What is Avastin?

Avastin is a new drug which interferes with tumor angiogenesis (blood vessel growth). Avastin interferes with an angiogenic process which is particularly strong in clear cell RCC, and has some modestly promising data in RCC. Because Avastin is FDA approved for another cancer, and is therefore available outside a clinical trial it is increasingly being prescribed for kidney cancer patients. There are also ongoing clinical trials involving Avastin.

Technically, "Avastin is a monoclonal antibody to Vascular Endothelial Growth Factor." Maybe you don't recognize that as English? Let's take it one step at a time, last things first:

• Vascular Endothelial Growth Factor: This is a natural substance secreted by your body's cells which encourages blood vessel growth. It is a central molecule in the growth of new blood vessels. The layer of cells that forms the inner lining of blood vessels is called "Endothelium".
  
  In more detail, VEGF is a signaling molecule which is secreted by cells which need a blood supply. Receptors for VEGF can be found on nearby blood vessels, and stimulation by VEGF causes new blood vessels to sprout and grow towards the source of the VEGF.
  
  One situation which causes cells to secrete VEGF is low oxygen which naturally indicates a need for a better blood supply. Tumors commonly secrete VEGF, but clear cell RCC secretes an unusual amount of VEGF. This is because most clear cell RCCs have lost the function of a gene called "VHL". VHL stands for Von Hippel-Lindau and refers to Von Hippel-Lindau syndrome, a genetic syndrome which comes with a strong disposition to develop clear cell RCC. Most people with clear cell RCC don't have VHL syndrome, but in most cases their cancer cells have still lost the ability to make functioning VHL protein.
  
  The VHL protein in effect acts like an oxygen sensor for the cell, and when VHL function is lost, the failure mode is such that the cell constantly thinks it is short of oxygen, even if it isn't. As a consequence, it secretes huge amounts of VEGF. This is why clear cell RCC tumors are so incredibly vascular. VEGF is a major driving force for RCC. This makes VEGF a particularly important target in clear cell RCC.
  
  
• Monoclonal Antibody: Antibodies are a class of special proteins produced by certain cells of your immune system which lock onto specific targets on invaders and foreign substances. We all have many different antibodies in our bloodstream which together can target a wide range of invaders. Scientists have figured out how to create an antibody to a specific substance of interest and produce just that one antibody in large amounts. This is called a monoclonal antibody. Avastin is a monocolonal antibody to VEGF which binds up VEGF and inactivates it.
  
  

Pedigree and AKAs or A Monoclonal Antibody By Any Other Name...

Avastin is made by Genetech, a large biotech pharmaceutical company. Commercialization in Europe is by Roche Pharmaceuticals.

The generic name is "bevacizumab" (I haven't ever dared to even attempt to pronounce that!). This term will work better than Avastin for searches of the technical medical literature. "Avastin" may work better for finding popular information on the web.

A name used during earlier phases of the research is "rhuMAb-VEGF". Informally, it's been called "Anti-VEGF" or "Anti-VEGF monoclonal antibody".

How Avastin is Given

Since Avastin is an antibody and antibodies are proteins, Avastin cannot be administered by mouth. The problem is that the body knows just exactly what to do with proteins that you eat. It digests them. That won't do. Therefore, like other protein drugs, Avastin must be administered by injection. In this case you want it in the circulation to travel to the site of the tumors so it is given intravenously (IV).

In the NCI kidney cancer trial (see just below), an infusion was given every two weeks, except that the treatment started with two infusions only a week apart.

Avastin and Kidney Cancer: The Evidence

There have been two significant trials of Avastin in kidney cancer. The first, from the National Cancer Institute tested Avastin by itself, the second trial tested Avastin in combination with another drug called Tarceva. I'll cover these two trials separately starting with the NCI trial.

The NCI Trial

The National Cancer Institute conducted a randomized placebo controlled trial of Avastin in patients with clear cell RCC who'd already tried IL-2 [Yang 2003]. This study is why I'm covering Avastin. The results show a real but very modest benefit for Avastin over placebo in the length of time before there is significant tumor growth (Time to progression).

The trial had three arms, placebo, low dose Avastin (3mg/kg body weight), and high dose (10 mg/kg). Patients in the placebo group were allowed to cross over to the low dose arm if their tumors grew. I understand supply of drug was an issue at the time the trial was conducted.

The trial measured time to progression - how long it takes for tumors to start growing again, but not designed to give a good measurement of survival due to the cross-over provision. You should be aware that in theory, a treatment which delays progression might not actually improve survival.

Key Results:

• Statistically significant but very modest improvement in median time to progression: The median time before there was significant tumor growth (progression) was 4.8 months in the high dose Avastin arm compared to 2.5 months in the placebo group - a difference of only about 9 weeks. At the same time, it was very unlikely this was due to chance ("p<0.001" - highly statistically significant). Although the difference in the median is small, the curves do show a clear separation throughout which if anything becomes more prominent after the median (50% point) is reached.
  
  

  Yang 2003 Figure 1A.
  Time To Progression: High Dose Avastin vs. Placebo

  
  
• All patients eventually did progress: The "survival" curve for TTP shows all patients progressing eventually in both the high dose and placebo arms (see below). It appears that almost all patients were followed long enough to see actual progression, so the data is fully mature and won't change hardly at all with further follow-up.
  
  
• A very few patients seemed to get significantly more benefit: The paper mentions two patients, one with a partial response, and one with minor shrinkage who had sustained their responses for the two year treatment period of the trial, and who both had progression after they stopped treatment. In both cases they had another shrinkage after treatment was restarted on a compassionate basis. This shows that the drug was actually keeping the tumors in check, and also shows that for at least a few patients, though this drug isn't a cure, it really can give a very meaningful respite. I believe one of these patients is actually a member of the KIDNEY-ONC E-Mail List and was doing well after three or four years.
  
  
• No observed improvement in survival: The survival curves for the various arms are so intertwined over most of their length as to be entirely indistinguishable. The end of the curve was slightly higher for the high dose group than for the low dose group, which was in turn higher than placebo, but I am sure the numbers are small enough that this isn't reliable. This issue is actually quite muddled because patients on placebo were allowed to "cross-over" to get low dose Avastin after their tumors progressed. Unfortunately, the authors don't report how many actually did cross-over so it's hard to get an idea of much cross-over might've obscured any survival benefit, but the trial clearly wasn't designed to measure survival. Therefore, that they observed no improvement doesn't really mean the treatment couldn't improve survival.
  
  
• Low rate of tumor shrinkage: There were no complete responses, but there were four partial responses, all in the high dose group, for an objective response rate of 10%. They don't report the percentage of patients with shrinkages less than 50%. It's my impression that other antiangiogenic drugs create minor shrinkages at a relatively high rate and my opinion that this is biologically significant. One prominent specialist told me that reporting these shrinkages is often not permitted by journal policy. I think this is a mistake.
  
  
• Generally slight side-effects: It appears that most patients on high dose Avastin had no or slight side effects (and side effects were much less in the low dose arm, which was also less effective). The most common side effects were high blood pressure and protein in the urine, which you don't normally actually feel. The protein in the urine could indicate the drug causes kidney damage which is a special concern in RCC since most of us only have one kidney, but no patient had impaired kidney function due to these side effects. The article says these effects very slowly got better after treatment was stopped, but that they couldn't document complete recovery. It is possible this treatment could cause high blood pressure long term, but that's something which can be treated. The other major side effect was "malaise" - feeling lousy - in about a third of high dose patients. There was an increased incidence of nose bleeds and blood in the urine as well - the treatment can apparently cause some weakness in small blood vessels, even normal ones. Bleeding from tumors as blood vessels break down as a result of treatment may be a risk. According to the paper one patient in an earlier study died from this, but in this study bleeding didn't cause any serious problems.

Summary: The very modest benefits seen in this trial are nonetheless very likely not due to chance. This treatment has some biological effect. It is definitely no cure as patients eventually do progress. This progression means the drug becomes ineffective. As far as I know the reason the drug eventually fails isn't known.

Avastin and Tarceva

Tarceva (Generic Name: erlotinib, Research Name: OSI-774) is another Genetech drug (originally developed by OSI Pharmaceuticals). It inhibits a totally different pathway in cancer cells called the Epidermal Growth Factor pathway. Epidermal Growth Factor signals cells to divide, and cancer cells including renal cancer overexpress the receptor for EGF. Tarceva is an inhibitor of the EGF receptor. Tarceva is an oral drug. Tarceva is an experimental drug currently not FDA approved except for use in clinical trials. I expect it may be approved some time in the not too distant future though based on results in lung cancer.

A phase II trial of these two drugs in 63 patients with exciting results was reported at the 2004 ASCO Meeting, [Hainsworth 2004].

This trial used same Avastin dose as in the high dose arm of the NCI Trial excepting that a higher initial loading dose was omitted in this trial and the full dose of Tarceva as used in other trials (150mg/day) was used here.

Key Results Include

• 21% Partial Response Rate (No complete responses)
• 21% Minor Response Rate
• 45% Stable Disease
• 13% Continued Progression

The median time to progression (significant tumor growth) was a year, compared to about 5 months in the NCI trial of Avastin alone. Also, most of the partial and minor responses and even stable disease cases were still holding though follow-up was still fairly short.

Side Effects: Rash and diarrhea were the most common side effects with nausea and vomiting and bleeding (mostly minor) also common. Rash and diarrhea are common side effects with EGF pathway inhibitors. Bleeding has been noted as a side effect of Avastin. Only one patient quit the trial due to side effects and some patients have been on treatment for more than a year.

Conclusion: These results are clearly exciting, but it's actually not clear that the results are actually better than with Avastin alone. This may seem like an odd thing to say considering that NCI had a 10% response rate with a 5 month median time to progression compared to 21% and a year for this combination. The thing is the those NCI results are based on only 39 patients so there is considerable uncertainty in the response rate there. Perhaps NCI was a bit unlucky or (less likely with more patients) this trial was a little lucky. At one year, 80% of the patients in the NCI Avastin trial had progressed compared to only 50% here. I think this is a little less likely to be luck, but there is still uncertainty. One certainty is that adding Tarceva added additional side effects typical of EGF inhibitors although only one patient had to actually stop treatment due to side effects.

In addition to the statistical uncertainties, Tarceva alone has had questionable to marginal results in kidney cancer [Beeram 2004]. Another EGF inhibitor, Iressa, was a failure in kidney cancer, though it has been FDA approved for lung cancer on what I think are rather thin results. I also don't know of a specific biological rationale why inhibiting the VEGF pathway would make EGF inhibitors more active, though anything is certainly possible. At best it's unclear just why adding a marginally active drug to Avastin would dramatically improve the results. If it turns out that Avastin alone is more active than the NCI trial indicated - well that wouldn't be a bad thing either!

There is currently a randomized trial of Avastin + Tarceva vs Avastin alone which will clarify these issues but which comes with the usual dilemmas for patients who want only the best chance.

Future of Avastin in Kidney Cancer

Obviously this isn't a cure, although even now it offers some hope for a meaningful respite. I think biologically it will be important to figure out why the drug becomes ineffective. If this is understood better it might be possible to design a therapy which will be effective for longer.

Obviously continuing to combine Avastin with other drugs, including other drugs already used in antiangiogenesis therapy and other antiangiogenic drugs, is an important route for the future. Even drugs targeted at other aspects of cancer biology might be tried. Historically, the most effective drug treatments for cancer have been combinations. If you block several important pathways at once, it will become much harder for the cells to develop resistance since they have to develop resistance to both drugs at the same time. Barring a "common mode failure" where cells find a way to overcome both drugs with a single change, this becomes quite unlikely.

Already a clinical trial comparing Interferon to Interferon + Avastin in metastatic RCC is in progress, and I expect to see one comparing Avastin + Tarceva to Avastin alone in the near future.

Finally, I think it would also be worthwhile to see if Avastin, which can be tolerated for lengthy periods by most patients, could be used to prevent recurrence after nephrectomy (adjuvant therapy). Avastin might be able to prevent tiny tumors from ever getting enough of a blood supply to grow into detectable metastases.

A Few Questions, Comments, and Speculations

I have some real questions as to how this all works. It's very possible the answers are in the literature and I may get the answers to some of them as I continue to research this treatment.

How Does It Really Work?

As I understand it, receptors for VEGF that are responsible for triggering new blood vessel growth are on the outside of blood vessels and so the VEGF you want to inhibit is outside the vessels where it can bind to these receptors. Avastin is injected into the veins which gives it access to the inside of blood vessels. Antibodies are relatively large molecules which one would think could not easily get out of the blood vessels to get to the VEGF outside the vessel. So how does this work? I do understand that tumor blood vessels are immature, chaotic, and leaky. Maybe Avastin is able to get outside the blood vessels to get at VEGF just where it is needed by virtue of blood vessels in the tumor being leaky. I don't know.

I find it very interesting that an IV infusion every two weeks could be effective. This suggests to me that the antibody is very slowly released from the blood stream into the tumor, yet in large enough quantities to still be effective.

Biologic Verification of Anti-Vascular Effects

The clinical trials I've looked at don't make any direct attempt to show that Avastin actually is binding up enough VEGF in the environs of the tumor to be effective. Now this might not be easy to do, but I'd like to know if this has been studied elsewhere. The NCI Kidney Cancer Paperdoes talk about measuring VEGF and Avastin concentrations in blood, but it's not clear to me why this is actually relevant (this is absolutely not a criticism - it's a question!).

Tumor Shrinkage Versus "Stability" - A Strength Estimator?

At first glance, stopping blood vessel growth should stop tumor growth but not cause actual tumor shrinkage. It turns out though, that new blood vessels are "immature" and require continued VEGF to survive. I would expect then that effective inhibition of VEGF should cause some tumor shrinkage as the very new vessels disintegrate due to lack of VEGF. Older vessels in the tumor may be fully mature and insensitive to anti-angiogenic therapy, so complete response in larger tumors well established should be rare.

In the NCI study of Avastin alone, the rate of partial response (50% or more shrinkage) was 10% in the high dose group (which means only 4/39 patients - very small numbers). We don't know how many patients had minor but still measurable shrinkage because it isn't reported. I think while minor shrinkage isn't traditionally reported, for anti-angiogenic drugs this may be a particularly prominent part of the pattern for effective drugs. Data I have seen for BAY 43-9006, another drug with possible anti-angiogenic effects, shows a spectrum of shrinkage including the minor responses. I think is a significant part of the picture. It's especially useful when the amount shrinkage/growth is histogrammed rather than just placed into the standard crude categories; progressive, stable, partial response, and complete response. I would like to see this kind of data reported for anti-angiogenic treatments.

Mechanism of Failure

That almost all patients in the NCI trial had tumor growth means the drug (by itself) fails eventually. Understanding the mechanism of failure could allow better treatments to be developed. I can think of several possible failure mechanisms. Perhaps the drug is cleared from the system differently after a while and no longer reaches its target. Possibly the tumors produce an increased amount of VEGF in response to the drug, or perhaps they produce other angiogenic factors. The blood vessels themselves could modulate their sensitivity to various angiogenic factors as well.

Avastin Status

In late February 2004, Genetech received FDA approval to market Avastin for treatment of metastatic colon cancer on the basis of a randomized trial [Hurwitz 2003] showing a significant benefit from adding Avastin to standard chemo.

Now that Avastin is approved for colon cancer, doctors can legally prescribe it for any purpose, including RCC. So called off label use of drugs is very common in oncology and the results with kidney cancer will provide a good basis for it, but getting insurance to pay for it may well be a problem. Better insurance plans will pay for drugs which have real evidence for them, especially favorable data from a randomized trial, even if the use is off label.

I also understand that Avastin is being developed in Europe by Roche. I am not sure what the exact status is there.

If you look for Avastin trials at ClinicalTrials.gov you will see that there is an unprecedented number of trials in all kinds of cancers and situations - presumably triggered by the success in colon cancer. This represents an enormous investment in time, money, and patient hopes.

Finally to be complete, you should know that not all results with Avastin have been favorable. A large randomized trial of Avastin + Capecitabine versus Capecitabine alone for third line treatment of metastatic breast cancer did not show benefit; a major failure. The study was presented in abstract form at the 2002 San Antonio Breast Cancer Meeting but unusually, the abstract there doesn't actually... describe the results which I am reduced to getting from various press releases. Apparently the combination arm had a higher response rate but the responses where shorter on average and survival or overall time to progression wasn't better.

Resources

• My review of a clinical trial of Avastin + Interferon Versus Interferon in KidneyCancerTrials.org
  
  
• TargetVEGF is all about VEGF especially in cancer treatment. It's intended for health professionals, but has some great graphics and detailed technical information. By no coincidence whatever it's sponsored by Genentech, maker of Avastin.

References

Beeram M, Rowinsky EK, Weiss GR, Syed S, Mita A, Patnaik A, Mita M, Goldston M, De Bono JS, Tolcher AW

Durable disease stabilization and antitumor activity with OSI-774 in renal cell carcinoma: A phase II, pharmacokinetic (PK) and biological correlative study with FDG-PET imaging

Proc Am Soc Clin Oncol (ASCO Meeting Abstract) 2004 23 : Abstract 3050 [You can find this abstract on the ASCO website (I can't link directly because ASCO won't provide a stable URL)]

Comment: This is the only trial of Tarceva in RCC I've been able to find. I'm not terribly impressed. I think that while three of 30 patients had very disease stabilization for a little more than a year, and nearly a quarter had stable disease for six months, this is a little hard to interpret given the variable natural course of kidney cancer. The partial response rate of one in 30 does not add to my confidence. I would say if this drug is active at all, the activity is rather weak.

Hainsworth JD, Sosman JA, Spigel DR, Schwert RC, Carrell DL, Hubbard F, Greco FA

Phase II trial of bevacizumab and erlotinib in patients with metastatic renal carcinoma (RCC).

Proc Am Soc Clin Oncol (ASCO Meeting Abstract) 2004 23 : Abstract 4502 [You can find this abstract on the ASCO website (I can't link directly because ASCO won't provide a stable URL)]

Comment: Note that Dr. Hainsworth's presentation including slides is also available on the ASCO site and is slightly more up to date than the abstract.

Hurwitz H, Fehrenbacher L, Cartwright T, Hainsworth J, Heim W, Berlin J, Griffing S, Novotny W, Holmgren E, Kabbinavar F

Bevacizumab (a monoclonal antibody to vascular endothelial growth factor) prolongs survival in first-line colorectal cancer (CRC): Results of a phase III trial of bevacizumab in combination with bolus IFL (irinotecan, 5-fluorouracil, leucovorin) as first-line therapy in subjects with metastatic CRC.

Proc Am Soc Clin Oncol (ASCO Meeting Abstract) 2003 22 : Abstract 3646 [You can find this abstract on the ASCO website (I can't link directly because ASCO won't provide a stable URL)]

Comment: A slide presentation showing much more data than the abstract is also available - look for a link on the same page as the abstract.

This is the study on which Avastin was approved for colon cancer. Now that it is approved for colon cancer, it can legally be used for any purpose including treating RCC.

The dose here was 5 mg/kg which is half that of the high dose arm in the study in RCC, but more than in the low dose arm of that trial. Side effects due to Avastin may have been less than the potential in RCC due to the lower dose.

This randomized trial showed improvements in survival as well as time to progression when Avastin was added to standard chemo. Interestingly, the TTP curves (from the slide show) converge in the end suggesting that the treatment only delays progression as appears to be the case in RCC as well. This apparently does lead to an improvement in survival here. Whether it does or not for RCC remains to be seen.

Yang JC, Haworth L, Sherry RM, Hwu P, Schwartzentruber DJ, Topalian SL, Steinberg SM, Chen HX, Rosenberg SA.

A randomized trial of bevacizumab, an anti-vascular endothelial growth factor antibody, for metastatic renal cancer.

N Engl J Med. 2003 Jul 31;349(5):427-34. [PubMed Abstract (will open in new window)]

Comment: Free full text of this article is available from the New England Journal of Medicine, but you have to register first. Registration is free.