Phase I study of fotemustine in pediatric patients with refractory brain tumors


  • Darren R. Hargrave M.D.,

    1. New Agents and Innovative Therapy Program, The Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada
    2. Pediatric Brain Tumor Program, The Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada
    Search for more papers by this author
  • Eric Bouffet M.D.,

    1. New Agents and Innovative Therapy Program, The Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada
    Search for more papers by this author
  • Janet Gammon R.N.,

    1. New Agents and Innovative Therapy Program, The Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada
    Search for more papers by this author
  • Nauman Tariq M.Sc.,

    1. Pediatric Brain Tumor Program, The Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada
    Search for more papers by this author
  • Ron M. Grant M.D.,

    1. Division of Haematology/Oncology, The Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada
    Search for more papers by this author
  • Sylvain Baruchel M.D.

    Corresponding author
    1. New Agents and Innovative Therapy Program, The Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada
    2. Pediatric Brain Tumor Program, The Hospital for Sick Children, University of Toronto, Toronto, Ontario, Canada
    • New Agents and Innovative Therapy Program, The Hospital for Sick Children, 555 University Avenue, Toronto, Ontario M5G 1X8, Canada
    Search for more papers by this author
    • Fax: 416-813-8024



Fotemustine is a nitrosourea with theoretic and preclinical advantages over the original analogs, carmustine and lomustine, in the treatment of brain tumors. This is the first pediatric Phase I study of fotemustine.


Patients younger than 21 with recurrent/resistant brain tumors were enrolled in a conventional Phase I study. Fotemustine was administered intravenously every 3 weeks at increasing dose levels starting at 100 mg/m2. Toxicity and response data were monitored closely.


Fifteen evaluable patients entered the study and received a total of 45 courses of fotemustine (dose range, 100–175 mg/m2). Myelosuppression was observed, with the dose-limiting toxicity being Grade 4 neutropenia and thrombocytopenia. Toxicity was delayed and cumulative. The maximum tolerated dose was 150 mg/m2 every 3 weeks. There were three documented radiologic responses (20% of patients) comprising one partial response and two minor responses in patients with a sarcoma, medulloblastoma, and ependymoma, respectively.


Fotemustine administered at a dose of 150 mg/m2 every 3 weeks is well tolerated in children and has antitumor activity in several brain tumors. This is the first dedicated Phase I study of a single agent nitrosourea in a pediatric population. More comparative studies should be undertaken to define the optimum nitrosourea analog for use in children with brain tumors. Cancer 2002;95:1294–301. © 2002 American Cancer Society.

DOI 10.1002/cncr.10814

Chloroethylnitrosoureas are a group of alkylating antineoplastic drugs with demonstrated activity against a number of solid tumors including malignant brain tumors. Since the development of the parent compound in 1959 (1-methyl-1-nitroso-3-nitro-guanidine [MNNG]) many new analogs have been synthesized and tested in adult Phase I studies, but there has been no dedicated pediatric Phase I study of a nitrosourea.

Fotemustine was developed by Servier Research Laboratories (Paris, France) and is characterized by high lipophilicity and low molecular weight with the chemical structure of a phosphonoalanine carrier group grafted onto a nitrosourea radical (Fig. 1). This structure improves penetration through the cell membrane and blood–brain barrier by using the amino acid transport system. In preclinical in vivo studies, fotemustine compared favorably with carmustine (1,3-bis(2 chloroethyl)-2-nitrosourea [BCNU]) and lomustine (1-2 (chloroethyl)-3cyclo-hexyl-1-nitrosourea [CCNU]) against several human tumor cell lines.1 This observation was confirmed in L1210 leukemia implanted in mouse ventricle and xenograft models of human glioblastoma2 and medulloblastoma.3 Animal pharmacokinetic studies using 14C-fotemustine show rapid distribution of the compound in all tissues including the brain.4 In permeability studies,5, 6 its diffusion coefficient in brain parenchyma implies high uptake in glial cells. Data from mouse, rat, dog, and monkey demonstrated myelosuppression as the main subacute and cumulative toxicity, with thrombocytopenia and leukopenia more pronounced than anemia.7

Figure 1.

Chemical structure of fotemustine (diethyl-1-[3-(2-chloroethyl)-3-nitrosoreido]-ethylphosphate).

Adult Phase I clinical studies showed that the maximum tolerated dose (MTD) was 100 mg/m2 per week with a dose-limiting toxicity (DLT) of thrombocytopenia.8 Subsequent Phase II studies in adults with metastatic melanoma used an induction dosing schedule of 100 mg/m2 per week by intravenous infusion over 1 hour for 3 consecutive weeks. After a 5-week rest, patients were maintained on a dose of 100 mg/m2 every 3 weeks. The overall objective response rate in a French multiinstitutional study9 of 153 patients was 24%. Twenty-five percent of patients with cerebral metastases had documented responses. In that study, the main toxicity was myelosuppression, with 25% of patients requiring dose reduction or delay of subsequent courses related directly to delayed toxicity from the weekly induction therapy. The same schedule was studied in 38 adults with recurrent malignant gliomas.10 An objective response rate of 26% was achieved and 47% of patients had stabilization of disease. In another Phase II study involving 22 patients with high-grade cerebral glioma,11 an 18% objective response rate was achieved and 32% of patients had stable disease. In both studies, thrombocytopenia and leukopenia were the main observed toxicities. Fotemustine has been studied in a number of trials either as high-dose therapy or in combination with radiotherapy or other chemotherapeutic agents in solid tumors including brain tumors. In a pediatric study of nine children with brain tumors, fotemustine (dose 100 or 150 mg/m2) was administered either alone or in combination with ifosfamide. The DLT was myelosuppression, with five of nine children having a complete response.12

There is significant evidence from preclinical and clinical experimental data for the activity of fotemustine in central nervous system (CNS) malignancy. A weekly schedule results in a significant need for dose reduction or delay, whereas a maintenance regimen of 100 mg/m2 every 3 weeks is well tolerated. To maximize dose intensity and maintain an active dose level close to the weekly induction schedule, it was proposed that 100 mg/m2 every 3 weeks should be the first dose level in this pediatric Phase I study of fotemustine in patients with recurrent brain tumors.



Patients younger than 21 years of age with a histologic diagnosis of a primary brain tumor or brain metastases (patients with brain stem glioma did not require a biopsy) for which no conventional treatment was available. Patients must have an Eastern Cooperative Oncology Group performance status of 2 or less and a life expectancy exceeding 8 weeks. In addition, patients were required to have adequate nutritional status (≥ 3rd percentile for weight/height), bone marrow (absolute neutrophil count [ANC] ≥ 1 × 109/L, platelet [Plt] ≥ 100 × 109/L, hemoglobin ≥ 9–10 g/dL), renal (normal creatinine or glomerular filtration rate ≥ 70 mL/min per 1.73 m2), hepatic (bilirubin < 1.5 mg/dL, aspartate aminotransferase < 2 × normal), pulmonary Grade 1 or lower, and cardiac (ejection fraction ≥ 27%) functions. Evidence of recovery from previous chemotherapy was required, with 6 or more months since bone marrow transplant and patients were to be off all growth factors for 1 week or more. For radiotherapy, 6 or more weeks had to elapse or 6 or more months for craniospinal radiotherapy/ total body irradiation. Written informed consent was obtained from all patients/parents as stipulated by the Institutional Review Board (IRB). Exclusion criteria included pregnancy/ breast-feeding, uncontrolled infection, or other concomitant anticancer therapy.

Data Collection

All study entrants provided a complete medical history and underwent a physical examination before each course. Laboratory tests of bone marrow, renal, and hepatic function were performed before the study and weekly thereafter. At the start of the study, all patients had a chest X-ray, echocardiogram, and appropriate neuroimaging (computed tomographic scan/magnetic resonance imaging). Neuroimaging was repeated after two courses (6 weeks) and subsequently as clinically indicated (at least every six weeks). All investigations were performed before and at the end of study. In the event of an adverse event or study death, the study coordinator and the IRB were informed immediately and an adverse drug reaction report was filed. The toxicity was assigned a grade as per the 1984 National Cancer Institute common toxicity criteria.

Treatment Plan and Modifications

Fotemustine was supplied by Servier as a freeze-dried powder to be reconstituted with the supplied solvent and diluted with 5% dextrose solution before a 1-hour intravenous infusion. Antiemetics were administered as appropriate. The initial starting dose was 100 mg/m2 every 3 weeks. Subsequent dose escalations included dose level 2, 125 mg/m2; dose level 3, 150 mg/m2; dose level 4, 175 mg/m2, and dose level 5, 200 mg/m2 every 3 weeks. Three patients were to be treated at each dose level with six patients at the MTD. The DLT was defined as the dose that caused Grade 3 or higher renal, cardiovascular, pulmonary, or CNS myelotoxicity, any Grade 4 nonmyelotoxicity, or any Grade 4 myelotoxicity lasting more than 1 week. The MTD was defined as the dose level immediately below the level at which two of three to six patients experienced DLT after the first course of fotemustine.

Patients had to have evidence of recovery from previous course toxicity before receiving additional therapy. If the ANC was less than 0.5 × 109/L or the Plt count was less than 25 × 109/L for more than 1 week, the subsequent course dosage was decreased by 25%, which was also the case in the event of other organ toxicity of Grade 3 or higher. There was no intrapatient dose escalation. If patients had progressive disease after 6 weeks (two courses), they were removed from the study. Patients with stable or responsive disease received a minimum of two courses. All patients who completed one or more courses were evaluable for toxicity.


Sixteen patients were entered into this single institution trial between June 1998 and April 2001. Patient characteristics are shown in Table 1. There were seven male and nine female patients with a median age of 5 (range, 1.8–14.5) years. Of these 16 patients, 15 were evaluable for toxicity. The majority of entrants had a performance status of 0 or 1 and the most common diagnoses were recurrent medulloblastoma, anaplastic ependymoma, and brain stem glioma (Table 1). All patients had surgery and histologic diagnosis but in four patients this was only a limited biopsy. All patients had received previous therapy with either radiotherapy alone (two patients), chemotherapy alone (six patients), or a combination of both modalities (eight patients) and three patients had received previous nitrosoureas. Concomitant drug therapy included five patients receiving anticonvulsants and seven receiving steroids. The 15 evaluable patients received a total of 45 courses with a median of 3 (range, 1–6) courses per individual. The dose range was 100–175 mg/m2 every 3 weeks.

Table 1. Patient Characteristics
  • ECOG: Eastern Cooperative Oncology Group; CNS: central nervous system; BMT: bone marrow transplantation.

  • a

    Patient not evaluable as he/she progressed before completing first course.

No. of patients entered16
No. of patients evaluable15a
Median age (range)5 (1.8–14.5) yrs
ECOG performance status 
 Anaplastic ependymoma3
 Brain stem glioma3
 Glioblastoma multiforme1
 CNS rhabdoid tumor1
 Radiation-induced CNS sarcoma1
 CNS metastatic osteosarcoma1a
Previous therapy 
 Surgery (biopsy only)12 (4)
 Chemotherapy (BMT)13 (1)
 Radiotherapy and chemotherapy9
No. of patients (evaluable courses) at each dose level (mg/m2) 
 1003 (12)
 1253 (10)
 1506 (17)
 1753 (6)
Total no. of evaluable patients (courses)15 patients (45)



The main observed toxicity was myelosuppression and the DLT was neutropenia (one of three patients) and thrombocytopenia (one of three patients), at Grade 4 for more than 1 week at a dose of 175 mg/m2. Therefore, the MTD was 150 mg/m2.

Hematologic toxicity

Anemia was not a major toxicity with no significant toxicity after the first course. Only one patient required a transfusion for Grade 4 anemia after four courses at a dose level of 175 mg/m2 (Table 2). Neutropenia was dose dependent with no toxicity after the first course at 100 mg/m2, only one Grade 1 neutropenia at 125 mg/m2, 50% of patients having Grade 2 or 3 at 150 mg/m2 and one DLT of Grade 4 at 175 mg/m2. Neutropenia was also cumulative at the higher dose levels of 150–175 mg/m2 with 50% of patients experiencing Grade 3 or 4 neutropenia. The median onset of Grade 3/4 neutropenia was Day 14 (range, 4–35), with a nadir on Day 30 (16–52) and recovery by Day 36 (20–65). Only two cases of febrile neutropenia requiring antibiotics occurred and there were no documented bacteremias. Thrombocytopenia after the first course occurred at a dose level of 175 mg/m2. However, it was severe with two of three patients suffering Grade 3 or 4 thrombocytopenia. One patient required a Plt transfusion. Cumulative toxicity was seen, with Grade 3 and 4 Plt toxicity occurring with increasing number of courses at all dose levels. One third of all patients required a Plt transfusion on one or more occasions. Grade 3/4 thrombocytopenia had a median onset on Day 7 (1–23), nadir on Day 32 (16–42), and recovery by Day 58 (31–84). No dose adjustments were necessary for myelosuppression but there was a course delay for 14 of 45 (31%) patients (thrombocytopenia in 10 and neutropenia in 4) with a median delay of 24 (range, 6–46) days. Delays were more common at higher dose levels (50% of courses at 175 mg/m2) and after several courses, due to the cumulative Plt toxicity.

Table 2. Hematologic Toxicity: Data after Initial Fotemustine Dose and after Subsequent Courses
Dose (mg/m2)100125150175   
  • WNL: within normal limits for age.

  • a

    All grades.

  • b

    Grades 3 and 4.

No. of patients (total no. of courses)3 (12)3 (10)6 (17)3 (6)
 No. of patients affected after first course/(% of total courses) affectedMedian day (range)
Anemia g/dL (grade)       
 WNL (0)3/(100)2/(90)3/(41)0   
 10-WNL (1)01/(10)2/(41)2/(33)   
 8–10 (2)001/(12)1/(17)7 (7–34)a15 (1–44)a35 (15–64)a
 6.5–7.9 (3)0000/(33)   
 < 6.5 (4)000/(6)0/(17)   
Neutropenia × 109/L (grade)       
 > 2 (0)3/(92)2/(40)3/(35)1/(33)   
 1.5–1.9 (1)01/(30)0/(6)0   
 1.0–1.4 (2)00/(30)1/(29)1/(17)9.5 (7–35)a18.5 (7–52)a26 (14–98)a
 0.5–0.9 (3)0/(8)02/(18)0/(17)   
 < 0.5(4)000/(12)1/(33)14 (4–35)b30 (16–52)b36 (20–65)b
Thrombocytopenia × 109/L (grade)       
 > 100 (0)3/(67)3/(80)6/(94)1/(33)   
 75–100 (1)00/(10)00   
 50–74.9 (2)0/(17)00010 (7–23)a28 (11–42)a56 (25–84)a
 25–75 (3)0/(8)0/(10)01/(50)   
 < 25 (4)0/(8)00/(6)1/(17)7 (7–23)b32 (16–42)b58 (31–84)b

Nonhematologic toxicity

With the use of standard prophylactic antiemetics, nausea and vomiting were not a major problem (Table 3). Only two patients at a dose level of 150 mg/m2 reported symptoms. One patient reported mild constipation during her four courses of therapy, and the other reported asymptomatic transaminitis, but this was not observed in the rest of the group. Two patients had moderate increases in liver transaminases that did not require alteration of the dosing schedule and they recovered spontaneously. There were three cases of skin toxicity. Twelve days after the first dose of fotemustine, one patient developed a rash consisting of nonitching, red flat-topped papules on the hands and face. After a nonspecific result from a skin biopsy, the diagnosis was Gianotti-Crosti syndrome (or papular acrodermatitis of childhood), a harmless self-resolving condition. The rash recurred 10 days after the second dose and, therefore, may have been related to the fotemustine administration. The second patient had transient inflammation around the intravenous site on two occasions and a third patient developed a Grade 2 phlebitis at the site of infusion on Day 16 after the fifth dose, which lasted for 57 days. No other nonhematologic toxicity was observed.

Table 3. Nonhematologic Toxicity
Toxicity grade1234
No. of patients (courses) affected
  1. ALT: alanine aminotransferase; AST: aspartate aminotransferase; GGT: gamma-glutamyl transpeptidase.

 Nausea1 (1)   
 Vomiting2 (3)   
 Constipation1 (4)   
 ALT 2 (2)  
 AST 1 (1)  
 GGT 1 (1)  
 Rash 1 (1)  
 Phlebitis 1 (1)  
 Intravenous site inflammation 1 (2)  


There were three radiologic responses (20% of patients), one partial response (≥ 50% reduction) and two minor responses (≥ 25% to ≤ 50% reduction) accompanied by associated clinical responses. The patient with a partial response had a radiation-induced undifferentiated brain sarcoma following previous therapy for acute lymphoblastic leukemia, which included CNS radiotherapy. Following diagnosis of his brain tumor, he had a subtotal resection, which was followed by more radiotherapy and chemotherapy (ifosfamide, carboplatin, and etoposide), but the brain tumor recurred. At this point, he commenced fotemustine at 125 mg/m2. After two courses, he had marked clinical and radiologic improvement (Fig. 2) and a reduction in tumor size. He received three more courses. Unfortunately, the tumor progressed (4 months after initial response) and the patient was eliminated from the study. The two minor responses (duration of 8 and 2 months) occurred in patients with recurrent metastatic medulloblastoma (postautologous bone marrow transplantation) and anaplastic ependymoma who were treated with 100 and 150 mg/m2, respectively. None of the three patients who had previously received a nitrosourea had a response documented with fotemustine. One patient died of progressive disease during the study and all other patients eventually were removed from study with progressive disease.

Figure 2.

Major response to fotemustine in patient with radiation-induced brain sarcoma. A) Baseline magnetic resonance image (MRI) before administration of fotemustine. B) MRI after two courses of 125 mg/m2 fotemustine, accompanied by resolution of right-side weakness.


This pediatric study of fotemustine is the first published Phase I study to investigate a single-agent nitrosourea in children. Since the synthesis of MNNG, the original parent compound, in 1959,13 many analogs have been developed and tested in adult Phase I studies,14–21 the best known and still most widely used being BCNU22 and CCNU.13, 23 However, despite numerous adult studies of single-agent nitrosoureas, there have been no published dedicated pediatric Phase I studies of nitrosourea monotherapy. BCNU and CCNU, which are used widely in pediatric oncology, were developed into pediatric Phase II studies solely by extrapolation from the adult MTD and toxicity data.24 This is a questionable alternative to dedicated pediatric Phase I studies due to the wide range of MTDs and toxicity profiles seen in children compared with adults.25–28

Fotemustine is a third-generation chloroethylnitrosourea with theoretic and preclinical evidence of improved cell membrane and blood–brain permeability compared with BCNU and CCNU.5, 6 This study documents responses in a heavily pretreated group of children with recurrent CNS tumors and this is encouraging for future development. Nonhematologic toxicity was minimal and evidence from adult data in a high-dose schedule suggests that pulmonary toxicity may be less severe with fotemustine29 than with BCNU-containing regimens.30, 31 This advantage is worth exploring in future trials. The main observed toxicity as with previous nitrosoureas was myelosuppression, with both neutropenia and thrombocytopenia as the acute DLT. The pattern of thrombocytopenia was characteristic of the nitrosourea group, as it was dose related, delayed, and cumulative.

The inclusion of fotemustine in a multiagent regimen with other myelosuppressive chemotherapy agents would require either schedule adjustment with a longer administration interval (4–6 weeks) or a specific mechanism to ameliorate bone marrow toxicity. One theoretically attractive example of a multidrug regimen would include fotemustine in conjunction with a methylating agent such as dacarbazine, procarbazine, or the newer analog, temozolomide. There is evidence that this combination reverses the known resistance mechanism mediated by O6-alkylguanine-DNA alkyltransferase (ATase).3, 32–34 This combination and pretreatment with the compound O6-benzylguanine, which depletes ATase, further reverses this particular resistance mechanism.35–37 Unfortunately, potentiation of cytotoxicity is not specific to tumor cells and unwanted toxicity is also increased including dose-limiting myelosuppression.38, 39 To circumvent this lack of specificity, others have investigated the strategy of selectively enhancing the repair capacity of bone marrow for ATase, via ex vivo transfer and in vivo expression of a mutant version of ATase in early hemapoietic progenitor cells.40–43 The transduction of this mutant version of the enzyme in mice hematologic progenitor cells has been achieved and these cells are refractory to ATase inactivation.44, 45 Another strategy to diminish the DLT of nitrosoureas (fotemustine) is to use chemoprotective agents such as amifostine, which was developed originally as a radioprotector against nuclear warfare.46 Both preclinical studies and clinical trials show that cytoprotection is selective and does not occur at the expense of the antitumor effects of anticancer agents.47 A recent Phase I study48 produced a recommended dosing schedule for the pediatric population. A study of amifostine and fotemustine in adults with metastatic melanoma demonstrated the chemoprotective abilities of this agent. No patients suffered Grade 3 or 4 myelosuppression compared with a historical control group.49

The nitrosoureas are an important group of agents in the treatment of pediatric CNS tumors and the development of improved new analogs is desirable. However, it is essential that new agents are evaluated properly in pediatric clinical trials. Monotherapy with fotemustine, 150 mg/m2 every 3 weeks, is tolerable and should be investigated further in Phase II/III studies alone or in a combination with other agents to determine the role of chemoprotective therapies. Ultimately, fotemustine should be tested in a randomized setting against the older established nitrosoureas that are used currently in pediatric brain tumor protocols.


The authors thank Servier Research Laboratories for technical support. They also thank Dr. M. Greenberg, Dr. M. Silva, and Dr. S. Weitzman for enrolling patients in the trial.