A phase 2 trial of irinotecan (CPT-11) in patients with recurrent malignant glioma: A North American Brain Tumor Consortium study

Abstract

The purpose of this study was to determine the response to CPT-11 administered every three weeks to adults with progressive malignant glioma, treated with or without enzyme-inducing antiepileptic drug (EIAED) therapy, at the recommended phase 2 dose determined from a previous phase 1 study. Adult patients age 18 or older with a KPS of 60 or higher who had measurable recurrent grade III anaplastic glioma (AG) or grade IV glioblastoma multiforme (GBM) were eligible. No more than one prior chemotherapy was allowed, either as adjuvant therapy or for recurrent disease. The CPT-11 dose was 350 mg/m² i.v. every three weeks in patients not on EIAED and 750 mg/m² in patients on EIAED therapy. Patients with stable or responding disease could be treated until tumor progression or a total of 12 months of therapy. The primary end point of the study was to determine whether CPT-11 could significantly delay tumor progression, using the rate of six-month progression-free survival (PFS-6). The trial was sized to be able to discriminate between a 15% and 35% rate for the GBM group alone and between a 20% and 40% rate for the entire cohort. There were 51 eligible patients, including 38 GBM and 13 AG patients, enrolled. The median age was 52 and 42 years, respectively. PFS-6 for the entire cohort was 17.6%. PFS-6 was 15.7% (95% confidence interval [CI], 0.07–0.31) for the GBM patients and 23% (95% CI, 0.07–0.52) for AG patients. Toxicity for the group included diarrhea and myelosuppression. We conclude that the recommended phase 2 dose of CPT-11 for patients with or without EIAED was ineffective on this schedule, in this patient population.

The treatment of recurrent malignant gliomas using available chemotherapy drugs is ineffective for the prolongation of life in the majority of patients. The median time to tumor progression after treatment for patients with recurrent glioblastoma multiforme (GBM)⁴ is only 8 to 12 weeks, and fewer than 20% of patients are progression free by six months (Hess et al., 1999). Patients with grade 3 anaplastic gliomas (AG) fare better, but fewer than half are progression free at six months. Irinotecan (CPT-11), a drug that is active against colon carcinoma, has been used to treat recurrent glioma. Preclinical activity of CPT-11 has been promising, with in vitro activity noted against a panel of xenografts derived from ependymoma, childhood and adult high-grade astrocytoma, and medulloblastoma (Hare et al., 1997). A phase 2 study of CPT-11 for malignant glioma was conducted at Duke University that used doses and schedules recommended for the treatment of colon cancer (Friedman et al., 1999). The results suggested clinical efficacy with less toxicity, particularly diarrhea, than has been seen in other cancer patient populations. This lack of toxicity in many cases may have been due to treatment of the glioma patients with enzyme-inducing antiepileptic drug (EIAED) therapy, which could alter the exposure to CPT-11 in this population. Preclinical and clinical studies have demonstrated that EIAEDs can alter the metabolism and elimination of CPT-11 and its metabolites (Haaz et al. 1998). Pretreatment of rats with the antiepileptic drug phenobarbital, an inducer of UGT-1A1 and CYP3A4, has been shown to enhance the formation of SN-38G and decrease the area under the concentration–time curve (AUC) for both CPT-11 and SN-38. However, the non-enzyme-inducing antiepileptic drug valproic acid increased the AUC of SN-38, presumably by inhibiting the glucuronidation of SN-38 (Gupta et al., 1997). These results have been confirmed in clinical trials of CPT-11 in patients with glioma (Reid et al., 2001).

Based on the aforementioned background, the North American Brain Tumor Consortium (NABTC) completed a phase 1 study of CPT-11 in patients on EIAED to determine the optimal phase 2 dose when given every three weeks, and to compare the pharmacokinetics with those in patients not on EIAED who are treated at the recommended phase 2 dose for other cancers (Prados et al., 2004). The recommended phase 2 dose was 750 mg/m² given every three weeks to patients on EIAED. A phase 2 study was then conducted based on the results of the phase 1 trial. Patients not on EIAED were treated at a dose of 350 mg/m². This report details the results of the phase 2 study.

Patients and Methods

Study Population and Patient Eligibility

Patients 18 years or older with a histologically confirmed diagnosis of a progressive or recurrent malignant glioma were eligible to participate, provided they had measurable disease, a KPS of ⩾60, and acceptable hematologic, liver, and renal function. The latter required an absolute neutrophil count of ⩾1500/mm³ and platelet count of ⩾100,000/mm³, serum creatinine level <1.5 mg/dl, serum bilirubin level <1.5 mg/dl, and aspartate amino-transferase level less than three times the institutional upper normal limits. Patients in this study had undergone no more than one prior chemotherapy regimen, including either one prior adjuvant therapy or one prior regimen for recurrent tumor. Patients not on EIAED (group A) received a dose of 350 mg/m², and patients on EIAED (group B) were treated at a dose of 750 mg/m². In both groups, the interval from prior irradiation or chemotherapy had to be at least four weeks, and six weeks if prior nitrosourea therapy had been used. All patients had a retrospective central neuropathology review.

Patients previously treated with CPT-11, topotecan, or other topoisomerase 1 inhibitors were excluded. All patients were provided with and had to sign a written informed consent approved by the local Institutional Review Board at each institution before treatment, informing them of the investigational nature of this study.

Treatment

Patients who met eligibility requirements received an intravenous infusion of CPT-11 over a 90-min period every three weeks. Patients continued to receive this dose throughout treatment until tumor progression, unacceptable toxicity, or completion of 12 treatment cycles. A treatment cycle was considered one infusion and a three-week evaluation period. No dose escalation was allowed. Dose reduction was allowed if during the prior infusion there had been evidence of toxicity as defined by the NCI Common Toxicity Criteria version 2.0 scale.

Patient Monitoring and Toxicity Assessment

Complete physical and neurological examinations, including KPS, were performed every other cycle. Weekly complete blood counts with differential and platelets were obtained throughout treatment. Levels of creatinine, blood urea nitrogen, total bilirubin, aspartate aminotransferase, and serum electrolytes were measured before each cycle. Magnetic resonance imaging of the brain was done every other cycle to assess response. Patients with stable or responding disease received the same dose at the next cycle or a reduced dose if adverse events were observed in the current cycle. If a patient experienced grade 3 or greater nonhematologic toxicity or grade 4 hematologic toxicity, the dose of the subsequent cycle was reduced by one level (50 mg/m²). If the toxicity was thought to be directly related to CPT-11, subsequent doses were not reescalated, even if there was minimal or no toxicity at the reduced dose. A new course of treatment could begin if the absolute neutrophil count was ⩾1500/mm², the platelet count was ⩾100,000/mm³, and any other treatment-related toxicities were ⩽grade 1. If, after a one-week delay, toxicities were ⩽grade 1, treatment resumed. If the toxicity did not resolve in one week, a second one-week delay was allowed, but retreatment required a reduction of one dose level. If retreatment had to be held off for more than two weeks, or if the administered dose would be ⩽200 mg/m², the patient was removed from the study.

Patients were removed from study if tumor progression was determined by MRI. The definition of progression was a 25% increase in the sum of the products of all measurable disease over the smallest sum observed, a clear worsening of any evaluable disease, or the appearance of any new lesion. Failure to return for evaluation due to death or deteriorating condition was considered to represent disease progression. Although this was not an imaging response study, claimed responses by the individual investigator were recorded as either partial response (decrease in cross-sectional area by 50% with stable or improving steroids) or stable disease (less than a partial response, but not qualified as progression). There was no central review of claimed partial responses.

Statistical Considerations

The goal of this study was to determine whether CPT-11 could significantly delay progression in patients with recurrent malignant glioma. The primary end point was progression-free survival at six months (PFS-6). The trial cohort includes both GBM and AG patients who were entered at an anticipated 2:1 ratio. The trial was sized to be able to discriminate between a 20% and 40% rate of PFS-6 for the entire cohort and a 15% and 35% rate for the GBM group alone. The GBM comparison is the one of primary concern. With accrual of 32 GBM patients, the trial would be considered a success if at least eight GBM patients showed PFS-6. This would give a 0.92 probability of detecting a 35% rate of PFS-6, with 0.9 probability of rejecting the agent if the PFS-6 were only 15%. Assuming an accrual of approximately 16 AG patients, there would be a reasonable power to discriminate between a 20% and 40% rate of PFS-6 for the entire group. For the group as a whole, CPT-11 would be considered effective if there was at least 30% PFS-6. This would give at least a 0.9 probability of detecting a 40% rate of PFS-6, with at least a 0.9 probability of rejecting the drug if the PFS-6 was only 20%. Because central neuropathology review was done retrospectively, a small increase in patient numbers (five additional patients) was allowed in order to have a sufficient number of eligible patients with GBM, the cohort of primary concern. As noted below, the final result of GBM enrollment was higher than anticipated.

Results

Patient Characteristics

A total of 53 patients were enrolled in the phase 2 study. One patient was declared ineligible after central pathology review revealed a grade II astrocytoma. An additional patient withdrew consent prior to any treatment. Thus, 51 eligible and treated cases were evaluated in this report (Table 1). The median age was 42 years and 52 years for AG and GBM, respectively. The majority of patients (38 of 51) had recurrent GBM, the remainder having recurrent grade III tumors.

Table 1.

Patient characteristics

Characteristic	Number of Patients*
Number of eligible patients	51
Glioblastoma multiforme	38
Anaplastic glioma	13
Anaplastic astrocytoma	10
Anaplastic oligodendroglioma	1
Anaplastic oligoastrocytoma	2
Males/females	35/16
Median age
Glioblastoma multiforme	52 (range, 27–68)
Anaplastic glioma	42 (range, 29–68)
KPS
100	17.65%
90	21.57%
80	37.25%
70	15.69%
60	7.84%

^*KPS is given as percent of patients.

Toxicity

Only four patients discontinued treatment because of toxicity; the remainder continued treatment until tumor progression or the completion of planned therapy. The majority of recorded toxicities were grade 1 and grade 2. Grade 3 and 4 toxicities were more frequent in patients not on EIAEDs (group A), including a higher rate of grades 3 and 4 granulocytopenia. Patients on EIAEDs (group B) had no grade 4 toxicities, and patients in group A had 10 grade 4 events (all hematologic).

Efficacy

The primary end point was the number of patients who were progression free at six months. Overall, the PFS-6 for the entire group was 17.6% (9 of 51 patients). Of the 13 AG patients, three were known to have time to progression longer than six months (95% confidence interval [CI], 0.07–0.52, response rate). Of the 38 grade 4 patients, six were known to have time to progression of longer than six months (95% CI, 0.07–0.31, response rate). Thus, even with a slight increase in eligible patients with GBM, only six of 38 met the PFS-6 target rather than the eight of 32 anticipated for calling this a successful regimen, and the upper bound of the 95% CI was less than 35%. For the entire group of 51 patients, the best claimed radiographic response was immediate disease progression in 30, stable disease in 17, and partial responses (PRs) in three (5.8% PR rate). One patient was not evaluable for response because the patient was treated with only one cycle of CPT-11 and was removed because of toxicity prior to MRI.

Discussion

The goal of this phase 2 study was to evaluate efficacy of CPT-11 with the use of the recommended phase 2 dose for patients based on EIAED use, which was based on a prior phase 1 toxicity and pharmacokinetic study. Previous phase 2 studies have been conducted with this agent in this patient population without modifying the dose, which were based on pharmacokinetic considerations, as well as phase 2 studies that evaluated different schedules of administration using doses based on EIAED use. The current study failed to show efficacy, measured as PFS-6, in patients with recurrent GBM treated at the time of first relapse by using a once-every-three-week schedule. We conclude that single-agent CPT-11 used at the maximum tolerated dose (MTD) is ineffective in recurrent malignant glioma.

A number of clinical trials have now been reported using CPT-11 in various schedules and doses. Friedman et al. (1999) reported 60 patients (48 GBM patients) treated at a dose of 125 mg/m² once weekly for four weeks with a two-week break. Of these, nine (15%) had a PR and 33 (55%) had stable disease. Median time to tumor progression was 18 weeks in the GBM group, but ranged from 6 to 68+ weeks at the time of the report. The PFS-6 was not described. Friedman et al. concluded that CPT-11 was active and likely to have more activity with strategies to increase drug exposure. Chamberlain (2002) reported a phase 2 study of 40 patients with recurrent GBM, treated with a starting dose of 400 mg/m², which was increased to 500 mg/m² every three weeks. Fifteen patients were not receiving EIAEDs. In this study, the disease of all patients progressed within six weeks, and the author concluded the drug was ineffective in this patient population. Cloughesy et al. (2003) reported on two studies using a three-week schedule. In the first group, 14 patients were treated at a starting dose of 300 mg/m², which was escalated ultimately to 350 mg/m². There were two PRs and two patients with stable disease, with median time to tumor progression of six weeks. The second group of 35 patients (26 with GBM) was treated with intrapatient dose escalation, increasing the dose by 100 mg/m² increments (based on toxicity of the prior infusion) every three weeks. Dose-limiting toxicity ranged from 400 to 1700 mg/m², with three cases of PR and 15 with stable disease. Median time to tumor progression was 2.7 months, and the estimated PFS-6 was 26% (confidence intervals not given). The authors concluded that the drug was active and that a higher level of efficacy might be possible if an MTD could be established for each patient. Raymond et al. (2003) reported a phase 2 study using a dose of 350 mg/m² every three weeks, with the preferred anticonvulsant being valproic acid. Twenty-five patients were treated prior to radiation, and the remaining patients were treated following disease progression after radiotherapy. All patients were chemotherapy naive. The overall response rate was 2.2%, estimated PFS-6 was 46%, and median time to tumor progression was 14.4 weeks in the patients whose disease recurred after radiotherapy. The authors concluded that CPT-11 has limited clinical activity as a single agent. Buckner et al. (2003) summarized the results of two sequential phase 2 studies conducted by the North Central Cancer Treatment Group. The first trial treated patients with a dose of either 125 mg/m² or 100 mg/m² weekly for four weeks with a two-week break, using the lower dose if prior nitrosourea therapy had been given. The second trial used a dose of 300 mg/m² or 250 mg/m² given once every three weeks, using the lower dose in cases of prior nitrosourea treatment. A total of 64 patients were enrolled in both studies. The response rate was 10%, but the duration of the response was not reported.

None of the aforementioned studies systematically investigated the efficacy of CPT-11 based on an MTD pharmacologically derived in patients on or not on EIAEDs. However, Batchelor et al. (2004) reported on a multi-center, two-stage, phase 2 study of weekly administration of CPT-11, using a dose based on the prior phase 1 study conducted by the New Approaches to Brain Tumor Treatment (NABTT) consortium. Patients on EIAEDs were treated with a dose of 411 mg/m² weekly for four weeks, and those not on EIAEDs were treated with a dose of 117 mg/m² weekly for four weeks, followed by a two-week break. The estimated PFS-6 was 56% (95% CI, 31%–79%). The response rate was 6% of the 18 patients (12 with GBM) treated in the first stage of this study. Because of the low response rate, the study was terminated after the first stage. The authors concluded that CPT-11 has minimal efficacy in this patient setting.

The current study, similar to the NABTT trials, sought to evaluate the clinical efficacy of CPT-11 based on the pharmacokinetics of this agent, and as in the former study, little benefit was found in use of the drug. It appears that either dose schedule (weekly for four weeks or every three weeks) is ineffective in patients with recurrent malignant glioma, both in terms of objective response rates or duration of response. However, given the observation that some patients had objective responses, the drug may be of some use in other treatment settings, such as in combination with other agents or radiotherapy. We would conclude this drug as a single agent has little efficacy and should not be used in patients with recurrent malignant glioma.

Institution	Investigators	NABTC Grant	GCRC Grant
University of California, San Francisco	Michael D. Prados* Susan Chang M. Kelly Nicholas	CA 62422	M01-RR00079
University of Texas M.D. Anderson	W.K.A. Yung* Kurt Jaeckle	CA 62412	None
University of Texas Southwestern	Karen Fink*	CA 62455	M01-RR00633
Dana Farber/Brigham and Women’s	Howard A. Fine* Patrick Y. Wen	CA 62407	None
University of Pittsburgh	David Schiff*	CA 62404	M01-RR00056
University of Texas at San Antonio	John Kuhn*	CA 62426	M01-RR0134
University of California, Los Angeles	Timothy Cloughesy*	CA 62399	M01-RR0865
University of Michigan	Harry Greenberg* Larry Junck	CA 62399	M01-RR00042
University of Wisconsin	Minesh Mehta* H. Ian Robins	CA 62421	M01-RR03186