 2 to 6.

Results: The dose-limiting toxicity in patients without marrow involvement was reversible neutropenia. Common nonhematologic toxicities included ≤ grade 2 nausea/vomiting and increased hepatic transaminases. A higher mean i.v. ClTB was observed (1.75 ± 1.0 L/h/kg) compared with adult reports, with a mean t1/2B of 16.5 ± 9.7 hours. Mean oral bioavailability was 28.5 ± 22.5%. The apparent oral clearance (12.1 ± 13.0 L/h/kg) and volume of distribution (69.4 ± 30.6 L/kg) were substantially higher than in adults given similar oral doses.

Conclusions: The maximum tolerated dose in children without bone marrow involvement was 30 mg/m2, similar to that reported in adults, with myelosuppression being the dose-limiting toxicity. Higher plasma clearance resulted in lower area under the plasma concentration-time curves at a given dose compared with that reported in adults.

Vinca alkaloids are an important class of cell cycle–dependent antimitotic agents with activity established in pediatric malignancies. The parent compounds, vincristine and vinblastine, are naturally occurring Catharanthus roseus (Madagascan periwinkle) alkaloids that have been used widely in chemotherapy regimens. Analogues of the parent compounds have been synthesized to broaden their therapeutic activity, improve therapeutic index, and identify candidates that have adequate and consistent bioavailability when administered orally. Several synthesized analogues with antitumor activity have failed because of unpredictable toxicities or have not been candidates for oral administration because of inadequate or erratic absorption (13).

Vinorelbine (Navelbine, GlaxoSmithKline, Research Triangle Park, NC) is a unique semisynthetic Vinca alkaloid with a broad activity spectrum, effective as a single agent and in combination therapy for advanced breast cancer and non–small-cell lung cancer (49). In animal models, vinorelbine has shown significant activity against some human childhood high-grade glioma xenografts (10). Vinorelbine differs from other Vinca in chemical structure, effects on microtubules, and toxicity profile. Unlike other synthetically modified Vinca, vinorelbine is synthesized by modification of the catharanthine ring versus the vindoline ring and has differential interactions with tubulin (11, 12). It is well tolerated when administered orally with leukopenia as the dose-limiting toxicity.

Vinca alkaloids inhibit microtubule assembly (depolymerization) and induce unwinding and spiral formation. Although vinorelbine has been as active as vincristine and vinblastine against tubulin assembly in vitro, it is inefficient at causing spiral formation. It preferentially binds to the mitotic spindle instead of axonal neurons, so vinorelbine is neuron-sparing, imparting less neurotoxicity than other compounds in this class. It was believed that vinorelbine may have superior efficacy for central nervous system tumors compared with vincristine, where dose-limiting neurotoxicity has hampered treatment in some patients.

In clinical trials of adult cancer patients, rapid but variable oral absorption was reported after administration of the initial hard, powder-filled vinorelbine capsule, with a mean fraction absorbed of 0.40 as determined by RIA analysis (13). Oral bioavailability of the newer soft, liquid-filled capsule has been reported as relatively low at 27%, which suggests a large first-pass effect, with moderate interindividual and intraindividual differences in absorption (14).

This study sought to determine safety, maximum tolerated dose, and pharmacokinetic behavior of vinorelbine given by short i.v. infusions at weekly intervals in pediatric patients and to evaluate bioavailability and tolerance of oral administration in children. Antitumor activity against pediatric malignancies also was evaluated within the confines of a phase I study.

Patient selection. Enrollment was open to subjects ≤21 years old with histologically verified malignancies refractory to conventional therapy (except for brain-stem tumors) and other therapies of higher priority. A performance status of ≤2 (modified Zubrod scale) and a life expectancy of at least 2 months were also required. Patients had to have been recovered fully from prior therapy with at least 2 weeks between their last chemotherapy dose (4 weeks for nitrosourea or mitomycin C) and the start of vinorelbine therapy and 1 week since their last dose of hematopoietic colony-stimulating factor. Previous treatment with vincristine or vinblastine was allowed given that cross-resistance of vinorelbine with other Vinca alkaloids had not been found. Patients without bone marrow involvement were required to have adequate bone marrow function [peripheral absolute neutrophil count (ANC) ≥1,000/mm3, transfusion-independent platelet counts ≥100,000/mm3, and hemoglobin ≥10.0 g/dL]. Patients with bone marrow involvement (solid tumor, leukemia or lymphoma who had granulocytopenia, anemia, and/or thrombocytopenia) were eligible but were stratified and dose escalated separately. Also required were adequate renal function (serum creatinine <2.0 × normal, or creatinine clearance ≥70 mL/min/1.73 m2, or radioisotope glomerular filtration rate ≥70 mL/min/1.73 m2) and liver function (total bilirubin ≤1.5 × normal and alanine aminotransferase or aspartate aminotransferase <2.5 × normal). Signed informed consent was obtained from all patients or their legal guardians before study entry according to the Declaration of Helsinki.

Treatment plan. Vinorelbine was administered every 7 days for 6 weeks. Each week of therapy was considered a cycle for the purposes of toxicity evaluation. Dosage was calculated using body surface area on the days of treatment. The first dose was given orally as liquid-filled capsules 8 hours after eating or 4 hours before eating, at a dose thrice the i.v. dose for a given patient. That dose was based on limited data that indicated ∼30% oral bioavailability for this drug form. Oral doses were rounded to the nearest 10 mg, with a maximum of 200 mg regardless of body surface area. Patients took capsules a few minutes apart, under observation by clinical staff. Administration of drugs known to alter gastrointestinal motility or absorption (i.e., metoclopromide, antacids, H2 antagonists, and anticholinergics) was prohibited on the day of oral vinorelbine treatment. For patients who could not swallow capsules or had gastrointestinal dysfunction, all doses were administered i.v.

I.v. doses were in 0.9% NaCl or D5W given over 20 minutes, followed by at least 100 mL of fluid on days 0 (for those who could not receive oral dosing), 7, 14, 21, 28, and 35. Other concomitant chemotherapy or immunomodulating agents were not allowed, including steroids and growth factors. Corticosteroid therapy was permissible if needed for treatment of increased intracranial pressure in patients with central nervous system tumors. Radiotherapy for painful localized lesions was allowed provided at least one measurable lesion was left not irradiated to assess tumor response.

Dose escalation. The study was designed to escalate dosage of vinorelbine in successive cohorts of children to assess toxicity and find the appropriate dose for subsequent phase II trials. The i.v. starting dose was 24 mg/m2 (80% of the adult maximum tolerated dose). Doses were escalated in groups of three in 25% increments over three levels up to 37.5 mg/m2 or until the maximum tolerated dose was determined. Intrapatient dose escalation was not permitted. When the maximum tolerated dose was exceeded, an intermediate dose level halfway between the last nontoxic and the toxic dose was studied in an attempt to identify a dose level that was below but closer to the unacceptable level of toxicity. Patients were entered and evaluated at this intermediate dose level as described for other dose levels.

The dose-limiting toxicity for nonhematologic toxicity was defined as grade 3 or 4 toxicity, excluding grade 3 nausea, vomiting, and fever, or a grade 3 hepatic toxicity that returned to grade 1 before the next scheduled treatment. Dose-limiting hematologic toxicity in patients without bone marrow involvement was defined as grade 4 toxicity that lasted ≥7 days, or was accompanied by any grade 3 or 4 nonhematologic toxicity. In those patients, vinorelbine was withheld if their ANCs were <200/μL or their platelet counts were <20,000/μL. Vinorelbine was resumed upon recovery above those levels. If their ANCs were 200 to 499/μL or their platelet counts were between 20,000 and 49,999/μL, the next vinorelbine dose was 33% lower. Patients with bone marrow involvement were not evaluated for hematologic toxicity and were escalated separately.

If none of the first three patients treated at a dose level had dose-limiting toxicity, the dose was escalated to the next level; if two or three of the first three patients had dose-limiting toxicity, that dose was considered too high and the dose was decreased to the previous level. If one of the first three patients at a particular dose level had dose-limiting toxicity, three more patients were enrolled at that level. If after six patients only one had dose-limiting toxicity, the dose was escalated. If two or more had dose-limiting toxicity, the dose was greater than the maximum tolerated dose. The maximum tolerated dose was the dose at which one of six patients at most had dose-limiting toxicity, and the next higher level had at least two patients who had dose-limiting toxicity. However, due to a large number of early patient withdrawals, the protocol was amended after the initial 10 patients were enrolled to include dose escalation in cohorts of four patients. Cohorts of four patients were enrolled to ensure that an adequate number of patients per cohort were available for evaluation of toxicity and response. The same criteria described above were used to determine dose-limiting toxicity and maximum tolerated dose in these patient cohorts. Any adverse drug reactions or toxic deaths believed to be treatment related were reported within 24 hours to the study chair. A National Cancer Institute Adverse Reaction Form for Investigational Agents was filed with the Children's Cancer Group (now Children's Oncology Group) Operations Office within 7 days and with the study sponsor within 10 days.

Patient evaluation. Each patient was to be observed for 30 days after their last dose or until resolution of any drug-related toxicity. Patients who completed therapy were observed every 6 months until death. Criteria for termination of protocol therapy included family request for removal, progressive disease, development of hypersensitivity reactions to vinorelbine infusions, nonhematologic grade 4 toxicity, or any nonreversible grade 3 or 4 toxicity. Grade 3 nonhematologic toxicity had to have resolved to grade 2 or better within 3 weeks of treatment for the patient to remain on study. The same criteria applied to grade 3 or 4 hematologic toxicities in patients without bone marrow involvement. An exception was made for hepatotoxicity. Grade 4 hepatotoxicity for ≤3 days was permitted. Those removed from protocol therapy were observed for survival at 3-month intervals until death.

Patients with measurable disease were also evaluated for response. Response criteria for patients with hematologic malignancies were as follows: complete response, M1 bone marrow (5% blasts) with no circulating blasts or extramedullary disease, with recovery of peripheral blood counts to ANC >1,000/mm3 and platelets >100,000/mm3 × 4 weeks; partial response, M2 marrow (<25% blasts), and no circulating blasts, with recovery of blood counts as described above; progressive disease, increase of ≥25% in the number of circulating or extramedullary leukemic cells; stable disease, not qualified for complete response, partial response, or progressive disease. Response criteria for solid tumor patients were as follows: complete response, complete disappearance of all disease for 4 weeks; partial response, 50% reduction of measurable tumor size for 4 weeks without new lesions or progression of any lesion; stable disease, <50% reduction of tumor size and an increase of <25% of the sums of the largest diameters of measurable lesions without new lesions; progressive disease, new lesions or an increase of >25% of measurable lesions (excluding bone). If clinical benefit was observed, as determined by the treating physician, patients were eligible to receive additional courses.

Patients who completed at least four of six cycles of vinorelbine or experienced a dose-limiting toxicity were considered in the toxicity evaluation for maximum tolerated dose assessment.

Pharmacokinetic evaluation. Pharmacokinetic evaluations of oral and i.v. vinorelbine were done on the first (oral) and second (i.v.) doses. For dose 1 (oral), 4 mL of blood were collected in heparinized tubes before the drug was given, at 15, 30, and 45 minutes, and at 1, 1.5, 2, 4, 6, 8, 16, 24, 48, 72, and 96 hours after administration. For dose 2 (i.v.), samples were collected, as described, before the drug was given, at 20, 25, 30, and 45 minutes, and at 1, 2, 6, 8, 16, 24, 48, 72, and 96 hours after administration. Samples were drawn through an in-dwelling catheter or heparin-lock in the contralateral arm from the infusion and analyzed for vinorelbine concentration by high-performance liquid chromatography using fluorescence detection (15).

Vinorelbine was extracted from plasma as follows: plasma samples (50 μL-1 mL) were alkalinized with 500 μL of 1 mol/L K+PO4 buffer (pH 10) and extracted with 8 mL of ethyl ether. If <1 mL of plasma was used, the volume was made up to 1 mL with blank plasma. Samples were shaken gently for 30 minutes and centrifuged for 10 minutes at ∼1,000 × g. Six milliliters of the organic layer were transferred to 12 mL graduated conical glass tubes and evaporated to dryness under nitrogen at room temperature. The residue was reconstituted with 200 μL mobile phase [acetonitrile/40 mmol/L ammonium acetate (pH 3), 55:45]. The extracted samples were then washed with 2 mL hexane and centrifuged for 5 minutes. One hundred microliters of the bottom layer were injected onto the chromatographic system and separation was achieved using a cyano column (5 μmol/L particle size, 4.6 × 150 mm) with a mobile phase flow rate of 1 mL/min. The effluent was monitored by fluorescence detection at an emission wavelength of 360 nm and excitation wavelength of 280 nm. A standard calibration curve (2-50 ng/mL) was generated by adding desired amounts of vinorelbine into 1 mL blank plasma and processing with the unknown samples. The lower limit of quantification of this assay was 2.0 ng/mL.

Pharmacokinetic variables for vinorelbine were determined from concentration-time data for oral and i.v. vinorelbine using a noncompartmental analysis tool in WinNonlin, version 3.0 (Pharsight Corporation, Mountain View, CA).

Study population. Forty-six patients were entered in this trial at 13 different Children's Cancer Group institutions. Twenty patients who completed at least five of six scheduled vinorelbine cycles and two who completed four of six scheduled cycles were evaluable for toxicity and disease response. Seven additional patients who experienced dose-limiting toxicity but received less than five cycles were eligible for toxicity evaluation only. Seventeen patients, not experiencing toxicity, were withdrawn from the study early due to progressive disease and were not considered eligible for further evaluation.

The clinical characteristics of patients evaluated are summarized in Table 1. Nineteen males and 10 females were included, who were between the ages of 2 and 17 years (median, 12 years). Most patients had central nervous system tumors, bone tumors, or soft-tissue sarcomas. Four patients with disease involvement in bone marrow were evaluable at the second dose (30 mg/m2), three with hematologic malignancies and one with sarcoma. All patients had a performance status of ≤2 (modified Zubrod scale). Most had been treated with one to three prior chemotherapy regimens and one to two prior radiation courses. Half had been exposed to more than six different chemotherapy agents.

Table 1.

Patient characteristics

CharacteristicTotal
No. evaluable patients (enrolled)* 29 (46) 
Sex (M/F) 19/10 
Median age, y (range) 12 (2-17) 
Previous cytotoxic therapy  
    No. radiation courses  
        0 
        1 13 
        2 
    No. chemotherapy regimens  
        1-3 24 
        4-6 
        8 
Disease or site  
    CNS 10 
    Sympathetic nervous system 
    Soft tissue (no. with + BM7 (1) 
    Bone 
    Hematologic (no. with + BM4 (3) 
    Nasopharyngeal carcinoma 
CharacteristicTotal
No. evaluable patients (enrolled)* 29 (46) 
Sex (M/F) 19/10 
Median age, y (range) 12 (2-17) 
Previous cytotoxic therapy  
    No. radiation courses  
        0 
        1 13 
        2 
    No. chemotherapy regimens  
        1-3 24 
        4-6 
        8 
Disease or site  
    CNS 10 
    Sympathetic nervous system 
    Soft tissue (no. with + BM7 (1) 
    Bone 
    Hematologic (no. with + BM4 (3) 
    Nasopharyngeal carcinoma 

Abbreviations: CNS, central nervous system; BM, bone marrow.

*

Twenty-nine of 46 patients enrolled received a full course of therapy (at least four cycles) or experienced a dose-limiting toxicity and were, therefore, eligible for inclusion in the evaluation of toxicity or response. Numbers reflect characteristics of the 29 evaluable patients.

No. with + BM = no. with disease involving the bone marrow.

One patient with Ki-1 non-Hodgkin's lymphoma.

Toxicity. No dose-limiting toxicities were observed in the four evaluable patients at 24 mg/m2. However, there was one death on study at this dose level. A 7-year-old patient with medulloblastoma and a history of syndrome of inappropriate antidiuresis experienced symptoms related to increased intracranial pressure and tumor progression after cycle 5 of vinorelbine. This was documented as a serious adverse event but was found to be related to tumor progression. One patient had a 33% dose reduction for course 4 secondary to persistent grade 2 neutropenia and was removed from the study thereafter due to progressive disease.

Dose-limiting toxicities observed in patients without bone marrow involvement at ≥30 mg/m2 were neutropenia and leukopenia. These toxicities were dose-dependent and cumulative with number of cycles administered but reversible (Table 2). Thrombocytopenia and anemia were mild to moderate in most patients and did not seem to be dose-dependent. All but one patient without bone marrow involvement had at least grade 1 to 2 anemia over the course of their treatment. Two patients had grade 4 anemia of short duration (<7 days). These two cases were complicated by progressive pulmonary involvement of metastatic Ewing sarcoma in one patient and by a history of anemia and preexisting grade 2 anemia at enrollment in the other. One patient had grade 4 thrombocytopenia lasting >7 days, which was associated with progressive disease.

Table 2.

Hematologic toxicity by dose level (highest-grade National Cancer Institute common toxicity criteria v. 1.0)

TotalDose level
1
2*
3
4
24 mg/m230 mg/m237.5 mg/m233.75 mg/m2
Toxicity 
Leukopenia 10 
Neutropenia 15 
Thrombocytopenia 16 
Anemia 
TotalDose level
1
2*
3
4
24 mg/m230 mg/m237.5 mg/m233.75 mg/m2
Toxicity 
Leukopenia 10 
Neutropenia 15 
Thrombocytopenia 16 
Anemia 
*

Four of nine evaluable patients had bone marrow involvement and were not eligible for evaluation of hematologic toxicity.

Intermediate dose level initiated subsequent to dose level 3.

The most common nonhematologic toxicities were grade 1 to 2 nausea and vomiting and increases in hepatic transaminases (Table 3). The majority of the gastrointestinal toxicities occurred after administration of the initial oral dose. Grade 3 to 4 increases in transaminases occurred in four patients, one at each dose level. Three cases were in patients without bone marrow involvement. Two of these resolved to grade 1 within 7 days and another was complicated by concurrent grade 3 hyperbilirubinemia due to progressive disease and was not thought to be drug related. The fourth case, in a patient with leukemia, was complicated by massive hepatosplenomegaly and suspected tumor lysis syndrome. Grade 4 hyperbilirubinemia was documented in one patient with leukemia at 30 mg/m2. This toxicity was observed concurrently with documented sepsis and an active lung abscess, which complicated assessment of the relation to study drug. Grade 4 mucositis was observed in two patients. Both cases were likely related to vinorelbine administration. One occurred concomitantly with grade 4 hematologic toxicity following oral administration (cycle 1) in a patient at 37.5 mg/m2 and resolved within 7 days. This patient did not continue therapy. The other occurred in a patient with leukemia at 30 mg/m2 after cycle 3, who refused treatment thereafter and was removed from study.

Table 3.

Major nonhematologic toxicities by dose level

Dose level (no. patients)
GradeTotal (N = 29)
1
2
4*
3
n (%)24 mg/m2 (n = 4)30 mg/m2 (n = 9)33.75 mg/m2 (n = 8)37.5 mg/m2 (n = 8)
Nausea/vomiting 8 (28) 
     
Increased transaminases 7 (24)  
   
     
Fever 6 (21)   
     
     
Diarrhea 5 (17)  
     
     
Headache 4 (14)   
     
Abdominal pain 3 (10)   
Increased bilirubin 3 (10)    
     
     
Cough 3 (10)   
     
Mucositis 2 (7)   
Dose level (no. patients)
GradeTotal (N = 29)
1
2
4*
3
n (%)24 mg/m2 (n = 4)30 mg/m2 (n = 9)33.75 mg/m2 (n = 8)37.5 mg/m2 (n = 8)
Nausea/vomiting 8 (28) 
     
Increased transaminases 7 (24)  
   
     
Fever 6 (21)   
     
     
Diarrhea 5 (17)  
     
     
Headache 4 (14)   
     
Abdominal pain 3 (10)   
Increased bilirubin 3 (10)    
     
     
Cough 3 (10)   
     
Mucositis 2 (7)   
*

Intermediate dose level initiated subsequent to dose level 3.

Headaches, fever, diarrhea, abdominal pain, and cough occurred in three to six patients each. Other toxicities noted in one to two patients each included constipation, hives, fatigue, increased blood urea nitrogen, tachypnea, hematuria, bone pain, and phlebitis at the injection site. Only two patients (one with a sympathetic nervous system tumor and one with a soft tissue sarcoma) had peripheral neuropathy of grade 2 to 3 severity, one at 30 mg/m2 and another at 33.75 mg/m2. Both cases were considered by the treating physician to be disease related and not attributed to the study drug.

At 30 mg/m2, mean (±SD) ANC nadirs were 3.1 ± 1.7 and 2.3 ± 3.2 103/μL at cycle 1 (n = 5) and cycle 6 (n = 4), respectively. Three of five patients evaluable for hematologic toxicity (four had bone marrow involvement and were not eligible for this evaluation) developed grade 4 neutropenia or leukopenia that lasted >7 days requiring a 33% dose reduction or a cycle to be skipped. In two of these patients, factors other than study drug were judged to have contributed to the toxicity. One was complicated by use of multiple antibiotic medications for fever and concomitant i.v. acyclovir for a shingles infection, requiring two cycles (cycles 3 and 4) to be skipped. The other occurred after completion of the last (sixth) cycle in a patient with reduced bone marrow reserves due to extensive prior therapy. Grade 4 neutropenia resolved within 14 days in all patients at this dose level.

At 37.5 mg/m2, neutropenia was dose limiting and the maximum tolerated dose for patients without bone marrow involvement was exceeded. Mean (±SD) ANC nadirs were 1.8 ± 1.5 and 1.2 ± 0.9 103/μL at cycle 1 (n = 6) and cycle 6 (n = 6), respectively. Four of eight evaluable patients had drug-related grade 4 neutropenia lasting >7 days, which required treatment delays or dose reductions. Three of these cases were observed at cycle 3 or later. One patient had treatment delayed for cycle 2 and 3. In all but one case, these grade 4 neutropenias resolved by day 14. In one patient, grade 4 neutropenia lasted 17 days after cycle 3 requiring 1-week delays in cycles 4 and 5 and a dose reduction at cycle 6.

Six of the eight evaluable patients at this dose level completed the six cycles of therapy. One patient was removed from protocol therapy after cycle 4 because of progressive disease. The other was removed from study following a single oral dose of vinorelbine. This patient was a 14-year-old male with extraosseous Ewing sarcoma/primitive neuroectodermal tumor who presented with grade 4 neutropenia and grade 4 mucositis and vomiting beginning 3 and 7 days, respectively, after a 150 mg (113 mg/m2) oral dose. Exacerbation of preexisting congestive heart failure due to dehydration was also observed in this patient. The event was considered partly related to vinorelbine administration. The patient recovered without sequelae but was removed from the study due to progressive disease and toxicity.

At the intermediate level (33.75 mg/m2), mean ANC nadirs were 3.6 ± 2.9 at cycle 1 (n = 7) and 1.0 ± 0.3 103/μL at cycle 6 (n = 4). Three of eight evaluable patients had grade 4 neutropenia and one had grade 4 thrombocytopenia of >7 days duration but all resolved within 14 days. The grade 4 thrombocytopenia was thought to be related to progressive disease. Five patients required either delays in subsequent cycles and/or a dose reduction. One patient each required a subsequent 66% dose reduction or had treatment delayed at cycle 6 from continued neutropenia. All but one received at least six cycles of therapy at this dose level. The other was removed from study after cycle 4 due to progressive disease.

Pharmacokinetics. Blood was collected to evaluate population pharmacokinetics in pediatric patients and was not mandatory. Samples were collected for analysis in a total of 31 patients over four dose levels. Pharmacokinetic evaluations could be done in 26 and 20 patients after i.v. and oral administration, respectively. Pharmacokinetic data were insufficient to determine reliable pharmacokinetic variables in 5 of 31 patients receiving i.v. drug. Pharmacokinetic data obtained following oral administration were insufficient in one patient. Three patients did not receive vinorelbine orally because of difficulty swallowing pills, herpetic stomatitis, or active oral candida infection. An additional seven did not receive oral vinorelbine due to early completion of the oral trial after indications of poor bioavailability from this formulation. The oral dosage form was generally well tolerated, with mild to moderate nausea, vomiting, or diarrhea observed in 40% of patients receiving an oral dose.

Pharmacokinetic variables for vinorelbine were determined from concentration-time data for oral and i.v. vinorelbine using noncompartmental analysis (Table 4). In each case (oral or i.v.), the plasma disposition curves were triphasic. Cmax was an average 7-fold higher after i.v. versus oral administration (691.7 and 89.5 ng/mL), respectively. The mean Tmax for oral vinorelbine was reached within 2 hours (range, 0.25-6 hours), with a mean oral fraction absorbed of 0.29 ± 0.23. The fraction absorbed was highly variable (coefficient of variation, 78.9%). Drug exposure (area under the plasma concentration-time curve) also varied within each dose level. Mean i.v. area under the plasma concentration-time curve at 37.5 mg/m2 was 1.5-fold that following equivalent oral doses (113 mg/m2). Mean oral and i.v. exposures were nearly equivalent at 30 mg/m2 i.v. (90 mg/m2 oral).

Table 4.

Mean ± SD pharmacokinetic variables for i.v. and orally administered vinorelbine

Parameteri.v. (n = 26)Oral (n = 20)
Dose, mg/m2 (range) 24-37.5 90-113 
Total dose, mg (range) 38 (17-66) 120 (50-200) 
Cmax (ng/mL) 691.7 ± 598.7 89.5 ± 65.3 
Tmax (h)  1.92 ± 1.74 
AUC0-∞ (ng/mL-h)* 676 ± 495 542 ± 437 
F  0.29 ± 0.23 
Cl (L/h/kg) 1.75 ± 1.0 12.1 ± 13.0 
t1/2 (h) 16.5 ± 9.7 10.3 ± 10.0 
Vdss (L/kg) 21.1 ± 12.2 69.4 ± 30.6 
Parameteri.v. (n = 26)Oral (n = 20)
Dose, mg/m2 (range) 24-37.5 90-113 
Total dose, mg (range) 38 (17-66) 120 (50-200) 
Cmax (ng/mL) 691.7 ± 598.7 89.5 ± 65.3 
Tmax (h)  1.92 ± 1.74 
AUC0-∞ (ng/mL-h)* 676 ± 495 542 ± 437 
F  0.29 ± 0.23 
Cl (L/h/kg) 1.75 ± 1.0 12.1 ± 13.0 
t1/2 (h) 16.5 ± 9.7 10.3 ± 10.0 
Vdss (L/kg) 21.1 ± 12.2 69.4 ± 30.6 

Abbreviations: Cmax, peak plasma concentration; Tmax, time to Cmax; AUC, area under the plasma concentration-time curve; F, fraction absorbed; Cl, plasma clearance; t1/2, half-life of terminal phase; Vdss, steady-state volume of distribution.

*

At 30 mg/m2 i.v. and 90 mg/m2 orally.

Apparent oral clearance.

Oral and i.v. vinorelbine doses were distributed widely with a mean oral apparent steady-state volume of distribution of 69.4 ± 30.6 L/kg and i.v. volume of distribution of 21.1 ± 12.2 L/kg. Mean plasma clearance after i.v. administration was 1.75 ± 1.0 L/h/kg. When stratified by dose level, means for i.v. ClTB were 0.99 ± 0.2 L/h/kg (24 mg/m2, n = 3), 1.91 ± 1.0 L/h/kg (30 mg/m2, n = 16), 2.1 ± 0.78 L/h/kg (33.75 mg/m2, n = 3), and 1.42 ± 1.1 L/h/kg (37.5 mg/m2, n = 4). Mean apparent oral clearance was substantially greater (12.1 ± 13.0 L/h/kg) than clearance after i.v. administration. The mean terminal half-life for i.v. and oral doses were 16.5 ± 9.7 and 10.3 ± 10.0 hours, respectively.

Antitumor response. Among 22 patients eligible for evaluation of response, disease control was maintained in five patients without bone marrow involvement who received the study prescribed course of six cycles of treatment. These patients continued treatment off protocol, on a compassionate basis, following removal from the study. A partial response was observed in a patient at the 33.75 mg/m2 dose level who had recurrent rhabdomyosarcoma and completed 16 weeks of therapy before progressive disease. This patient had a very high exposure to vincristine before the trial of vinorelbine.

Four patients had stable disease. One patient with astrocytoma who received 36 weekly cycles of vinorelbine at 37.5 mg/m2 had stable disease for 3 years. Two others, one with Ki-1 non-Hodgkin lymphoma (33.75 mg/m2) and another with meningioma (30 mg/m2), received 36 cycles and had disease control maintained during those periods. One heavily pretreated patient with recurrent ependymoma received 15 courses of vinorelbine at 24 mg/m2 with stable disease that lasted 17 weeks.

Vinca alkaloids are an important part of many curative pediatric chemotherapeutic regimens. When new agents from this class of drugs become available and have possible benefits, the pediatric population is an obvious place for their development.

Phase I and II studies of vinorelbine in adult cancer patients identified the dose-limiting toxicity of vinorelbine as hematologic, specifically leukopenia and neutropenia (49, 16). The i.v. single-dose maximum tolerated dose for vinorelbine in those studies was 30 to 36 mg/m2/wk. Our study sought to determine the safety and maximum tolerated dose of single-agent vinorelbine given i.v. in pediatric patients with previously treated leukemia, lymphoma, or solid tumors. We also attempted to determine the oral bioavailability and pharmacokinetic behavior of vinorelbine given orally and i.v. in children.

Myelosuppression was the dose-limiting toxicity in this study in patients without bone marrow involvement. Similar to that reported in adult trials, neutropenia occurred in >90% of those patients and was moderate to severe (grades 3 to 4) in 72% of patients. At doses ≥30 mg/m2, recovery from grade 4 neutropenia took between 7 and 14 days, which should allow weekly × 6 dosing. Dose-limiting toxicity was observed at 37.5 mg/m2 where half the patients (four of eight) eligible for evaluation of toxicity had prolonged myelosuppression (recovery to grade 3 in >7 days). In this multicenter trial, the determination of whether neutropenia was drug-related or disease-related was made by the treating physician. In the first cohort enrolled at 37.5 mg/m2, neutropenia was confirmed as drug-related in two of four patients only after enrollment of a second four-patient cohort. These finding instigated establishment of an intermediate dose level to further refine the maximum tolerated dose. Six of eight patients subsequently entered at 33.75 mg/m2 had grade 4 neutropenia, with four who experienced a delay in hematologic recovery >7 days. The maximum tolerated dose in this trial was reached at 30 mg/m2 for patients without disease in the marrow.

The mean total body clearance following i.v. administration in this pediatric population was ∼40% higher than that reported in most studies in adult cancer patients who received similar doses (1720). In the current study, the largest individual group in whom pharmacokinetic analysis could be done was at the 30 mg/m2 dose level (n = 16). At this dose level, ClTB was 1.91 ± 1.0 L/h/kg compared with 1.0 to 1.3 L/h/kg reported for adults who received that dose (1820). In our study, systemic exposure was lower at a given dose compared with that reported in previous adult studies but was proportionate to clearance. The calculated area under the plasma concentration-time curve in this study at 30 mg/m2 was 676 ng/mL-h compared with 800 to 900 ng/mL-h reported in adult patients (19, 20). Thus, we believe that a weekly schedule of i.v. vinorelbine is safe and provides adequate dose intensity.

Oral vinorelbine was generally well-tolerated in this pediatric population. Only 9 of 23 patients who received initial oral doses of vinorelbine had gastrointestinal toxicity, which was primarily characterized as grades 1 to 3 nausea/vomiting or grade 1 diarrhea. In this study, 10 and 40 mg liquid-filled gelatin capsules were used to administer an oral dose of vinorelbine at thrice the i.v. dose. That dosage form showed low bioavailability (29%) in our population. Whereas adult trials have reported an apparent oral clearance of 0.43 to 1.45 L/h/kg (13, 14, 21), mean apparent oral clearance in this study was substantially higher at 12.0 ± 13 L/h/kg. The high oral clearance we observed was likely caused by poor bioavailability of this formulation in children, resulting in a large value estimated for apparent volume of distribution. The mean oral steady-state volume of distribution was 69.4 ± 30.6 L/kg, which is ∼1-fold larger than that reported in adults. The relatively low and highly variable bioavailability of the oral capsules in this trial precluded further development of that dosage formulation.

Vinca alkaloids are important components in primary treatment regimens of pediatric malignancies. Weekly vinorelbine administration seems to have manageable toxicity in the pediatric population, with an maximum tolerated dose of 30 mg/m2 i.v. in patients without bone marrow disease involvement. A sufficient number of patients were not evaluable for establishment of an maximum tolerated dose in the marrow involvement arm. We have shown that treatment with vinorelbine has resulted in prolonged disease control among patients with prior exposure to Vinca alkaloids, specifically vincristine. This suggests to us that vinorelbine may show activity among patients who have tumors refractory to treatment regimens containing Vinca alkaloids. Using the current weekly schedule, the subsequent phase II trial is testing vinorelbine among such patients with relapsed sarcoma, neuroblastoma, and primary brain tumors. We hope to further define a therapeutic role for vinorelbine in pediatric malignancies.

Grant support: National Cancer Institute grant U10 CA13539.

Note: This study was conducted as a Children's Cancer Group (now Children's Oncology Group) study by the Phase I Working Group.

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

1
Takasugi BJ, Jones SE, Robertone AB. Phase II trial of vinzolidine, an oral Vinca alkaloid in Hodgkin's disease and non-Hodgkin's lymphoma.
Cancer Treat Rep
1984
;
68
:
1399
–401.
2
Budman DR, Schulman P, Marks M, et al. Phase I trial of vinzolidine.
Cancer Treat Rep
1984
;
68
:
979
–82.
3
Sarna G, Mitsuyasu R, Figlin R, Ambersley J, Groopman J. Oral vinzolidine as therapy for Kaposi's sarcoma and carcinomas of lung, breast, and colon/rectum.
Cancer Chemother Pharmacol
1985
;
14
:
12
–4.
4
Tortoriello A, Facchini G, Caponigro F, et al. Phase I/II study of paclitaxel and vinorelbine in metastatic breast cancer.
Breast Cancer Res Treat
1998
;
47
:
91
–7.
5
Depierre A, Lemarie E, Dabouis G, Garnier G, Jacoulet P, Dalphin JC. Efficacy of Navelbine (NVB) in non-small cell lung cancer (NSCLC).
Semin Oncol
1989
;
16
:
26
–9.
6
Winer EP, Chu L, Spicer DV. Oral vinorelbine (Navelbine) in the treatment of advanced breast cancer.
Semin Oncol
1995
;
22
:
72
–9.
7
Miller VA, Krug LM, Ng KK, et al. Phase II trial of docetaxel and vinorelbine in patients with advanced non-small-cell lung cancer.
J Clin Oncol
2000
;
18
:
1346
–50.
8
Palmeri S, Leonardi V, Gebbia V, et al. Gemcitabine plus vinorelbine in stage IIIB or IV non-small cell lung cancer (NSCLC): a multicentre phase II clinical trial.
Lung Cancer
2001
;
34
:
115
–23.
9
Canobbio L, Boccardo F, Pastorino G, et al. Phase-II study of Navelbine in advanced breast cancer.
Semin Oncol
1989
;
16
:
33
–6.
10
Hanley ML, Elion GB, Colvin OM, et al. Therapeutic efficacy of vinorelbine against pediatric and adult central nervous system tumors.
Cancer Chemother Pharmacol
1998
;
42
:
479
–82.
11
Fellous A, Ohayan R, Vacassin T, et al. Biochemical effects of Navelbine on tubulin and associated proteins.
Semin Oncol
1989
;
16
:
9
–14.
12
Ngan VK, Bellman K, Hill BT, Wilson L, Jordan MA. Mechanism of mitotic block and inhibition of cell proliferation by the semisynthetic Vinca alkaloids vinorelbine and its newer derivative vinflunine.
Mol Pharmacol
2001
;
60
:
225
–32.
13
Rahmani R, Zhou XJ, Bore P, Cantfort JV, Focan C, Cano JP. Oral administration of [3H]navelbine in patients: comparative pharmacokinetics using radioactive and radioimmunologic determination methods.
Anticancer Drugs
1991
;
2
:
405
–10.
14
Rowinsky EK, Noe DA, Trump DL, et al. Pharmacokinetic, bioavailability, and feasibility study of oral vinorelbine in patients with solid tumors.
J Clin Oncol
1994
;
12
:
1754
–63.
15
Rowinsky EK, Lucas VS, Hsieh AL, et al. The effects of food and divided dosing on the bioavailability of oral vinorelbine.
Cancer Chemother Pharmacol
1996
;
39
:
9
–16.
16
Mathe G, Reizenstein P. Phase I pharmacologic study of a new Vinca alkaloid: Navelbine.
Cancer Lett
1985
;
27
:
285
–93.
17
Schilling T, Fiebig HH, Kerpel-Fronius S, et al. Clinical phase I and pharmacokinetic trial of vinorelbine administered as single intravenous bolus every 21 days in cancer patients.
Invest New Drugs
1996
;
14
:
371
–8.
18
Jehl F, Quoix E, Leveque D, et al. Pharmacokinetic and preliminary metabolic fate of navelbine in humans as determined by high performance liquid chromatography.
Cancer Res
1991
;
51
:
2073
–6.
19
Marquet P, Lachatre G, Debord J, Eichler B, Bonnaud F, Nicot G. Pharmacokinetics of vinorelbine in man.
Eur J Clin Pharmacol
1992
;
42
:
545
–7.
20
Sabot C, Marquet P, Debord J, Carpentier N, Merle L, Lachatre G. Bayesian pharmacokinetic estimation of vinorelbine in non-small-cell lung cancer patients.
Eur J Clin Pharmacol
1998
;
54
:
171
–5.
21
Zhou XJ, Bore P, Monjanel S, et al. Pharmacokinetics of navelbine after oral administration in cancer patients.
Cancer Chemother Pharmacol
1991
;
29
:
66
–70.