Frontline bevacizumab fails to boost glioblastoma survival in key trials

CHICAGO – Adding bevacizumab to frontline radiation and temozolomide therapy for glioblastoma produced disappointing results in a double-blind, placebo-controlled phase III trial.

Bevacizumab (Avastin) extended progression-free survival, but did not improve overall survival in the Radiation Therapy Oncology Group (RTOG) 0825 study, and investigators reported that the angiogenesis inhibitor was associated with worse neurocognitve and quality of life outcomes.

Among 637 patients who were randomized, median overall survival reached 16.1 months in those assigned to radiation, temozolomide (Temodar), and placebo, compared with 15.7 months in patients assigned to radiation, temozolomide, and bevacizumab, Dr. Mark R. Gilbert reported in the plenary session at the annual meeting of the American Society of Clinical Oncology.

Median progression-free survival reached 10.7 months in the bevacizumab arm, vs. 7.3 months in the placebo arm, said Dr. Gilbert, professor of neuro-oncology at the University of Texas M.D. Anderson Cancer Center in Houston. The P value was .0007, but this difference was not statistically significant because of an unusual trial design that split criteria for statistical significance between the two primary outcome measures of overall and progression-free survival. In addition, patients in the bevacizumab arm had significantly worse neurocognitive and overall symptom scores over time.

"We feel that bevacizumab remains an important therapy for our patients with glioblastoma, but the results of this study do not support its frontline use. Rather, it can be reserved as a later treatment," he said at a press briefing prior to his presentation.

Outcomes contrast with Avaglio study

Some results of the publicly funded RTOG 0825 trial run counter to those from the Avaglio study, an industry-sponsored trial whose results were also presented at the ASCO annual meeting.

The Avaglio investigators found that adding bevacizumab to radiotherapy and temozolomide "achieved a clinically meaningful, statistically significant progression-free survival improvement in patients with glioblastoma," Dr. Warren Mason, a neuro-oncologist at the Princess Margaret Hospital in Toronto, reported at another session.

The Avaglio trial, which ran parallel to RTOG 0825 and was very similar in design, showed a quality of life benefit for patients on bevacizumab, but did not look at neurocognitive outcomes.

Median progression-free survival in Avaglio reached 10.6 months in 458 patients in its radiation, temozolomide, and bevacizumab arm, vs. 6.2 months in 463 patients in its radiation, temozolomide, and placebo arm, and the difference was significant (hazard ratio, 0.64; P less than .0001).

However, as in RTOG 0825, median overall survival was virtually identical in both Avaglio trial arms, at 16.8 and 16.7 months, respectively.

Bevacizumab misses RTOG targets

The RTOG trial enrolled neurologically stable patients with glioblastoma who had a Karnofsky performance score of at least 70 and who had tumor tissue samples available for evaluation. They were randomized to receive standard chemoradiation with temozolomide for 3 weeks followed by chemoradiation with temozolomide and either bevacizumab or placebo for 6-12 maintenance cycles.

The treatment type was unblinded at progression, and patients were allowed to cross over, or continue on, bevacizumab. Progression was defined as a greater than 25% increase in tumor area without recent steroid reduction, or worsening of neurologic symptoms.

In a pooled analysis including both study arms, patients with methylation of the DNA repair enzyme O6-methylguanine-DNA methyltransferase (MGMT) had better median and overall survival than patients with MGMT unmethylated (methylated MGMT is a marker for favorable prognosis). Median overall survival for the MGMT methylated group was 23.2 months, compared with 14.3 months for the unmethylated group (HR in unmethylated tumors, 2.10; P less than .001). Respective median progression-free survival was 14.1 months vs. 8.2 months (HR, 1.67; P less than .001).

(The results of the GLARIUS trial, also reported at this meeting, showed that bevacizumab combined with irinotecan (Camptosar) in chemoradiation was associated with longer progression-free survival than temozolomide chemoradiation in newly diagnosed patients with MGMT-unmethylated glioblastoma. In a preliminary finding from that study, there was a suggestion of an overall survival benefit for the bevacizumab-irinotecan combination.)

As noted, in RTOG 0825, patients in the bevacizumab arm scored significantly worse on a clinical trial battery composite of cognitive function (P = .038), on overall symptom interference with daily activity (P less than .001), and on quality of life cognitive function (P less than .009).

In contrast, patients on bevacizumab in Avaglio had improved quality of life, prolonged preservation of performance scores, and reduced steroid doses compared with controls.

Why the differences?

Dr. Howard A. Fine, director of the brain tumor center at New York University, the invited discussant for RTOG 0825, said that the differences in patient-reported outcomes between the trials may reflect differences in statistical analysis, possibly different analytical methodologies used to query the data, or the lack of neurocognitive data in Avaglio.

Another possibility to account for the worsening neurocognitive function with bevacizumab is the drug’s documented ability to stabilize the blood-brain barrier and decrease MRI gadolinium enhancement, he said. This could mean that patients experience radiographic occult disease progression that does not show up well on brain scans.

Dr. Gilbert and Dr. Fine commented that theoretically at least, glioblastoma should be an ideal target for angiogenesis inhibitors such as bevacizumab, because they display a high degree of vascularity, and endothelial proliferation is part of the pathologic definition of the disease.

But although bevacizumab has been demonstrated to improve response rates and delay progression in recurrent glioblastoma, its relative lack of efficacy in the first line is puzzling, Dr. Fine said.

Potential explanations include the crossover designs of both the RTOG and Avaglio studies, suboptimal delivery of bevacizumab to the tumor, or the possibility that the vascular endothelial growth factor receptor may not be the best target in glioblastoma.

"Despite these new data demonstrating its limitation, I feel very strongly that bevacizumab represents the single most important agent in glioblastoma since temozolomide, and maybe even more so. Ongoing and future trials will better define how and when it should be optimally used in these patients that have such limited therapeutic options," Dr. Fine concluded.

The RTOG 0285 study was supported by the National Cancer Institute, with additional support from Genentech. Avaglio was supported by Roche. Dr. Gilbert disclosed consulting for, and receiving honoraria and research support from, Genentech. Dr. Mason disclosed being a consultant/adviser to Hoffman-La Roche. Dr. Fine reported having no relevant financial disclosures.

CHICAGO – Adding bevacizumab to frontline radiation and temozolomide therapy for glioblastoma produced disappointing results in a double-blind, placebo-controlled phase III trial.

Bevacizumab (Avastin) extended progression-free survival, but did not improve overall survival in the Radiation Therapy Oncology Group (RTOG) 0825 study, and investigators reported that the angiogenesis inhibitor was associated with worse neurocognitve and quality of life outcomes.

Among 637 patients who were randomized, median overall survival reached 16.1 months in those assigned to radiation, temozolomide (Temodar), and placebo, compared with 15.7 months in patients assigned to radiation, temozolomide, and bevacizumab, Dr. Mark R. Gilbert reported in the plenary session at the annual meeting of the American Society of Clinical Oncology.

Median progression-free survival reached 10.7 months in the bevacizumab arm, vs. 7.3 months in the placebo arm, said Dr. Gilbert, professor of neuro-oncology at the University of Texas M.D. Anderson Cancer Center in Houston. The P value was .0007, but this difference was not statistically significant because of an unusual trial design that split criteria for statistical significance between the two primary outcome measures of overall and progression-free survival. In addition, patients in the bevacizumab arm had significantly worse neurocognitive and overall symptom scores over time.

"We feel that bevacizumab remains an important therapy for our patients with glioblastoma, but the results of this study do not support its frontline use. Rather, it can be reserved as a later treatment," he said at a press briefing prior to his presentation.

Outcomes contrast with Avaglio study

Some results of the publicly funded RTOG 0825 trial run counter to those from the Avaglio study, an industry-sponsored trial whose results were also presented at the ASCO annual meeting.

The Avaglio investigators found that adding bevacizumab to radiotherapy and temozolomide "achieved a clinically meaningful, statistically significant progression-free survival improvement in patients with glioblastoma," Dr. Warren Mason, a neuro-oncologist at the Princess Margaret Hospital in Toronto, reported at another session.

The Avaglio trial, which ran parallel to RTOG 0825 and was very similar in design, showed a quality of life benefit for patients on bevacizumab, but did not look at neurocognitive outcomes.

Median progression-free survival in Avaglio reached 10.6 months in 458 patients in its radiation, temozolomide, and bevacizumab arm, vs. 6.2 months in 463 patients in its radiation, temozolomide, and placebo arm, and the difference was significant (hazard ratio, 0.64; P less than .0001).

However, as in RTOG 0825, median overall survival was virtually identical in both Avaglio trial arms, at 16.8 and 16.7 months, respectively.

Bevacizumab misses RTOG targets

The RTOG trial enrolled neurologically stable patients with glioblastoma who had a Karnofsky performance score of at least 70 and who had tumor tissue samples available for evaluation. They were randomized to receive standard chemoradiation with temozolomide for 3 weeks followed by chemoradiation with temozolomide and either bevacizumab or placebo for 6-12 maintenance cycles.

The treatment type was unblinded at progression, and patients were allowed to cross over, or continue on, bevacizumab. Progression was defined as a greater than 25% increase in tumor area without recent steroid reduction, or worsening of neurologic symptoms.

In a pooled analysis including both study arms, patients with methylation of the DNA repair enzyme O6-methylguanine-DNA methyltransferase (MGMT) had better median and overall survival than patients with MGMT unmethylated (methylated MGMT is a marker for favorable prognosis). Median overall survival for the MGMT methylated group was 23.2 months, compared with 14.3 months for the unmethylated group (HR in unmethylated tumors, 2.10; P less than .001). Respective median progression-free survival was 14.1 months vs. 8.2 months (HR, 1.67; P less than .001).

(The results of the GLARIUS trial, also reported at this meeting, showed that bevacizumab combined with irinotecan (Camptosar) in chemoradiation was associated with longer progression-free survival than temozolomide chemoradiation in newly diagnosed patients with MGMT-unmethylated glioblastoma. In a preliminary finding from that study, there was a suggestion of an overall survival benefit for the bevacizumab-irinotecan combination.)

As noted, in RTOG 0825, patients in the bevacizumab arm scored significantly worse on a clinical trial battery composite of cognitive function (P = .038), on overall symptom interference with daily activity (P less than .001), and on quality of life cognitive function (P less than .009).

In contrast, patients on bevacizumab in Avaglio had improved quality of life, prolonged preservation of performance scores, and reduced steroid doses compared with controls.

Why the differences?

Dr. Howard A. Fine, director of the brain tumor center at New York University, the invited discussant for RTOG 0825, said that the differences in patient-reported outcomes between the trials may reflect differences in statistical analysis, possibly different analytical methodologies used to query the data, or the lack of neurocognitive data in Avaglio.

Another possibility to account for the worsening neurocognitive function with bevacizumab is the drug’s documented ability to stabilize the blood-brain barrier and decrease MRI gadolinium enhancement, he said. This could mean that patients experience radiographic occult disease progression that does not show up well on brain scans.

Dr. Gilbert and Dr. Fine commented that theoretically at least, glioblastoma should be an ideal target for angiogenesis inhibitors such as bevacizumab, because they display a high degree of vascularity, and endothelial proliferation is part of the pathologic definition of the disease.

But although bevacizumab has been demonstrated to improve response rates and delay progression in recurrent glioblastoma, its relative lack of efficacy in the first line is puzzling, Dr. Fine said.

Potential explanations include the crossover designs of both the RTOG and Avaglio studies, suboptimal delivery of bevacizumab to the tumor, or the possibility that the vascular endothelial growth factor receptor may not be the best target in glioblastoma.

"Despite these new data demonstrating its limitation, I feel very strongly that bevacizumab represents the single most important agent in glioblastoma since temozolomide, and maybe even more so. Ongoing and future trials will better define how and when it should be optimally used in these patients that have such limited therapeutic options," Dr. Fine concluded.

The RTOG 0285 study was supported by the National Cancer Institute, with additional support from Genentech. Avaglio was supported by Roche. Dr. Gilbert disclosed consulting for, and receiving honoraria and research support from, Genentech. Dr. Mason disclosed being a consultant/adviser to Hoffman-La Roche. Dr. Fine reported having no relevant financial disclosures.

CHICAGO – Adding bevacizumab to frontline radiation and temozolomide therapy for glioblastoma produced disappointing results in a double-blind, placebo-controlled phase III trial.

Bevacizumab (Avastin) extended progression-free survival, but did not improve overall survival in the Radiation Therapy Oncology Group (RTOG) 0825 study, and investigators reported that the angiogenesis inhibitor was associated with worse neurocognitve and quality of life outcomes.

Among 637 patients who were randomized, median overall survival reached 16.1 months in those assigned to radiation, temozolomide (Temodar), and placebo, compared with 15.7 months in patients assigned to radiation, temozolomide, and bevacizumab, Dr. Mark R. Gilbert reported in the plenary session at the annual meeting of the American Society of Clinical Oncology.

Median progression-free survival reached 10.7 months in the bevacizumab arm, vs. 7.3 months in the placebo arm, said Dr. Gilbert, professor of neuro-oncology at the University of Texas M.D. Anderson Cancer Center in Houston. The P value was .0007, but this difference was not statistically significant because of an unusual trial design that split criteria for statistical significance between the two primary outcome measures of overall and progression-free survival. In addition, patients in the bevacizumab arm had significantly worse neurocognitive and overall symptom scores over time.

"We feel that bevacizumab remains an important therapy for our patients with glioblastoma, but the results of this study do not support its frontline use. Rather, it can be reserved as a later treatment," he said at a press briefing prior to his presentation.

Outcomes contrast with Avaglio study

Some results of the publicly funded RTOG 0825 trial run counter to those from the Avaglio study, an industry-sponsored trial whose results were also presented at the ASCO annual meeting.

The Avaglio investigators found that adding bevacizumab to radiotherapy and temozolomide "achieved a clinically meaningful, statistically significant progression-free survival improvement in patients with glioblastoma," Dr. Warren Mason, a neuro-oncologist at the Princess Margaret Hospital in Toronto, reported at another session.

The Avaglio trial, which ran parallel to RTOG 0825 and was very similar in design, showed a quality of life benefit for patients on bevacizumab, but did not look at neurocognitive outcomes.

Median progression-free survival in Avaglio reached 10.6 months in 458 patients in its radiation, temozolomide, and bevacizumab arm, vs. 6.2 months in 463 patients in its radiation, temozolomide, and placebo arm, and the difference was significant (hazard ratio, 0.64; P less than .0001).

However, as in RTOG 0825, median overall survival was virtually identical in both Avaglio trial arms, at 16.8 and 16.7 months, respectively.

Bevacizumab misses RTOG targets

The RTOG trial enrolled neurologically stable patients with glioblastoma who had a Karnofsky performance score of at least 70 and who had tumor tissue samples available for evaluation. They were randomized to receive standard chemoradiation with temozolomide for 3 weeks followed by chemoradiation with temozolomide and either bevacizumab or placebo for 6-12 maintenance cycles.

The treatment type was unblinded at progression, and patients were allowed to cross over, or continue on, bevacizumab. Progression was defined as a greater than 25% increase in tumor area without recent steroid reduction, or worsening of neurologic symptoms.

In a pooled analysis including both study arms, patients with methylation of the DNA repair enzyme O6-methylguanine-DNA methyltransferase (MGMT) had better median and overall survival than patients with MGMT unmethylated (methylated MGMT is a marker for favorable prognosis). Median overall survival for the MGMT methylated group was 23.2 months, compared with 14.3 months for the unmethylated group (HR in unmethylated tumors, 2.10; P less than .001). Respective median progression-free survival was 14.1 months vs. 8.2 months (HR, 1.67; P less than .001).

(The results of the GLARIUS trial, also reported at this meeting, showed that bevacizumab combined with irinotecan (Camptosar) in chemoradiation was associated with longer progression-free survival than temozolomide chemoradiation in newly diagnosed patients with MGMT-unmethylated glioblastoma. In a preliminary finding from that study, there was a suggestion of an overall survival benefit for the bevacizumab-irinotecan combination.)

As noted, in RTOG 0825, patients in the bevacizumab arm scored significantly worse on a clinical trial battery composite of cognitive function (P = .038), on overall symptom interference with daily activity (P less than .001), and on quality of life cognitive function (P less than .009).

In contrast, patients on bevacizumab in Avaglio had improved quality of life, prolonged preservation of performance scores, and reduced steroid doses compared with controls.

Why the differences?

Dr. Howard A. Fine, director of the brain tumor center at New York University, the invited discussant for RTOG 0825, said that the differences in patient-reported outcomes between the trials may reflect differences in statistical analysis, possibly different analytical methodologies used to query the data, or the lack of neurocognitive data in Avaglio.

Another possibility to account for the worsening neurocognitive function with bevacizumab is the drug’s documented ability to stabilize the blood-brain barrier and decrease MRI gadolinium enhancement, he said. This could mean that patients experience radiographic occult disease progression that does not show up well on brain scans.

Dr. Gilbert and Dr. Fine commented that theoretically at least, glioblastoma should be an ideal target for angiogenesis inhibitors such as bevacizumab, because they display a high degree of vascularity, and endothelial proliferation is part of the pathologic definition of the disease.

But although bevacizumab has been demonstrated to improve response rates and delay progression in recurrent glioblastoma, its relative lack of efficacy in the first line is puzzling, Dr. Fine said.

Potential explanations include the crossover designs of both the RTOG and Avaglio studies, suboptimal delivery of bevacizumab to the tumor, or the possibility that the vascular endothelial growth factor receptor may not be the best target in glioblastoma.

"Despite these new data demonstrating its limitation, I feel very strongly that bevacizumab represents the single most important agent in glioblastoma since temozolomide, and maybe even more so. Ongoing and future trials will better define how and when it should be optimally used in these patients that have such limited therapeutic options," Dr. Fine concluded.

The RTOG 0285 study was supported by the National Cancer Institute, with additional support from Genentech. Avaglio was supported by Roche. Dr. Gilbert disclosed consulting for, and receiving honoraria and research support from, Genentech. Dr. Mason disclosed being a consultant/adviser to Hoffman-La Roche. Dr. Fine reported having no relevant financial disclosures.