A new report was published on-line this afternoon by the Journal of Clinical Oncology (JCO). It covers a Phase III (randomized) clinical trial of Avastin (Bevacizumab) in women with metastatic BC. Over 1200 patients were included in the analysis, all with Her2 negative disease.
The design of the randomized study protocol was a bit unusual, in that the treating physicians could choose among a few, standard chemo options to give their patients – the so-called “backbone” for treatment for each cohort in the trial, along with hormonal treatment and the study drug: Avastin or a placebo. Avastin is a monoclonal antibody that binds to the vascular endothelial growth factor (VEGF). It’s manufactured by Roche and is quite costly.
What the investigators report, now, is that women who received Avastin and any of the chemo regimens did better – in terms of what’s called progression free survival – than did those who received the same chemo and placebo treatment. The difference was a matter of a few months, on average, and there were no measurable change in overall survival. What this means is that in some women with metastatic breast cancer, Avastin appears to help keep the disease in check.
The study is called RIBBON-1, which I learned this evening would be for the first study of Regimens in Bevacizumab for Breast Oncology. Sounds lame, I know, but believe me — it’s hard for oncologists to keep trials straight without acronyms. Even with the acronyms.
It happens I know some women with triple negative BC who have benefited from Avastin. These women may be outliers on the curve, but they are real and they exist and I know them personally, in what should be the middle of their lives.
Maybe we, and the FDA, shouldn’t give up so fast on Avastin.
—
It looks like the publication has confirmed what we already know from previous conference presentations — ie PFS but not OS is improved.
Here’s the problem — Avastin was approved through the fast track mechanism based on phase II data, on condition that the primary endpoint, overall survival (ie patients live longer) is significantly improved in the phase III trials.
In neither the RIBBON1, not the AVADO trials, was there an improvement in OS. By definition, it didn’t meant the full approval criteria as the FDA pointed out in their decision to recommend withdrawal in breast cancer.
The challenge is that it’s easy to look at individual patients subjectively and say well they benefitted, but in totality, the data tells us objectively that the downside is that many did not, the overall benefit was small and the side effects are not minimal. Is it therefore reasonable to expose large numbers of women to the systemic side effects for a few who might possibly benefit when the majority did not?
A better way to look at this would be to have a biomarker to preselect patients who might benefit most from the therapy, but we don’t. Until that time, I can totally understand the FDA’s position on this issue given the regulatory framework we have in place.
Hi Sally, I appreciate your input on this.
I agree that a biomarker to identify likely responders would be helpful, but unfortunately — and by contrast to treatments that work by blocking a particular oncogene turned on in some cancer cases, for example — there is none for Avastin. If there are, say, 4 percent of metastatic BC patients who respond well to this drug, they should have the chance to get the drug and stay alive. The number of women living metastatic BC in the U.S. is estimated to be between 160,000 — 200,000.
The problem is that oncologists sometimes give/push drugs that aren’t working, for various reasons including hope and income. Hope is tricky and not necessarily something we’d want to “solve.” But the income/conflict-of-interest problem could be fixed by having doctors’ salaries disconnected from the number and types of treatments they prescribe.
If oncologists had nothing to gain by giving Avastin (or any other cancer drug) to their patients, they’d be far more likely to recommend continued prescription only in cases where it’s clearly helping more than it’s hurting.
Hi Elaine,
Totally agree with your comments re: hope and income, I wish it wasn’t so conflicted.
The issue I have with say 4% (or whatever the number who benefit is), we need better ways to figure out who should get it upfront do different subtypes benefit eg triple negative as you mention, but that has not been proven in a large randomised phase III trial that I know of. We can only go on the RIBBON1 and AVADO trial data, which do not meet the agreed criteria for approval. Why should Avastin be a special case?
Do we really need to treat everyone with breast cancer so that 4% can benefit? Is that truly ethical or responsible? We should remember that if the median benefit is say 2 months, then 50% will do better, but 50% will do worse and therein lies the fundamental crux of medicine — first do no harm.
I don’t think Avastin should be a special case. Doctors give drugs all the time for unapproved uses, and most cancer drugs only work in a fraction of patients with any given disease.
What some commentators don’t appreciate is that for a small percent of BC patients the difference with Avastin may be a matter of years. For those individuals, the drug is extremely worthwhile and should be available. The problem remains as to how to identify those patients. Since there’s no biomarker for Avastin responsiveness, we have to rely on doctors’/patients’ judgment.