Reports

This report is based on medical evidence presented at sanctioned medical congress, from peer reviewed literature or opinion provided by a qualified healthcare practitioner. The consumption of the information contained within this report is intended for qualified Canadian healthcare practitioners only.

MEDICAL OPTIONS in Neuro-Oncology

Recurrence Patterns in Gliomas

Dr. Normand Laperrière, University of Toronto

Management of Recurrent Glioblastoma Multiforme

Dr. Rolando Del Maestro, McGill University

Rechallenge with Temozolomide: Dosing Strategies

Dr. James R. Perry, University of Toronto

Clinical Trials in Glioblastoma Multiforme

Dr. Warren P. Mason, University of Toronto

RECURRENCE PATTERNS IN GLIOMAS

Editorial Overview:

Normand Laperrière, MD, FRCPC

Department of Radiation Oncology, Princess Margaret Hospital/University Health Network, Associate Professor of Medicine, University of Toronto, Toronto, Ontario

Imaging studies of glioma/glioblastoma multiforme (GBM) recurrence patterns may be helpful in individualizing patient management and possibly improving tumour control. Magnetic resonance image (MRI) mapping studies of the location and extent of initial presenting and recurrent tumours, as well as diffusion tensor-weighted (DTI) MRI studies, may provide non-invasive means of predicting recurrence patterns, thus suggesting areas to target at initial treatment for possible therapeutic benefit. At present, however, these studies are of more academic than practical interest due to their costs, technical complexity and uncertain clinical benefits.

The potential for relapse and recurrence of malignant gliomas is very high, and disease progression is nearly inevitable in the majority of patients with these aggressive tumours. The combination of surgery, chemoradiation and adjuvant chemotherapy has been effective at improving survival in patients with glioblastoma, but the concept of cure is still largely elusive.

MRI studies of glioma/GBM both before and after surgical resection raise the interesting prospect of more individualized treatment of brain tumours through identification of specific patterns of recurrence. In this review, we will examine two studies that attempt to find order in the seemingly chaotic patterns of tumour recurrence.

As Price and colleagues noted, gliomas display marked heterogeneity, with tumours of similar histology showing a wide-ranging variation in their pathological features and growth patterns (Eur Radiol 2007;17:1675-84). With such variability, it is difficult to devise a comprehensive, standardized treatment strategy that will be appropriate for all patients.

Longitudinal Study

With an eye to improving management of patients with glioblastoma, investigators with the European Organisation for Research and Treatment of Cancer (EORTC) and the National Cancer Institute of Canada Clinical Trials Group (NCIC-CTG) are developing and working to validate a novel tool for analyzing MRI data in order to examine clinical recurrence patterns of the malignancies (Wick et al. Neuro Oncol 2008 Aug 4;Epub ahead of print).

The authors conducted a nested longitudinal study of patients with newly diagnosed GBM who were enrolled in the EORTC 26981/22981/ NCIC CE.3 trial, a phase III randomized study of radiotherapy with or without the oral alkylating agent temozolomide. Their goal was to determine whether the addition of concomitant and adjuvant temozolomide to radiotherapy altered the pattern of recurrence compared with radiotherapy alone.

Table 1.

Because temozolomide treats the whole brain, there is a question of whether chemoradiotherapy and adjuvant therapy with the alkylating agent might reduce the incidence of distant recurrence. Alternatively, it was theorized that temozolomide-induced radiosensitivity of tumours and prolonged survival might foster distant recurrences.

Borrowing a technique employed in stroke research, the authors used MRIcro software to map the location and extent of tumours prior to surgery and after recurrence, and transferred the images of individual patients into the same template image of a stereotactic space. This technique allows investigators to identify areas of overlapping tumour involvement from one to the next by subtracting images of overlap at recurrence from those of overlap at baseline.

The authors carried out the comparison studies on a total of 63 patients, 33 of whom had been treated with radiotherapy alone and 30 of whom received radiation with concomitant temozolomide followed by adjuvant temozolomide. There were no significant between-group differences in terms of extent of resection or of methylguanine methyltransferase (MGMT) promoter methylation status (Table 1). In both groups the median time between baseline MRI and first histological diagnosis was 0.2 months. The median time from MRI at baseline to MRI demonstrating recurrence was 5.4 months among patients treated with radiotherapy alone, compared with 7.3 months in the radiotherapy/temozolomide group.

In a group-wise analysis of recurrence patterns, the authors found that in each treatment group, the tumours showed a geographic heterogeneity, affecting the temporal, parietal and frontal cortices in both the left and right hemisphere without apparent discrimination.

Also, the addition of temozolomide did not appear to affect the anatomical shift of recurrent tumours, nor did it change their size or recurrence patterns compared with radiotherapy alone. There were also no significant between-group differences in the frequency of distant recurrences, which were seen in 23% of patients on radiotherapy alone vs. 18% who received radiochemotherapy (P=0.056).

This study provides some important insights into tumour recurrence patterns in the chemoradiation/adjuvant therapy era, because it relies on data from a randomized study rather than uncontrolled observational data. The experience with brachytherapy for gliomas in the 1990s is illustrative of the need for such data. Although our centre in Toronto and the University of California, San Francisco, published data showing that brachytherapy did not appear to alter recurrence patterns, a third centre in Boston published disparate results suggesting that brachytherapy did alter recurrence patterns—a conflict that was not resolved until the completion of two randomized controlled studies provided level I evidence that the treatment did not have an effect on tumour recurrence patterns.

So what can we learn from this study? Within the limits of this small sample, temozolomide does not seem to have influenced the recurrence pattern in patients who are managed with the Stupp protocol as part of their initial management compared with patients who received radiation treatment only.

DTI-weighted MRI

Dr. Stephen Price, Cambridge University Hospitals, UK, and colleagues have tried a different approach to predicting tumour recurrence patterns for individualized therapy, in this case using DTI. DTI allows for the measurement of water diffusion in multiple directions and for the fractional anisotropy to be calculated for each voxel.

To see whether DTI could be used to predict patterns of glioma recurrence, the authors compared DTI- to T2-weighted images of patients with a histologically confirmed diagnosis of glioma. All patients had an initial DTI study and then had follow-up imaging performed either more than two years after the initial study or at the time of symptomatic recurrence.

It is difficult to draw clear inferences from this study, because it captured patients with a mixture of tumour types who were at various stages of treatment. In all, 12 of the patients had World Health Organization (WHO) grade IV tumours, five had WHO grade III lesions (one anaplastic astrocytoma, two anaplastic oligodendroglioma and two anaplastic oligoastrocytoma) and eight had WHO grade II tumours (six diffuse astrocytomas, one oligodendroglioma and one oligoastrocytoma).

Seventeen of the 25 patients had been treated previously with surgery and radiotherapy, and of this group, 14 were studied at tumour recurrence and three during follow-up when they had stable disease.

Some of the patients underwent radiotherapy, chemotherapy or a combination between the two imaging studies, but six of the 25 patients received no additional therapies between studies.

The authors determined that there were three basic patterns of the DTI images relative to the T2-weighted images among the 25 patients:

• A diffuse pattern of abnormality, seen in 12 patients, where the DTI images revealed a disruption of white matter pathways diffusely around the T2-weighted images, corresponding in most cases to generalized tumour growth at recurrence.

• A localized pattern of abnormality seen in eight patients, where the DTI images demonstrated one localized area of white matter pathway disruption relative to the T2-weighted images, a finding that was associated with recurrent tumour growth in that particular area.

• A minimal pattern of abnormality, where there was no significant DTI abnormality beyond the T2-weighted abnormality, seen in five patients, which may be a predictor of more indolent tumour growth and improved
suggest.

Table 2.

<img2605|center>

The authors concluded that DTI, which can reveal occult tumour or patterns of tumour infiltration that may be missed on conventional (T1 gadolinium-enhanced or T2) MRI sequences, might have a role in future in treatment planning. They suggest that the technique, when used preoperatively, has the potential for guiding biopsies and for targeting local chemotherapy, and may be useful in devising more targeted radiotherapy plans.

This study provides some interesting information about a subgroup of gliomas that are more locally confined with less infiltration of the adjacent white matter pathways. Patients with such tumours were a minority in this study, but conceivably could benefit from more aggressive local therapies in view of the minimal extent of progression into adjacent brain.

The heterogeneity of the cohort (Table 2) would imply that the study primarily provided a snapshot rather than a comprehensive overview of the pathology and recurrence pattern of specific tumour types. It should also be noted that DTI analysis is an expensive and labour-intensive proposition. Although the technique does not add significantly to image acquisition times, there is a great deal of post-processing involved, requiring the services of a physicist who is experienced with the software and who can perform the fine-tuning necessary to derive the optimum level of detail and data from DTI-weighted studies.

Summary

The studies by Wick et al. and Price and colleagues suggest that in addition to their vital role in diagnosis, staging and surgical planning, additional MRI techniques may in future have a role in better delineation of tumour infiltration patterns and prediction of recurrence patterns. This may lead to more targeted local therapies that may impact on survival. What is needed, ultimately, are more prospective cohort studies limited to specific tumours (e.g. GBM, low-grade gliomas, anaplastic, etc.) with uniform criteria and longitudinal follow-up that hopefully can reveal the secrets of these challenging aggressive malignancies.

MANAGEMENT OF RECURRENT GLIOBLASTOMA MULTIFORME

Editorial Overview:

Rolando Del Maestro, MD, PhD, FRCS(C), FACS, DABNS

William Feindel Chair in Neuro-Oncology Director, Brain Tumour Research Centre Montreal Neurological Institute and Hospital Professor, Division of Neurosurgery and Oncology, McGill University, Montreal, Quebec

The Canadian recommendations for the treatment of newly diagnosed glioblastoma multiforme (GBM) patients involve the use of radiotherapy plus concomitant temozolomide followed by adjuvant temozolomide chemotherapy (Stupp et al. N Engl J Med 2005;352:987-96, Mason et al. Curr Oncol 2007;14(3):110-7). Unfortunately, in the majority of these patients the tumour recurs. This communication will outline a number of the pertinent issues involved in the management of patients with recurrent GBM. The treatment of each patient must be individualized and is best carried out by an interdisciplinary team using a multidisciplinary approach. Presentation of these cases at a formal Brain Tumour Board allows a broad range of expertise to be utilized in making recommendations for further treatment. It is critical to define true tumour progression and differentiate it from what has been termed pseudoprogression. A number of options can be discussed with the patients and their families and important issues such as medical and neurological status, imaging studies and other chemotherapy treatment options outlined. Treatment options can include further surgical excision, participation in clinical trials, altering existing temozolomide schedules or re-instituting them. A combination of temozolomide with other alkylating agents such as procarbazine and newer chemotherapeutic agents can also be used.

Differentiating True Tumour Progression from Pseudoprogression

Following treatment with concomitant radiotherapy and temozolomide some patients develop early necrosis within the radiation treatment field. This has been termed pseudoprogression, since, based on the imaging studies, it is impossible to differentiate this from true tumour progression without a biopsy (Chamberlain et al. Neurooncol 2007;82:81-3). Pseudoprogression was described originally by Hoffman et al. (J Neurosurg 1979;50:624-8) and was further outlined as radiation injury associated with alkylating agents by de Witt et al. (Neurology 2004;63:535-7).

In a study carried out by Brandes et al. (J Clin Oncol 2008;26:2192-7), about 50% of patients had an enlarged lesion on the first MRI following concomitant treatment. Sixtyfour per cent of these patients who continued temozolomide therapy had stable or reduced changes on an MRI carried out three months later. Interestingly, 66% of these patients had MGMT promoter methylation and had an improved survival. These results support the Canadian guidelines which outline that if increased radiological changes are seen on the first MR after concomitant therapy that temozolomide be continued for another three months, and that the following MR scan be used to help determine the presence of true tumour progression.

Recurrent Glioblastoma: Canadian Recommendations

The management of recurrent GBM may involve a combination of therapeutic modalities, ranging from surgical excision in select patients to chemotherapy with a variety of agents, to temozolomide rechallenge, to temozolomide in combination with other agents.

Surgery

Current Canadian recommendations for patients with recurrent GBM recommend that if the patient has adequate performance status and has reasonable levels of cognitive and other functions, repeat resection should be considered. In many cases, reoperation can reduce the patient’s need for corticosteroids and decrease or ameliorate neurological dysfunction. It should be noted that there is currently no Level 1 evidence to support repeat resection, as there have been no randomized, controlled clinical trials comparing repeat resection head-to-head with other modalities such as chemotherapy in patients with recurrent glioblastoma. Nonetheless, the general consensus among neurosurgeons and oncologists is that if an operation can be performed safely and can either relieve symptoms or provide further information—for example, discriminating between pseudo-progression and actual tumour progression—then surgery may be a suitable option (Mason et al. Curr Oncol 2007;14(3):110-7).

Chemotherapy

As noted above, there are currently no standard recommendations for the use of chemotherapy in patients with recurrent glioblastoma. In our centre, the preference is, whenever possible, to enrol patients in clinical trials. For patients not eligible for clinical trials for medical or practical reasons (e.g. lack of access to clinical trials), there are several therapeutic options. These include lowdose continuous temozolomide, nitrosureas, etoposide, carboplatin, irinotecan, imatinib and bevacizumab.

In our centre, the first choice for treatment of recurrent GBM is frequently low-dose continuous temozolomide (50 mg/m2 for 28/28 days each cycle). Perry et al. reported preliminary data from the RESCUE trial at the ASCO 2008 annual meeting. They found that among 90 patients for whom follow-up data were available, six-month progressionfree survival (PFS) among patients in early adjuvant failure (failure occurring during adjuvant temozolomide) was 28.6%. Among patients with extended adjuvant failure (after at least six months of ongoing adjuvant temozolomide), the six-month PFS rate was 9.5%. Among patients who had completed temozolomide (concomitant and adjuvant), 30.4% had six-month PFS. Epigenetic methylation of the MGMT promoter has been associated with improved survival in GBM patients treated with radiosurgery and concomitant temozolomide (Hegi et al. N Engl J Med 2005;252:997-1003). The role of the methylation of MGMT promoter in response to low-dose temozolomide in the RESCUE trial is presently being ascertained.

Another option for patients not eligible for clinical trials is the combination of procarbazine and temozolomide (Huang et al. Can J Neurol Sci 2008;35(2):192-7), and for a select group of patients who can afford it, bevacizumab along with another cytotoxic agent (Vredenburgh et al. J Clin Oncol 2007;25(30):4722-9). Currently, bevacizumab is not approved for first- or second-line treatment of GBM or any other type of brain tumour in any province or territory. Therefore, this agent is limited for use in either private-paying or privately insured patients, or as compassionate use on a case-by-case basis.

There is variability across Canada as to the availability of temozolomide (Khoo et al. Cancer Drug Access, Part 3, The New Wave of Cancer Drugs, Report Card on Cancer in Canada, 2007, 40-52, www.canceradvocacy. ca). In Alberta, patients with GBM receive temozolomide concomitant with radiation therapy, followed by a maximum of six cycles of adjuvant temozolomide according to the protocol by Stupp et al. (N Engl J Med 2005;352:987-96). In Quebec, in consultation with physicians, patient
eive temozolomide as long as the tumour is responding.

Table 1. <img2606|center>

In addition to the temozolomide/ procarbazine combination mentioned above, other combination chemotherapy regimens for recurrent GBM have been explored (Table 1).

Despite the widespread use of intensive chemotherapy in patients with recurrent glioblastoma, slightly less than one-third of all patients will have objective responses (Brandes et al. Crit Rev Oncol Hematol 2008;67(2):139-52). The nitrosoureas carmustine (BCNU), lomustine (CCNU) and streptozocin and other liposoluble alkylating agents such as temozolomide have been the backbone of first-line chemotherapy for GBM that recurs after surgery and radiotherapy/ chemoradiotherapy.

Temozolomide combinations under exploration have all shown similar results. These include:

• Temozolomide plus the platinum compound cisplatin which in one study produced six-month PFS of 34%.

• Temozolomide plus the matrix metalloproteinase inhibitor marimastat, which was shown to produce a six-month PFS of 39%, with a median PFS of 17 weeks.

• Temozolomide plus 13-cis-retinoic acid, which was associated with a six-month PFS of 32% and a median PFS of 16 weeks (Brandes et al. 2008).

Re-irradiation

The Canadian recommendations do not address the issue of re-irradiation of patients with recurrent glioblastoma, since clinical evidence to support this practice is lacking. Re-irradiation can be technically complex and has the potential for significantly greater toxicities than radiotherapy for newly diagnosed malignancies (Brandes et al. Crit Rev Oncol Hematol 2008). The use of fractionated stereotactic re-irradiation and stereotactic radiosurgery has been reported, but these studies were uncontrolled and the results may have been skewed by selection bias (Wen PY, Kesari S. N Engl J Med 2008;359:492-507). In highly selected patients, in whom tumour recurrence occurs outside of the original radiation field, re-radiation can be a consideration.

Pathological Grading Issues

Management of patients with recurrent glioblastoma is highly individualized, and is tailored according to various factors, including tumour histology and molecular genetic signatures that help clinicians make more accurate diagnosis and prognosis.

One of the difficulties in the diagnosis and staging of tumours, however, is the problem of tumour heterogeneity, a well-documented feature of gliomas and glioblastomas. If, during surgery, only a small portion of the tumour is recovered for pathology studies, it is possible that higher grade of tumour may be missed, leading to the possibility of under-treatment.

In other words, some tumours determined by histolopathology to be WHO grade III may be harbouring elsewhere in the unsampled tumour tissue features that are more typical of GBM. This distinction is important, due to evidence and clinical experience that has demonstrated a benefit of adding temozolomide to radiotherapy in patients with GBM.

Burnet and colleagues from the University of Cambridge, UK, have pointed out that some patients have tumours that according to pathology studies are WHO grade III, yet have radiographic evidence of features suggestive of more aggressive glioblastoma, a category they define as grade III/IV gliomas (Burnet et al. Radiother Oncol 2007;85:371-8). To determine whether there were differences in outcome in patients with such tumours compared with patients with tumours that more closely matched standard grade III criteria, they undertook a retrospective review of all patients with high-grade gliomas who were treated with radical radiotherapy at their centre.

The review included 245 patients, 52 of whom had grade III tumours, 18 of whom had grade III/IV, and 175 of whom had GBM. The cases were analyzed with a Cox proportional hazards model to evaluate the groups alongside Kaplan-Meier curves.

The authors found that there were no significant differences in survival between those patients graded III/IV and patients with GBM, and that patients with grade III/IV had a nearly fourfold greater risk of death than patients with grade III (hazard ratio 3.7, 95% confidence interval 1.9-7.0)

When they conducted a pathology review, they found that all 18 grade III/IV tumours displayed pathological features suggestive of aggressive behaviour. Among these tumours, seven had oligodendroglial components with both necrosis and microvascular proliferation— features that are both seen in GBM. Of the most recently diagnosed eight tumours, for which counts of the tumour proliferation marker MIB-1 were present, all were above 14%, the lower threshold for suspicion of GBM, and in half of all cases, the count was 30% or higher.

The results suggest that in patients with histological grade III tumours with radiographic features suggestive of more aggressive tumours consideration be given, in consultation with the patient and the interdisciplinary team involved in the patient’s care, to using radiotherapy with concomitant temozolomide and adjuvant chemotherapy.

Summary

Progress in the treatment of recurrent GBM has been incremental, with small but meaningful improvements in our ability to prolong patient lives and delay disease progression. Clinical trials with new targeted agents, in combination with new therapeutic uses of existing agents (e.g. continuous temozolomide), offer the greatest prospect for advancement, and the best hope for our patients with high-grade gliomas.

RECHALLENGE WITH TEMOZOLOMIDE: DOSING STRATEGIES

Editorial Overview:

James R. Perry, MD, FRCPC

Head, Division of Neurology, Tony Crolla Chair in Brain Tumour Research, Sunnybrook Health Sciences Centre, Associate Professor of Medicine, University of Toronto, Toronto, Ontario

A recently published retrospective study and early results from an uncontrolled phase II trial indicated that continuous daily treatment with temozolomide may provide additional benefit even in patients with previous exposure to the alkylating agent. The challenge to providing the agent in this setting revolves around extended dosing schedules that may enhance the agent’s activity against recurrent gliomas. Retrospective data showed that patients who received continuous daily temozolomide (50 mg/m²) after a first tumour recurrence had six-month progression-free survival (PFS) above 50%. Six-month PFS was 42% among patients treated with continuous temozolomide fol lowing a second relapse. Similar results have been seen in an ongoing clinical trial. The concept of temozolomide rechallenge shows early promise for extending PFS in patients with these aggressive recurrent tumours.

Rechallenging After Failure

The oral alkylating agent temozolomide has been shown to offer significantly better PFS than procarbazine in patients with recurrent or progressive glioblastoma, and to produce sixmonth PFS rates approaching 50% (Hou et al. Neurosurg Focus 2006;20(4):E5).

In the first-line setting, concurrent chemoradiotherapy with temozolomide has been shown to improve two-year and fouryear survival rates among patients with newly diagnosed GBM (Stupp et al. N Engl J Med 2005:352:987-96, Mirimanoff et al. Int J Radiat Oncol Biol Phys 2007:69(suppl 5):52)

Nonetheless, GBM remains an intractable disease, and despite aggressive intervention, will nearly always resume progression within a relatively short time. Second-line therapies— nitrosoureas, carboplatin, etoposide, irinotecan, or a combination of these agents—have only limited efficacy and at best only modest duration of effect.

Oncologists are accustomed to switching therapeutic gears when an initially efficacious first- or second-line therapy fails, and thus it seems counterintuitive to suggest rechallenging the tumour with a “failed” agent. Yet there is intriguing, albeit preliminary, clinical evidence to support the idea that temozolomide delivered in an extended dosing schedule has activity in patients with recurrent malignant gliomas (Perry et al. Cancer 2008;113:2152-7).

Therapeutic Rationale

There is currently no standard second-line regimen for malignant glioma following failure of the upfront chemoradiation strategy using temozolomide during radiotherapy. The rationale for changing regimens or rechallenging with continuous temozolomide after failure of conventional therapy with the agent is based on several factors, including action of the drug against MGMT, a DNA repair enzyme. Intriguingly, preliminary studies suggest that chemotherapy, when administered continuously on a daily basis, may have antiangiogenic properties by inhibiting recovery of microvascular endothelium and by suppressing mobilization and activity of bone marrow-derived circulating endothelial precursor cells.

In contrast to other agents used in firstline and/or salvage therapy of patients with recurrent GBM, temozolomide can be safely and effectively delivered in a number of different dosing schedules, which appears to make the agent a good choice for recurrent gliomas.

Clinical Experience

We retrospectively reviewed our experience with continuous temozolomide delivered at a daily dose of 50 mg/m² in patients with GBM progression occurring after conventional chemotherapy with temozolomide. Daily chemotherapy with temozolomide was delivered to three groups of patients with malignant gliomas:

• Group 1: Patients with histologically confined GBM who had surgery and conventional external beam radiation therapy (RT). These patients, who were treated before temozolomide chemoradiation became standard practice locally, received conventional temozolomide, and when they progressed, they were offered continuous temozolomide. This group consisted of 21 patients, median age 54 years, range 33-68.

• Group 2: Patients with GBM who underwent surgery and RT combined with concurrent temozolomide at first recurrence after completion of standard concomitant and adjuvant temozolomide (14 patients, median age 54, range 38-62)

• Group 3: Patients with other anaplastic gliomas at second relapse on conventional temozolomide (14 patients, median age 49; range, 34-56).

Patients in group 1 with second progression after a median of three cycles (range, two to 12) of temozolomide were then put on continuous therapy at 50 mg/m2 until further progression. In these patients, the overall clinical benefit rate—a composite of complete response, partial response, and stable disease—was 47%, with six-month PFS of 17%.

In group 2, patients who had received initial chemoradiotherapy with temozolomide and at least six cycles of adjuvant temozolomide received continuous temozolomide at progression, which occurred at a median of three months (range, two to 10) following adjuvant temozolomide. These patients received a median of five cycles of continuous temozolomide 50 mg/m². In this group, two patients (14%) had partial responses, nine (64%) had stable disease, and the six-month PFS from the time of first relapse was 57%.

To patients in group 3 (other anaplastic gliomas at second relapse), all of whom had been treated with surgery and radiation followed by adjuvant temozolomide at first progression, continuous temozolomide was offered until further progression. There were two (14%) partial responses, and six patients (43%) had stable disease. The overall clinical benefit rate was 57%, and six-month PFS from the time of second relapse was 42%.

The only observed toxicities during continuous temozolomide therapy were hematologic, primarily lymphopenia. There were no grade 4 toxicities (Table 1) (Perry et al. Cancer 2008).

RESCUE Study

Given the encouraging results seen in the retrospective study, we decided to undertake a phase II study of continuous dose-intense temozolomide for up to one year in patients with high-grade glioma who failed the standard five-day adjuvant regimen.

Patients were enrolled in the RESCUE trial in a two-stage design, under which the next stage could not be initiated until at least one of the 15 patients enrolled had achieved a six-month PFS.

Patients had GBM at first relapse after standard chemoradiation with 60 Gy delivered in 30 fractions, and concurrent temozolomide 75 mg/m<sup&g
eks, plus adjuvant temozolomide 150 to 200 mg/m² given five days out of 28.

Table 1.

<img2607|center>

Three groups of patients were identified a priori:

• Group B1: Those in early adjuvant failure (i.e. during temozolomide months 3 to 6).

• Group B2: Those with extended adjuvant failure (during temozolomide therapy but after at least six months). This group included some patients who had been on adjuvant temozolomide for more than three years.

• Group B3: Those in failure following completed adjuvant therapy of at least six months of temozolomide.

The groups were identified for separate analysis because it was thought they might be biologically different and that the effectiveness of temozolomide would be different based on the level of prior exposure. A fourth group of patients with anaplastic gliomas was included. These patients had completed two or more cycles of conventional temozolomide.

The primary end point was six-month PFS. Secondary end points included objective response rate, time and duration of response, overall survival at 12 months, correlation between efficacy and duration of prior temozolomide therapy, and safety.

A total of 120 patients—30 per group— were enrolled in 12 centres across Canada. A preliminary analysis of the first 90 patients was presented at the annual meeting of the American Society of Clinical Oncology in Chicago, Illinois, in June 2008 (Perry et al. ASCO 2008, abstract 2010).

The investigators found that six-month PFS among patients in group B1 (early failure) was 28.6%. In group B2 (extended adjuvant failure), the six-month PFS rate was 9.5%. Among patients who had completed temozolomide (group B3), 30.4% had six-month PFS, and among patients with anaplastic tumours, the six-month PFS was 42.1%. Among all glioblastoma patients combined, the six-month PFS rate was 24%.

Non-hematological toxicities were generally not severe with no grade 4 toxicities; grade 3 toxicities (each occurring in two patients) were limited to fatigue, headache, muscle weakness, seizure, aphasia, and confusion. Lymphopenia was the most common hematological toxicity, occurring in nearly half of all patients, with grade 3 events occurring in about 3%. Despite the frequency of lymphopenia, however, there were no opportunistic infections reported by investigators as of June 2008. No prophylaxis for Pneumocystis carinii pneumonia was required.

The Ongoing RESCUE study

The RESCUE trial is ongoing, with six-month PFS data on the remaining patients and overall survival data still to be reported. Future directions for investigation include further exploration of alternative temozolomide regimens and analysis of MGMT status. Models of the potential anti-angiogenic activity of these regimens are to be undertaken. In addition, controlled studies comparing these extended temozolomide regimens to other agents are required to determine the efficacy of this approach.

CLINICAL TRIALS IN GLIOBLASTOMA MULTIFORME

Editorial Overview:

Warren P. Mason, MD, FRCPC

Medical Director, The Gerry & Nancy Pencer Brain Tumor Centre, UHN-Princess Margaret Hospital, Associate Professor of Medicine, University of Toronto, Toronto, Ontario

The poor prognosis of patients with recurrent high-grade gliomas demands therapeutic innovation aimed at prolonging survival while helping patients retain at least a fair quality of life. Investigational options for the treatment of high-grade gliomas include alternating one-weekon, one-week-off schedules of dosedense temozolomide, and accelerated hypofractionated, intensity-modulated radiation therapy with concurrent and adjuvant temozolomide. In both cases, the therapy is well tolerated and appears efficacious. Other options that are being explored include antiangiogenic agents and novel therapeutic compounds.

Despite extensive resection, radiation therapy and chemotherapy, the majority of glioblastoma mutiforme (GBM) tumours will recur, with a median time to progression after radiation therapy and concomitant temozolomide of 6.9 months (Wen PY, Kesari S. N Engl J Med 2008;359:492-507). The poor prognosis for patients with recurrent GBM has driven a nearly constant stream of clinical trials aimed at evaluating emerging therapies, and at refining existing front-line therapies in hopes of preventing or delaying recurrence and improving both PFS and overall survival (OS) rates (Table 1).

Trials with Existing Therapies for Newly Diagnosed Patients

Temozolomide

In addition to a recently published retrospective study of the strategy of rechallenging patients with recurrent glioblastoma with continuous temozolomide dosing (Perry et al. Cancer 2008;113:2152-7), other ongoing clinical trials are investigating novel dosing strategies with temozolomide for first-line therapy of glioblastoma.

Dr. Stefano Dall’Oglio, Ospedale Civile Maggiore, Verona, Italy, and colleagues have published results of a phase II trial observing the efficacy of a weekly alternating temozolomide schedule after surgery and concomitant chemoradiotherapy in patients with newly diagnosed high-grade gliomas (J Neurooncol 2008:90(3):315-9). They enrolled 34 patients (21 men, 13 women; age range, 30 to 70 years; mean age, 53 years), 32 of whom had GBM and two of whom had anaplastic astrocytoma. Each patient underwent surgery and concomitant chemoradiotherapy. After a four-week break, patients were placed on an alternating one-week-on, one-week-off schedule of temozolomide for a total of 12 cycles. For the first cycle, patients received temozolomide 75 mg/m², for the second, 100 mg/m², for the third, 125 mg/m², and for the fourth through the 12th cycl
;2. Patients were monitored for hematological toxicity every week during concomitant chemoradiotherapy, and then every four weeks.

Table 1.

<img2608|center>

The authors found that 12 months from the end of radiotherapy, the OS rate was 59%, with a median of 13 months. Six of the 34 patients had hematological toxicities, with grade 1 neutropenia occurring in four, grade 2 thrombocytopenia in one, and grade 4 thrombocytopenia plus grade 1 neutropenia in one. There was one case of Pneumocystis carinii pneumonia, but no other opportunistic infections were observed. The authors concluded that the results seemed encouraging for high-grade glioma patients with good prognostic factors (i.e. recursive partitioning analysis [RPA] Class 1 patients).

Accelerated hypofractionated IMRT

Investigators from McGill University in Montreal, Quebec, have initiated a protocol in which accelerated hypofractionated intensity-modulated radiotherapy (hypo- IMRT) is delivered with concomitant and adjuvant temozolomide to patients with newly diagnosed GBM. (Panet-Raymond et al. Int J Radiation Oncology Biol Phys 2008 Jun 11;Epub ahead of print).

Patients received radiotherapy over four weeks with a concomitant boost technique, for a total dose of 60 Gy delivered to the periphery of the gross tumour volume, and 40 Gy to the planning target, with dose divided into 20 fractions. Temozolomide was delivered concomitantly during radiotherapy and adjuvantly thereafter, according to the Stupp regimen (N Engl J Med 2005;352:987-96).

A retrospective analysis of 35 patients who underwent treatment under the protocol showed that after a median follow-up of 12.6 months, 29 (82.8%) had completed the combined modality treatment, with 25 (71.4%) receiving a median of four cycles of adjuvant temozolomide. The median OS was 14.4 months, and median disease-free survival was 7.7 months. Median survival was significantly better among patients who underwent partial or total tumour resection compared with biopsy only, at 16.1 months for the resected patients, compared with 7.1 months for the biopsy-only patients (P=0.035).

Methylation status also made a difference: patients whose tumours had methylated 0-6-methylguanine-DNA methyltransferase (MGMT) promoters had a median survival of 14.4 months, compared with 8.7 months for those whose tumours had unmethylated MGMT promoters (P=0.049). There was a predominantly central pattern of failure within 2 cm of the initial gross tumour volume.

The treatment appeared to be very well tolerated, with grade 3-4 toxicity limited to a single patient who experienced nausea and emesis during adjuvant temozolomide administration.

The authors noted that although threefourths of the patients had poor prognosis disease (RPA class V or VI), the median survival in the study was comparable to that seen with standard radiotherapy fractionation schedules plus temozolomide.

Clinical Trials for Recurrent Glioblastoma: Antiangiogenic Agents

The most active area of clinical investigation in the prevention or treatment of glioblastoma recurrence is antiangiogenesis. Agents that target vascular growth factors or growthfactor receptors, which have shown activity against a variety of solid tumours, are currently being explored in mid- to late-stage clinical trials.

Cediranib

Cediranib (AZD2171) is a potent oral vascular endothelial growth factor (VEGF) receptor tyrosine kinase inhibitor (TKI) that selectively inhibits VEGF receptors 1, 2 and 3. In a phase II study of 31 patients with GBM who had failed previous therapy with radiation, surgery and chemotherapy, the six-month PFS was 25.8%, compared with 15% for historic controls, reported Dr. Tracy Batchelor, Massachusetts General Hospital, Boston, and colleagues at the 2008 annual meeting of the American Association for Cancer Research (Batchelor et al. AACR 2008, Abstract LB-247). Median OS was 221 days, and median PFS was 117 days.

The investigators had previously shown that cediranib’s primary target, VEGFR-2, is expressed on glioblastoma vascular endothelia. In a study published last year, Batchelor et al. used MRI techniques to show prolonged but reversible normalization of tumour vessels of recurrent glioblastomas with once-daily administration of the drug (Batchelor et al. Cancer Cell 2007;11(1):83-95). The normalization effect lasted for a minimum of four weeks, and persisted for as long as four months. Cediranib also alleviated tumour edema.

Bevacizumab

Bevacizumab is a humanized monoclonal IgG1 antibody that binds to and inhibits the biological activity of human VEGF-A. It is approved in Canada for the treatment of metastatic cancer of the colon or rectum in combination with fluoropyrimidine-based chemotherapy.

In 2007, Dr. James J. Vrendenburgh, Duke University, Durham, North Carolina, and colleagues published results of a phase II study of bevacizumab in combination with the topoisomerase-1 inhibitor irinotecan in 32 adults with recurrent grade III-IV glioma, 23 of whom had grade IV tumours (Vrendenburgh et al. Clin Cancer Res 2007;13(4):1253-9). The patients received bevacizumab 10 mg/kg and irinotecan 125-340 mg/m2 intravenously (depending on patient use of enzyme-inducing antiepileptic drugs) every two weeks of a sixweek cycle. Median PFS for patients with grade IV disease was 20 weeks, and 30 weeks for patients with grade III tumours (overall PFS 23 weeks). The probability of six-month PFS was 38%, and the probability of sixmonth OS was 72%. There were no central nervous system hemorrhages, although there were three cases of deep venous thrombosis or pulmonary emboli. One patient had an arterial ischemic stroke.

In a randomized phase II, non-comparative trial of bevacizumab alone or with irinotecan, Cloughesy and colleagues at centres in the US reported favourable six-month PFS and overall survival with bevacizumab alone or in combination with irinotecan, with the added benefit of a reduction in corticosteroid use (Cloughesy et al. ASCO 2008, abstract 2010b). Median OS among 85 patients receiving bevacizumab only was 9.2 months and 8.7 months for 82 patients on the combination. Six-month PFS was 42.61% among patients in the bevacizumab-only group and 50.2% in the combination group. The median duration of response was 5.6 months for bevacizumab alone and 4.3 months for the combination. Grade 3 or greater adverse events included venous thromboembolism in three patients on bevacizumab alone and in seven on the combination, and infections in eight and 11 patients, respectively. Other adverse events occurring in one or two patients in each group include arterial thromboembolic events, wound-healing complications, proteinuria (one patient on the combination) and hemorrhage (two patients on the combination).

Other Strategies

Research into more effective and less toxic therapies for recurrent GBM continues apace. Neyns and colleagues from the Netherlands are investigating sunitinib, a small molecule that inhibits VEGF receptors and other receptor kinases. This agent has shown activity against renal cell carcinomas and gastrointestinal stromal tumours, but preliminary results in high-grade gliomas have been suboptimal.

Other strategies that have been explored include convection-enhanced delivery of diphtheria toxin or Pseudomonas exotoxin directly to tumour; re-irradiation; and the use of agents that have shown efficacy against other tumours in the laboratory or in the clinic, including mTOR inhibitors such as temsirolimus and everolimus, and other TKIs, including imatinib, erlotinib, and gefitinib, but none of these agents or strategies have thus far approached therapies currently in use. The search continues.