Multicentre phase II study and pharmacokinetic analysis of irinotecan in chemotherapy-naïve patients with glioblastoma

E. Raymond1,+, M. Fabbro2, V. Boige1, O. Rixe1, M. Frenay3, G. Vassal1, S. Faivre1, E. Sicard1, C. Germa4, J. M. Rodier1, L. Vernillet4 and J. P. Armand1

1 Institut Gustave Roussy, Villejuif; 2 CRLC Val D’aurelle, Montpellier; 3 Centre Antoine Lacassagne, Nice; 4 Aventis, Paris, France

Received 20 June 2002; revised 12 December 2002; accepted 13 December 2002


    Abstract
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Background:

To assess the antitumour activity and safety profile of irinotecan and its pharmacokinetic interactions with anticonvulsants in patients with glioblastoma multiforme.

Patients and methods:

This multicentre phase II and pharmacokinetic study investigated the effects of irinotecan 350 mg/m2 given as a 90-min infusion every 3 weeks either prior to (group A) or after relapse following radiotherapy (group B) in chemotherapy-naïve patients with glioblastoma. Preferred concomitant medication for seizure prevention was valproic acid. Pharmacokinetic analysis of irinotecan and its main metabolites (SN-38, SN-38-G, APC and NPC) was performed during cycle 1. An independent panel of experts reviewed the activity data.

Results:

Fifty-two patients (25 patients in group A and 27 patients in group B) received a total of 191 cycles of irinotecan. Forty-six patients (22 patients in group A and 24 patients in group B) were evaluable and externally reviewed for activity. According to external review, one partial response (group B), seven minor responses (three in group A and four in group B), 12 disease stabilisations (seven in group A and five in group B) were observed. This resulted in an overall response rate of only 2.2% (95% confidence interval 0.2% to 6.5%). The median time to tumour progression was 9 weeks in group A and 14.4 weeks in group B. Six-month progression-free survival rates were 26% in group A and 43% in group B. Grade 3–4 toxicities (percentage of patients in groups A and B) consisted of neutropenia (12.5% and 25.9%), diarrhoea (8.3% and 7.4%), asthenia (12.5% and 7.4%) and vomiting (0% and 7.4%). The clearance of irinotecan was 12.4 and 14.4 l/h/m2 in two patients who received no anticonvulsant. In patients receiving valproic acid, the clearance of irinotecan was 17.2 ± 4.4 l/h/m2.

Conclusions:

Irinotecan given at the dose of 350 mg/m2 every 3 weeks has limited clinical activity as a single agent in patients with newly diagnosed and recurrent glioblastoma after radiotherapy. The toxicity profile and plasma disposition of irinotecan and SN-38 were not strongly influenced by anticonvulsant valproic acid therapy. Although the response rate of irinotecan as a single agent was limited, it remains an attractive drug for combination studies in patients with glioblastoma.

Key words: anticonvulsants, APC, irinotecan, NPC, SN-38, valproic acid


    Introduction
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Malignant gliomas represent ~60% of primary malignant tumours of the central nervous system [15]. Nitrosoureas (including BCNU and CCNU) [69] and more recently temozolomide [10] are alkylating agents classically used in patients with glioblastoma. However, newer active anticancer agents that have a different mechanism of action are urgently required to improve the outcome of patients with gliomas.

Irinotecan has been shown to exert antitumour activity against several human tumour types and was recently approved in combination with 5-fluorouracil (5-FU)/folinic acid as first-line chemotherapy for metastatic colorectal cancer [11]. In animal studies, irinotecan displayed marked antitumour activity against a broad panel of subcutaneous and intracranial human glioblastoma multiforme, ependymoma and medulloblastoma xenografts [1214]. After i.v. injection, irinotecan has a very complex metabolism that mainly takes place in the liver. Irinotecan can be converted into two inactive metabolites, 7-ethyl-10-[4-N-(5-aminopentanoic acid)-1-piperidino]-carbonyloxycamptothecin (APC) and 7-ethyl-10-[4-(1-piperidino)-1-amino]-carbonyloxycamptothecin (NPC), by the CYP3A4 enzyme and into an active metabolite, 7-ethyl-10-hydroxycamptothecin (SN-38), by carboxylesterase enzymes in the liver [15, 16]. SN-38 is further metabolised to SN-38 glucuronide {7-ethyl-10-[3,4,5-trihydroxy-pyran-2-carboxylic acid]-camptothecin (the ß-glucuronide conjugate of SN-38) (SN-38-G)} through conjugation by uridine diphosphate glucuronosyltransferase (UGT1A1) [17]. SN-38 has a low molecular weight and presents lipophilic characteristics but has shown poor penetration of central nervous fluid (CSF) in comparison with other camptothecins [18].

Based on these data, this multicentre phase II study was conducted to evaluate the antitumoral activity and toxicity profile of irinotecan administered every 3 weeks in chemotherapy-naïve patients with glioblastoma either previously treated by or not treated by radiotherapy. Supportive care in the management of patients with glioblastoma frequently includes use of corticosteroids to control cerebral oedema and oral anticonvulsants to prevent epilepsy. Concomitant medications may modify the metabolism of irinotecan and could affect the efficacy and toxicity profile of the drug. Valproic acid (2-propylpentanoic acid) is a frequently used anticonvulsant that is extensively glucuroconjugated with potential interactions with the disposition of irinotecan [19]. In this study, we investigated the pharmacokinetics and pharmacodynamics of irinotecan and its main metabolites, SN-38, SN-38-G, APC and NPC, with regard to concomitant medications (predominantly valproic acid) used during the trial.


    Patients and methods
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Inclusion criteria
Inclusion criteria were as follows: aged >=18 years; histologically confirmed recurrent glioblastoma multiforme (grade 4 astrocytoma) measurable on contrast enhanced magnetic resonance imaging (MRI) performed within 2 weeks before study initiation; no previous chemotherapy; moderate MRC less than or equal to two neurological symptoms; and performance status ranging from 0 to 3. A minimum interval of 3 weeks between prior surgery or 6 weeks prior radiotherapy and enrolment must have elapsed. Other criteria were neutrophils >=2 x 109/l, platelets >=100 x 109/l, haemoglobin >=10 g/dl, total bilirubin <1.25 x the institutional upper normal limit (UNL), alkaline phosphatases and ASAT <=2.5 x UNL; prothrombin time >=50% and creatinine <=120 µmol/l. For patients with corticosteroids, a stable dose for 20 days was required before study entry. Exclusion criteria were past or present history of chronic diarrhoea, current uncontrolled infection, other investigational drugs, pregnancy and lactating women of child-bearing age had to take contraceptive measures. All patients gave written informed consent prior to registration for the study.

Drug administration
Aventis Pharma (Paris, France) supplied irinotecan in 5 ml vials (100 mg of active product). Irinotecan was diluted in 250 ml (0.9% NaCl) and administered as a 90 min i.v. infusion every 21 days at the dose of 350 mg/m2 [20].

Two groups of patients were to be successively considered:

Group A: patients with inoperable or incompletely resected newly diagnosed radiotherapy-naïve glioblastoma received irinotecan prior to radiotherapy. They were scheduled to receive three cycles of irinotecan followed by radiotherapy in case of stabilisation or tumour progression.
Group B: patients relapsing after radiotherapy. Patients in group B were scheduled to receive up to six cycles of irinotecan, according to the investigator efficacy assessment.

Radiotherapy consisted of a conformational administration of 50–60 Grays into the tumour volume given over a period of 6–7 weeks.

Dose adjustments
Dose adjustment of irinotecan was based on the worst toxicity observed during the previous cycle. The next cycle was delayed until the neutrophil and platelet counts were >=1500/µl and >=100 000/µl, respectively, and treatment toxicity was fully resolved. If this exceeded a 2-week delay, treatment with irinotecan was discontinued. In cases of grade 4 thrombocytopenia (grade 3–4 neutropenia lasting >7 days) and febrile neutropenia, the dose of irinotecan was reduced to 300 mg/m2 and, if necessary, could be subsequently reduced to 250 mg/m2. Patients were withdrawn from the study if they required more than two dose reductions.

Concomitant medications
Boluses of methylprednisolone 80 mg with standard doses of ondansetron were given to prevent vomiting. Atropine (1–2 mg) was administered for treatment of cholinergic syndromes [21] and then, if necessary, given prophylactically at subsequent cycles. Patients began antidiarrhoeal treatment for delayed diarrhoea occurring more than 24 h after irinotecan administration [22]. The prophylactic use of loperamide was not allowed. Patients presenting with severe vomiting, blood in their faeces or severe diarrhoea after 48 h were hospitalised. Haematopoietic growth factors were not allowed.

To minimise interaction with the metabolism of irinotecan, patients taking anti-epileptic treatments were recommended to use valproic acid or carbamazepine 2 weeks prior to irinotecan. Valproic acid was given orally at a daily dose ranging from 20 to 30 mg/kg/day (<60 mg/kg/day). Concentrations of anticonvulsants were monitored during the study as necessary to maintain therapeutic levels. Chronic oral administration of corticosteroids was used as needed with a careful monitoring of doses.

Baseline and follow-up examinations
Patients underwent physical and neurological examination within 2 weeks prior to entering the study, immediately prior to the first irinotecan injection and then at least once before each subsequent drug infusion. Complete blood cell with differential and platelet counts, serum biochemistry and hepatic parameters were assessed weekly. Toxicity was graded according to National Cancer Institute-Common Toxicity Criteria.

Response determination was based on both the comparison of the baseline brain contrast-enhanced MRI or CT scan done 2 weeks prior to irinotecan with those performed every two to three infusions. Treatment was discontinued in case of tumour progression, unacceptable toxicity or patient refusal. Patients were followed-up monthly (MRI or CT scan every 2–3 months) after the completion of the study until death.

Evaluation of response to therapy
The primary end point of this study was to assess the anti-tumour activity according to the criteria of MacDonald et al. [23] and this was reviewed by a panel of independent experts. Patients were considered evaluable for efficacy if they had received at least two cycles of irinotecan. In addition to classical response parameters [23], a minor response was defined as a >25% but a <50% reduction from baseline in the size of enhancing tumour on either the CT or MRI scan. Follow-up examinations for tumour evaluation were performed using the same method as that used at baseline. A CT or MRI scan was performed every two cycles or every three cycles in patients entering groups A and B, respectively. Partial and minor responses had to be confirmed using the same technique, either a CT or MRI scan, after a period of at least 4 weeks had elapsed since the first onset of response. Tumours were considered stable if they met the criteria for tumour stabilisation for at least two cycles.

Pharmacokinetic evaluation
Pharmacokinetic analysis was performed during the first infusion. Blood specimens (2 ml heparinised tubes) were drawn 5 min before, 45 min during and at the end of the infusion, then 1 h, 6 h and 22.5 h after the end of the infusion (4°C). Plasma was harvested immediately by centrifugation at 1200 g for 30 min and stored at –80°C until analysis.

Total forms of CPT-11 and its metabolites (SN-38, SN-38-G, APC and NPC) were assayed using high-performance liquid chromatography with fluorescence detection as described by Rivory et al. [24], and with slight modifications [25]. Calibration standard responses were linear. Limits of quantification in plasma were 10 ng/ml for all compounds. Irinotecan concentrations were expressed in free base units for pharmacokinetic analyses. Irinotecan, SN-38, SN-38-G, APC and NPC plasma concentration data were analysed by non-compartmental methods. Peak plasma concentrations (Cmax) were determined for irinotecan, SN-38, SN-38-G, APC and NPC from concentration–time curves. Areas under the plasma concentration–time curves (AUC0–24h) were calculated using the linear trapezoidal rule from time zero to the last sampling with quantifiable drug concentrations.

Metabolite ratios were calculated as follows:

[AUC0–{infty} (metabolite)/AUC0–{infty} (irinotecan)] x100 for SN-38, SN-38-G and NPC;

[AUC0–last sampling (APC)/AUC0–last sampling (irinotecan)] x100 for APC.

The ratio of SN-38 glucuronidation was defined as follows:

[AUC0–last sampling (SN-38-G)/AUC0–last sampling (SN-38)] x100.

Statistical analysis
The primary end point was the objective response rate according to the expert panel. Secondary objectives were progression-free survival, duration of response and overall survival (Kaplan–Meier analysis).

The number of patients in each group was determined by Gehan test with a type error I ({alpha} = 5%) and a type error II (ß = 10%) to have a control rate of 20%. For the first stage, 14 patients were enrolled in each group. Regarding the disease control rate observed on these 14 patients, additional patients could be included for the second stage. In total, 25 patients per group had to be enrolled in this study.

The statistical (SAS software, ver. 6.12®) and pharmacokinetic analyses (WinNonlin; Scientific Consulting, Cary, NC, USA) used non-parametric Mann–Whitney and Kruskal–Wallis tests displayed using GraphPad-InStat 3.00® (GraphPad Software, San Diego, CA, USA). A two-sided value of P <0.05 was considered significant.


    Results
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
Patient characteristics
A total of 52 patients with confirmed glioblastoma were enrolled in this study (25 patients in group A and 27 patients in group B). All patients received at least one cycle of chemotherapy. One patient withdrew his consent immediately after the first cycle and was not evaluable for activity and toxicity. Baseline patient characteristics are summarised in Table 1. Most of the patients entering this trial had concomitant anticonvulsant therapy that included valproic acid, carbamazepine, phenytoin and/or phenobarbital in 40, seven, four and/or two patients, respectively. Forty patients had concomitant treatment with corticosteroids at study entry. Only one patient had neither anticonvulsant therapy nor treatment with corticosteroids at study entry. Other concomitant medications likely to interfere with the CYP3A4 metabolism of irinotecan were registered.


View this table:
[in this window]
[in a new window]
 
Table 1. Patient characteristics
 
Treatment delivery
A total of 191 cycles were administered (69 cycles in group A and 122 cycles in group B). The median number of cycles were three (range 1–8) and four (range 1–10) in groups A and B, respectively. Short treatment delays (ranging from 3 to 6 days) were reported in two patients (two cycles) in group A and in seven patients (eight cycles) in group B. These delays were mainly due to the scheduling of cycles and were caused by irinotecan toxicity in only one patient (asthenia). Longer delays (>=7 days) were reported in three patients (five cycles) in group A and four patients (seven cycles) in group B. Longer delays were due to toxicity in two patients (four cycles). Cycles were given at the dose of 350 mg/m2 in 68 cycles (98.6%) in group A and 107 cycles (87.7%) in group B. Dose reduction to 300 mg/m2 irinotecan was performed for three patients (one in group A and two in group B) and 14 cycles. The median relative dose intensity of irinotecan was 0.98 (range 0.83–1.02) in group A and 0.97 (range 0.72–1.02) in group B.

Antitumour activity
The antitumour activity of irinotecan in patients with glioblastoma was evaluated using CT scans and MRI of the brain in 12 and 40 patients, respectively (Table 2).


View this table:
[in this window]
[in a new window]
 
Table 2. Antitumour activity of irinotecan in patients with glioblastoma
 
According to the investigators, 49 patients were evaluable for activity. One partial response (in group A), four minor responses (two in group A and two in group B) and 16 tumour stabilisations (six in group A and 10 in group B) were observed, resulting in an overall control rate of 43% [95% confidence interval (CI) 28.5% to 57%].

According to expert panel review, 46 patients were considered evaluable for activity. Six patients (three in group A and three in group B) were not evaluable (in four cases, images from baseline were not technically comparable with images used for the documentation of response, one patient withdrew his consent before tumour evaluation and one died prematurely from a concomitant disease). After review, one partial response lasting 35 weeks was observed in group B, seven minor responses (three in group A and four in group B), 12 tumour stabilisations (seven in group A and five in group B) and 26 tumour progressions were reported. This resulted in an overall response rate of 2.2% (95% CI 0.2% to 6.5%). Minor and partial responses were observed in patients concomitantly treated with valproic acid. The median time-to-progression was 9 weeks (range 3.6–53.1; 95% CI 8.1–22.4) in group A and 14.4 weeks (range 5.5–36.8; 95% CI 9.0–21.1) in group B. The 6-month progression-free survival rate was 26% in group A and 43% in group B (Figure 1; no statistical difference was observed between groups A and B; Cox proportional hazard ratio = 1.34, p = 0.38). The median overall survival was 5.8 months (range 1.5–13.0; 95% CI 3.1–9.6) in group A and 6.8 months (range 3.0–14.4; 95% CI 5.5–11.5) in group B.



View larger version (18K):
[in this window]
[in a new window]
 
Figure 1. Progression-free survival (Kaplan–Meier analysis) of patients with glioblastoma treated with irinotecan either prior to (group A) or after radiotherapy (group B). No statistical difference was observed between groups A and B (Cox proportional hazard ratio: 1.34, P = 0.38).

 
Safety
Fifty-one patients and 190 cycles were evaluable for safety. Overall safety data are presented in Table 3. No clinically relevant difference in toxicity was observed in the two groups of patients.


View this table:
[in this window]
[in a new window]
 
Table 3. Grade 3–4 toxicity of irinotecan per patient and cycle
 
Grade 3–4 cholinergic syndrome with acute diarrhoea, abdominal pain, nausea, vomiting, hypersalivation, sweating and asthenia occurred in five patients (10%) and six cycles (3%).

Mild to moderate delayed diarrhoea was frequently observed. Grade 1 diarrhoea was observed in 13 patients (25.5%) and in 44 of 190 cycles (23%). Grade 2 diarrhoea was observed in 21 patients (41.2%) and 28 cycles (14.7%). Grade 3 diarrhoea was observed in four patients (7.9%) and five cycles (2.6%), requiring hospitalisation for hydration in three patients. No grade 4 diarrhoea was reported. Two patients (two cycles) experienced severe diarrhoea in group A and two patients (three cycles) in group B. Diarrhoea was efficiently treated with oral loperamide.

Two patients, a 60-year-old male and a 63-year-old female, who experienced febrile grade 2–3 diarrhoea without neutropenia were adequately treated with oral loperamide and ciprofloxacine. These primary events subsequently led, in both cases while the diarrhoea had resolved, to more severe sepsis and patient death. In one of these patients, who developed an acute respiratory distress syndrome, Enterococcus faecalis, Staphylococcus faecalis and Staphylococcus maltophilia were identified from blood cultures. No bacterial documentation was possible in the other patient. Pharmacokinetic data were not available for these two patients.

Severe grade 3–4 neutropenia was reported in a total of 10 patients (19.6%) and 21 cycles (11%). In group A, grade 4 neutropenia was observed in three patients (12.5%) with one of them experiencing febrile neutropenia, whilst in group B, seven patients (25.9%) presented with grade 3–4 neutropenia including three patients (11.1%) with febrile grade 3–4 neutropenia. In total, four patients had episodes of febrile grade 3–4 neutropenia (7.8% of patients and 2.1% of cycles). One patient treated in group A presented with grade 3 anaemia. No grade 3–4 thrombocytopenia was observed.

Other toxicities were as follows: grade 3 asthenia was observed in three and two patients in groups A and B, respectively; grade 3–4 vomiting in two patients in group B; and grade 1–2 alopecia in six and nine patients in groups A and B, respectively.

Transient grade 3 elevations of transaminases were observed in three patients. One of them had a grade 3 elevation of ALAT when carbamazepine was substituted for valproic acid. Another patient had grade 3 elevation of ASAT while receiving concomitant treatment with valproic acid for a month. In these two patients, valproic acid and irinotecan were maintained with no recurrence of grade 3 hepatic toxicity at subsequent cycles. The last patient had a grade 3 elevation of ASAT while he was concomitantly treated with phenytoin and paracetamol. In this patient, recovery was obtained after withdrawal of phenytoin and paracetamol.

Pharmacokinetics of irinotecan
Pharmacokinetic sampling was performed in 27 patients (14 patients in group A and 13 patients in group B). A typical pharmacokinetic profile is shown in Figure 2 and the main parameters of irinotecan and its metabolites are presented in Table 4. The mean AUC0–{infty} of irinotecan was 18 518 ± 5323 ng·h/ml [coefficient of variation (CV): 28.7%] with a mean clearance of 17.6 ± 4.7 l/h/m2 (CV: 26.5%). The mean AUC0–{infty} of SN-38 was 471.6 ± 224.3 ng·h/ml (CV: 47.6%). The mean metabolic ratio of SN-38/irinotecan was 0.6 ± 0.7 % (CV: 108%). Mean AUC0–{infty} of SN-38-G was 1243 ± 687 ng·h/ml (CV: 55.2%). The ratio of SN-38-G/irinotecan was 4.7 ± 2.0% (CV: 43%) with a glucuronidation ratio of 16.5 ± 14.3% (CV: 86.2%). Mean AUCs of APC and NPC were 5407 ± 2598 ng·h/ml (CV: 48%) and 533 ± 263 ng·h/ml (CV: 49%), respectively. Metabolic ratios of APC and NPC were 27.5 ± 13% (CV: 47.5%) and 1.5 ± 1% (CV: 66.6%), respectively.



View larger version (16K):
[in this window]
[in a new window]
 
Figure 2. Typical pharmacokinetic profile of irinotecan and its main metabolites in a patient with glioblastoma. This patient received valproic acid concomitantly to irinotecan.

 

View this table:
[in this window]
[in a new window]
 
Table 4. Comparison of pharmacokinetic parameters of irinotecan and its metabolites in patients with glioblastoma or other tumour types
 
Toxicity and pharmacokinetics of irinotecan according to concomitant medications
All patients were receiving concomitant medications (median number, 5; range 2–10) with several drugs. We found no influence of the doses and duration of exposure to corticosteroids and omeprazole on the toxicity and the pharmacokinetic parameters of irinotecan and its metabolites (data not shown). In 40 patients treated with valproic acid for seizure prevention, grade 3–4 diarrhoea, neutropenia and asthenia were observed in nine (22.5%), 13 (32.5%) and 14 (35%) patients, respectively. Neither grade 3–4 diarrhoea, neutropenia, nor asthenia was observed in patients concomitantly treated with carbamazepine, phenytoin and/or phenobarbital.

For pharmacokinetic drug interaction analysis, three groups of patients (Figure 3) were defined: (i) patients without anti-epileptic drug (two patients); (ii) patients receiving valproic acid only (20 patients); and (iii) patients concomitantly treated with carbamazepine or phenobarbital (five patients). None of the patients analysed in the pharmacokinetic study received phenytoin. Clearances of irinotecan were 12.4 and 14.2 l/h/m2 in two patients who received no anti-epileptic drugs, 20.9 ± 5.7 l/h/m2 (range 11–25.3 l/h/m2) under phenobarbital/carbamazepine and 17.2 ± 4.4 l/h/m2 (range 9.5–27.5 l/h/m2) under valproic acid (P >0.05). Pharmacokinetic parameters of SN-38 (as well as other metabolites) were not significantly different in patients receiving no anticonvulsants (AUCSN-38: 131.9 ± 35 ng·h/ml) as compared with those receiving valproic acid (AUCSN-38: 115.3 ± 136 ng·h/ml, P >0.05). Metabolic ratios of metabolites with regards to concomitant anticonvulsants are shown in Figure 4.



View larger version (13K):
[in this window]
[in a new window]
 
Figure 3. Clearances of irinotecan in patients with glioblastoma. Three groups of patients were defined to assess possible drug interactions: (group 1, squares) patients without anticonvulsant (two patients), (group 2, triangles) valproic acid (20 patients) and others (group 3, inverted triangles) including carbamazepine or phenobarbital (five patients).

 


View larger version (16K):
[in this window]
[in a new window]
 
Figure 4. Metabolic ratios of irinotecan metabolites in patients with glioblastoma. Metabolite ratios of SN-38, SN-38-G, APC and NPC were calculated as follows: [AUC0–{infty} (metabolite)/AUC0–{infty} (irinotecan)] x 100 for SN-38, SN-38-G and NPC. [AUC0–last sampling (APC)/AUC0–last sampling (irinotecan)] x 100 for APC. Individual (plots) and means ratios (horizontal lines) are presented in patients taking no anticonvulsant (squares), valproic acid (triangles) and other drugs (inverted triangles) for seizure prevention (carbamazepine and phenobarbital).

 

    Discussion
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
In human glioma xenografts in nude mice, irinotecan displayed marked antitumour activity [1214], which then translated into promising antitumour activity in patients with malignant gliomas [2628]. However, clinical trials in patients with gliomas also revealed that the toxicity profile and the pharmacokinetic parameters of irinotecan were strongly modified by concomitant medications such as anticonvulsants. Seizure prevention using carbamazepine, phenytoin and phenobarbital were associated with lower exposure to SN-38; this subsequently led to the resumption of phase I/II clinical trials using increasing doses of irinotecan [2628].

In our study, irinotecan (350 mg/m2 every 3 weeks) displayed a good safety profile which allowed the planned dose intensity to be maintained with limited occurrences of dose reduction and treatment delays. Grade 3–4 toxicity consisted of neutropenia (19.6% per patient) and delayed diarrhoea (7.9% per patient). This toxicity profile was consistent with toxicities reported in phase III studies including patients with colorectal cancer with the same schedule (neutropenia ranged from 14% to 22% and delayed diarrhoea in 22%) [2931]. Similarly, cholinergic syndrome and asthenia were reported in 10% and 9.8% of patients and were comparable with that previously reported in patients with colorectal cancer [30, 31]. Grade 3–4 vomiting was observed in only 3.9% of patients with gliomas and was lower than that in colon cancer patients (14%). This might be related to the antiemetic properties of chronic administration of corticosteroids combined with the effects of the antiemetic regimens used in this study. Two patients experienced severe infections of digestive origin that resulted in toxic death. In these patients, chronic administration of a high dose of corticosteroids could have led to immunosuppression.

In our study, immediate post-operative imaging was not mandatory for patients in group A. However, for patients with complete or partial resection, contrast enhanced tumour masses of >2 cm were required at study entry if these patients were to be considered measurable and therefore eligible for this study. It was considered sufficient to exclude most of the changes in post-surgical radiological enhancement, as these could sometimes be confounded with radiological responses. However, the overall response rate was lower than that reported for anticancer drugs currently used in the treatment of glioblastoma. In addition, our results seem to be lower than those reported by Friedman et al. [26], who showed eight objective responses among 48 patients with glioblastoma using a weekly administration of irinotecan. Poor prognosis factors in our patient population (large tumour sizes, multiple lesions, age, performance status and number of complete or partial resections) might have influenced the result [3537]. For instance, the median age of patients in our study was about 10 years older than in most of the recent chemotherapy trials (median age 53 versus 44 [34], 46 [33] or 46 [26] in other clinical trials). More recently, Cloughesy et al. [38] reported the results of a phase II study where irinotecan was given at a dose of 300 mg/m2 as a 90-min infusion every 3 weeks, which was then escalated to 350 mg/m2 in the absence of severe haematological toxicity. In that study, two patients (14%) presented a partial response and two additional patients had tumour stabilisation with a median time-to-progression of 6 weeks. These data appeared to be comparable with those reported in our study. In our study, the median time-to-progression was 9 weeks (range 3.6–53.1; 95% CI 8.1–22.4) in group A and 14.4 weeks (range 5.5–36.8; 95% CI 9.0–21.1) in group B, and the 6-month progression-free survival rate was 26% in group A and 43% in group B. In this study, the 6-month control rate appears to be similar to that of nitrosourea [32] and temozolomide [33, 34].

The use of experimental agents prior to radiotherapy in newly diagnosed glioblastoma multiforme has become increasingly popular in recent years. This methodology of early screening for new compounds was used recently to optimise the chance of detecting activity that would otherwise be undetectable given the limited survival of patients relapsing after radiotherapy. However, one might also consider that participation in such a study could have negative effects on the outcome of patients by delaying radiation therapy. Grossman et al. [39] recently reported the survival of 368 patients with newly diagnosed glioblastoma treated with investigational new drugs either prior to or following radiotherapy. In their study, no significant difference in survival was detected between the two groups. However, the authors stressed that careful monitoring of tumour progression in patients treated with new agents prior to radiotherapy is mandatory to avoid delaying salvage radiotherapy. In our study, we observed slightly higher progression-free and overall survival rates in group B patients, which could either be due to the effectiveness of radiotherapy or to selection bias that excludes patients with rapidly progressing tumours from receiving radiotherapy. In the absence of postoperative randomisation, it was impossible to compare groups A and B, although both had very similar survival outcomes.

Medications commonly used by patients with malignant gliomas (corticosteroids, phenytoin, carbamazepine and phenobarbital) affect the CYP3A4 enzyme system [26, 4045], cause an increase in the clearance of paclitaxel [40] and irinotecan, and require the use of larger doses of chemotherapy [4245]. In our study, the clearance of irinotecan was increased in four of five patients exposed to phenobarbital or carbamazepine, corroborating previous studies [46].

Interestingly, valproic acid is also registered for seizure prevention and does not interact strongly with the CYP3A4 enzyme. Valproic acid is conjugated in humans, reversibly inhibits hepatic UGT1A1 conjugation, both by competitive and noncompetitive mechanisms [46, 47], and interacts with drugs requiring glucuronidation [48]. In addition, valproic acid and its metabolites exert choleretic effects in animal models [49]. In Wistar rats, concomitant administration of valproic acid 200 mg/kg with irinotecan 20 mg/kg inhibited SN-38-G formation and increased AUCSN-38 by 270% as compared with control rats receiving irinotecan alone [19]. Therefore, valproic acid was expected to result in increased SN-38 exposure and intestinal toxicity in humans. However, our study shows that valproic acid did not increase the rates of diarrhoea and neutropenia as compared with previous studies in patients treated without anticonvulsants [30, 31]. In previously published pharmacokinetic data from phase I/II studies in patients without brain tumours [5053] (Table 4), we found that in patients without gliomas the mean clearances of irinotecan ranged from 11.0 to 21.1 l/h/m2 (Table 4). In our study, the clearance of patients receiving valproic acid (17.2 l/h/m2, range 9.5–27.5) appears to be similar to that observed in patients treated without anticonvulsants (14.2 l/h/m2, range 12.4–16). In our study, the exposure to SN-38 (471.6 ± 224.3 ng·h/ml) was comparable with that previously reported by Abigerges et al. [50] and Rivory et al. [54] in phase I trials using similar doses of irinotecan. This is in contrast with the study by Friedman et al. [26] where the clearance of irinotecan in patients receiving phenytoin, carbamazepine and/or phenobarbital was 30.4 ± 8.3 l/h/m2 and was correlated with low exposures to SN-38 and SN-38-G.

Based on our data, the glucuronidation of SN-38 did not seem to be completely inhibited in patients treated with valproic acid. Discrepancies between species might explain the absence of drug interactions between valproic acid and irinotecan in humans. In addition, we have recently shown that concomitant medications with valproic acid generate new metabolites by oxidation of the camptothecin backbone or the piperidinylpiperidine lateral chain of irinotecan [25]. Other factors that might help to explain the absence of inhibition of UGT1A1 in humans might be related to the schedule of dosage of valproic acid. In humans, valproic acid was given as a chronic daily oral administration starting several days before irinotecan, while in animal experiments a single i.v. valproic acid infusion was given immediately before the administration of irinotecan. Finally, chronic oral corticosteroids might help to counteract the effects of valproic acid by inducing UGT1A1 enzymes.

In summary, irinotecan (350 mg/m2 every 3 weeks) has an effect that results in a median time-to-progression of 9 weeks in group A and 14.4 weeks in group B, with 6-month progression-free survival rates of 26% and 43% in groups A and B, respectively. The toxicity profile of irinotecan requires careful clinical follow-up in patients with glioblastoma. We showed that valproic acid does not increase the exposure to SN-38 and the toxicity of irinotecan in patients with glioblastoma. Although the response rate of irinotecan as a single agent was limited, it remains an attractive drug for use in combination with nitrosourea, temozolomide and radiotherapy in patients with glioblastoma.


    Acknowledgements
 
We gratefully acknowledge the assistance of P. Dielenseger, M. Granier and D. Leleu with the pharmacokinetic sampling carried out in this study. We are grateful to A. Hua and M.L. Risse for their help in the preparation of this manuscript.


    Footnotes
 
+ Correspondence to: Dr Eric Raymond, Department of Medicine, Institute Gustave-Roussy, 39 Rue Camille Desmoulins, 94815 Villejuif, Cedex, France. Tel: +33-1-4211-4289; Fax: +33-1-4211-5217; E-mail: raymond{at}igr.fr Back


    References
 Top
 Abstract
 Introduction
 Patients and methods
 Results
 Discussion
 References
 
1. Laws ER Jr, Thapan K. Brain tumors. CA Cancer J Clin 1993; 43: 263–271.[Abstract/Free Full Text]

2. Sant M, van der Sanden G, Capocaccia R. Survival rates for primary malignant brain tumours in Europe. Eur J Cancer 1998; 34: 2241–2247.[CrossRef][ISI]

3. Burger PC, Vogel F, Green SB, Strike TA. Glioblastoma multiforme and anaplastic, pathologic criteria and prognostic implications. Cancer 1985; 56: 1106–1111.[ISI][Medline]

4. Lesser GJ, Grossman S. The chemotherapy of high-grade astrocytomas. Semin Oncol 1994; 21: 220–235.[ISI][Medline]

5. De Vita VT Jr. Principles of chemotherapy. In De Vita VT Jr, Hellman S, Rosenberg SA (eds): Cancer—Principles and Practice of Oncology. Philadelphia, PA: Lippincott 1993; 276–292.

6. Fine HA, Dear KBG, Loeffler JS et al. Meta-analysis of radiation therapy with and without adjuvant chemotherapy for malignant gliomas in adults. Cancer 1993; 71: 2585–2597.[ISI][Medline]

7. Stenning SP, Friedman LS, Bleehen NM. An overview of published results from randomized studies of nitrosoureas in primary high grade malignant gliomas. Br J Cancer 1987; 56: 89–90.[ISI][Medline]

8. Walker MD, Green SB, Byar DP et al. Randomized comparisons of radiotherapy and nitrosoureas for the treatment of malignant glioma after surgery. N Engl J Med 1980; 303: 1323–1329.[Abstract]

9. Green SB, Byar DP, Walker MD et al. Comparison of carmustine, procarbazine, and high-dose methylprednisolone as additions to surgery and radiotherapy for the treatment of malignant glioma. Cancer Treat Rep 1983; 67: 121–132.[ISI][Medline]

10. Yung WK, Prados MD, Yaya-Tur R et al. Multicenter phase II trial of temozolomide in patients with anaplastic astrocytoma or anaplastic oligoastrocytoma at first relapse. J Clin Oncol 1999; 17: 2762–2771.[Abstract/Free Full Text]

11. Vanhoefer U, Harstrick A, Achterrath W et al. Irinotecan in the treatment of colorectal cancer: clinical overview. J Clin Oncol 2001; 19: 1501–1518.[Abstract/Free Full Text]

12. Vassal G, Boland I, Santos A et al. Potent therapeutic activity of irinotecan (CPT-11) and its schedule dependency in medulloblastoma xenografts in nude mice. Int J Cancer 1997; 3: 156–163.

13. Houghton PG, Cheshire PJ, Hallman JD et al. Efficacy of topoisomerase I inhibitors, topotecan and irinotecan, administered at low dose levels in protracted schedules to mice bearing xenografts of human tumors. Cancer Chemother Pharmacol 1995; 36: 393–403.[CrossRef][ISI][Medline]

14. Coggins CA, Elion GB, Houghton PJ et al. Enhancement of irinotecan (CPT-11) activity against central nervous system tumor xenografts by alkylating agents. Cancer Chemother Pharmacol 1998; 41: 485–490.[CrossRef][ISI][Medline]

15. Mathijssen RHJ, van Alphen RJ, Verweij J et al. Clinical pharmacokinetics and metabolism of irinotecan (CPT-11). Clin Cancer Res 2001; 7: 2182–2194.[Abstract/Free Full Text]

16. Haaz MC, Rivory L, Riche C et al. Metabolism of irinotecan (CPT-11) by human hepatic microsomes: participation of cytochrome P-450 3A and drug interactions. Cancer Res 1998; 58: 468–472.[Abstract]

17. Iyer L, King CD, Whitington PF et al. Genetic predisposition to the metabolism of irinotecan (CPT-11). J Clin Invest 1998; 101: 84–94.

18. Blaney SM, Takimoto C, Murry DJ et al. Plasma and cerebrospinal fluid pharmacokinetics of 9-aminocamptothecin (9-AC), irinotecan (CPT-11) and SN-38 in non-human primates. Cancer Chemother Pharmacol 1998; 41: 464–468.[CrossRef][ISI][Medline]

19. Gupta E, Wang X, Ramirez J, Ratain MJ. Modulation of glucuronidation of SN-38, the active metabolite of irinotecan, by valproic acid and phenobarbital. Cancer Chemother Pharmacol 1996; 39: 440–444.[CrossRef][ISI]

20. Armand JP, Extra Y, Catimel G et al. Rationale for the dosage and schedule of CPT11 (irinotecan) selected for phase II studies, as determined by European phase I studies. Ann Oncol 1996; 7: 837–842.[Abstract]

21. Gandia D, Abigerges D, Armand JP et al. CPT11 induced cholinergic effects in cancer patients. J Clin Oncol 1993; 11: 196–197.[ISI][Medline]

22. Abigerges D, Armand JP, Chabot G et al. Irinotecan (CPT 11) high-dose escalation using intensive high-dose loperamide to control diarrhea. J Natl Cancer Inst 1994; 86: 446–449.[Abstract]

23. MacDonald DR, Cascino TL, Schold SC et al. Response criteria for phase II studies of supratentorial malignant glioma. J Clin Oncol 1990; 8: 1277–1280.[Abstract]

24. Rivory LP, Robert J. Reversed-phase high-performance liquid chromatographic method for the simultaneous quantitation of the carboxylate and lactone forms of the camptothecin derivative irinotecan, CPT-11, and its metabolite SN-38 in plasma. J Chromatogr B Biomed Appl 1994; 661: 133–141.[CrossRef][Medline]

25. Santos A, Zanetta S, Cresteil T et al. Metabolism of irinotecan (CPT-11) by CYP3A4 and CYP3A5 in humans. Clin Cancer Res 2000; 6: 2012–2020.[Abstract/Free Full Text]

26. Friedman HS, Petros WP, Friedman AH et al. Irinotecan therapy in adults with recurrent or progressive malignant glioma. J Clin Oncol 1999; 17: 1516–1525.[Abstract/Free Full Text]

27. Buckner J, Reid J, Schaaf L et al. A phase II trial of irinotecan (CPT-11) in recurrent glioma. Proc Am Soc Clin Oncol 2000; 19: (Abstr 679a).

28. Gilbert MR, Supko J, Grosman SA et al. Dose requirements, pharmacology and activity of CPT-11 in patients with recurrent high grade glioma. A NABTT CNS Consortium trial. Proc Am Soc Clin Oncol 2000; 19: (Abstr 622).

29. Eckhardt SG. Irinotecan: a review of the initial phase I trials. Oncology 1998; 12 (8 Suppl 6): 31–38.

30. Rougier P, Van Cutsem E, Bajetta E et al. Randomised trial of irinotecan versus fluorouracil by continuous infusion after fluorouracil failure in patients with metastatic colorectal cancer. Lancet 1998; 352: 1407–1412.[CrossRef][ISI][Medline]

31. Cunningham D, Pyrhönen S, James RD et al. Randomised trial of irinotecan plus supportive care versus supportive care alone after fluorouracil failure for patients with metastatic colorectal cancer. Lancet 1998; 352; 1413–1418.[CrossRef][ISI][Medline]

32. Levin VA, Edwards MS, Wright DC et al. Modified procarbazine, CCNU and vincristine (PCV-3) combination chemotherapy in the treatment of malignant brain tumors. Cancer Treat Rep 1980; 64: 237–241.[ISI][Medline]

33. Newlands ES, O’Reilly SM, Glaser MG et al. The Charing Cross Hospital experience with temozolomide in patients with gliomas. Eur J Cancer 1996; 32A: 2236–2246.[CrossRef]

34. Bower M, Newlands ES, Bleehen NM et al. Multicentre CRC phase II trial of temozolomide in recurrent or progressive high-grade glioma. Cancer Chemother Pharmacol 1997; 40: 484–488.[CrossRef][ISI][Medline]

35. Grant R, Liang BC, Page MA et al. Age influences chemotherapy response in astrocytomas. Neurology 1995; 45: 929–933.[Abstract]

36. Grant R, Liang BC, Slattery MSC et al. Chemotherapy response criteria in malignant glioma. Neurology 1997; 48: 1336–1340.[Abstract]

37. Prados MD, Gutin PH, Phillips TL et al. Highly anaplasic astrocytoma: a review of 357 patients treated between 1977 and 1989. Int J Radiat Oncol Biol Phys 1992; 23: 3–8.[ISI][Medline]

38. Cloughesy TF, Filka E, Nelson G et al. Irinotecan treatment for recurrent malignant glioma using an every-3-week regimen. Am J Clin Oncol 2002; 25: 204–208.[CrossRef][ISI][Medline]

39. Grossman SA, Carson K, Piantadosi S, Fisher J. Survival of adults with newly diagnosed glioblastoma multiforme (GBM) treated with experimental agents and delayed radiation therapy (RT) versus experimental agents and immediate radiation: the experience of the NABTT CNS Consortium. Proc Am Soc Clin Oncol 2002; 21: 71a (Abstr 282).

40. Fetell MR, Grossman SA, Fisher JD et al. Preirradiation paclitaxel in glioblastoma multiforme: efficacy, pharmacology, and drug interactions. J Clin Oncol 1997; 15: 3121–3128.[Abstract]

41. Grossman SA, Hochberg F, Fisher J et al. Increased 9-aminocamptothecin dose requirements in patients on anticonvulsants. Cancer Chemother Pharmacol 1998; 42: 118–126.[CrossRef][ISI][Medline]

42. Reid JM, Buckner JC, Scaaf LJ et al. Anticonvulsants alter the pharmacokinetics of irinotecan (CPT-11) in patients with recurrent glioma. Proc Am Soc Clin Oncol 2000; 19: (Abstr 620).

43. Cloughesy T, Filka E, Friedman H et al. A phase I (intrapatient dose escalation) open-label study of irinotecan (CPT-11) in patients with recurrent or progressive malignant glioma. Proc Am Soc Clin Oncol 2000; 19: (Abstr 624).

44. Prados M, Kuhn J, Yung WKA et al. A phase-I study of CPT-11 given every three weeks to patients with recurrent malignant glioma. A North American Brain Tumor Consortium (NATBC) study. Proc Am Soc Clin Oncol 2000; 19: (Abstr 627).

45. Friedman HS, Cokgor I, Tourt-Uhlig S et al. Phase I trial of CPT-11 plus BCNU in malignant glioma. Proc Am Soc Clin Oncol 2000; 19: (Abstr 659).

46. Reid JM, Cha S, Buckner JC et al. Pharmacokinetics of CPT-11 in glioma patients: pooled analysis of data from four NCI-sponsored trials (Duke Universtity, NABTC, NABTT, NCCTG). Proceedings of the AACR–NCI–EORTC International Conference 415; 2001 (Abstr).

47. Howell SR, Hazelton GA, Klaassen CD. Depletion of hepatic UDP-glucuronic acid by drugs that are glucuronidated. J Pharmacol Exp Ther 1986; 236: 610–614.[Abstract]

48. Taburet AM, Aymard P. Valproate glucuronidation by rat liver microsomes. Interaction with parahydroxyphenobarbital. Biochem Pharmacol 1983; 32: 3859–3861.[CrossRef][ISI][Medline]

49. Watkins JB, Klaassen CD. Effects of inducers and inhibitors of glucuronidation on biliary excretion and choleretic action of valproic acid in rat. J Pharmacol Exp Ther 1982; 220: 305–310.[Abstract]

50. Abigerges D, Chabot GG, Armand JP et al. Phase I and pharmacokinetic studies of the camptothecin analog irinotecan administered every 3 weeks in cancer patients. J Clin Oncol 1995; 13: 210–221.[Abstract]

51. Rothenberg ML, Kuhn JG, Burris HA et al. Phase I and pharmacokinetic trial of weekly CPT-11. J Clin Oncol 1993; 11: 2194–2204.[Abstract]

52. Rowinski EK, Grochow LB, Ettinger DS et al. Phase I and pharmacological study of the novel topoisomerase I inhibitor 7-ethyl-10-[4-(1-piperidino)-1-piperidino] carbonyloxycamptothecin (CPT-11) administered as a ninety-minute infusion every 3 weeks. Cancer Res 1994; 54: 427–436.[Abstract]

53. Gupta E, Mick R, Ramirez J et al. Pharmacokinetic and pharmacodynamic evaluation of topoisomerase inhibitor irinotecan in cancer patients. J Clin Oncol 1997; 15: 1502–1510.[Abstract]

54. Rivory LP, Haaz MC, Canal P et al. Pharmacokinetic interrelationships of irinotecan (CPT-11) and its three major metabolites in patients enrolled in phase I/II trials. Clin Cancer Res 1997; 3: 1261–1266.[Abstract]