Vinorelbine in previously treated advanced childhood sarcomas

Evidence of activity in rhadbomyosarcoma




Vinca alkaloids have proved active against a number of pediatric malignancies. The aim of this study was to assess the feasibility and effectiveness of using vinorelbine in previously treated pediatric patients with advanced sarcomas.


From September 1998 to August 2001, 33 previously treated patients with progressive sarcoma were treated: 13 had rhabdomyosarcomas, 5 had other soft tissue sarcomas, 9 had Ewing sarcomas, and 6 had osteosarcomas. Vinorelbine was given intravenously on Days 1 and 8 of a 21-day treatment cycle. Four patients with uncontrolled pain or central nervous system invasion received concurrent radiotherapy and were only evaluated for toxicity.


One hundred seventy-eight treatment cycles were administered (median of four per patient, range 1–20). Grade 3 to 4 neutropenia occurred in 63% of patients, Grade 3 anemia in 9%, and Grade 3 thrombocytopenia in 3%. Nonhematological toxicity was mild or moderate, i.e., always lower than Grade 3, with the exception of one child who experienced paralytic ileus. Twenty-eight patients were assessable for response. Eight patients had a partial response, one patient had a minor response, and nine patients had stable disease. Objective responses were observed in 6 of 12 patients with rhadbomyosarcomas (five of six of the alveolar subtype), in one of five patients with osteosarcomas, and in one of seven patients with Ewing sarcomas. The median duration of response was 10 months (range, 3+ to 20).


Vinorelbine has a favorable toxicity profile with evidence of biological activity in already heavily treated pediatric patients with sarcomas. In particular, the objective response rate obtained for patients with alveolar rhabdomyosarcoma seems very promising. Due to the few cases considered here, further Phase II studies are needed to establish a potential role of vinorelbine in the treatment of these tumors. Cancer 2002;94:3263–8. © 2002 American Cancer Society.

DOI 10.1002/cncr.10600

Vinorelbine is a semisynthetic vinca alkaloid synthesized from catharantine and vindoline extracted from Vinca rosea leaves. It differs from other vinca alkaloids in the substitution on the catharantine rather than on the vindoline ring of the vinca alkaloid molecule. Like other alkaloids, however, vinorelbine induces cytotoxicity by inhibiting microtubule assembly. However, it may be a more selective inhibitor of the mitotic than of the neural axonal microtubule formation, suggesting a lesser likelihood of neurotoxicity.1 In adult Phase I clinical trials, the dose-limiting toxic reactions were neutropenia and leukopenia, with peripheral neuropathy and constipation at higher doses. The recommended dose for further clinical evaluation was 30 mg/m2 per week.2, 3 For patients with substantial exposure to previous chemotherapy or radiation, it has been4 suggested that single-agent therapy start at a lower dosage. The Phase I trial performed by the Children's Cancer Group demonstrated that the spectrum of hematologic and other types of toxicity in pediatric patients was much the same as in adults and that the systemic clearance of vinorelbine is substantially higher in children.5 Vinorelbine has undergone clinical evaluation in a number of malignancies,6 but little experience is available on its use in pediatric patients with sarcomas. Fidias et al.7 reported their experience at the Dana-Faber Cancer Institute in Boston with 38 adult sarcoma patients. They observed one complete response in a patient suffering from angiosarcoma with pulmonary metastases, three mixed responses, and six cases of stable disease. The Italian Cooperative Group on AIDS and Tumors found that vinorelbine is a safe and effective drug in the treatment of patients with AIDS-related Kaposi sarcoma previously treated with one or more chemotherapy regimens, leading to an overall objective response rate of 43%.8 Epelman et al.9 reported a Phase II study of 12 cases of refractory or recurrent pediatric cancer with a response rate of 35%.9 In 1999, at the SIOP (International Society of Paediatric Oncology) meeting, the same group presented the results of a Phase II study of 17 patients with rhabdomyosarcoma or other soft tissue sarcomas, reporting an impressive response rate of 47% (five patients had complete remission of disease [four with rhabomyosarcomas and one with angiosarcoma] and three had partial remission of disease).10

To establish the potential role of vinorelbine in the treatment of pediatric sarcomas, this study was conducted on advanced sarcomas observed over a 3-year period at the Pediatric Unit of the Istituto Nazionale Tumori of Milan. Vinorelbine was administered at a dose of 30 mg/m2 on Days 1 and 8 every 3 weeks, given that all patients enrolled had previously been heavily treated with chemotherapy and radiotherapy.


Patients with histologically proven sarcoma were eligible for the study when they were shown to have advanced, progressive measurable disease for which there was no known cure. Patients had to have demonstrated recovery from all toxicities associated with any previous therapy, with a minimum of 2 weeks since the last dose of chemotherapy (4 weeks in the case of therapy containing nitrosurea). At least 6 months should have elapsed since bone marrow transplantation or radiation therapy using fields that included the craniospinal axis, spine, and/or greater than 50% of the bony pelvis. Other eligibility criteria included Eastern Cooperative Oncology Group performance status of 2 or less and a life expectancy of at least 8 weeks. Normal liver function (bilirubin level < 1.5 × normal and alanine aminotransferase and aspartate aminotransferase < 2.5 × normal) and normal renal function for age were required. The absolute granulocyte count had to be 1000 or more per microliter and the platelet count had to be 100,000 or more per microliter.

Written informed consent conforming to institutional and national guidelines, indicating that patients and/or their parents or legal guardians were aware of the investigational nature of the study, was also required.

The patients' characteristics are listed in Table 1. From September 1998 to August 2001, 33 children eligible for the study were treated at our center. Of the 33 children, 13 had rhabdomyosarcomas (six embryonal type and seven alveolar type), 5 had other soft tissue sarcomas, 9 had Ewing sarcomas, and 6 had osteosarcomas. The patients' median age was 16 years (range 2–29). All patients showed progressive disease at accrual.

Table 1. Patients' Characteristics at Baseline
CharacteristicNo. of patients
Total no. of patients33
Age (yrs) 
 Ewing sarcoma9
 Synovial sarcoma3
No. of lines of previous chemotherapy regimens 
Disease localizations 
 Soft tissue12
 Lymph nodes5

There was a median of two previous chemotherapy regimens (range 1–4) for all patients. Fourteen patients (42%) had undergone previous high-dose chemotherapy with stem cell support (seven with rhabdomyosarcomas and seven with Ewing sarcomas). All 33 patients had received previous chemotherapy with vincristine. Twenty-seven patients (82%) had also been given radiotherapy in their previous treatment, including hemi-body irradiation in one case.

Vinorelbine was given intravenously at a dose of 30 mg/m2 on Days 1 and 8. The cycle was repeated every 3 weeks. The drug was given on an outpatient basis, diluted in isotonic solution to a concentration between 1.5 and 3 mg/dL, and infused over 6–10 minutes into either a large central vein or a free-flowing infusion of 0.9% sodium chloride or 5% dextrose into a fixed peripheral venous infusion device. In patients who received vinorelbine in a peripheral vein, the vein was then flushed with a rapid infusion of at least 75–125 mL of normal saline solution to reduce the risk of chemical phlebitis.

Dose modifications were based on observed toxicity. A hemogram was obtained every week and neutrophil and platelet counts of 1.0 and 100 × 109cells per liter, respectively, were required before each new injection. Vinorelbine administration was delayed until hematologic recovery; the dose was subsequently reduced by 25% if two consecutive injections had to be deferred. The same rule applied in cases of neutropenic fever or nonhematologic toxicity higher than Grade 2. The treatment was abandoned in the event of severe neuropathy or ileus to avoid persistent detrimental effects on the patient's quality of life.

The treatment was repeated as planned until progression of disease or unacceptable toxicity (as defined above) was observed, or until it was discontinued at the patient's request. Adverse effects were evaluated according to the first version of the National Cancer Institute's common toxicity criteria,11 based on clinical and laboratory evaluations at each cycle. World Health Organization criteria were used to assess partial response (PR) and complete response. Minor response (MR) and stable disease (SD) were also considered. MR was defined as a 25–50% decrease in the product of the diameters of measurable target lesions. SD was defined as a less than 25% decrease or increase in the measurable lesion lasting at least 2 months. The response rate was calculated with its exact binomial confidence interval (CI). The duration of response was measured from the beginning of treatment until progression of disease became evident. Progression-free survival was analyzed from the beginning of treatment until the patient either had documented disease progression or died.

The purpose of this study was to assess the feasibility and efficiency of vinorelbine administered according to the above schedule in pediatric patients with advanced sarcomas. The primary endpoints of the trial were to determine the objective response rate (for the whole series and in different histology groups), the delivered dose intensity expressed as a percentage of the intended schedule (60 mg/m2 every 3 weeks), and the toxicity of vinorelbine as a single agent in these already heavily treated patients. Secondary endpoints were time to treatment failure and survival.


Treatment Exposure

In all, 178 treatment cycles were administered. A median of four cycles (range 1–20) were administered per patient. In six patients, treatment was still being administered 3, 3, 3.5, 4, 7, and 11 months after initiation. It was terminated early due to chemotherapy-related adverse events in only one patient, who experienced paralytic ileus after the first cycle (the patient recovered completely after discontinuing the chemotherapy, however). Eight patients (24%) received two cycles or less due to early progression of disease. Because of delays and dose reductions, only 16 patients (48%) received the treatment as scheduled. The dose intensity was between 99% and 90%in four patients, between 89% and 75% in seven patients, and less than 75% but greater than 50% in six patients.

Concurrent radiotherapy was administered to four patients. This was due to brain metastases in two cases (one child with alveolar rhabdomyosarcoma, the other with synovial sarcoma) and to pain in bone sites in two cases (both with metastatic Ewing sarcoma).


All 33 treated patients were assessed for toxicity. Grade 3 and 4 neutropenia occurred in 9 (27%) and 12 (36%) patients, respectively. Grade 4 neutropenia was generally short-lived (median 3 days) and only 8 (24%) patients received granulocyte–colony-stimulating factor for febrile neutropenia. None required hospitalisation. Grade 3 to 4 neutropenia was evident in 7 of the 14 patients (50%) who had previously received high-dose chemotherapy with stem cell rescue, and in 74% of those who had not. Grade 3 anemia was observed in three (9%) patients, with two requiring red blood cell transfusion. Thrombocytopenia was rare. Grade 3 toxicity on platelets was observed in one patient with bone marrow involvement and Grade 2 thrombocytopenia was observed in another two cases. Hematologic toxicity was not cumulative and was readily reversible on discontinuation of the drug. Grade 1 to 2 constipation with abdominal pain was observed in nine patients (apart from the child who developed a paralytic ileus), eight patients developed Grade 1 to 2 jaw pain, and seven patients had a flu-like syndrome. Mucositis and myalgia were less frequent (9% and 6%, respectively) and were always no higher than Grade 2. No life-threatening adverse effects were observed. Phlebitis developed in 5 of 16 patients (31%) who had vinorelbine injected in a peripheral vein. Alopecia never occurred. There was no evidence of any increment in toxicity in patients receiving concurrent radiotherapy.


As of November 1, 2001, 28 patients were assessable for response. Five patients were not evaluable: one had discontinued the chemotherapy after the first cycle due to its toxicity and four received concurrent radiotherapy.

Overall, there were eight patients with PR (28%), one with MR, and nine with SD lasting more than 2 months. The characteristics of the responders are shown in Table 2. Ten patients showed progression of disease after one to three cycles (median 2). The median duration of PR was 10 months (range 3+ to 20), the MR lasted 2.5 months, and the median duration of SD was 3.5 months (range 2.5–6). The median duration of progression-free survival for all patients was 3.5 months (range 0.5–20), whereas the median overall survival was 6 months (range 1.5–24+ months), with eight patients surviving more than a year after starting treatment.

Table 2. Characteristics of Responding Patients
GenderAgeHistologyNo. of previous CT regimensDisease sitesDuration of response (mos)Status
  1. CT = chemotherapy; OS: osteosarcoma; ES: Ewing sarcoma; RMS-A: alveolar rhabdomyosarcoma; RMS-E: embryonal rhabdomyosarcoma; CR: complete response; AWD: alive with disease; NED: no evidence of disease; DOD: dead of disease.

M13OS3Lung20Postsurgical CR, AWD at 24 mos
F27ES2Bone/soft tissue4+Under treatment
F20RMS-A2Soft tissue15Postradiotherapy CR, AWD at 20 mos
F17RMS-A2Soft tissue13Postradiotherapy CR, DOD at 20 mos
M13RMS-E1Liver/soft tissue/lymph nodes9DOD at 14 mos
F9RMS-A2Soft tissue11+Postsurgical CR, NED at 11+ mos
F24RMS-A3Soft tissue/lymph nodes4AWD at 5 mos
M29RMS-A2Lung/soft tissue/lymph nodes3.5+Under treatment

Table 3 shows responses to the treatment according to histology. Of the 12 rhabdomyosarcoma patients assessable for response, 6 had an objective response (response rate 50% with exact 95% CI: 21–79%), i.e., in five of the six patients with alveolar rhabdomyosarcoma (83%, exact 95% CI: 36–99%) and in one of the six patients with embryonal subtype. Two patients with previously unresectable disease underwent complete macroscopic resection after achieving a partial remission of disese with four and eight cycles of treatment, respectively. One of them was a boy with osteosarcoma who underwent pulmonary metastasectomy, the other was a girl with rhabdomyosarcoma of the leg who received compartmental surgery. Both of these patients subsequently received consolidation radiotherapy. Two other patients with rhabdomyosarcoma achieved complete remission with radiotherapy after obtaining a PR to vinorelbine; these two girls had a primary tumor at the extremities and multiple subcutaneous nodules. Chemotherapy with vinorelbine was not discontinued during the administration of radiation therapy. All four patients who achieved complete remission (with either surgery or radiotherapy) received vinorelbine in an adjuvant setting. At the time of this report, one was still receiving treatment 11 months after it was first implemented. In the other three, vinorelbine was discontinued after 10, 13, and 16 months, respectively, of therapy. They had recurrence of disease 2–4 months after discontinuation of chemotherapy.

Table 3. Response to Vinorelbine According to Histology
ResponseRhabdomyosarcoma (%)Other soft tissue sarcoma (%)Ewing sarcoma (%)Osteosarcoma (%)Total
  1. PR: partial response; MR: minor response; SD: stable disease; PD: progressive disease.

PR6 (50)1 (14.3)1 (20)8 (28.6)
MR1 (8.3)1 (3.6)
SD1 (8.3)3 (75)3 (42.8)2 (40)9 (32.1)
PD4 (33.3)1 (25)3 (42.8)2 (40)10 (35.7)
Total12 (100)4 (100)7 (100)5 (100)28 (100)

Among the four patients not assessable for response because they received concomitant radiotherapy were two boys, one with brain metastases from a synovial sarcoma and one with multiple bone metastases from a Ewing sarcoma. In the first boy, his disease is in complete remission after treatment with panencephalic radiotherapy and concomitant vinorelbine. He is still receiving treatment 7 months after it was initiated. In the second boy, his disease remained stable for 13 months.


The management of children and adolescents with disseminated or recurrent sarcomas remains an unsolved problem for pediatric oncologists. This is particularly so in the case of rhabdomyosarcoma, which is the most common soft tissue sarcoma in pediatric patients.

With a multimodality therapeutic approach comprising surgery, multiagent chemotherapy, and radiotherapy, the outcome for patients with localized rhabdomyosarcoma has improved dramatically in the last three decades. Nowadays, more than 70% of children can be cured.12, 13 However, the prognosis for patients with extensive regional tumors, metastatic lesions, and recurrent disease remains unsatisfactory despite the intensification of therapy. Therefore, it is important to identify new agents to incorporate in multiagent chemotherapeutic protocols to improve the outcome for high-risk patients.

Clinical investigations in classic Phase II trials may underestimate the effectiveness of new agents when tested in already heavily treated patients with a lower tolerance of additional chemotherapy and with tumors resistant to multiple antineoplastic agents. Several new agents evaluated in patients with recurrent rhabdomyosarcoma and other pediatric sarcomas have demonstrated little effect. Unlike the classic Phase II trials, the up-front window therapy approach investigates new agents in previously untreated patients at high risk of treatment failure. That is, promising new drugs are assessed in patients with newly diagnosed tumors with a lesser likelihood of multiple drug resistance.

The large number of clinical responses achieved with vinorelbine in our patients with recurrent rhabdomysarcoma (particularly the alveolar histotype) is highly significant because it was obtained in already heavily treated patients. Moreover, all the patients had previously been treated with the analog vincristine, suggesting that there is no cross-resistance. Fifty percent of rhabdomyosarcoma patients and 83% of those with the alveolar subtype achieved a PR. This response rate is better than that obtained in previously treated patients with ifosfamide, melphalan, and topotecan, currently considered the most promising new agents for treating patients with rhabdomyosarcoma. Due to the small number of patients enrolled in our study, confidence limits of response rates are extremely wide, but the lower values could be of worth. The response rate to ifosfamide as a single agent was 18% in previously treated patients with rhabdomyosarcomas and 86% in up-front window therapy for untreated patients;14 for melphalan, it was 8% in previously treated patients and 76% in patients with newly diagnosed high-risk rhabdomyosarcomas;15 and for topotecan, it was 45% in untreated patients with metastatic rhabdomyosarcomas, whereas no response was observed in patients with refractory tumors.16, 17

As emphasized by Smith and Anderson,18 antitumor activity detected in previously treated patients with recurrent disease is potentially more significant than such activity assessed in window studies. The recurrent, treatment-resistant tumor cells are the target that must be destroyed by the new drug to improve outcome.

The number of cases in our trial is too small for any conclusion to be drawn and further studies are needed to confirm our apparently excellent results and to assess the use of vinorelbine in multiagent chemotherapy. However, our data are supported by the findings of Epelman et al.10 Disease stabilization (obtained in 32% of the group as a whole) is indirect and minor evidence of the antitumoral effect of vinorelbine, because all patients had documented disease progression at the start of the trial. The results obtained in patients with bone and soft tissue sarcomas other than rhabdomyosarcoma also warrant further investigations. The results reported by Fidias et al.7 and by Epelman et al.,10 who observed a complete response to vinorelbine in patients with angiosarcoma, and the data from the Italian Cooperative Group on AIDS and Tumors, which demonstrated that vinorelbine is effective in treating Kaposi sarcoma,8 should prompt investigation, particularly on vascular malignancies, which were not included in our series.

In our study, the response to the drug lasted a median of 10 months. Three patients with disease in complete remission (obtained with chemotherapy plus surgery or radiotherapy) had recurrence of disease 2–4 months after discontinuation of vinorelbine therapy. Similar data were reported by Fidias et al.7 for a patient with angiosarcoma who maintained complete remission of disease for 12 months. The disease recurred after treatment was suspended. We do not know whether this would suggest a role for vinorelbine in controlling microscopic residual disease via a cytostatic mechanism.

In our series, the toxicity profile is also favorable. Our data indicate that vinorelbine is safe and well tolerated in outpatients. As reported elsewhere, granulocytopenia was the dose-limiting toxic reaction. Although the incidence of Grade 3–4 neutropenia was high (63% of patients), it was not cumulative and was associated rarely with severe complications. Tolerance was also good in patients who had previously received high-dose chemotherapy with stem cell support and who might have been expected to have lower hematologic reserves. Cases of nonhematologic toxicities were mostly mild or moderate. Drug-associated neurotoxicity occurred less often than with other commonly used alkaloids compounds.

In conclusion, our study demonstrates that vinorelbine as a single agent is feasible and potentially effective in already heavily treated pediatric patients with recurrent sarcomas. In particular, it has a significant clinical effect in patients with alveolar rhabdomyosarcoma. The small numbers in our series limit the value of our results, suggesting that further trials are needed to confirm our data and ascertain the role of vinorelbine in the treatment of pediatric patients with rhabdomyosarcoma and other sarcomas.