US panel endorses Roche's Avastin for brain cancer
This article was originally published in Scrip
The US FDA's outside experts unanimously endorsed accelerated approval of Roche's (formerly Genentech) anti-VEGF monoclonal antibody Avastin (bevacizumab) in previously treated glioblastoma, finding the anticancer agent was reasonably likely to have a clinically meaningful benefit in this difficult to treat population.
The oncologic drugs advisory committee voted 10 to zero that objective response rates of 19-26% and a four-month median duration of response with bevacizumab were sufficiently meaningful to serve as a surrogate for expected clinical benefit. If licensed, Avastin would be the first glioblastoma therapy to receive accelerated approval based on response rates rather than a survival benefit.
Committee members called the bevacizumab response rates "dramatic" compared with historical rates of 4-7% in other glioblastoma trials. However, it remained unclear to panellists whether the improvements resulted from bevacizumab's antitumour effect, its ability to reduce oedema and radiation-induced necrosis in brain tumour patients, or a combination of both.
confirmatory study
Several panellists said they believed a planned Phase III confirmatory study in the first-line glioblastoma setting would demonstrate a survival benefit. Nevertheless, they urged Roche and the FDA to include quality-of-life improvement measures in the confirmatory study that could potentially serve as the basis for full approval if a survival benefit is not demonstrated.
The Phase III, 920-patient global study will evaluate the efficacy and safety of bevacizumab in combination with radiotherapy and temozolomide for patients with newly diagnosed glioblastoma. Results are expected in 2014.
Expected to begin within the next several months, the study is currently designed with the co-primary endpoints of overall survival and progression-free survival. Although neurocognitive function is a secondary endpoint, it is not powered to the degree that would satisfy the committee, said Dr David Schenkein, Genentech's senior vice-president of clinical haematology and oncology.
"I think we've heard you. We'll discuss this with the sponsor," FDA director of oncology drug products Dr Richard Pazdur told the committee.
In a statement issued after the meeting, Dr Schenkein said: "We look forward to working with the FDA to potentially provide people with this devastating disease the first new treatment in more than a decade." The sBLA was submitted on October 31st and is undergoing a priority review, with a May 5th user fee date.
Glioblastoma affects about 10,000 people annually in the US. There are few approved treatments, and the FDA has not accepted objective response or progression-free survival as surrogate endpoints to support accelerated approval in this disease.
Oral lomustine and intravenous carmustine were approved in the 1970s for primary and metastic brain tumours based on tumour response. The carmustine wafer, Eisai's Gliadel, was approved in 1996 for recurrent glioblastoma based upon overall survival; an indication for first-line treatment of high-grade malignant glioma was added in 2003. Temozolomide (Schering-Plough's Temodar) was granted accelerated approval in 1999 on the basis of durable objective response in patients with refractory anaplastic astrocytoma, but it was not approved for refractory glioblastoma. Regular approval for first-line glioblastoma was granted in 2005 after a confirmatory trial showed a survival benefit.
Avastin is currently approved in first- and second-line metastatic colorectal cancer and first-line metastatic non-squamous, non-small cell lung cancer. It also holds accelerated approval for first-line metastatic breast cancer.
BRAIN study
Genentech sought accelerated approval based on the Phase II BRAIN (AVF3708g) study, in which 167 relapsed glioblastoma patients were randomised to either bevacizumab alone or in combination with irinotecan. For US regulatory purposes, the primary efficacy endpoint was objective response rate, as determined by an independent review facility, in the bevacizumab monotherapy arm. The secondary endpoint was duration of response.
The company presented the BRAIN data at the American Society for Clinical Oncology annual meeting last year. 28.2% of 85 patients in the bevacizumab monotherapy arm experienced an objective response, which was defined as a complete or partial response confirmed on two consecutive assessments at least four weeks apart. The median duration of response was 5.6 months, and progression-free survival at six months was 43%. The reported objective response and progression free survival rates were significantly higher than historical controls.
The FDA found slightly lower efficacy in its analysis: a 25.9% objective response, a 4.2-month median duration of response, and a 36% six-month progression free survival.
At the FDA's request, Genentech also submitted data from a US National Cancer Institute, single-arm study of bevacizumab in 56 patients with recurrent glioblastoma. The objective response rate was 19.6%, and the median duration of response was 3.9 months. No complete responses were seen in either study, the FDA said.
The incidence of bevacizumab-induced adverse events did not appear to be significantly increased in glioblastoma patients in the two studies. There were two deaths possibly related to treatment.
The FDA questioned the value of using objective response determined by standard MRI as a surrogate endpoint for survival in glioblastoma. Tumour measurement in this disease is difficult, particularly after prior surgery and radiation therapy, the agency pointed out. Bevacizumab's mechanism of action also can decrease the ability to detect an antitumour effect in imaging studies.
The monoclonal antibody neutralises VEGF-induced vascular permeability, which stabilises the blood-brain barrier. This results in decreased extravasation of fluid into brain parenchyma and improvements in oedema, which translates into a decrease in gadolinium enhancement in standard MRI scans. Bevacizumab has also been shown to reduce radiation necrosis. "It is unclear whether the radiographic improvement accompanied by decreased requirement in steroids reported in this application is the result of an antitumour effect of bevacizumab or represents radiographic improvement due to improvement in tumour necrosis and tumour associated brain oedema," the agency said in briefing documents.