Current and future developments in the use of temozolomide for the treatment of brain tumours

Current and future developments in the use of temozolomide for the treatment of brain tumours Roger Stupp, Marc Gander, Serge Leyvraz, and Edward ewlands Brain tumours comprise only 2% of all adult cancers, but they are among the most debilitating malignant diseases. Temozolomide, an alkylating agent that can be administered orally, has been approved for the treatment of recurrent malignant glioma on a daily schedule for 5-day cycles. Continuous administration schedules with a higher dose intensity are being explored, but an improvement in efficiency remains to be shown. The benefit from temozolomide given as a single agent in recurrent disease will be several weeks at best. This drug is therefore now undergoing clinical testing as neoadjuvant chemotherapy or with concomitant radiotherapy in patients with newly diagnosed glioma. Several phase I trials are investigating the combination of temozolomide with other agents active against brain tumours. This review briefly summarises the pharmacological background and clinical development of temozolomide and focuses on current and future clinical exploration of this drug for the treatment of brain tumours. Lancet Oncol 2001; 2: 552 60 Primary brain tumours comprise only about 2% of all adult cancers. Malignant gliomas are the most common type of brain tumours, and high-grade gliomas are among the most rapidly growing and devastating neoplasms. Despite surgery and radiotherapy, these tumours almost invariably recur, rapidly leading to death. The median survival of patients with glioblastoma (Figure 1) is less than 1 year from diagnosis. For anaplastic astrocytoma, median survival is about 18 months. Many of the commonly used chemotherapy agents have limited activity in these malignant diseases. The role of the blood brain barrier as an obstacle to chemotherapy efficiency is debatable, because it is commonly disrupted in aggressive glioma. Temozolomide was developed in the 1980s, through rational drug design, by the UK Cancer Research Campaign. 1,2 It has some advantages over dacarbazine, with which it shares the active metabolite 5-(3-methyl)1-triazen-1-yl-imidazole- 4-carboxamide (MTIC). 3 It is rapidly and completely absorbed after oral administration and spontaneously converts into the active metabolite without the need for enzymatic demethylation in the liver. Temozolomide has excellent penetration into all body tissues, including the brain. One mechanism of resistance to this agent is mediated through the enzyme O6-alkylguanine transferase (AGT). 552 Figure 1. Giant-cell glioblastoma (x 300). Most of the tumour-cell bodies express glial fibrillary acidic protein. Cellular and nuclear polymorphism is shown. Many brain tumours express low concentrations of this enzyme. 4,5 Pharmacology Temozolomide is one of a series of imidazotetrazinone derivatives. The development of mitozolomide, the first compound of this class to show broad-spectrum antitumour activity in preclinical models, 6,7 was abandoned because of the unpredictable and long-lasting haematological toxic effects seen in early clinical trials. Studies of the structure activity relation suggested that 3-substituted derivatives of mitozolomide could cleave to form the linear triazene MTIC. 8 Preclinical data confirmed that temozolomide had a very different toxicity profile from mitozolomide in various tumour models, including astrocytomas, glioblastomas, and ependymoma. 9,10 Whereas dacarbazine requires enzymatic demethylation in the liver to be converted into the reactive species, temozolomide is spontaneously activated into MTIC at physiological ph in aqueous solution (Figure 2). Preclinical studies revealed good bioavailability after oral administration, schedule-dependent antitumour activity, and good tissue penetration, including into the brain. 8 RS, MG, and SL are at the Multidisciplinary Oncology Center, University Hospital CHUV, Lausanne, Switzerland, and E is at the Department of Cancer Medicine, Charing Cross Hospital, London, UK. Correspondence: Dr Roger Stupp, Multidisciplinary Oncology Center, University Hospital CHUV, 46, Rue du Bugnon, 1011 Lausanne, Switzerland. Tel: +41 21 314 0156. Fax: +41 21 314 0737. Email: Roger.Stupp@chuv.hospvd.ch THE LACET Oncology Vol 2 September 2001 Courtesy of R Janzer

Review Pharmacokinetics Temozolomide shows linear pharmacokinetics with the area under the concentration time curve (AUC) increasing in proportion to the dose. The excellent bioavailability allows oral administration of the drug. Temozolomide is rapidly absorbed after oral dosing, 11 with moderate variability between individuals and some influence of the fasting state. 12 Maximum plasma concentrations are measured 30 90 minutes after administration. In a daily schedule for 5 days, no accumulation of temozolomide was observed. When taken in conjunction with food, the drug s absorption is delayed, leading to a lower peak concentration and a decrease in the AUC of about 10%. 13 In most clinical trials temozolomide was given with fasting for at least 1 hour before administration and for 1 hour afterwards. On current data, the small change in the AUC with food is unlikely to be clinically relevant. Excretion of temozolomide is thought to occur mainly via the kidney, with a plasma half-life of 1.8 hours. 14 There are, however, no recommendations for dose reduction in patients with renal failure. Published data on penetration into the cerebrospinal fluid in human beings are scarce. We prospectively measured cerebrospinal-fluid and plasma concentrations of temozolomide in 33 patients. The equilibrium was reached 4 hours after drug intake with a mean ratio of 40% (SD 11). 15 O O Temozolomide H 2 H H 2 O Dacarbazine CH 3 CH 3 CH 3 Spontaneous hydrolysis (at physiological ph) O H 2 H CH 3 MTIC Active demethylation (liver) Figure 2. Temozolomide and dacarbazine and their common active metabolite, MTIC. H Mechanism of action Reaction of water with the electropositive C4 atom of temozolomide opens the heterocyclic ring, releasing MTIC and carbon dioxide. 16 MTIC is unstable and degrades into methyldiazonium ion, a reactive methylating compound. Like the chloroethylnitrosoureas, with which they have a common range of preclinical activity, imidazotetrazinones act as major groove-directed DA-alkylating agents. 17 They are base-selective and preferentially bind the middle guanine residue of a GGG sequence. The sites of methylation on DA are the 7 atoms on guanine, O3 on adenine, and O6 on guanine. 18 Although O6-methylguanine represents only a minority of adducts formed by temozolomide, it has a critical role in the cytotoxic action of the drug. The O6 atom of guanine is the initial site of attack on DA of other active agents against malignant gliomas, such as the cross-linking chloroethylnitrosoureas. O6-methylguanine in itself is not lethal to cells; it does not inhibit processes such as DA replication and transcription. However, the preferred base pairing during DA replication results in incorporation of thymine instead of cytosine opposite O6-methylguanine. The mismatch repair pathway of the cell, which excises the aberrant thymine residue in the daughter strand, recognises this mismatch. However, unless the methyl adduct is removed from the guanine, thymine is likely to be reinserted opposite the lesion. The MutS branch of the mismatch repair pathway has a key role in signalling the initiation of apoptosis in response to O6-methylguanine. 19 Repetitive futile rounds of mismatch repair are thought to result in a state of chronic strand breaks, which triggers an apoptotic response. 20 Resistance mechanisms Resistance to temozolomide is thought to be mediated through two independent repair mechanisms AGT and the DA mismatch repair proteins. O6-methylguanine is subject to a single-step error-free reversal of damage reaction, in which the methyl adduct is transferred to a cysteine residue on AGT. The cellular concentration of this enzyme is a major determinant of temozolomide cytotoxicity in vitro. 21,22 Interest in AGT cellular concentrations in malignant glioma arose from the enzyme s role in the cytotoxicity of chloroethylnitrosoureas. Concentrations were higher in benign and in chemoresistant central-nervous-system tumours such as meningiomas and neurinomas and lower in more chemosensitive tumours such as oligodendrogliomas and anaplastic oligodendrogliomas. 23 In a series of 167 samples of glioma and anaplastic astrocytoma, 26% had no measurable activity of AGT. 24 In that study, patients with glioblastomas and anaplastic astrocytomas with low concentrations of AGT had significantly better survival than those with high concentrations of the enzyme. In a series of 38 patients with newly diagnosed glioblastoma and anaplastic astrocytoma treated with temozolomide after surgery and before radiotherapy, low expression of AGT (detection in less than 20% of cells) was found in 70% of tumours. 5 The response rate to temozolomide was 60% in the low-agt group and only 10% in the group with AGT expression in more than 20% of tumour cells. A strong association between AGT activity and response and survival was found with a functional assay measuring methylation of the AGT-gene promoter. Patients with glioma who had a methylated promoter, and thus inactivated AGT, had significantly longer survival when receiving carmustine chemotherapy than patients with a non-methylated AGT promoter (36 versus 22 months; p<0.001, univariate analysis). 25 Prospective validation on a larger number of patients is now needed. Colon cancer cell lines with mismatch repair mutations THE LACET Oncology Vol 2 September 2001 553

are resistant to temozolomide, and the resistance cannot be overcome by AGT depletion. 26 In a panel of xenograft tumours from children, a defect in MLH1, a key mismatch repair protein, was an important determinant of resistance to temozolomide. 27 The DA mismatch repair proteins MSH2 and MLH1 were detected by immunohistochemistry in at least 60% of tumour cells in more than 80% of glioblastomas and anaplastic astrocytomas. 5 However, in that series, the rates of response to temozolomide of tumours with high and low expression of these proteins were similar. A combination of low concentrations of AGT and high concentrations of mismatch repair proteins might be expected to identify the majority of responders to temozolomide therapy. However, in the study by Friedman and colleagues, 5 previous knowledge of these variables predicted only 66% of responses. These negative results do not disprove the role of AGT or mismatch proteins, but rather suggest that immunohistochemical determination may not reflect the functionality of these pathways. Further investigation with functional assays is required. Another cell target that can potentiate the cytotoxicity of temozolomide is the nuclear enzyme poly-adp-ribose polymerase. This enzyme recognises DA strand breaks and is involved in their repair. An inhibitor of poly-adp-ribose polymerase, 3-aminobenzamide, potentiates temozolomide in vitro. 28 More recent work with two novel inhibitors of this enzyme (U-1025 and U-1005) showed that they can potentiate both temozolomide and topotecan in vitro. 29 Further clinical development of these compounds is planned. Approved and investigational schedules of temozolomide administration On the basis of preclinical models suggesting increased activity with repeated exposure to temozolomide, the schedule of daily treatment for 5 days was developed. 8 A daily dose of 200 mg/m 2 (150 mg/m 2 for the first cycle in patients who have received chemotherapy previously) was established as the maximally tolerated dose and recommended for phase II testing. 11,13 The dose-limiting toxic effect with the intermittent schedule has been thrombocytopenia. At the recommended dose of 200 mg/m 2 daily, grade 4 thrombocytopenia is observed in less than 10% of patients, and grade 3/4 neutropenia in less than 5%. ausea can easily be prevented by standard antiemetics or paediatric doses of 5- HT 3 antagonists. Some patients complain of fatigue in the week after therapy. Several other schedules of temozolomide administration have been explored. In a phase I trial, the feasibility of continuous daily administration of temozolomide over 6 7 Table 1. Schedules and dosing of temozolomide Schedule Dose Dose intensity Ref mg/m 2 mg/m 2 /week Daily x 5 days, every 28 days 200 250 Approved dosing Daily x 7 days, every 14 days 150 525 31 Daily x 21 days, every 28 days 100 525 32 Daily x 42 days, every 70 days 75 315 30 weeks was assessed. Doses were increased gradually from 50 mg/m 2 to 100 mg/m 2 daily. 31 Myelosuppression was identified as the dose-limiting toxic effect; other side-effects with continuous administration are minor. Antiemetic prophylaxis is not required in most cases. The recommended dose for further studies using continuous administration is 75 mg/m 2 given daily for 6 7 weeks. This schedule allows the administration of 2.1 g/m 2 for each 4 weeks, which is twice the dose that is administered when temozolomide is given in cycles of daily doses for 5 days. Whether continuous low-dose administration and higher dose intensity lead to improved efficacy remains unclear. Other schedules of continuous daily administration are being investigated (Table 1). With administration of temozolomide for 7 days every 2 weeks (7 days on, 7 days off) myelosuppression and thrombocytopenia were again dose limiting at 175 mg/m 2, and a daily dose of 150 mg/m 2 is recommended for further studies. 31 A schedule of daily temozolomide for 21 days in each 28 days was explored, and a dose of 100 mg/m 2 recommended for phase II studies. 32 In all trials, myelosuppression with neutropenia and thrombocytopenia is the dose-limiting side-effect. Continuous administration of temozolomide is associated with a high frequency of lymphocytopenia. 30 To date, there has been no comparative trial of the different schedules, and it is unclear whether a continuous administration schedule confers meaningful clinical advantage. Temozolomide for recurrent malignant glioma The initial trials of temozolomide were carried out by the UK Cancer Research Campaign. 2 In 1993, the drug was licensed to the Schering-Plough Research Institute. Three pivotal trials led to the licensing of temozolomide in glioblastoma and anaplastic astrocytoma (Table 2). 33 35 Two large single-group phase II trials were done in parallel for anaplastic astrocytoma and glioblastoma. The third trial was a randomised phase II trial for patients with recurrent glioblastoma. For the first Table 2. Pivotal trials in recurrent malignant glioma Histology umber of Therapy Response Progression Median survival PFS at Ref patients rate (%) rate (%) (months) 6 months (%) GBM 138 TMZ 8 47 5.4 18 35 AA (ITT) 162 TMZ 35 38 13.6 46 33 GBM 112 TMZ 5 54 7.3 21 34 versus 5 113 PCZ 67 5.8 8 GBM, glioblastoma multiforme; AA, anaplastic astrocytoma; PFS, progression-free survival; ITT, intention-to-treat; TMZ, temozolomide; PCZ, procarazine 554 THE LACET Oncology Vol 2 September 2001

Review time, patients with malignant glioma were enrolled into separate trials for anaplastic astrocytoma (WHO grade III) or glioblastoma multiforme (WHO grade IV). Previous trials included both patients with grade III and grade IV tumours in a single trial, and results were commonly reported for both histological variants combined. Among patients with malignant glioma, 70 85% have grade IV tumours at diagnosis or at recurrence. All trials included a specifically developed quality-of-life module and the EORTC QOL30 questionnaire. Histology and magnetic resonance images were centrally reviewed. A minimum Karnofsky performance status of 70% was required for inclusion in these studies. The primary endpoint in all trials was 6-month progression-free survival. In brain tumours responses are rare and delayed, so progression-free survival may better reflect a meaningful clinical benefit than the usual response criteria. 36 This unusual endpoint now requires prospective validation in randomised trials. A total of 162 patients were included in the phase II trial for anaplastic astrocytoma. 33 However, on central pathology review the histology could be confirmed in only 111 patients, indicating the difficulties and uncertainties of histopathological classification of lower-grade gliomas. evertheless, there was no significant difference in outcome between the eligible patients and an intention-to-treat analysis. Objective responses with independent radiological confirmation were observed in 35% of patients, with disease stabilisation in another 27%. At 6 months, 46% of patients were progression free, and the 1-year survival was 56%. Patients without progression at 6 months showed improved quality-of-life scores, which were maintained until about 4 weeks before disease progression. 37 In glioblastoma, a phase II study showed an objective response rate of only 8%, but temporary disease stabilisation was achieved in 53% of patients, suggesting a potential clinical benefit in 61%. A second trial was designed as a randomised phase II trial conducted mainly in the USA. The 225 patients were assigned either procarbazine (125 150 mg/m 2 daily by mouth for 28 days, in cycles of 56 days) or temozolomide (150 200 mg/m 2 daily for 5 days, in cycles of 28 days). All patients had had recurrences after previous surgery, radiotherapy, or both. Substantial proportions of the groups (65% and 68%, respectively) had received chemotherapy previously. At 6 months, 21% of the temozolomide group were progression free, compared with 8% of patients assigned procarbazine; 34 however, overall survival did not differ significantly between the groups. Subgroup analysis suggested a potential benefit mainly in chemotherapynaïve patients. Temozolomide was better tolerated than procarbazine, with a lower rate of serious adverse events. This benefit translated into improved quality-of-life scores for the responding or non-progressing patients. 38 On the basis of these three phase II trials, temozolomide was approved for the treatment of malignant glioma. However, approval in the USA was granted for recurrent anaplastic astrocytoma only. The European Medicinal Evaluation Agency approved temozolomide for recurrent glioblastoma and anaplastic astrocytoma, whereas in Switzerland temozolomide is registered only for recurrent glioblastoma. This difference in interpretation of the data on temozolomide therapy reflects the remaining doubt on the validity of the progression-free endpoint and the absence of a true phase III trial. Temozolomide has never been compared with standard nitrosourea-based chemotherapy, such as procarbazine, lomustine, and vincristine). evertheless, in patients with recurrent high-grade glioma, the benefit in time to progression and possibly survival will be a few months at best. Efforts should therefore concentrate on integrating this agent into a multimodality treatment approach for malignant glioma. Another unresolved issue is treatment duration. In recurrent disease the protocol foresaw treatment until progression or for a maximum of 12 cycles. Subsequently, patients were allowed to continue therapy for up to 24 months. o cumulative toxic effects or late effects of exposure to the alkylating agent were observed, but the number of patients exposed to long-term therapy is very small. Temozolomide in newly diagnosed high-grade glioma The true value of new treatments or modalities is best explored in previously untreated patients. The activity of temozolomide shown in patients with recurrent anaplastic astrocytoma and glioblastoma together with the strong in vitro activity against several glioma cell lines and central- Figure 3. Primary chemotherapy with temozolomide before radiation. Partial response after 2 cycles. THE LACET Oncology Vol 2 September 2001 555

Table 3. ovel treatments with temozolomide for primary brain tumours Investigator Ref Histology Treatment CR+PR CR PR SD PFS OS Remarks Year (%) (%) (%) (%) Friedman 5 GBM 33 eoadjuvant 52 3 4 12 months Six patients without 1998 TMZ x 5 for responding. progression at patients 12 months Gilbert 2000 39 GBM 33 eoadjuvant 39 3 10 TMZ x 5 AA 18 39 2 5 Stupp 2000 43 GBM 45 ewly /A 1 year: 65% Continuous daily diagnosed: administration during XRT + TMZ radiotherapy ewlands Unpubl ewly diagnosed: XRT + TMZ ewlands Unpubl Recurrent: TMZ + PCZ Chinot 2001 59 AOD 48 Recurrent: 443 8 13 19 12 months: 46% Recurrent after prior and AOA TMZ x 5 25% XRT and PCV Brada 2000 65 Astro II 13 eoadjuvant: MR 85 11 MR TMZ x 5 Oligo II 7 3 1 MR GBM, glioblastoma multiforme; AA, anaplastic astrocytoma; AOD, anaplastic oligodendroglioma; AOA, anaplastic oligoastrocytoma; astro II, astrocytoma WHO grade II; oligo II, oligodendroglioma WHO grade II; TMZ, temozolomide; PCZ, procarbazine; XRT, radiotherapy; MR, minor response; CR: complete response; PR, partial response; PFS: progressionfree survival; OS: overall survival; SD, stable disease; PVC, procarbazine, lomustine and vincristine nervous-system tumour xenografts are the basis for evaluating this chemotherapy in newly diagnosed patients. Temozolomide chemotherapy may be given as neoadjuvant treatment before definitive radiotherapy, concomitantly with radiotherapy, or as adjuvant therapy after completion of radiotherapy (Figure 3). eoadjuvant therapy was explored by Friedman and colleagues in 33 patients with glioblastomas and five with anaplastic astrocytomas (Table 3). 5 All patients had histologically confirmed high-grade glioma, with the last surgical intervention 14 28 days before the start of therapy. All patients had to have residual disease measurable on magnetic resonance imaging to be eligible for the trial. About two-thirds of patients had undergone subtotal surgical resection, and the others had undergone biopsy only. Treatment consisted of up to four cycles of temozolomide (200 mg/m 2 daily for 5 days) every 4 weeks. Magnetic resonance imaging was repeated before each cycle. Treatment continued in the absence of progressive disease. An objective response was shown in 17 of 33 patients with glioblastomas, including three complete remissions The response lasted for at least 4 months in 13 patients, and four patients showed progression during the neoadjuvant treatment. Disease stabilisation was achieved in four patients. Primary tumour progression occurred in 12 patients. The median survival for the responding patients was 12 months, and six patients are alive without disease progression 18 24 months from diagnosis. Similar results were reported by Gilbert and colleagues. 39 Their patients with glioblastomas (33) or anaplastic astrocytomas (18) were treated with temozolomide 200 mg/m 2 for a maximum of four cycles before radiotherapy. Complete responses were seen in 10% of patients, and partial responses in 29%, irrespective of the histology. A third of patients progressed before radiotherapy and nearly half of the patients received the four planned 556 courses of neoadjuvant therapy. Response to neoadjuvant chemotherapy may identify patients who might benefit from more intensive treatment with concomitant chemoradiotherapy or additional adjuvant chemotherapy. Only a randomised trial will be able to show the real impact and benefit from such a strategy. Temozolomide and radiation Synergy may be achieved when temozolomide chemotherapy and radiotherapy are given concomitantly. Additive growth inhibition was shown in vitro when the glioblastoma cell line U373MG, with low AGT expression, was incubated with temozolomide and treated with a single fraction of radiotherapy. By contrast, the colorectal Mawi cell line, which has high AGT expression, showed additive growth inhibition only after inhibition of AGT with O6-benzylguanine. 40 Fractionated radiotherapy and concomitant treatment with temozolomide was explored by van Rijn and colleagues in two glioma cell lines. 41 Increased cytotoxicity was shown in the D384 glioma cell line, with reduced cell survival and disappearance of the usual dose-effect relation. In the similar U251 cell line, however, no increase in cytotoxicity could be shown. There is no apparent explanation for this difference and further investigations are required. The suggestion of increased cytotoxicity together with the known activity of temozolomide in recurrent and newly diagnosed high-grade glioma are the basis for introducing this agent in newly diagnosed patients at the same time as radiotherapy. All criteria for combining chemotherapy and radiation, as previously established, 42 are met: temozolomide may eliminate microscopic disease outside the radiation field (spatial cooperation); temozolomide increases the radiation effects (radiosensitisation); the modalities have different toxicity profiles (toxicity independence); and temozolomide may be selectively active on glioma cells with THE LACET Oncology Vol 2 September 2001

Review low amounts of AGT (protection of normal tissues). We have investigated the concomitant administration of temozolomide and standard fractionated radiotherapy in a prospective phase II trial. Temozolomide (75 mg/m 2 ) was given daily (7 days per week) for 6 7 weeks during the entire period of radiotherapy, which was administered once daily (Monday to Friday) at 2 Gy per fraction for a total dose of 60 Gy. After completion of therapy, patients were scheduled to receive six additional cycles of adjuvant temozolomide (200 mg/m 2 daily for 5 days). Results on the first 45 patients treated 43 showed that standard radiotherapy can be added to daily temozolomide at the previously established dose of 75 mg/m 2. All but four patients were able to complete the prescribed dose of radiation, and temozolomide was discontinued early in two patients owing to thrombocytopenia. However, two patients developed Pneumocystis carinii pneumonia, perhaps due to the frequent use of corticosteroids in these patients and the lymphocytopenia caused by continuous temozolomide administration. This treatment is being tested in a randomised European Organisation for Research and Treatment of Cancer (EORTC)/ational Cancer Institute of Canada (CIC) trial comparing the combination of temozolomide and radiotherapy with radiotherapy alone in newly diagnosed patients with glioblastomas (Figure 4). Temozolomide in combination Temozolomide and nitrosoureas Chloroethylnitrosoureas and temozolomide alkylate DA at the O6 position of guanine. AGT, which can reverse the alkylation, is inactivated in this process, and its depletion can reverse the drug resistance. Similarly, O6-benzylguanine, a non-cytotoxic substrate of AGT, will deplete the AGT pool. Previous therapy with O6-benzylguanine sensitises tumour cells to alkylation by temozolomide or chloroethylnitrosoureas. However, AGT depletion will occur in all tissues, and systemic toxicity may necessitate dose reduction of nitrosourea or temozolomide. On the basis of the observed progressive and cumulative depletion of AGT after a daily schedule of temozolomide, 44 we have explored the sequential combination of temozolomide and the chloroethylnitrosourea fotemustine in a phase I trial. 45 Temozolomide was given in two daily doses, followed by fotemustine (100 mg/m 2 ) 4 hours after the second dose. At doses of 400 mg/m 2 or more of temozolomide (200 mg/m 2 on days 1 and 2) myelosuppression with grade 4 leucocytopenia and thrombocytopenia was dose limiting. AGT depletion in R Radiotherapy with concomitant and adjuvant temozolomide Temozolomide 75 mg/m 2 every day 6 weeks Eligibility criteria Temozolomide 200 mg/m 2 every day x 5 day repeat every 28 days; for 6 cycles... Focal radiotherapy (30 2 Gy, 60 Gy) Tumour volume with 2 3 cm margin ewly diagnosed and histologically proven glioblastoma multiforme Age 18 70 years (inclusive) WHO PS 2 6 weeks (42 days) since biopsy or resection Adequate bone marrow, hepatic and renal function o other severe underlying disease Written informed consent Stratification factors: peripheral-blood lymphocytes was not related to the dose of temozolomide. In our current protocol we give a low dose of temozolomide (100 mg/m 2 on days 1 and 2) before fotemustine (100 mg/m 2 ) on day 2. A dose of 100 mg/m 2 O6-benzylguanine has been shown to suppress AGT activity effectively in malignant glioma. 46 In a phase I trial in malignant glioma O6-benzylguanine was combined with carmustine. As predicted, dose-limiting toxicity (myelosuppression) was reached at far lower doses of carmustine than after administration of that drug alone. 47 Similarly, temozolomide has been combined with carmustine in a randomised phase I trial evaluating the sequence dependence of the administration of these two drugs. 48 This trial suggested greater toxicity when temozolomide preceded carmustine. A similar trial established a dose of 150 mg/m 2 carmustine on day 1 followed by 100 mg/m 2 temozolomide on days 1 5, but only 80 mg/m 2 temozolomide on days 1 5, when followed by 100 mg/m 2 carmustine on day 5, suggesting greater AGT depletion with previous exposure to 5 days of temozolomide. 49 Again, temozolomide doses that can be given in combination with carmustine are well below established single-agent doses. Temozolomide has been combined with thioguanine, and a dose of 40 mg THE LACET Oncology Vol 2 September 2001 557 Age PS 0/1 versus 2 Radiotherapy only Resection versus biopsy Institution Figure 4. Study design for the EORTC/CIC randomised trial in newly diagnosed glioblastoma multiforme. PS, performance status weeks

thioguanine 6 hours before and after temozolomide intake combined with 150 mg/m 2 temozolomide for 5 days is considered safe. However, no data on AGT depletion are available. Trials combining O6-benzylguanine and temozolomide are underway. In all trials, the doses of the nitrosourea or temozolomide, or both, had to be substantially lower than those for singleagent therapy. o data on efficacy with these lower doses are available, and the possibility that such a combination will lead to clinically meaningful increased activity remains doubtful. Temozolomide and procarbazine A phase I trial combining temozolomide and procarbazine is in progress on the basis that both these drugs inhibit AGT and have activity against malignant gliomas. The trial is designed to use procarbazine as an AGT inhibitor given 1 hour before temozolomide, with both drugs given daily for 5 days in cycles of 28 days. Temozolomide has been given in the standard dose of 200 mg/m 2, then combined in subsequent courses with procarbazine and given in escalating doses (50, 75, then 100 mg/m 2 ). This trial is continuing and is near to the dose-limiting toxic effect of thrombocytopenia. The combination has been well tolerated, and responses have been seen in patients with both low-grade and high-grade gliomas (ewlands and colleagues, unpublished). Temozolomide and irinotecan Irinotecan is also active against malignant glioma. 50 Patients with recurrent glioma after radiotherapy (53) and/or chemotherapy (41) were treated with irinotecan (125 mg/m 2 weekly for 4 weeks repeated after 6 weeks). An overall response rate of 15% was reported, with a further 55% of patients having stable disease for longer than 12 weeks. Owing to enzyme induction by concomitant antiepileptic treatment, the AUC of irinotecan and its active metabolites was up to 40% lower than previously reported in patients with colorectal cancer. 50,51 In a subsequent phase I study, the dose of irinotecan could be increased up to 1200 mg/m 2 every 3 weeks, compared with the standard dose of 350 mg/m 2 every 3 weeks in patients with colorectal cancer. 52 Studies combining these two active agents with only partially overlapping toxicity need to be done. In vitro studies suggest a schedule-dependent synergy between the two agents with an increased effect of irinotecan after exposure to temozolomide. 53 Patients requiring antiepileptic therapy should be considered for treatment with one of the newer non-enzyme-inducing antiepileptic drugs. 54 Anaplastic oligodendroglioma and oligoastrocytoma Anaplastic oligodendroglioma and mixed oligoastrocytoma are very chemoresponsive, with response rates of 60 75% to combination treatment with procarbazine, lomustine, and vincristine. 55,56 Patients who had previously received radiotherapy were as likely to respond as those who had not. Specific chromosomal aberrations on chromosomes 1p and 19q were found to be predictive of chemoresponsiveness. 57 eoadjuvant therapy with this combination of drugs in anaplastic oligodendroglioma and oligoastrocytoma has been explored. 58 558 Temozolomide has shown activity as second-line chemotherapy in patients previously treated by radiotherapy and procarbazine, lomustine, and vincristine chemotherapy. 59 Patients (median age 42 years) with anaplastic oligodendroglioma (38) and oligoastrocytoma (nine) were treated with 150 200 mg temozolomide daily for 5 days. Complete responses were seen in seven, and partial regression in 13 patients, for an overall response rate of 43%. Only eight patients (17%) progressed initially. At 12 months, 34% of patients were progression free. These data suggest that there is no major cross-resistance between the combination regimen and temozolomide. A confirmatory phase II trial is continuing. The EORTC is also evaluating first-line chemotherapy with temozolomide in patients with recurrent oligoastrocytoma and anaplastic oligodendroglioma after radiotherapy. Temozolomide for low-grade glioma Treatment of low-grade glioma remains controversial. 60,61 Radiotherapy is commonly given and may prolong time to progression. Prospective and retrospective analyses have shown, however, that deferred radiotherapy does not compromise overall survival. 62,63 Radiotherapy is associated with a significant risk of late neurological sequelae. In a retrospective analysis, radiation necrosis was reported in nine of 62 patients, and delayed cognitive impairment (diagnosed without prospective neuropsychometric testing) in 13. 64 These findings suggest that at a median observation time of 6 years, over a third of the patients had clinically significant sequelae from the previous radiotherapy. Patients with lowgrade astrocytoma, and in particular the more chemosensitive oligodendroglioma and mixed oligoastrocytoma, have median survival of longer than 10 years. Primary chemotherapy with temozolomide may be an alternative to immediate radiotherapy. 61 In particular, the novel continuous exposure schedules have some theoretical advantages in the treatment of slowly proliferating diseases. Preliminary results of a continuing trial for newly diagnosed patients with low-grade glioma were recently reported. 65 Patients with grade II astrocytoma or oligodendroglioma were treated with the daily 5-day schedule for 6 12 months. Minor responses were seen in 11 of the 13 patients with astrocytoma, and three of the seven with oligodendroglioma had partial remissions and one a minor response. o patient progressed. Investigators at Duke University, C, USA, reported on 16 evaluable patients with progressive, low-grade glioma treated with the intermittent schedule of temozolomide. 66 Disease stabilisation was observed in 13 patients, and three patients progressed. A large randomised trial is needed to assess the true value of this approach and to assess the price in terms of toxicity of primary chemotherapy in low-grade disease. Late myelodysplastic syndrome and second malignant disorders are of concern after chronic exposure to an alkylating agent. 67,68 Temozolomide for other brain tumours and metastases Primary temozolomide chemotherapy or chemoradiotherapy for metastatic brain tumours is currently under investigation. This approach may be of particular interest in diseases for THE LACET Oncology Vol 2 September 2001

Review Search strategy and selection criteria We briefly summarise the pertinent preclinical data and present a summary of the pivotal trials that led to licensing of this compound. We focus on the future development and potential of temozolomide as a single agent and in combination with other chemotherapy agents and with radiation. Published and unpublished data were identified by searches of PubMed, review of the proceedings of the last 3 years of American Society of Clinical Oncology, American Association of Cancer Research, and European Society of Medical Oncology, and by screening the references of relevant articles. Papers were selected on the basis of their relevance to the topic and novelty of approach. With the focus on future potential development, unpublished reports and abstracts of small and ongoing trials were included. In contrast, for recurrent disease we limit the review to three large pivotal trials. which temozolomide or dacarbazine have shown activity outside the brain. Responses to temozolomide have been reported for patients with non-small-cell lung cancer and primary central-nervous-system lymphoma; disease stabilisation has been reported for metastatic melanoma. 69 71 In a small randomised trial of only 45 evaluable patients with brain metastases (mainly from lung cancer), an increased response rate was observed when standard-dose temozolomide was combined with radiotherapy (2 Gy per fraction, 40 Gy total) compared with radiotherapy alone (96% versus 66%), but no difference in overall survival. Time to local failure has not been reported. 72 Conclusions Temozolomide has some, generally modest, activity against high-grade glioma and has shown activity in relapsed oligodendroglioma after nitrosourea chemotherapy. However, no formal comparative trial against nitrosoureabased chemotherapy has been done. To define the true role of the drug in the treatment of brain tumours, further exploration within well-designed and controlled clinical trials is necessary. Assessment of the activity of new drugs in brain tumours is a difficult task, because measurement of response rates according to conventional criteria may not reflect the real clinical benefit. 73 ew endpoints, such as progression-free survival at 6 months, require prospective validation and should be supported by quality-of-life analysis. Acknowledgments Roger Stupp and Edward ewlands have served as consultants to Schering-Plough Research Institute and received research support. References 1 Stevens MF, ewlands ES. From triazines and triazenes to temozolomide. Eur J Cancer 1993; 7: 1045 47. 2 ewlands ES, Stevens MF, Wedge SR, et al. Temozolomide: a review of its discovery, chemical properties, pre-clinical development and clinical trials. Cancer Treat Rev 1997; 23: 35 61. 3 Loo T. Triazenoimidazole derivatives. In: Sartorelli A, Johns D, eds. Antineoplastic and immunosuppressive agents. Berlin: Springer Verlag, 1975: 553 54. 4 Belanich M, Randall T, Pastor MA, et al. Intracellular localization and intercellular heterogeneity of the human DA repair protein O(6)-methylguanine-DA methyltransferase. Cancer Chemother Pharmacol 1996; 37: 547 55. 5 Friedman H, McLendon R, Kerby T, et al. DA mismatch repair and O6-aklylguanine-DA alkyltransferase analysis and response to temodal in newly diagnosed malignant glioma. J Clin Oncol 1998; 16: 3851 57. 6 Stevens MF, Hickman JA, Stone R, et al. Antitumor imidazotetrazines: 1, synthesis and chemistry of 8-carbamoyl-3-(2- chloroethyl)imidazo[5,1-d]-1,2,3,5-tetrazin-4(3 H)-one, a novel broad-spectrum antitumor agent. J Med Chem 1984; 27: 196 201. 7 Gibson W, Hickman JA, Erickson LC. DA cross-linking and cytotoxicity in normal and transformed human cells treated in vitro with 8-carbamoyl-3-(2-chloroethyl)imidazo[5,1-d] -1,2,3,5-tetrazin- 4(3H)-one. Cancer Res 1984; 44: 1772 75. 8 Stevens MF, Hickman JA, Langdon SP, et al. Antitumor activity and pharmacokinetics in mice of 8-carbamoyl-3-methyl- imidazo[5,1-d]- 1,2,3,5-tetrazin-4(3H)-one (CCRG 81045; M & B 39831), a novel drug with potential as an alternative to dacarbazine. Cancer Res 1987; 47: 5846 52. 9 Friedman HS, Dolan ME, Pegg AE, et al. Activity of temozolomide in the treatment of central nervous system tumor xenografts. Cancer Res 1995; 55: 2853 57. 10 Plowman J, Waud WR, Koutsoukos AD, et al. Preclinical antitumor activity of temozolomide in mice: efficacy against human brain tumor xenografts and synergism with 1,3-bis(2- chloroethyl)-1- nitrosourea. Cancer Res 1994; 54: 3793 99. 11 ewlands ES, Blackledge GRP, Slack JA, et al. Phase I trial of temozolomide (CCRG 81045: M and B 39831: SC 362856). Br J Cancer 1992; 65: 287 91. 12 Marzolini C, Decosterd LA, Shen F, et al. Pharmacokinetics of temozolomide in association with fotemustine in malignant melanoma and malignant glioma patients: comparison of oral, intravenous, and hepatic intra-arterial administration. Cancer Chemother Pharmacol 1998; 42: 433 40. 13 Brada M, Judson I, Beale P, et al. Phase I dose-escalation and pharmacokinetic study of temozolomide (SCH 52365) for refractory or relapsing malignancies. Br J Cancer 1999; 81: 1022 30. 14 Tsang LL, Farmer PB, Gescher A, Slack JA. Characterisation of urinary metabolites of temozolomide in humans and mice and evaluation of their cytotoxicity. Cancer Chemother Pharmacol 1990; 26: 429 36. 15 Stupp R, Ostermann S, Leyvraz S, et al. Cerebrospinal fluid levels of temozolomide as a surrogate marker for brain penetration. Proc Am Soc Clin Oncol 2001; 20: 59a (abstr 232). 16 Clark AS, Deans B, Stevens MF, et al. Antitumor imidazotetrazines: synthesis of novel imidazotetrazinones and related bicyclic heterocycles to probe the mode of action of the antitumor drug temozolomide. J Med Chem 1995; 38: 1493 504. 17 Clark AS, Stevens MF, Sansom CE, Schwalbe CH. Anti-tumour imidazotetrazines, part XXI: mitozolomide and temozolomide: probes for the major groove of DA. Anticancer Drug Des 1990; 5: 63 68. 18 Denny BJ, Wheelhouse RT, Stevens MF, et al. MR and molecular modeling investigation of the mechanism of activation of the antitumor drug temozolomide and its interaction with DA. Biochemistry 1994; 33: 9045 51. 19 Hickman MJ, Samson LD. Role of DA mismatch repair and p53 in signaling induction of apoptosis by alkylating agents. Proc atl Acad Sci USA 1999; 96: 10764 69. 20 Karran P, Hampson R. Genomic instability and tolerance to alkylating agents. Cancer Surv 1996; 28: 69 85. 21 Wedge SR, Porteous JK, May BL, ewlands ES. Potentiation of temozolomide and BCU cytotoxicity by O(6)-benzylguanine: a comparative study in vitro. Br J Cancer 1996; 73: 482 90. 22 Graziani G, Faraoni I, Grohmann U, et al. O 6 -alkylguanine-da alkyltransferase attenuates triazene-induced cytotoxicity and tumor cell immunogenicity in murine L1210 leukemia. Cancer Res 1995; 55: 6231 36. 23 Wiestler O, Kleihues P, Pegg AE. O 6 -alkylguanine-da alkyltransferase activity in human brain and brain tumors. Carcinogenesis 1984; 5: 121 24. 24 Belanich M, Pastor M, Randall T, et al. Retrospective study of the correlation between the DA repair protein alkyltransferase and survival of brain tumor patients treated with carmustine. Cancer Res 1996; 56: 783 88. 25 Esteller M, Garcia-Foncillas J, Andion E, et al. Inactivation of the DA-repair gene MGMT and the clinical response of gliomas to alkylating agents. Engl J Med 2000; 343: 1350 54. 26 Liu L, Markowitz S, Gerson SL. Mismatch repair mutations override THE LACET Oncology Vol 2 September 2001 559

alkyltransferase in conferring resistance to temozolomide but not to 1,3-bis(2-chloroethyl)nitrosourea. Cancer Res 1996; 56: 5375 79. 27 Middlemas DS, Stewart CF, Kirstein M, et al. Biochemical correlates of temozolomide sensitivity in pediatric solid tumor xenograft models. Clin Cancer Res 2000; 6: 998 1007. 28 Wedge SR, Porteous JK, ewlands ES. 3-Aminobenzamide and/or O(6)-benzylguanine evaluated as an adjuvant to temozolomide or BCU treatment in cell lines of variable mismatch repair status and O(6)-alkylguanine-DA alkyltransferase activity. Br J Cancer 1996; 74: 1030 36. 29 Delaney CA, Wang LZ, Kyle S, et al. Potentiation of temozolomide and topotecan growth inhibition and cytotoxicity by novel poly(adenosine diphosphoribose) polymerase inhibitors in a panel of human tumor cell lines. Clin Cancer Res 2000; 6: 2860 67. 30 Brock CS, ewlands ES, Wedge SR, et al. Phase I trial of temozolomide using an extended continuous oral schedule. Cancer Res 1998; 58: 4363 67. 31 Figueroa J, Tolcher A, Denis L, et al. Protracted cyclic administration of temozolomide is feasible; a phase I, pharmacokinetic and pharmacodynamic study. Proc Am Soc Clin Oncol 2000; 19: 222a (abstr 868). 32 Denis L, Tolcher A, Figueroa J, et al. Protracted daily administration of temozolomide is feasible: a phase I and pharmacokineticpharmacodynamic study. Proc Am Soc Clin Oncol 2000; 19: 222a (abstr 786). 33 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: Temodal Brain Tumor Group. J Clin Oncol 1999; 17: 2762 71. 34 Yung WK, Albright RE, Olson J, et al. A phase II study of temozolomide vs. procarbazine in patients with glioblastoma multiforme at first relapse. Br J Cancer 2000; 83: 588 93. 35 Brada M, Hoang-Xuang K, Rampling R, et al. Multicenter phase II trial of temozolomide in patients with glioblastoma multiforme at first relapse. Ann Oncol 2001; 12: 259 66. 36 Macdonald DR, Cascino TL, Schold SC, Cairncross JG. Response criteria for phase II studies of supratentorial malignant glioma. J Clin Oncol 1990; 8: 1277 80. 37 Osoba D, Brada M, Yung WK, Prados MD. Health-related quality of life in patients with anaplastic astrocytoma during treatment with temozolomide. Eur J Cancer 2000; 36: 1788 95. 38 Osoba D, Brada M, Yung W, Prados M. Health-related quality of life in patients treated with temozolomide versus procarbazine for recurrent glioblastoma multiforme. J Clin Oncol 2000; 18: 1481 91. 39 Gilbert M, Olson J, Yung W, et al. Preradiation treatment of newly diagnosed anaplastic astrocytomas and glioblastoma multiforme using temozolomide. euro-oncology 2000; 2: 264 (abstr 77). 40 Wedge SR, Porteous JK, Glaser MG, et al. In vitro evaluation of temozolomide combined with X-irradiation. Anticancer Drugs 1997; 8: 92 97. 41 van Rijn J, Heimans JJ, van den Berg J, van der Valk P, Slotman BJ. Survival of human glioma cells treated with various combination of temozolomide and X-rays. Int J Radiat Oncol Biol Phys 2000; 47: 779 84. 42 Steel GG, Peckham MJ. Exploitable mechansisms in combined radiotherapy-chemotherapy: the concept of additivity. Int J Radiat Oncol Biol Phys 1979; 5: 85 91. 43 Stupp R, Ostermann-Kraljevic S, Pica A, et al. Promising survival with concomitant and adjuvant temozolomide for newly diagnosed glioblastoma multiforme. Proc Am Soc Clin Oncol 2000; 19: 163a (abstr 632). 44 Lee SM, Thatcher, Crowther D, Margison GP. Inactivation of O 6 - alkylguanine-da alkyltransferase in human peripheral blood mononuclear cells by temozolomide. Br J Cancer 1994; 69: 452 56. 45 Gander M, Leyvraz S, Decosterd L, et al. Sequential administration of temozolomide and fotemustine: depletion of O 6 -alkyl guanine-da transferase in blood lymphocytes and in tumours. Ann Oncol 1999; 10: 831 38. 46 Friedman HS, Kokkinakis DM, Pluda J, et al. Phase I trial of O 6 - benzylguanine for patients undergoing surgery for malignant glioma. J Clin Oncol 1998; 16: 3570 75. 47 Friedman HS, Pluda J, Quinn JA, et al. Phase I trial of carmustine plus O 6 -benzylguanine for patients with recurrent or progressive malignant glioma. J Clin Oncol 2000; 18: 3522 28. 48 Schold Jr S, Kuhn J, Chang S, et al. A phase I trial of 1,3-bis (2-chloroethyl)-1-nitrosourea plus temozolomide: a orth American Brain Tumor Consortium study. euro-oncology 2000; 2: 34 39. 49 Hammond L, Eckardt J, Kuhn J, et al. A phase I and pharmacokinetic trial of 2 sequences of BCU and temozolomide in patients with advanced solid tumors. euro-oncology 2000; 2: 267 (abstr 90). 50 Friedman HS, Petros WP, Friedman AH, et al. Irinotecan therapy in adults with recurrent or progressive malignant glioma. J Clin Oncol 1999; 17: 1516 25. 51 Reid J, Buckner J, Schaaf L, et al. Anticonvulsants alter the pharmacokinetics of irinotecan (CPT-11) in patients with recurrent glioma. Proc Am Soc Clin Oncol 2000; 19: 160a (abstr 620). 52 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: 161a (abstr 624). 53 Houghton PJ, Stewart CF, Cheshire PJ, et al. Antitumor activity of temozolomide combined with irinotecan is partly independent of O 6 -methylguanine-da methyltransferase and mismatch repair phenotypes in xenograft models. Clin Cancer Res 2000; 6: 4110 18. 54 Dichter MA, Brodie MJ. ew antiepileptic drugs. Engl J Med 1996; 334: 1583 90. 55 Cairncross G, Macdonald D, Ludwin S, et al. Chemotherapy for anaplastic oligodendroglioma: ational Cancer Institute of Canada Clinical Trials Group. J Clin Oncol 1994; 12: 2013 21. 56 van den Bent MJ, Kros JM, Heimans JJ, et al. Response rate and prognostic factors of recurrent oligodendroglioma treated with procarbazine, CCU, and vincristine chemotherapy: Dutch eurooncology Group. eurology 1998; 51: 1140 45. 57 Cairncross JG, Ueki K, Zlatescu MC, et al. Specific genetic predictors of chemotherapeutic response and survival in patients with anaplastic oligodendrogliomas. J atl Cancer Inst 1998; 90: 1473 79. 58 Paleologos A, Macdonald DR, Vick A, Cairncross JG. eoadjuvant procarbazine, CCU, and vincristine for anaplastic and aggressive oligodendroglioma. eurology 1999; 53: 1141 43. 59 Chinot Jr O, Honore S, Barrie Jr M, et al. Safety and efficacy of temozolomide in patients with recurrent anaplastic oligodendrogliomas after standard radiotherapy and chemotherapy. J Clin Oncol 2001; 19: 2449 55. 60 Shaw E. Looking through the retrospectoscope in the era of evidence-based medicine. J Clin Oncol 1997; 15: 1289 90. 61 Cairncross JG. Understanding low-grade glioma: a decade of progress. eurology 2000; 54: 1402 03. 62 Karim A, Cornu P, Bleehen, et al. Immediate postoperative radiotherapy in low grade glioma improves progression free survival, but not overall survival: preliminary results on an EORTC/MRC randomized phase III study. Proc Am Soc Clin Oncol 1998; 17: 400a (abstr 1544). 63 Leighton C, Fisher B, Bauman G, et al. Supratentorial low-grade glioma in adults: an analysis of prognostic factors and timing of radiation. J Clin Oncol 1997; 15: 1294 301. 64 Olson JD, Riedel E, DeAngelis LM. Long-term outcome of low-grade oligodendroglioma and mixed glioma. eurology 2000; 54: 1442 48. 65 Viviers L, Brada M, Hines F, et al. A phase II trial of primary chemotherapy with temozolomide in patients with low-grade cerebral gliomas. Proc Am Soc Clin Oncol 2000; 19: 167a (abstr 648). 66 Friedman A, Cokgor I, Edwards S, et al. Phase II treatment of anaplastic oligodendroglioma and low grade glioma with temozolomide. Proc Am Soc Clin Oncol 2000; 19: 169a (abstr 654). 67 Perry JR, Brown MT, Gockerman JP. Acute leukemia following treatment of malignant glioma. J euro-oncol 1998; 40: 39 46. 68 Pedersen-Bjergaard J. Long-term complications of cancer chemotherapy. J Clin Oncol 1995; 13: 1534 36. 69 Christodoulou C, Bafaloukos D, Kosmidis P, et al. Phase II study of temozolomide in heavily pretreated cancer patients with brain metastases. Ann Oncol 2001; 12: 249 54. 70 Reni M, Ferreri AJ, Landoni C, Villa E. Salvage therapy with temozolomide in an immunocompetent patient with primary brain lymphoma. (letter) J atl Cancer Inst 2000; 92: 575 76. 71 Abrey L, Olson J, Boutros D, et al. A phase II study of temozolomide for recurrent brain metastases. Proc Am Soc Clin Oncol 2000; 19: 166a (abstr 643). 72 Antonadou D, Paraskevaides M, Coliarakis, et al. Temozolomide enhances radiation treatment efficacy in brain metastases: a randomized phase II study. euro-oncology 2000; 2: 247 (abstr 6). 73 Gilbert MR. euro-oncology clinical trials: promise and pitfalls. Ann Oncol 2001; 12: 149 50. 560 THE LACET Oncology Vol 2 September 2001