www.elsevier.com/locate/jocn

Journal of Clinical Neuroscience 13 (2006) 18–22

Clinical study

Phase II study of two-weekly temozolomidein patients with high-grade gliomas

Shirley Wong a,*, Mark A. Rosenthal a, Anthony Dowling b, Ross Jennens a,Anne-Marie Woods a, David Ashley c, Lawrence Cher d

a Department of Clinical Haematology and Medical Oncology, Royal Melbourne Hospital, Melbourne, Australiab Department of Medical Oncology, St Vincent�s Hospital, Melbourne, Australia

c Department of Medical Oncology and Haematology, Royal Children�s Hospital, Melbourne, Australiad Department of Medical Oncology, Austin Health, Melbourne, Australia

Received 11 October 2004; accepted 20 October 2004

Abstract

Palliative chemotherapy has an increasing role in the management of recurrent high-grade gliomas. Temozolomide is a well-toleratedagent that results in objective responses and stabilisation of disease. Theoretically, temozolomide may be more effective when given in aprolonged schedule rather than the standard 5 days-monthly schedule. This Phase II study examined the efficacy and toxicity of temozol-omide when given in a two-weekly schedule. Twenty-five patients received 150 mg/m2 temozolomide daily for seven days alternating withseven days of no treatment. One cycle of temozolomide was a total of two weeks treatment in every 28 days. Of the 25 evaluable patients,there was one complete response (4%), four partial responses (16%) and 10 patients had disease stablisation (40%). The progression freesurvival at 6 months was 56%. Two-weekly temozolomide was well tolerated with only four episodes of Grade 3 thrombocytopenia.Overall, two-weekly temozolomide is an active and well tolerated schedule, but does not appear to improve on the activity of temozol-omide using the standard 5-day schedule.� 2005 Elsevier Ltd. All rights reserved.

Keywords: Glioma; Temozolomide

1. Introduction

Nearly all patients with high-grade gliomas progress de-spite treatment and only approximately 15% survive be-yond 2 years.1 The median overall survival from thediagnosis of relapsed disease is less than 9 months.1,2

New strategies are therefore required including the use ofnovel agents, novel combinations of agents and novelscheduling of proven agents.

Temozolomide is a well-established chemotherapy agentwith activity in high-grade gliomas. The standard schedule(200 mg/m2 days 1–5 every 28 days) results in objective re-

0967-5868/$ - see front matter � 2005 Elsevier Ltd. All rights reserved.

doi:10.1016/j.jocn.2004.10.018

* Corresponding author. Department of Medical Oncology, WesternHospital, Footscray, Victoria, Australia. Tel.: +61 3 83456666; fax: +61 383456445.

E-mail address: shirleys.wong@mh.org.au (S. Wong).

sponse rates of between 10–30% with disease stabilisationoccuring in a further 30% of patients.3,4,5 Despite this activ-ity, novel schedules of temozolomide may theoretically re-sult in greater efficacy and less toxicity.

An elevated level of O6-alkylguanine DNA alkyltrans-ferease has been implicated in conferring resistance to tem-ozolomide. Administration of temozolomide results indecreased O6-alkylguanine DNA alkyltransferease activitywithin four hours of an oral dose, which persists in tissuesfor at least 24 hours.6 Daily administration results inprogressive depletion of this activity over the standardfive-day treatment course.6 Preclinical studies have demon-strated progressive depletion following prolonged exposureto temozolomide.7 Hence, prolonged exposure of patientsto temozolomide may decrease drug resistance and enhancecytotoxicity.

S. Wong et al. / Journal of Clinical Neuroscience 13 (2006) 18–22 19

Schedule options include: ‘‘two-weekly’’(7 days on/7days off), 21 days on/7 days off and 42 days on. Phase Istudies demonstrated tolerability with a substantive in-crease in dose intensity.8–10 This Phase II study examinedthe efficacy and tolerability of temozolomide when givenin a two-weekly schedule (7 days on/7 days off). This sche-dule may not only decrease drug resistance but it also re-sults in a potential 2.8-fold increase in dose intensity.11

2. Patients and methods

2.1. Eligibility

Patients with recurrent or progressive high-grade glio-mas whose initial pathological diagnosis was either glio-blastoma multiforme, anaplastic astrocytoma, mixedoligoastrocytoma or gliosarcoma were eligible. Eligible pa-tients had evaluable disease on a contrast-enhanced MRIor CT. Other eligibility criteria included age P 18 years,Eastern Cooperative Oncology Group (ECOG) perfor-mance status 0–2, serum creatinine 6 1.5 times upper limitof normal, absolute neutrophil countP 1.5 · 109/L, plate-let count P 100 · 109/L, AST 6 2.5 times upper limit ofnormal, total bilirubin within normal limits and comple-tion of radiotherapy, any chemotherapy or investigationalagents more than 4 weeks previously. Written informedconsent was obtained from all patients, and the protocolwas approved by the respective institutional ethicscommittees.

Patients were excluded if they had severe intercurrent ill-ness, had not recovered from surgery, or were pregnant orbreastfeeding. Women of childbearing potential had to beusing effective methods of contraception while on study.Patients had to have a contrast enhanced imaging studywithin 3 weeks prior to study entry.

2.2. Study design

This was a multi-center, open-label, non-randomizedphase II study. Temozolomide was administred orally inthe morning, in a fasting state, at a dose of 150 mg/m2 dailyfor seven consecutive days of every other week. One 28-daycycle represented treatment with temozolomide on days 1–7 and days 15–21, with no treatment given during days 8–14 and days 22–28. Temozolomide capsules were availablein 5 mg, 20 mg, 100 mg and 250 mg strengths and all doseswere rounded up to the nearest 20 mg to accommodateavailable capsule strengths. Temozolomide was taken witha glass of water over as short a time as was possible. Pa-tients were instructed to swallow whole capsules, in rapidsuccession and not to chew capsules. No re-dosing of thepatient was allowed before the next schedule dose if vomit-ing occurred during the course of treatment.

Treatment was repeated every 28 days. In the absence ofdisease progression or unacceptable toxicity, patients con-tinued to receive treatment every other week for up toone year from initial dose. Monitoring consisted of a full

blood examination weekly during the first two cycles, thenon days 1 and 15 of all other cycles; liver and renal functiontests and clinical neurological evaluation was performedevery 4 weeks. Toxicity and dose modifications were basedon National Cancer Institute (NCI) Common Toxicity Cri-teria Version 2. Criteria for re-treatment were absoluteneutrophil count > 1.5 · 109/L, platelet count 100 · 109/L, haemoglobin >100 · 109/L, AST < 2.5 times upper limitof normal, total bilirubin within normal limit, serum creat-inine < 1.5 times upper limit of normal and all other toxic-ities resolved to baseline or grade 1. Disease status wasreassessed clinically each cycle and radiologically afterevery second cycle.

2.3. Dose modifications

The dose of temozolomide administred for the subse-quent cycles were determined by the nadir absolute neutro-phil count and platelet count on day 1. There were threedose levels in the dose schedule of 150 mg/m2 (the com-mencement dose), 125 mg/m2 and 100 mg/m2. The doseof temozolomide was reduced to the next lower dose levelif the absolute neutrophil count was < 0.5 · 109/L for 5days, if there was febrile neutropenia, if the platelet countwas < 25 · 109/L or if there was delayed hematologicrecovery (P2 weeks). Treatment delay of up to 4 weekswas allowed for the recovery of absolute neutrophil countto > 1.5 · 109/L and platelet count to > 100 · 109/L.

For grade 3 and 4 non-hematological toxicity, temozol-omide was withheld until the toxicity resolved to baselineor grade 1 and dose reduction to the next lower dose levelwas recommended. All patients who developed grade 4 tox-icity or required dose reduction to less than 100 mg/m2 dai-ly were removed from the study.

Patients received a serotonin antagonist as an anti-emetic prior to the temozolomide. Other standard anti-emetic treatments as required were acceptable. Prophylactichaematopoietic growth factors were not permitted, butcould be used at physician discretion for episodes of neu-tropenic sepsis.

2.4. Evaluation of response

A minimum of one cycle of temozolomide was requiredfor patients to be evaluable for response. Performance sta-tus evaluation, clinical examination, neurological evalua-tion and laboratory assessments were performed monthly.Patients were evaluated radiologically every 2 monthsusing the same contrast enhanced imaging study, as atbaseline. Measurement of response was based on the prod-uct of the two largest perpendicular diameters of the con-trast-enhancing lesion or, if multiple lesions were present,the sum of the products of the perpendicular diameters.12

A complete response (CR) was defined as complete resolu-tion of all symptoms and radiological disappearance of allenhancing tumours at least one month apart. A partial re-sponse (PR) was defined as stable or improved neurological

20 S. Wong et al. / Journal of Clinical Neuroscience 13 (2006) 18–22

status and a decrease of P 50% total tumour load radio-logically. Progressive disease (PD) was defined as worsen-ing of neurological status and P 25% increase in totaltumour load or any new tumour radiologically. All otherassessments were considered stable disease (SD).

2.5. Statistical methods

This study was designed to evaluate the safety and effi-cacy of two-weekly dosing schedule of temozolamide inthe treatment of high grade gliomas. The primary responsevariable was the response rate. Sample size was based onthe Minimax design proposed by Simon for a two-stagePhase II trial.13 Parameters which must be specified for thisdesign are P0, a response rate of no interest; P1, a desirableresponse rate; alpha, the probability of accepting an inef-fective drug; and beta, the probability of rejecting an effec-tive drug. The following set of parameters were selected bytaking into consideration that standard dosing schedule oftemozolamide has an objective response rate (CR + PR) of20–30% in patients with high-grade gliomas: P0 = 0.30,P1 = 0.5, alpha = 0.10, beta = 0.10. These choices ofparameters lead to the following sample sizes: 25 patientswere to be enrolled in the first stage, and the trial was tostop if there were less than eight responses. Otherwise,the trial was to continue until a total of 48 patients were en-rolled. Ninety-five percent confidence intervals were calcu-lated using the exact binomial distribution.

3. Results

3.1. Patient characteristics

Patient characteristics are shown in Table 1. All patientshad previously undergone radiotherapy and three had pre-viously received adjuvant chemotherapy. No patient had

Table 1Characteristics of 25 evaluable patients

Characteristics Number of patients

Number of patients 25SexMale 15Female 10

Age (years)Median (range) 54 (34–77)

ECOG performance status0 51 142 6

HistologyGlioblastoma multiforme 11Anaplastic astrocytoma 8Mixed oligoastrocytoma 5Gliosarcoma 1

Prior treatmentSurgery 25Chemotherapy 3Radiotherapy 25

received prior chemotherapy for recurrent disease.Twenty-five patients were enrolled onto the study fromJanuary 2004 to September 2004. Three patients did not re-ceive a second cycle of temozolomide. One patient was en-rolled into the study; but only received five days oftreatment because of rapid clinical deterioration. One pa-tient died of an unrelated illness (perforated diverticulumon day 4 of cycle one). One patient was removed fromthe study during cycle 1 due to grade 3 thrombocytopeniaand a severe intercurrent illness. All patients have been in-cluded in an intent to treat analysis of response but the twopatients who died early in cycle 1 from causes unrelated tothe study drug have been excluded from the toxicity evalu-ation. Thus 25 patients were evaluable for response and 23patients were evaluable for toxicity.

3.2. Response and survival

The response to treatment is shown in Table 2. Of the 25evaluable patients, one patient (4%) achieved a CR andfour patients (16%) achieved a PR. The overall objective re-sponse was 20% (95% confidence interval 5–40%). An addi-tional 10 patients (40%) achieved SD. The completeresponse duration was five months and those of the partialresponses were 2, 3, 3 and 6 months respectively. The sixmonths progression free survival was 56%. The medianoverall survival was 6 months (range 1.5–18 months).

3.3. Toxicity

The two-weekly dosing schedule of temozolomide regi-men was well tolerated. There were no grade 4 toxicities,no episodes of neutropenic sepsis nor treatment-relateddeaths. Thrombocytopenia was the most common haema-tological toxicity (Table 3). Three patients had a total offour episodes of grade 3 thrombocytopenia (44% of treat-ment cycles). All these three patients required temozolo-

Table 2Response rate in 25 evaluable patients

GBM(n = 11 )

AA(n = 8 )

OA(n = 5 )

Oligosarcoma(n = 1 )

Total(n = 25)

CR 1 0 0 0 1 (4%)PR 1 1 1 1 4 (16%)SD 6 2 2 0 10 (40%)PD 3 5 2 0 10 (40%)

Table 3NCI CTC Version 2 toxicity in 23 evaluable patients

Grade 1–2 (%) Grade 3–4 (%)

Thrombocytopenia 1 (4.5%) 3 (14%)Neutropenia 0 (0%) 2 (9%)Anaemia 1 (4.5%) 0 (0%)Lethargy 5 (23%) 0 (0%)Nausea/vomiting 4 (18%) 0 (0%)Constipation 4 (18%) 0 (0%)

S. Wong et al. / Journal of Clinical Neuroscience 13 (2006) 18–22 21

mide dose modifications; and one was withdrawn from thestudy after the third cycle of treatment due to prolongedthrombocytopenia. Two patients had grade 3 neutropeniaand one of these was withdrawn from the study after six cy-cles of treatment due to prolonged neutropenia. The mostcommon non-haematological toxicity was lethargy thatwas generally mild and self-limiting. Other non-haemato-logical adverse events such as nausea, vomiting and consti-pation were also reported during the study, but these eventsdid not require dose modification or discontinuation oftemozolamide.

Two patients developed unrelated events and one ofthem was discontinued from the study. One patient hadpulmonary embolism during the first cycle of treatmentbut continued on the study for a total of four cycles atwhich time they developed disease progression. The otherpatient developed a subdural haemorrhage after two cyclesof treatment but continued on the study for twelve cycles.

4. Discussion

Temozolomide has been proven to have a well-established role in the treatment of high-grade gliomas.The standard once-daily dosing schedule (150–200 mg/m2)has proven activity and is well tolerated.3–5,12 However,numerous investigators have evaluated alternative dosingschedules to increase efficacy, improve safety and reducedrug resistance.9–11,13,14

An elevated level of O6-alkylguanine DNA alkyltrans-ferease is one of the identified mechanisms for the develop-ment of temozolomide resistance. The alternative dosingschedules of �two-weekly� temozolomide allows prolongeddrug exposure, produced O6-alkylguanine DNA alkyl-transferease depletion and hence may improve drugeffectiveness.11,13

This study examined the efficacy and tolerability of atwo-weekly temozolomide schedule. Of the 25 patients, five(20%) achieved complete or partial response and 10 (40%)achieved stable disease. We observed a median overall sur-vival of six months and a six-month progression-free sur-vival of 56%. This schedule of temozolomide wasgenerally well tolerated. These findings are comparablewith older studies using the standard schedule as well asnewer studies applying novel treatment schedules.3–5,9,10,15–17 Overall, the two-weekly dosing schedule of tem-ozolomide was active and well tolerated but did not appearto have greater activity than that of the standard five-daytemozolomide schedule. Of additional interest, one patientdeveloped rapid neurological deterioration on day 5 of cy-cle 1. This may represent a flare response as previously re-ported by our group.18

There are two obvious limitations with this study. First,radiological assessment of response in glioma treatment isrecognised as complex and fraught with error. The assess-ment of radiological responses may be variable betweendifferent reviewers and these variable assessments havebeen reported by other investigators.19,20 Second, the num-

ber of patients enrolled was relatively small. However, pro-spective statistical measures were identified and theresponse rate did not warrant further accrual to the study.

In conclusion, this study reports the use of a two-weeklyschedule of temozolomide in the treatment of recurrenthigh-grade gliomas. The regimen is active and well toler-ated. However, the results of the study suggest that thereis no substantive clinical benefit over and above that ofthe standard five-day schedule. Future studies should ex-plore novel agents and the combination of these agentswith temozolomide.

References

1. Burger PC, Vogel FS, Green SB, et al. Glioblastoma multiforme andanaplastic astrocytoma: pathologic criteria and prognostic implica-tions. Cancer 1985;56:1011–106.

2. Burton E, Prados M. New chemotherapy options for the treatment ofmalignant gliomas. Curr Opin Oncol 1999;11:157–61.

3. Yung WK, Prados MD, Yaya-Tur R, et al. Multicenter phase II trialof temozolomide in patients with anaplastic astrocytoma or anaplasticoligoastrocytoma at first relapse: Temodal Brain Tumor Group. JClin Oncol 1999;17:2762–71.

4. Brada M, Hoang-Xuan K, Rampling R, et al. Multicentre phase IItrial of temozolomide in patients with glioblastoma multiforme at firstrelapse. Ann Oncol 2001;12:259–66.

5. Harris MT, Rosenthal MA, Ashley D, Cher L. An Australianexperience with temozolomide for the treatment of recurrent high-grade gliomas. J Clin Neurosci 2001;8:325–7.

6. Lee SM, Thatcher N, Crowther D, Margison GP. Inactivation ofO6-methyleguanine-DNA alkyltransferase in human peripheralblood mononuclear cells by temozolomide. Br J Cancer 1994;69:452–6.

7. Wedge SR, Porteus JK, May BL, Newlands ES. Potentiation oftemozolomide and BCNU cytotoxicity by O(6)-benzylguanine: acomparative study in vitro. Br J Cancer 1996;73:482–90.

8. Friedman HS, Kerby T, Calvert H. Temozolomide and treatment ofmalignant glioma. Clin Cancer Res 2000;6:2585–97.

9. Newlands E, Blackledge G, Slack J, et al. Phase I trial of temozol-omide. Br J Cancer 1992;65:287–91.

10. Brock CS, Newlands ES, Wedge SR, et al. Phase I trial oftemozolomide using an extended continuous oral schedule. CancerRes 1998;58:4363–7.

11. Simon R. Optimal two-stage designs for the phase II clinical trials.Control Clin Trials 1989;10:1–10.

12. DeBono J. Extended temozolomide dosing schedules permit theadministration of higher temozolomide dose intensities and inhibitthe DNA repair enzymes O6-methyleguanine-DNA alkyltransferase.Abstract 11th Annual European Conference on Cancer (ECCO)2001.

13. Chinot OL, Honore S, Dufour H, et al. Safety and efficacy oftemozolomide in patients with recurrent anaplastic oligodendroglio-mas after standard radiotherapy and chemotherapy. J Clin Oncol

2001;19:2449–55.14. Van Den Bent MJ, Chinot O, Boogerd W, et al. Second-line

chemotherapy with temozolomide in recurrent oligodendrogliomaafter PCV chemotherapy: EORTC Brain Tumour Group phase IIstudy 26972. Ann Oncol 2003;14:599–602.

15. Vera K, Djafari L, Faivre S, et al. Dose-dense regimen of temozol-omide given every other week in patients with primary central nervoussystem tumours. Ann Oncol 2004;15:161–71.

16. Rosenthal MA, Gruber ML, Glass J, et al. Phase II study ofcombination taxol and estramustine phosphate in the treatmentof recurrent glioblastoma multiforme. J Neuro-Oncol 2000;47:59–63.

22 S. Wong et al. / Journal of Clinical Neuroscience 13 (2006) 18–22

17. Chua S, Rosenthal MA, Wong S, et al. Phase II study of temozol-omide and caelyx in patients with recurrent glioblastoma multiforme.Neuro-oncol 2004;6:38–43.

18. Rosenthal MA, Ashley DL, Cher L. Temozolamide-induced flare inhigh-grade gliomas: a new clinical entity. Intern Med J 2002;32:346–8.

19. Macdonald DR, Cascino TL, Schold Jr SC, Cairncross JG. Responsecriteria for Phase II studies of supratentorial malignant gliomas. JClin Oncol 1990;8:1277–80.

20. Gilbert MR. Neuro-oncology clinical trials: promise and pitfalls. AnnOncol 2001;12:149–55.