About this Author
DBL%20Hendrix%20small.png College chemistry, 1983

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png After 10 years of blogging. . .

Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases. To contact Derek email him directly: Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
Not Voodoo

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
A New Merck, Reviewed
Liberal Arts Chemistry
Electron Pusher
All Things Metathesis
C&E News Blogs
Chemiotics II
Chemical Space
Noel O'Blog
In Vivo Blog
Terra Sigilatta
BBSRC/Douglas Kell
Realizations in Biostatistics
ChemSpider Blog
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Eye on FDA
Chemical Forums
Symyx Blog
Sceptical Chymist
Lamentations on Chemistry
Computational Organic Chemistry
Mining Drugs
Henry Rzepa

Science Blogs and News:
Bad Science
The Loom
Uncertain Principles
Fierce Biotech
Blogs for Industry
Omics! Omics!
Young Female Scientist
Notional Slurry
Nobel Intent
SciTech Daily
Science Blog
Gene Expression (I)
Gene Expression (II)
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net

Medical Blogs
DB's Medical Rants
Science-Based Medicine
Respectful Insolence
Diabetes Mine

Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem

Politics / Current Events
Virginia Postrel
Belmont Club
Mickey Kaus

Belles Lettres
Uncouth Reflections
Arts and Letters Daily
In the Pipeline: Don't miss Derek Lowe's excellent commentary on drug discovery and the pharma industry in general at In the Pipeline

In the Pipeline

« Bapineuzumab Still Does Not Work Against Alzheimer's | Main | Closure at the Neurosciences Institute? »

August 8, 2012

Does Aveo's Tivozanib Work, or Not?

Email This Entry

Posted by Derek

What makes a cancer drug effective? What if it stops cancer from spreading when you give it to patients - is that effective, or not? This topic has come up around here before, but there may be a rather stark example of it unfolding with Aveo Pharmaceuticals and their drug tivozanib.

Earlier this year, the company announced results of a trial in renal cell carcinoma of their drug versus the Bayer/Onyx drug Nexavar (sorafenib), which is the standard of care. It's not like Nexavar does a great job in that indication, though - when it was going through clinical trials, it ran in RCC patients versus placebo, since - you guessed it - placebo was the standard of care at the time. And while Nexavar did show a benefit under those conditions, there are still plenty of patients that don't respond. Thus tivozanib, and its window of opportunity. The compound itself is in the same broad chemical class (bi-aryl ureas) as sorafenib.

The Phase III results for the Aveo drug showed an improvement in progression-free survival - tracking the time it takes for the cancer to start spreading again. But progression-free survival does not necessarily mean "survival", not in the sense that cancer patients and their relatives really care about. Dying in the same amount of time, albeit with redistributed tumor tissue, is not the endpoint that people are waiting for.

The company is, of course, monitoring the patients that it's treated. And there's the problem: the current data show, after one year, that 77% of the tivozanib-treated patients are still alive. But 81% of the sorafenib patients have survived, and the FDA has officially expressed concern about the way things are going. That sent Aveo's stock down sharply the other day, as well it might. But there could be a way out:

Aveo said in today’s statement that basically it’s possible the preliminary survival data could be misleading. That’s because in cancer trials like this one, cancer patients whose disease worsens on one drug can then go on to get a second drug which may help them. In this case, Aveo said 53 percent of the patients who were randomly assigned to get the Bayer/Onyx drug went on to get subsequent therapy after their disease worsened—and “nearly all” of them were given Aveo’s tivozanib. By contrast, only 17 percent of the patients who were randomly assigned to initially get the Aveo drug went on to get a subsequent therapy. So it’s possible that the patients in the Bayer/Onyx control group may be ending up living longer at least partly because of the Aveo drug they got later on.

We'll have to wait for more data to sort all this out. Until that point, Aveo (and its shareholders) are probably in for a bumpy ride. But it's worth remembering that renal cell carcinoma patients are having a rather harder time of it than anyone else in this story, and they're the people who will be watching this most closely of all. . .

Comments (13) + TrackBacks (0) | Category: Cancer | Clinical Trials


1. johnnyboy on August 8, 2012 9:43 AM writes...

At the risk of being a repetitive ranter, i'll have to say it again: we will only make significant strides against cancer when as much effort and resources are put into early detection than are currently invested in developing drugs for refractory, metastatic tumors. The fact that a few months of progression-free survival is considered worthy of spending billions in drug development is not a victory in the battle against cancer, it's an exercise in futility. Where are the labs and companies trying to develop urinary markers for early detection of renal cell carcinoma, for instance ?

Permalink to Comment

2. NP1986 on August 8, 2012 10:49 AM writes...

The tivozanib trial wasn't designed to demonstrate a difference in overall survival. Patients in the sorafenib arm were able to cross-over and receive tivozanib after progression, and thus this is likely to confound the overall survival results.

Permalink to Comment

3. Anon on August 8, 2012 12:59 PM writes...

What are your thoughts on Depinho (new president of MDA) using a similar model for the Institute for Applied Cancer Science in Houston?

Permalink to Comment

4. NoDrugsNoJobs on August 8, 2012 1:08 PM writes...

JohnnyBoy - Its hard to market and make big bucks on a test. The US Supreme Ct, in its great wisdom and much to the concurrence of many on this board, felt it was inappropriate to patent the taking of a measured relationship where the relationship occurs in the body and making that the primary subject of a claimed invention. As a result, the claims must be narrowly tailored to a specific method of doing the analysis and of course, the more specific claims that would meet the Court's muster are also easier to design around. So for example, if you discovered that a particular marker was upregulated in the afflicted person's body, you could not patent the marker nor a diagnostic that hinged on looking at the marker and making the correlation. You might patent a specific assay to the marker but since there are many design arounds, you might never recoup your huge investment nor capitalize on the true scope of your discovery. All because the marker is produced as an act of nature in the body...(see the Prometheus case). While many medchemists felt this was common sense and couldn't understand the uproar, many diagnostic researchers felt it as a true body blow. It is a problem when a generalist court (USct) gets involved in highly technical issues. It is also a problem when we as individuals attempt common sense approaches to very complex areas outside our own understanding. The simpler the answer seems, the more we ought scrutinize it!

Permalink to Comment

5. Another Kevin on August 8, 2012 2:07 PM writes...

The problem with "early detection" is that it often doesn't work. Consider the history of PSA screening in men: its real effect seems to be detecting indolent tumours that the affected men would have otherwise carried to their graves without knowing they were there. These patients may even have their lives shortened by aggressive therapy against disease that would not have progressed.

I also wouldn't be quick to claim that "progression free survival" is always the wrong endpoint to measure. If the patient's survival time is not improved, but the patient can be maintained symptom-free for a greater portion of the remaining time, that's at least a small victory. It's a tragically tiny one, but it's something.

[Disclaimer: I work in support of diagnostic development.]

Permalink to Comment

6. johnnyboy on August 8, 2012 5:24 PM writes...

@4 - point taken, though there seems to be enough of an incentive for certain companies to invest in dx tests - ie Exact Sciences, which is running Ph3 studies of a colorectal cancer test.

@5 - well, one could also say "The problem with cancer drugs is that often they don't work" - which hasn't seemed to be an impediment on billions spent in their development. PSA is just one test for one tumor site, generalizing its limits to all possible dx markers seems a bit reductionist.

Permalink to Comment

7. NP1986 on August 8, 2012 5:58 PM writes...


For rarer cancers like kidney cancer, I wonder whether the economics of screening would be favorable. Even if you limit testing to older individuals, you would probably have to test over a 1,000 people to detect a single case of kidney cancer assuming 100% sensitivity of the test.

Permalink to Comment

8. barry on August 8, 2012 6:08 PM writes...

it is implicit in the "multifactorial etiology of cancer" that cancers are heterogeneous.
the cells (or descendents of the cells) that had acquired only some of the five or six mutations needed to become full blown monsters will usually be present along with those which have achieved immortality and invasiveness and independence from growth factors and resistance to death factors. By the time "cancer" is diagnosed and all those hallmarks are checked off, the likelihood of any single drug curing the whole disease is slim. We need early detection not only of cancer, but of the precursor states, that can be addressed with single drugs. And we need those precursor states to be recognized as diseases so that insurers will pay for their treatment.
But all of this is moot if industrial Drug Discovery efforts continue on the current slope to oblivion.

Permalink to Comment

9. Andy on August 8, 2012 6:32 PM writes...

This issue of patient crossover confounding OS data in trials designed to measure PFS endpoints, especially in early lines of therapy, is par for the course in oncology drug development. This is why it is so crucial to determine what your regulatory endpoint is before you start the trial, and then stick to that. If they want to use OS as a secondary endpoint (as they do in the TIVO-1 trial) they should explicitly exclude patients that crossover from one arm to the other, or explicitly design the study to allow (and take into account) such crossover patients.

Permalink to Comment

10. anonymous on August 8, 2012 7:37 PM writes...

As someone who watched a loved one die an even more rapid (and more unpleasant???) death from fruitless (and unbelievably expensive) RCC "therapy", I can tell you that a drug that doesn't ultimately extend "lifespan" but significantly improves the quality of the afflicted's "remaining lifespan" would be a priceless "gift". But then, who's thinking about this....

Permalink to Comment

11. Andy on August 8, 2012 8:55 PM writes...

@#10: We have a HOPE (health outcomes and pharmacoeconomics) group that cares about quality of life issues and related. We'd think even more about this if the FDA were to agree to make improved quality of life without increased PFS or OS a registration endpoint for our oncology drugs. Currently this is a very tough regulatory slog.

Permalink to Comment

12. bad_deal on August 8, 2012 11:00 PM writes...

"By contrast, only 17 percent of the patients who were randomly assigned to initially get the Aveo drug went on to get a subsequent therapy."

Their explanation could be partially plausible, if Bayer/Onyx drug fails quickly, and so those who do not respond get a chance to try something else. Of course, this also means that you'd never try Aveo drug first, since it may keep you feeling OK longer than you'd want, so that when it fails, you are pretty much done and cannot try anything else.

So then their market share would always include only ~53% of patients who took Onyx drug first, and you'd always want to take Onyx first, if the trial results are to be believed.

The most positive spin is still fairly negative, not a total failure, but not where they'd want to be that is for sure.

Permalink to Comment

13. NP1986 on August 9, 2012 9:32 AM writes...


Aveo's explanation for the low utilization of second line therapy in the tivozanib arm is that most patients were enrolled in Eastern Europe and other regions where access to other therapies is limited. In the US, patients would have access to other TKIs as well as Afinitor for second line therapy.

Also, your conclusion that tivozanib would work better as second line therapy does not follow from the data.

Permalink to Comment


Remember Me?


Email this entry to:

Your email address:

Message (optional):

The Last Post
The GSK Layoffs Continue, By Proxy
The Move is Nigh
Another Alzheimer's IPO
Cutbacks at C&E News
Sanofi Pays to Get Back Into Oncology
An Irresponsible Statement About Curing Cancer
Oliver Sacks on Turning Back to Chemistry