A Review of Dose-dense Temozolomide Alone and in Combination with Bevacizumab in Patients with First Relapse of Glioblastoma

Current Standard Treatment

The recommended standard treatment has long included surgery and radiation therapy. Traditionally, chemotherapy has been saved for second-line treatment. Surgery is usually performed soon after initial radiological diagnosis, while its aim is to obtain material for a conclusive diagnosis, as well as ro remove the tumor mass. If extensive surgery is not possible, at least a biopsy can most often be obtained to establish a histological diagnosis. Surgery is then usually followed by external radiation therapy. Despite intensive research in the field, further advancement in prolonging survival was not successful until it was shown that the oral chemotherapeutic agent temozolomide, used concomitantly with radiotherapy and in the adjuvant setting, significantly prolonged median survival from 12.1 to 14.6 months for patients with newly diagnosed glioblastoma (5). In the 5-year analysis, 9.8% of patients in the temozolomide arm were still alive, while in the control arm this figure was only 1.9% (6).

Treatments at Relapse

Previously, chemotherapy was used only at relapse, which invariably occurs. The most studied and recommended treatments were historically based on nitrosourea, often lomustine, given in combination with procarbazine and vincristine (PCV) (7). Several other chemotherapeutic combinations have been investigated, resulting in an objective response rate of less than 10% and a PFS of 2-3 months independently of which drug was being tested (3). Until 2009, no significant advances were seen in the relapse setting, but then a combination of bevacizumab and irinotecan suggested a significant clinical improvement (4). In a recent phase III study (8), 447 patients with high-grade glioma at relapse were randomly assigned to six cycles of PCV or to either of the temozolomide arms (200 mg/m² for 5 out of 28 days or 100 mg/m² for 21 out of 28 days). Treatment in all arms continued for up to 9 months, or until progression. When comparing the combined temozolomide arms with PCV-treated patients, no difference in survival was seen between the treatments. Notably, this study commenced before the results of the study with concomitant and adjuvant temozolomide were published. Therefore, since the patients in this study were chemotherapy-naïve at relapse and most patients are now treated with chemotherapy in the adjuvant setting, the results might not be fully applicable to the current population with relapsed glioblastoma.

Rationale for Current Dosing of Temozolomide

Temozolomide is an alkylating agent given in primary treatment at a standard dose of 150-200 mg/m² for 5 consecutive days in a 28-day cycle. In the initial study, this dose was found to be tolerable with an acceptable toxicity, but myelosuppression was dose-limiting (10). Another study concluded that a daily dose of 75 mg/m² temozolomide is also generally well-tolerated and that treatment can last for up to 7 weeks without increased toxicity. Use of this dose increases the cumulative dose two-fold compared to use of the standard dose. At higher daily doses (100 mg/m²) grade IV leucopenia and thrombocytopenia were seen in 25% of patients (11). These doses are now used in the standard treatment, i.e. 75 mg/m² daily concomitant with radiation therapy and 150-200 mg/m² for 5 out of 28 days in six cycles in the adjuvant setting after radiation therapy.

Efficacy of Temozolomide

The cytotoxic effect of temozolomide and other alkylating agents consists primarily of transferring a methyl group to the O-6 position of guanine in the DNA molecule, causing defective DNA replication and tumour cells to enter apoptosis. However, this type of lesion is normally repaired by the DNA repair enzyme methylguanine-DNA methyltransferase (MGMT). Based on the repair mechanism, it can be estimated that the main cytotoxic effect of alkylating agents is dependent on the MGMT enzyme. A cell containing more MGMT (expression is up-regulated) is more resistant to the effect of the alkylator, while a lower expression renders the cell sensitive. Expression of MGMT, like many other genes, is dependent on the methylation status of the promoter, to which transcription factors bind and commence transcription and synthesis of proteins. Methylation of the promoter interferes with the transcription factors and inhibits transcription of the gene associated with that promoter. Promoter methylation effectively silences the gene and the protein levels are consequently low. There are numerous reports of a correlation between MGMT methylation status and response to temozolomide treatment and also survival, where the silenced gene correlates to better prognosis (12, 13). Patients with low MGMT expression receiving temozolomide treatment had a significantly longer survival than those with a high expression of MGMT (13). In fact, the strongest predictor of outcome and benefit of temozolomide treatment was MGMT methylation status (6). It should be emphasized that when comparing patients with methylated and those with unmethylated promoters receiving the same treatment, i.e. radiotherapy alone or its combination with temozolomide, OS was lower in the patients with unmethylated promoter in both treatment groups, possibly reflecting differences in tumour biology. This might also be an effect of alkylating agents as salvage treatment, which will possibly further increase the gap between methylated and unmethylated groups as patients with an elevated expression of MGMT are not likely to respond to such treatment in the salvage setting either (13).

MGMT Depletion

MGMT repairs DNA by transferring the alkyl group added by an alkylating agent from the O-6 position of guanine in the DNA molecule to an internal cysteine residue in the active site of MGMT. Thereby, the DNA is repaired and replication can proceed. When this happens, the MGMT enzyme is irreversibly inactivated and the enzyme is marked for destruction in the ubiquitin proteolytic pathway. To restore its function, de novo protein synthesis of MGMT is required. This has led to the hypothesis that MGMT can be consumed faster than it is synthesized. Thus, the depletion of MGMT will reduce the ability of the cells to recover from treatment with alkylating agents and this might also enable patients with high expression of MGMT to benefit from temozolomide treatment. Depletion of MGMT can hypothetically be achieved by introducing a false substrate for MGMT, such as O-6-benzylguanine, which will consume MGMT at a high rate. The drawback of this method is that it also has the risk of increasing toxicity as MGMT in normal cells is also inactivated. Another option is to add more alkylating agents in order to introduce more alkyl groups, thereby saturating the MGMT system.

MGMT depletion has been studied intensely, mostly in peripheral blood monocytic cells (PBMCs). These are used as a surrogate marker for tumour MGMT depletion but it should be emphasized that further evidence is needed regarding the correlation between MGMT levels in PBMCs and tumour tissue. Tolcher et al. measured the levels of MGMT in PBMCs from patients under different regimens of temozolomide treatment and found temozolomide to be effective in depleting MGMT, even at relatively low doses. One arm, receiving temozolomide for 7 out of 14 days at a dose of 75-175 mg/m², showed a marked decrease in MGMT activity after 7 days of treatment, which was partly restored after 7 days off-treatment. The resulting decrease was approximately 55% of baseline levels. The other arm, receiving 85-125 mg/m² temozolomide for 21 out of 28 days, had a slower decline but the decrease was similar to that of the first arm (14). This implies that a dose-dense regimen of temozolomide might be efficient in depleting MGMT.

Dose-dense Temozolomide

A few clinical trials have been performed on a limited number of patients, where an altered dosing regimen of temozolomide has been evaluated to clarify whether it is possible to increase the cumulative dose and effect without increasing toxicity by depleting the MGMT levels. So far, the concept has only been evaluated in small phase II studies, where the results must be interpreted with precaution as the level of evidence is relatively weak. Two regimens have been investigated more thoroughly; one with temozolomide at 100-150 mg/m² for 21 out of 28 days and one with 150 mg/m² for 7 out of 14 days (7 days on/7 days off).

The 7 days on/7 days off regimen was evaluated in 21 patients with recurrent glioblastoma and objective response was seen in 10% of the patients whereas PFS at 6 months was 48%. The treatment was well-tolerated (15) and further analyzed in a non-randomized phase II study of 90 patients with recurrent glioblastoma, where PFS at 6 months was found to be 44%. The MGMT promoter methylation status of the tumours was analyzed and no differences were seen in survival depending on methylation status. The overall response rate was 10% and the authors concluded that a possible depletion of MGMT was accomplished with the used regimen (16). The 21/28 days schedule, evaluated in 18 patients, was also shown to be feasible with a response rate of 22%. MGMT promoter methylation status was assessed only in seven cases, which was insufficient for any firm conclusions (17). In another study, where the concept was to evaluate the efficacy of temozolomide compared to the previously established recurrent therapy with PCV in glioblastoma, two different temozolomide treatment arms were used. The treatment was given in either the current standard scheme (TMZ-5) or a dose-dense scheme of treatment with 100 mg/m² for 21 out of 28 days (TMZ-21). The results of this study indicated an inferiority of the dose-dense regimen, and the authors speculated that the cumulative dose may be less important than the peak doses accomplished with the higher daily dose in the TMZ-5 arm. In this respect, the other option of dose-dense temozolomide with 7 days on, 7 days off seems more appealing (8). In a recent phase II study by Abacioglu et al., the efficacy and tolerability of protracted, dose-dense temozolomide therapy (100 mg/m2 for 21 consecutive days of a 28-day cycle) was evaluated in patients with recurrent glioblastoma or grade III glioma who had previously received standard therapy. Out of 25 patients included, two patients had partial responses and 10 had stable disease (60% overall clinical benefit). The median PFS was 3 months (95% confidence interval, CI=1.8-4.2) and the median overall survival was 7 months (95% CI=5.1-8.9 months). The regimen was well-tolerated and the authors concluded that protracted, dose-dense temozolomide had modest activity with manageable toxicity in these patients (9). In a compelling study, the dose-dense scheme (7 days on, 7 days off) was compared with a metronomic regimen of 50 mg/m² daily continuously in the adjuvant setting for newly-diagnosed glioblastoma (18). To facilitate benchmarking of the results with the EORTC/NCIC phase III study (5), treatment was given in the same way, but adjuvant temozolomide was given in either a dose-dense or a metronomic scheme. Adjuvant treatment was given in six cycles in both arms. Each arm consisted of approximately 40 patients. The results showed no benefit of administering temozolomide in the metronomic regimen, as no sign of supremacy against historical data of the EORTC/NCIC-study was seen. Moreover, the metronomic regimen was also considered more costly. However, the dose-dense scheme possibly led to a median survival time (17.1 months) superior to the compared results and the authors proposed further studies in a phase III setting. In particular, patients with an unmethylated MGMT promoter (which results in high MGMT levels) seemed to benefit from the treatment when compared with historical data (overall survival of 15.4 months, compared to 12.7 months reported from the EORTC/NCIC study) (13). This possibly suggests that the dose-dense scheme of temozolomide was able to deplete MGMT in tumour tissue (18); however, as can be judged from the results achieved so far, depletion of MGMT seems to add minor improvement to the treatment of high-grade glioma.

Bevacizumab

Rationale. A growing tumour is dependent on oxygen and nutrients. When the tumour is small, sufficient supplies are maintained with simple diffusion of nutrients from surrounding capillaries. To sufficiently support a growing tumour, a crucial step is when the cell acquires the ability to attract blood vessels, the ‘angiogenic switch’. When a cell is hypoxic, i.e. insufficient oxygen supply, the expression of angiogenic vascular endothelial growth factor (VEGF) is induced to attract surrounding vessels. Endothelial cells and pericytes in the vessel walls respond to the signal by growing in the direction of the signal. Tumour cells, which are often hypoxic as they grow uncontrolled beyond the limits of simple diffusion, also secrete VEGF. A high level of VEGF secretion correlated to dense vascularization and also a high grade of malignancy in glioblastomas (19). However, a high expression of VEGF does not induce the formation of reliable vessels. The tumour-induced vessels vary in size, with calibre shifts and excessive branching, contain shunts and are also much more permeable than is normal. These properties affect the transportation of nutrients in the vessels, which is impaired as the blood flow is not optimal. The possibility of delivering blood-borne drugs effectively is consequently also reduced.

Such malformed vessels leak fluids into surrounding tissues as a consequence of the increased vascular permeability, giving rise to oedema in the surrounding brain tissue. The oedema raises the intracranial pressure, causing severe symptoms in the patients such as headache, nausea and neurological deficits. Oedema can often be relieved by treatment with corticosteroids. As corticosteroids have multiple unwanted side-effects, it must be emphasized that it is of utmost importance to keep the dose to a minimum.

Efficacy. Upon treatment with the anti-angiogenic anti-VEGF antibody bevacizumab, a rapid change in the configuration of the blood vessels has been observed, leading to a decrease in permeability, with a consequent lowering of interstitial fluid pressure and normalized blood flow. As oxygen is supplied to the tumour tissue, sensitivity to radiation therapy may increase when more free oxygen radicals are formed. The delivery of chemotherapeutic agents to the tumour tissue may also be facilitated with normalized blood vessels and reduced interstitial fluid pressure. Anti-angiogenic therapy can therefore be considered to create a treatment opportunity window, where tumour cells are more sensitive to treatment. It is reasonable to believe that anti-angiogenic treatment is most effective when combined with cytotoxic treatment.

If the anti-angiogenic treatment is too intense, the vessels are further diminished increasing the risk of even worse hypoxia and consequent tumour necrosis (20). Furthermore, the hypoxic environment created by an intense anti-angiogenic therapy may also lead to selection of tumour cells insensitive to hypoxia, which are often more malignant than their oxygen-consuming counterparts. Bevacizumab has been studied as a single treatment, and in combination with irinotecan for recurrent glioblastoma. One hundred and sixty-seven patients were enrolled in a randomized phase II study, half of them received bevacizumab at 10 mg/kg and the other half the same dose of bevacizumab, but in combination with irinotecan at 340 mg/m² or 125 mg/m² (with or without enzyme-inducing antiepileptic drugs, respectively). Bevacizumab was given every two weeks. The regimen was concluded as being tolerable and effective, as objective response rates were seen in 28% and 38% in the bevacizumab and in the combination arm, respectively. PFS at six months was estimated to be 42.6% and 50.3%, respectively, which are impressive numbers compared to historical data. A trend for decreasing corticosteroid use was also seen. The median OS was 8.7 months for the combination arm and 9.2 months for the bevacizumab-alone arm (4).

Combination of Bevacizumab and Temozolomide

Both in vitro and in vivo studies of the combination of bevacizumab and temozolomide give encouraging results with a superior survival in animal xenograft studies, when compared to the agents used as single treatment (21). Recently, a phase II study was published where 70 patients with newly diagnosed glioblastoma were treated according to the current standard treatment with concomitant and adjuvant temozolomide, but with the addition of bevacizumab (10 mg/kg) every 14 days (22). The study was a single-arm study, but a reference population of patients treated at the same site according to the current standard protocol served as a comparison. The results were also benchmarked with the results of the EORTC-NCIC (5) study regarding PFS and OS. OS was 19.6 months in the study group, and 21.1 months in the control group (p=0.06), while PFS was 13.6 months and 7.6 months (p=0.005), respectively. Regarding OS, both groups were superior to the EORTC-NCIC study, possibly suggesting that new treatments in the salvage setting are of value. In the reference group, 51% received bevacizumab at recurrence. When comparing PFS, the study group was superior to both the EORTC-NCIC study and the reference group. A possible interpretation of these results might be that bevacizumab prolongs survival, either given as first-line therapy or at recurrence. If given adjuvantly, the time-to- recurrence is increased, and when given at recurrence, an equal effect is seen as OS does not differ. The toxicity observed was acceptable and anticipated for bevacizumab. Currently, a randomized phase III study is evaluating the addition of bevacizumab to the first-line standard treatment (23).

In the relapse setting, there are ongoing studies with bevacizumab at 10 mg/kg every second week in combination with 7 days on/7 days off temozolomide at a daily dose of 100 mg/m². Results published so far consist of five accrued patients out of 30 planned and conclude that it is a well-tolerated and effective regimen (24). In another study, daily temozolomide (50 mg/m²) in combination with bevacizumab (10 mg/kg) was evaluated in 32 patients with recurrent glioblastoma. All patients had undergone standard radiotherapy and 63% had progressed on a 5-day temozolomide regime. The authors found acceptable toxicity and a response rate of 37.5% (25).