Journal of Neuro-Oncology

, Volume 111, Issue 3, pp 347–353

A phase II trial of oral gimatecan for recurrent glioblastoma

Authors

    • Johnnie L. Cochran Jr. Brain Tumor Center
  • Patrick Y. Wen
    • Center For Neuro-Oncology, Dana-Farber Cancer Institute
  • Lauren E. Abrey
    • Department of Neurology, University of Zurich
  • Camilo E. Fadul
    • Norris Cotton Cancer Center, Dartmouth Hitchcock Medical Center
  • Jan Drappatz
    • UPMC Cancer Pavilion
  • Nadia Salem
    • Sigma-Tau Pharmaceuticals Inc.
  • Jeffrey G. Supko
    • Clinical Pharmacology LaboratoryMassachusetts General Hospital
  • Fred Hochberg
    • Stephen E. & Catherine Pappas Center for Neuro-Oncology, Massachusetts General Hospital Cancer CenterHarvard Medical School
Clinical Study

DOI: 10.1007/s11060-012-1023-0

Cite this article as:
Hu, J., Wen, P.Y., Abrey, L.E. et al. J Neurooncol (2013) 111: 347. doi:10.1007/s11060-012-1023-0

Abstract

Gimatecan is a lipophilic oral camptothecin analogue with preclinical activity in glioma models. We conducted a multicenter phase II trial to evaluate the efficacy of gimatecan in adults with recurrent glioblastoma. Eligibility criteria included ≤1 prior treatment for recurrent disease, age ≥18, Eastern Cooperative Oncology Group performance status 0–1, and normal organ function. Patients taking enzyme-inducing anti-seizure medications were excluded. Gimatecan 1.22 mg/m2 was given orally once daily for 5 consecutive days during each 28-day cycle. The primary endpoint was progression-free survival at 6 months. A Simon 2-stage optimal design was used in which 19 patients were evaluated in the 1st stage, with an additional 36 patients accrued if >4 patients in stage 1 achieved PFS at 6 months. 29 patients were enrolled in the study, with median age of 58 years (range, 25–77 years); 58.6 % female. All patients received prior surgery, radiation therapy, and at least one chemotherapy regimen. The daily dose was reduced to 1.0 mg/m2 after four of the first 10 patients experienced grade 4 hematologic toxicity. Treatment-related grade 3/4 toxicities included thrombocytopenia (17.2 %), leukopenia (17.2 %) and neutropenia (10.3 %). None of the 19 patients treated at 1.0 mg/m2/day experienced grade 4 hematologic toxicity. One patient had a partial radiographic response by modified Macdonald criteria. Only 3 patients (12 %) were progression-free at 6 months. Median time to progression was 12.0 weeks (7.0, 17.0).Treatment with gimatecan 1.0 mg/m2/day for 5 days, repeated every 28-days showed minimal efficacy.

Keywords

GimatecanGlioblastomaClinical trialCamptothecinBrain tumorGlioma

Introduction

Despite advances over the past decade, prognosis for patients with glioblastoma remains poor. Standard therapy with radiation and temozolomide carries a median survival of 14.6 months, and median progression-free survival of 6.9 months [1]. Prognosis at recurrence is less clearly defined but no less dire. Patients treated on the lomustine control arm of the negative phase III clinical trial of enzastaurin had a median survival of 7.1 months, with 6-month progression-free survival (PFS-6) of 19 % [2]. The cohort of patients treated with bevacizumab monotherapy on the phase II clinical trial that led to FDA approval of bevacizumab for recurrent glioblastoma had a median survival of 9.2 months, with PFS-6 of 42.6 % [3]. More effective therapies are clearly needed.

As a class, camptothecin derivatives have demonstrated anti-neoplastic activity against a wide range of malignancies. By disrupting the function of DNA topoisomerase I, these agents act as cell cycle specific inhibitors of DNA transcription, replication, and repair. Two camptothecin analogues are currently approved for clinical practice today–irinotecan and topotecan. Irinotecan is approved for treatment of advanced colorectal cancer, both as first-line therapy in combination with 5-FU and as salvage treatment in 5-FU refractory disease. Topotecan is used as second-line therapy for advanced ovarian, cervical, and small cell lung cancer [4].

Although both topotecan and irinotecan demonstrated activity against malignant gliomas in preclinical models, clinical trials of these agents have yielded disappointing results [57]. Though irinotecan is still occasionally used off-label in combination with bevacizumab, such use has declined since a phase II trial did not detect any significant difference in survival between patients treated with bevacizumab alone vs combination therapy, even though the study was not designed to detect such a difference as a primary objective [3]. More recently, convection-enhanced delivery of topotecan improved survival in a platelet-derived growth factor (PDGF)-driven rat glioblastoma model, but clinical trial data has not been reported [8].

Irinotecan and topotecan were initially selected for development because their water solubility helps facilitate intravenous administration. It is now recognized, however, that lipophilic derivatives provide better tumor penetration, particularly for brain tumors [9]. Increased lipophilicity also promotes rapid intracellular accumulation and helps trap the DNA-topoisomerase I complex in a normally transient conformation, thereby initiating the cytotoxic response [10].

Gimatecan (7-[(E)-tert-butyloxyminomethyl]-camptothecin) is a highly lipophilic oral camptothecin analogue [9]. Preclinical studies of gimatecan demonstrate potent activity against a wide array of tumor types, including orthotopic human glioma xenografts [11]. Over 400 patients have received gimatecan in trials for advanced solid tumors, colorectal, breast, and gynecologic cancer, as well as in the phase I portion of this study [1216]. Disease activity was demonstrated in a phase II trial for patients with recurrent epithelial ovarian, fallopian tube, or peritoneal cancer, with partial response by combined CA-125 and RECIST criteria occurring in 24.6 % of patients and disease stabilization in 31.9 % [16]. Cumulative clinical experience indicates that myelosuppression (thrombocytopenia and neutropenia) is the dose limiting toxicity. Non-hematological toxicities include nausea, vomiting, and fatigue.

In the phase I portion of this study, oral gimatecan was given for 5 consecutive days every 28 day cycle in escalating doses. The dose was escalated independently in 2 study arms–patients taking enzyme-inducing anti-seizure drugs (EIASDs), and patients who were not. For patients not taking EIASDs, the maximum tolerated dose (MTD) was established as 1.22 mg/m2/day for 5 days each cycle. Dose limiting toxicities were thrombocytopenia and neutropenia. The study arm of patients taking EIASDs was discontinued at a dose of 2.99 mg/m2/day for 5 days each cycle prior to determination of the MTD, as further dose escalation would have required taking a prohibitively large number of pills. Concurrent use of EIASDs has also been found to significantly increase total body clearance for other camptothecin analogues [5, 1719]. Based on these results, the phase II portion of this trial was restricted to patients not taking EIASDs, and the starting dose was initially set at 1.22 mg/m2/day for 5 days each cycle.

Methods

Objective

The primary endpoint of this single-arm phase II trial was to evaluate the efficacy of oral gimatecan for the treatment of recurrent glioblastoma as determined by 6 month progression-free survival. Secondary endpoints included tolerability, toxicity, and radiographic assessment.

Patient eligibility

All patients enrolled in this trial were greater than 18 years of age with histologically confirmed diagnoses of progressive or recurrent glioblastoma as determined by contrast-enhanced imaging obtained within 2 weeks prior to starting the study drug. In addition to radiation therapy and adjuvant chemotherapy administered at initial diagnosis, patients were allowed to have 1 prior chemotherapy regimen for progressive or recurrent disease, as long as these therapies were discontinued 4 weeks before trial entry (6 weeks if prior therapy was nitrosourea-based). Patients were required to have acceptable hematological, kidney, and liver function at the time of trial enrollment (including absolute neutrophil count (ANC) >1.5 × 109/L; platelets >100 × 109/L; creatinine <1.5; alanine aminotransferase (ALT) and aspartate aminotransferase (AST) <1.5 times the upper limit of normal), with life expectancy of at least 3 months and an Eastern Cooperative Oncology Group (ECOG) performance status of 0–1. Patients were not allowed to take EIASDs for at least 14 days prior to study enrollment. Patients taking corticosteroids were required to be on a stable dose at least 5 days before drug administration. Pregnant women, and patients with uncontrolled intercurrent illness, a thromboembolic event within the past 6 months, or a medical condition that could potentially interfere with drug metabolism were excluded from the study. The protocol and consent for this trial were approved by the local institutional review board at each participating institution and all patients reviewed, signed, and provided written informed consent before enrollment.

Treatment regimen

Gimatecan was supplied by Sigma-Tau Research (Gaithersburg, MD) as oral capsules containing 0.1, 0.25, or 0.5 mg of active drug. The original protocol called for gimatecan 1.22 mg/m2/day to be given orally once daily for 5 consecutive days during each 28 day cycle, as this was the MTD defined in the prior phase I trial. However, after a planned review following treatment of 10 patients at this dose revealed significant myelosuppression (detailed below), the starting daily dose was reduced one level to 1.0 mg/m2/day at the recommendation of the data safety monitoring board. Treatment with gimatecan was to be continued up to 1 year (12 cycles), or until disease progression, unacceptable toxicity, or withdrawal of consent, whichever came first.

Patient monitoring and dose modifications

Patients were closely monitored throughout therapy for drug-related toxicity, and all adverse events were recorded and graded according to the National Cancer Institute Common Toxicity Criteria, version 2.0. Physical and neurologic examinations were performed at the start of each 28 day cycle, and laboratory tests were drawn on days 1 and 22 of each cycle. Patients with stable or responding disease were allowed to proceed with the subsequent cycle of therapy, provided that laboratory assessments were within an acceptable range. Dose delay of at least 1 week was mandated for grade 2 thrombocytopenia and grade 3 neutropenia; more severe hematologic toxicity resulted in dose reductions to predetermined levels (0.8 mg/m2/day or 0.64 mg/m2/day for patients who started at 1 mg/m2/day; 1.0 mg/m2/day, 0.8 mg/m2/day, or 0.64 mg/m2/day for patients who started at 1.22 mg/m2/day).

Imaging and response assessment

Gadolinium-enhanced magnetic resonance imaging (MRI) was obtained at baseline and repeated after every other cycle (8 week intervals) to assess response. Computed tomography (CT) imaging was allowable only for patients with a contraindication to obtaining an MRI. Imaging responses were assessed by modified Macdonald response criteria [20]. Complete response (CR) was defined as the complete disappearance of all measurable and assessable disease. Partial response (PR) was defined as a 50 % decrease in the sum of products of perpendicular diameters of all measurable lesions compared to baseline. Complete responders were required to have discontinued corticosteroid use except if needed for physiologic maintenance. Partial responders were required to have stable or improved neurologic examinations on equal or lesser corticosteroid dose. Progressive disease (PD) was defined as a 25 % increase in the sum of the products of all measurable disease over the smallest sum observed, clear worsening of any assessable disease, or the appearance of any new lesion. Stable disease (SD) or no response was defined as those patients with a tumor status that did not qualify for CR, PR, or PD. Failure to return for evaluation because of death or deteriorating condition was considered to represent PD.

Drug level monitoring

Blood samples were obtained to monitor the plasma concentration of gimatecan just before the fifth daily dose was taken during cycles 1 and 3. Blood (5 mL) was drawn from a peripheral arm vein within 5 min of dosing into a collection tube with freeze dried sodium heparin. Plasma was afforded by centrifugation (1,100–1,300×g, 10 min, 4 °C) and stored at −70°C until assayed. Total gimatecan was determined by high-performance liquid chromatography with fluorescence detection as previously reported [12].

Study design and statistical considerations

This study utilized a Simon two-stage optimal trial design with PFS-6 as the primary endpoint [22]. Time to progression (TTP) was defined as the time from initiation of therapy to the date of disease progression or death. P0 was set at 0.15, based on a historical PFS-6 of 15 % for glioblastoma patients enrolled in 8 phase II treatment trials of agents which were considered to be ineffective [22]. P1 was set at 0.30 to assess for a 15 % absolute improvement in PFS-6 over the historical value, accepting a false-positive rate (α) ≤5 % and a false-negative rate (ß) ≤20 % (i.e, power = 80 %). Using these parameters, 19 patients were to be enrolled in the first stage, with an additional 36 patients to be accrued in the second stage if four or more patients in the first stage achieved PFS-6.

Results

Patient characteristics

From October 2005 to April 2007, a total of 29 patients were enrolled in this phase II study. Patient characteristics are summarized in Table 1. Median patient age was 58 years (range, 25–77 years). 58.6 % of enrollees were women, which contrasts with the 55 % male predominance for glioblastoma in the general population. Twenty-three (79.3 %) of the patients had a baseline ECOG score of 1; the remainder had a baseline ECOG score of 0. All patients had prior surgery, radiation therapy, and chemotherapy with temozolomide. Eighteen patients (62.1 %) were treated with temozolomide as the sole prior chemotherapy. The other 11 patients (37.9 %) received a variety of prior chemotherapeutic treatments, including two patients who were treated with antiangiogenic therapy (one patient treated with bevacizumab, another with cediranib).
Table 1

Patient characteristics, including age, gender, ECOG performance status on enrollment, and prior treatment

# of patients

 

 ITT population

29

 EE population

25

Median Age (range)

58 (25–77)

Gender (%)

 

 Male

12 (41.4 %)

 Female

17 (58.6 %)

Baseline ECOG (%)

 

 0

6 (20.7 %)

 1

23 (79.3 %)

Prior treatment (%)

 

 Surgery (%)

29 (100 %)

 RT (%)

 

  IFRT

29 (100 %)

  SRS

2 (6.9 %)

 Chemo (%)

29 (100 %)

  Temozolomide

29 (100 %)

  Cediranib

1 (3.4 %)

  Erlotinib/rapamycin

3 (10.3 %)

  Talampanel

1 (3.4 %)

  BCNU

1 (3.4 %)

  Bevacizumab

1 (3.4 %)

  BCNU wafers

2 (6.9 %)

  Thalidomide

1 (3.4 %)

  Irinotecan

2 (6.9 %)

 # of prior chemo (%)

 

  1

18 (62.1 %)

  2

7 (24.1 %)

  3

4 (13.8 %)

Concomitant Meds (%)

 

 Anti-seizure

20 (69 %)

 Steroids

20 (69 %)

ITT intent-to-treat, EE efficacy evaluable, ECOG Eastern Cooperative Oncology Group performance status scale, IFRT involved-field radiation therapy, SRS stereotactic radiosurgery

Toxicity

The first 10 patients enrolled in the trial were treated at the 1.22 mg/m2/day dose level, which was determined to be the MTD in the phase I portion of this study. Four of these 10 patients developed grade 3/4 thrombocytopenia, 3 developed grade 3/4 neutropenia, and 3 developed grade 3/4 leukopenia. (Table 2) Based on the recommendation of the data safety monitoring board, the starting dose was lowered one dose level to 1.0 mg/m2/day in an attempt to decrease hematologic toxicity to a more acceptable level. None of the 19 patients enrolled at the 1.0 mg/m2/day dose experienced grade 4 hematologic toxicity. One patient developed grade 3 thrombocytopenia, and 2 patients developed grade 3 leukopenia without grade 3 neutropenia. Only 1 patient withdrew from the study because of a drug-related adverse event (grade 3 thrombocytopenia and leukopenia). Treatment delay occurred in 11 patients (38 %) and dose reduction was necessary in 8 patients (28 %). For the entire cohort, treatment-related grade 3/4 toxicities included thrombocytopenia (17.2 %), leukopenia (17.2 %) and neutropenia (10.3 %). As expected, non-hematologic toxicity consisted primarily of fatigue (18 patients, 62.1 %–all with grade 1/2 toxicity), nausea and vomiting (7 patients, 24.1 %–6 patients with grade 1/2 toxicity and 1 patient with grade 3 toxicity), and diarrhea (7 patients 24.1 %–all with grade 1/2 toxicity). Non-hematologic toxicity was not dose limiting, and occurred in equal frequency in the two dose groups.
Table 2

Grade 2/3/4 hematologic toxicity

  

CTC Grade

  

2

3

4

1.22 (mg/m2/day)

Thrombocytopenia

0

3

1

(n = 10) 

Neutropenia

0

1

2

 

Leukopenia

2

1

2

1.0 (mg/m2/day)

Thrombocytopenia

2

1

0

(n = 19) 

Neutropenia

2

0

0

 

Leukopenia

4

2

0

Excess hematologic toxicity at the original starting dose of 1.22 mg/m2/day prompted reduction of the starting dose to 1.0 mg/m2/day

CTC–NCI common toxicitiy criteria

Drug level monitoring

The 18 % reduction in the daily dose from 1.22 to 1.0 mg/m2 resulted in a 20 % decrease in the geometric mean (±SD) concentration of total gimatecan in plasma prior to administration of the fifth daily dose in cycle 1 (56 ± 23 ng/mL, n = 8 vs. 45 ± 20 ng/mL, n = 13).

Efficacy

The primary endpoint of this phase II trial was progression-free survival at 6 months. Time to progression was determined for 27 of the 29 patients. (Figure 1) The remaining two patients were still receiving active treatment at the time of data lock. Four patients were censored from efficacy analysis at 6 months due to lack of tumor assessment at that time point. Of the 25 patients who were evaluable, only 3 patients (12 % of the evaluable patients, 10.3 % of the intention-to-treat population) reached PFS-6. All three of these patients were treated at the 1.0 mg/m2/day dose. Median time to progression was 12.0 weeks (95 % CI: 7.0, 17.0).
https://static-content.springer.com/image/art%3A10.1007%2Fs11060-012-1023-0/MediaObjects/11060_2012_1023_Fig1_HTML.gif
Fig. 1

Kaplan–Meier curve of progression-free survival for the ITT population. PFS-6 was 10.3 %. Median time to progression was 12 weeks=censored

Imaging response was assessed using modified Macdonald criteria. (Table 3) No patients had a complete response. One patient (in the 1.0 mg/m2/day cohort) had a partial imaging response, while stable disease was the best imaging response seen in 15 patients. Eleven patients had progressive disease after 2 cycles of therapy, and 2 patients were not assessed.
Table 3

Radiographic response evaluated by MRI every second cycle (i.e, every 8 weeks)

 

1.22 mg/m2/day cohort

1.0 mg/m2/day cohort

All patients

Not evaluable

2

0

2

CR

0

0

0

PR

0

1

1

SD

4

11

15

PD

4

7

11

Total:

10

19

29

27 patients were assessable for radiographic response, which was determined by modified Macdonald criteria

CR complete response, PR partial response, SD stable disease, PD progressive disease

Discussion

This study is the first phase II trial to evaluate the efficacy of gimatecan for the treatment of glioblastoma. The primary endpoint of this study, PFS-6, is an established metric for evaluating the efficacy of new treatments for patients with recurrent glioblastoma [23]. In order to minimize the number of patients treated with a potentially ineffective drug yet maintain sufficient power to detect improvement over currently available treatments, this trial utilized a Simon optimal two-stage design. Nineteen patients were to be accrued in the first stage of this trial, with opening of the second stage contingent upon demonstrating PFS-6 in at least 4 patients. However, a pre-planned safety analysis performed after the first 10 patients were enrolled demonstrated unacceptable myelosuppression at the dose that was defined as the MTD in the phase I portion of this trial (1.22 mg/m2/day). The decision was therefore made to enroll an additional 19 patients into the first stage of the trial at the next lower dose level (1.00 mg/m2/day), resulting in a total enrollment of 29 patients in the first stage of this trial. Of this group (25 of whom were evaluable for efficacy), only 3 patients (12 % of evaluable patients; 10.3 % of the intention-to-treat population) achieved PFS-6, and thus this study was terminated after the first stage. Radiographic responses were similarly unimpressive, as no patients achieved CR, and only one patient reached PR. These results mirror those of a phase I trial for another lipophilic camptothecin analogue, karenitecin, for patients with recurrent malignant glioma that also failed to demonstrate clinical benefit [17].

As a novel agent, the optimal dosing of gimatecan has not yet been defined. The dose schedule utilized in this trial consisted of 5 consecutive treatment days every 28 day cycle, as this regimen was found to be safe in a phase I trial for advanced solid tumors [13]. However, from a pharmacodynamic standpoint, as an S-phase specific agent, low-dose daily dosing may be more effective, as prolonged drug exposure increases the proportion of cells that reach S-phase [24]. Indeed, amongst a series of dose regimens used to treat mice orthotopically transplanted with human glioblastoma xenografts, median survival was most prolonged by the regimen that utilized daily gimatecan dosing [11]. As an orally bioavailable drug, gimatecan lends itself to frequent dosing more than topotecan and irinotecan, which are usually given intravenously. Therefore, it is reasonable to wonder whether a different dosing schedule would have produced better efficacy results, although toxicity could potentially be correspondingly greater as well. In a phase I trial for patients with advanced solid tumors, weekly administration of oral gimatecan resulted in continuous systemic exposure to potentially effective concentrations of the active intact lactone form of the drug. However, brain tumor patients were not evaluated in this trial.

The role of gimatecan in combination therapy has also not been determined. Indeed, preclinical studies show that continuous low-dose gimatecan has an antiangiogenic effect, suggesting possible synergism with other antiangiogenic therapies, and also highlighting the potential benefit of daily dosing [25]. Another potential combination strategy is to administer gimatecan with other DNA-damaging agents or inhibitors of DNA repair. Poly adenosine diphosphate-ribose polymerase (PARP) inhibitors, for example, have been shown to sensitize cells to camptothecin-mediated DNA damage, and preclinical studies of a leukemia cell line that developed resistance to camptothecin treatment exhibited upregulation of topoisomerase II activity, which could potentially be targeted by several currently available drugs [26, 27]. A synergistic antitumor effect was also seen in a preclinical study of gimatecan and the tyrosine kinase inhibitor imatinib in a malignant glioma xenograft model [28].

Another limitation of this study is the lack of biomarker analysis. Biomarker analysis is playing an increasingly important role both in drug development and clinical decision-making. Studies of advanced colorectal cancer and ovarian cancer have demonstrated a correlation between higher topoisomerase I levels and improved survival following treatment with irinotecan [29, 30]. Topoisomerase I activity was not assessed as part of this study, and it is possible that greater clinical benefit would have been seen if enrollment was restricted to glioblastoma patients whose tumors demonstrated increased topoisomerase I activity. Other potential markers of camptothecin activity include γ-H2AX, CHK2 phosphorylation by ATM and DNA-PK, and ATM autophosphorylation [31].

In summary, PFS-6 for patients with recurrent glioblastoma taking gimatecan monotherapy administered on a 5 day every 28 day cycle was not significantly better than the historical PFS-6 for patients with recurrent glioblastoma who were given treatments that were deemed no more effective than currently available therapies. As expected, the main toxicity of gimatecan was myelosuppression. It remains to be seen whether a different dosing schedule or combination therapy would have been more effective, but given the lack of efficacy seen in this study, further trials of gimatecan for patients with glioblastoma are not planned at this time.

Acknowledgments

Grant support from Sigma-Tau Research.

Conflict of interest

N. Salem is employed by Sigma-Tau.

Ethical standards

This study complies with the current laws of the United States of America.

Copyright information

© Springer Science+Business Media New York 2012