Originally published Aug. 23.
By Turna Ray
After an independent panel of experts recommended last month that the US Food and Drug Administration revoke approval of Avastin for use in treating metastatic breast cancer due to its unfavorable risk/benefit ratio, patient advocacy organizations called for the use of personalized medicine strategies to keep the drug on the market for those who are most likely to benefit from treatment.
In a joint letter to the FDA and key legislators in the US Congress earlier this month, Susan G. Komen for the Cure and the Ovarian Cancer National Alliance urged for more research to identify biomarkers that will allow doctors to pinpoint which patients will respond to the drug with limited side effects.
"We recognize the benefits of Avastin overall are modest for women with metastatic breast cancer," said Nancy Brinker, Komen's founder and CEO, in the letter to the FDA. "However, we do know that for some women, Avastin offers a greater than modest benefit. We hope that this decision will not restrict access to Avastin to all patients." Komen for the Cure has awarded $60 million in grants to fund breast cancer research, and at least one of these projects is looking at guiding Avastin treatment with pharmacogenomic data.
The letter follows a 12-1 decision from the FDA's Oncology Drugs Advisory Committee July 20, recommending that the FDA should rescind its approval of Avastin in the metastatic breast cancer setting. The recommendation was based on results from two studies submitted by Genentech that showed Avastin did not improve overall survival and also increased life-threatening toxicity.
ODAC's recommendation has elicited criticism from patient advocates who argue for giving patients access to as many treatments as possible. A revocation of FDA approval would still allow doctors to prescribe the drug off-label in breast cancer patients, but insurers are unlikely to cover the treatment — estimated to cost around $8,000 per month — for a non-FDA approved indication.
The topic of pharmacogenomics has not figured prominently among those who oppose ODAC's recommendation, however, even though researchers identified PGx markers in the very trial Genentech used to garner accelerated approval for Avastin in breast cancer two years ago.
Avastin, approved by the FDA in 2008 for metastatic breast cancer, is a monoclonal antibody that inhibits vascular endothelial growth factor activity. Within the pivotal study, called E2100, researchers identified gene markers that were associated with longer survival while on the drug and appeared to protect patients against drug-induced hypertension.
In reviewing the two Avastin trials submitted by Genentech, advisory committee members expressed concern about the fact that overall survival appeared to favor the placebo arm. Additionally, despite the drug's improvement of progression-free survivial, the advisory committee didn't like the fact that the addition of Avastin increased the rate of grade 3-5 adverse events in both studies
During the July meeting, although FDA and ODAC members enumerated many concerns with the design of E2100 and the two studies considered by the advisory committee, pharmacogenomics didn't factor into the public discussion. Furthermore, there is little indication from Genentech that it is considering a PGx strategy as a means of keeping the Avastin on the market in a subpopulation of patients.
A Genentech spokesperson told Pharmacogenomics Reporter that the company did not submit PGx data to the advisory committee. However, "we're looking at biomarkers for angiogenesis as part of our independent work as well as our work with academic centers," the spokesperson said. "As part of that work we're continuing to evaluate findings in E2100 and other clinical trials … [and] we're continuing discussion with the FDA about the use of Avastin in breast cancer."
[ pagebreak ]
The spokesperson could not provide any specific information about PGx work being done related to Avastin. According to the company representative, the topic of personalized medicine may have been brought up briefly during the advisory committee meeting, but only in passing. PGx Reporter could not find any mention of the topic in a search of the meeting transcript.
E2100, in addition to showing a 52 percent increase in progression-free survival in the Avastin/paclitaxel arm compared to paclitaxel alone, also reported preliminary PGx data linking VEGF-1154 AA and -2578 AA genotypes to an improvement in median overall survival. Additionally, E2100 showed that patients with VEGF-634 CC and -1498 TT genotypes had protection from grade 3-4 hypertension, a common side effect of Avastin.
Bryan Schneider, Indiana University assistant professor and lead author of E2100, was quoted last year in an article in Clinical Cancer Research, saying "if validated, these [PGx] findings could help direct which subgroup of patients should receive bevacizumab."
Although Genentech has not indicated that it is internally advancing the PGx leads from E2100, Schneider appears to be doing this on his own in a Phase III study called E5103. With a $5.8 million grant from Komen for the Cure, Schneider is studying whether certain genetic markers will improve disease-free and overall survival for women being treated with Avastin and standard chemotherapy.
Furthermore, just because Genentech isn't talking about PGx doesn't necessarily mean that the company isn't doing work quietly in this area.
George Sledge, another Indiana University professor and a co-investigator in E2100, doesn't feel that the preliminary study gave Genentech sufficient data on the biomarkers to change the labeling for Avastin. "But the analysis we did is good science," Sledge told PGx Reporter in June. "It is hypothesis generating but needs to be validated."
To Sledge's knowledge Genentech has not elucidated plans to conduct confirmatory PGx trials for Avastin internally based on the results of E2100. "But Genentech is interested in PGx and could very well be doing internal investigations," he added.
Sledge attempted to lead a biomarker-driven study for Avastin but the study was halted due to lack of patient accrual. The trial — called "Predicting Response and Toxicity in Patients Receiving Paclitaxel and Avastin for Breast Cancer: A Multicenter Genomic, Proteomic and Pharmacogenomic Correlative Study" — began in 2007 but had to be terminated in 2009.
The 36-month study was being conducted by researchers at the Hoosier Oncology Group, the Department of Defense, Indiana University and others, and would have attempted to correlate tumor gene expression with response to paclitaxel and Avastin in patients with advanced breast cancer. Additionally, the study would have attempted to link serum and tumor proteomic profiles with response, compare genomic and proteomic profiles of patients, as well as correlate toxicity and response with drug-related PGx targets.
"To date no other trials have analyzed gene and protein expression at the same time points in the same patient, combined with clinical outcome," researchers wrote in the summary description of the study in Clinicaltrials.gov. "Similar to previous attempts to predict response based on expression of a single gene or protein, we expect that neither genomic or proteomic profiling alone will be sufficient to optimize therapy," they hypothesized, adding that they expected "an iterative process that combines information gleaned from both platforms, modified to avoid toxicity based on pharmacogenomics."
Although FDA isn't bound to the recommendations of its advisory committees, the agency usually does follow the guidance of its expert panels. One of the few times FDA didn't fall in line with advisory committee votes was when it approved Avastin as a treatment for metastatic breast cancer in 2008, despite a 5-4 ODAC vote in favor of denying marketing approval for the drug.
[ pagebreak ]
The FDA is slated to decide whether to take Avastin off the market on Sept. 17. If the agency chooses to revoke its approval for the drug in the metastatic breast cancer market, Avastin will still remain available for other FDA-approved indications in colorectal cancer, lung cancer, kidney cancer, and glioblastoma.
Although the agency has not given any indication in public discussions, the FDA may be internally mulling whether pharmacogenomics might be used to reintroduce Avastin in a subpopulation of metastatic breast cancer patients. If the FDA is factoring in PGx leads for Avastin, the agency could deal with Avastin as it did with AstraZeneca's Iressa. In 2004, FDA decided to limit marketing of Iressa after it failed to show a survival advantage in clinical trials. The agency had previously granted accelerated approval to the drug based on a 10 percent tumor shrinkage seen in NSCLC patients receiving Iressa in addition to standard treatments.
However, despite limited survival data, instead of taking the drug off the market as a treatment for NSCLC, FDA placed Iressa in a restricted-access program that meant that only patients already on the drug and benefiting from it could maintain access to it. No new patients could receive the drug, however.
In Europe, where AstraZeneca withdrew its marketing application for Iressa in 2005 after lackluster study results, the company has recently used pharmacogenomics to launch the drug in patients who test positive for epidermal growth factor receptor-tyrosine kinase mutations.
Additionally, after reviewing the company's PGx data, the UK's National Institute for Health and Clinical Excellence recommended in May that the National Health Service should pay for Iressa in the first-line treatment setting if patients test positive for EGFR mutations and if the drug developer provided the drug at a fixed price (PGx Reporter 06/02/10).
AstraZeneca has not announced any plans to introduce Iressa with PGx data in the US. But since the company has already launched the drug in a genetically targeted population in Europe, the US market may be next pending additional studies.
According to ClinicalTrials.gov, in May 2008, AstraZeneca completed a Phase I pharmacokinetic and pharmacodynamic study of Iressa in combination with Merck's Erbitux in patients with advanced cancers, including NSCLC patients. In the study, researchers attempted to correlate the pharmacogenomic profiles of study patients with the efficacy of the combination treatment.
Additionally, the National Cancer Institute of Canada and the US Intergroup are working on a PGx correlative study of post-operative treatment with Iressa compared to placebo in NSCLC patients. This study, which has not yet begun recruiting, could potentially expand on the PGx findings from a Phase III study (BR.19), which investigated outcomes in early-stage NSCLC patients randomized to Iressa or placebo after complete tumor resection. That trial, funded by AstraZeneca, concluded that adjuvant treatment with Iressa "did not confer disease-free survival or [an] overall survival advantage in the overall population," and "KRAS and EGFR copy were neither prognostic nor predictive" of drug benefit.
AstraZeneca's experience with Iressa could provide a model for Roche/Genentech's Avastin in terms of using PGx to keep the drug on the market in a subpopulation of patients.
According to Roche's pipeline, Avastin is being studied in two Phase III clinical trials for the adjuvant treatment of HER 2-positive and HER 2-negative breast cancer. The company expects to submit a marketing application for the HER 2-positive indication in 2013, and plans to file for the HER 2-negative indication after 2013. Additionally, Roche is investigating Avastin in combination with Herceptin as a first-line combination treatment in metastatic breast cancer. For this indication, the expected filing date is next year.