August 26, 2007
Anti-Angiogenesis Compounds Against Cancer Might Pose Heart Attack Risk
Angiogenesis inhibitors work against cancer by blocking the development of new blood vessels into tumors so that tumors can't grow. The hope with angiogenesis inhibitors is that they will have less side effects than conventional chemotherapy and radiation. Given that chemo and radiation cause severe side effects a treatment can be better and still have pretty substantial side effects. Angiogenesis inhibitors turn to out to be capable of causing lethal side effects.
Angiogenesis inhibitors, drugs that block a tumor's development of an independent blood supply, have been touted as effective cancer fighters that result in fewer side effects than traditional chemotherapy. However, a new study by researchers at UCLA's Jonsson Cancer Center has shown that one method of blocking blood-supply development could result in serious and potentially deadly side effects.
The UCLA researchers blocked angiogenesis factor VEGF in a type of cell in mice and found that the mice eventually suffered heart attacks as a result.
Several newly developed angiogenesis inhibitors work by blocking vascular endothelial growth factor (VEGF), an important signaling protein that spurs the growth of new blood vessels. Avastin, an angiogenesis inhibitor approved by the Food and Drug Administration for colon and lung cancers, inhibits angiogenesis by blocking VEGF signaling from outside the cell. UCLA researchers wanted to know what happened when VEGF signaling was blocked from within endothelial cells, a mechanism used by some small-molecule drugs currently being tested in late-phase clinical trials.
The result was unexpected and sobering. More than half of the mice in the study suffered heart attacks and fatal strokes, while those that remained alive developed serious systemic vascular illness, said Luisa Iruela-Arispe, a professor of molecular, cell and developmental biology and director of the Cancer Cell Biology program at UCLA's Jonsson Cancer Center.
The study appears Aug. 24 in the prestigious peer-reviewed journal Cell.
This result suggests humans might suffer more complications from the long term use of anti-angiogenesis compounds.
"This was an extremely surprising result," said Iruela-Arispe, past president of the North American Vascular Biology Organization and a national expert on angiogenesis. "I think this study is cause for some caution in the use of angiogenesis inhibitors in patients for very long periods of time and in particular for use of those inhibitors that block VEGF signaling from inside the cell."
Anti-angiogenesis compounds block the growth of cancers and therefore in many cases won't cure cancer. So people who take anti-angiogenesis compounds have to take them for many years. Hence long term side effects eventually become an issue. However, if you have an otherwise fatal cancer running a risk of blood clots or even a heart attack can be a price worth paying.
About 5 percent of patients taking Avastin develop blood clot-related side effects, Iruela-Arispe said. But because Avastin was approved only three years ago, it is unclear what side effects may occur when patients remain on the drug for many years, she said.
Anti-angiogenesis drugs side effects are acceptable only because those side effects are milder compared to even more undesirable alternatives such as dying sooner or experiencing even greater side effects from chemotherapy and still dying. We need anti-cancer treatments that are far milder while at the same time producing much higher cure rates.
To cure cancer the key piece we need is much more selective ways to deliver treatments aimed precisely at cancer cells while leaving normal cells alone. That's the huge challenge that cancer researchers have been wrestling with for decades. Anti-angiogenesis drugs currently are delivered into the entire body. So they suffer from the same problem that chemotherapy suffers from: They affect too large a range of cells. A more selective way to deliver anti-angiogenesis compounds might involve attaching them to monoclonal antibodies that are targeted to cancer cells. That would get anti-angiogenesis compounds to concentrate near cancer cells. But the anti-angiogenesis compounds need to work their effects against blood vessel cells and blood vessel stem cells that are near cancer cells. A fancier monoclonal antibody mechanism of delivery might activate and release an anti-angiogenesis drug only once the monoclonal antibody attaches to a cancer cell.
I had such high-hopes! I was hopeful that the differential growth patterns between regular tissues and cancer would allow this strategy to succeed long-term. Unfortunately, this is another of Dr. Bob's Laws coming home to roost: It is impossible to selectively turn-off general systems in an organism and not have dire consequences.
They got clobbered on the matrix metalloproteases.
They got nailed on COX-2.
They just got spanked on angiogenesis.
This is in accordance with another of Dr. Bob's Laws which is: Linear thinking about concentration-based computers [cells] and their aggregates [organisms] is a sure fire way to fail spectacularly.
One technologically available alternative here, would be to get a catheter up by the tumor -- hit it with an immense anti-angiogenesis load and then have another apheresis catheter in place with a filter to pull out the drug.
Burned once again by Dr. Bob's Laws. Bummer dude.
Catheter by a tumour and another to filter out the drug: Where would this work where surgery to remove the tumour wouldn't be possible? It would have to have not metastasized yet and still be inoperable.
We can't cure cancer without making treatments that only turn on in cancer cells or which only get attached the cancer cells. We need specificity.
The laws started out as musings by Dr. H. Teager during an independent study I had with the good doctor. So I guess I inherited the musings and finalized them. Dr. Teager unfortunately perished from metastazied lung cancer.
I concur with your thoughts Randall. There is a real need for specificity -- this is why I suspect we need to take a long hard look again at the immunotherapies. The immune system is the ONLY biologically compatible system we know of that can attack specific cells and destroy them. We need to coopt its mechanisms.
Additionally we need to personalize treatment. Take a sample of the cancer cells, find the biologic differences between those cells AND the PARTICULAR patient, and tailor accordingly.
And last but not least, we need the freedom to try. Right now between the funding, FDA and hospital IRB, locked into an 18th century model of medicine, we can't even evolve therapies with patients that have nothing to lose (and willing to try a Hail Mary pass at the end zone!).
What about gene therapies?
In theory (and I'll ignore the mammoth problem of delivery mechanism for the moment) one could write a piece of genetic code to put into a large number of cancerous and normal cells. In normal cells the code would start "executing", make some proteins, detect that the cell is normal, and do nothing more. But in cancerous cells the code would execute, detect aberration, and then do an assortment of things that would kill or fix the cell.
Writing such code would be difficult. Also, the code needed might be lengthy.
Now, I'll stop ignoring the delivery problem. First off, lots of people are chasing gene therapy with much less ambitious goals and failing left and right. The immune system objects. The genes actually cause cancer. You are probably more familiar with the research literature than I am.
Plus, the ratio of normal cells to cancerous cells is enormous. 1 million to 1? Higher? If one had no specificity in terms of which cells the gene therapy vector would need to enter then the amount of gene therapy that would need to be delivered would be enormous.
Also, when delivering such huge quantities of gene therapy genes we'll get many copies in some cells and no copies in other cells. That's bad.
If we could develop some specificity of delivery then the picture brightens to whatever extent we can achieve specificity. But we still have the problem of needing to get into almost all the cancer cells (some cancer cells are too sick to divide or will die when not surrounded by other cancer cells).
If one can achieve perfect specificity of delivery then one doesn't need to deliver gene therapy. If one can only hit cancer cells then one can deliver a payload that will kill whichever cells one hits. Even if one has a hit ratio of 1 cancer cell to 9 normal cells it would be okay in some cases to kill all the cells one hits - as long as one hits 100% of the cancer cells. Some collateral damage is acceptable - particularly once we get cell therapies that can repair damage.
We need the greatest degree of specificity for brain cancers. Don't want to burn out our brains trying to hit cancer cells.
Immuno techniques: Can we achieve the needed specificity based on what cancer cells look like on the outside?
Agreed that Judah Folkman deserves a Nobel. I do not understand why he hasn't gotten it yet.
Yeah -- why didn't he get a Nobel yet? While I understand his goal was cancer eradication via this, getting to the core of the process of angiogenesis is a breathtaking feat all on its own. As an applied researcher I love when fundamental researchers open up new vistas like this. Gives me something new to try to fit in all the day-to-day puzzles.
Gene therapies: We have some interesting advantages given to us by local immunosuppression. If I inject a pathogen into a rat with cancer the pathogen is rapidly cleared from the body except at the site of tumor. This effect can be used to deliver the gene therapy only where we want it with relatively high specificity. I would have to chat with some UCSD and Northwestern folk about exactly how high we could get it, but back-of-envelope I think we could get better than million-to-1 in our favor. This doesn't solve the delivery problem but it gets us from mammoth-level to horse-level.
Gene code: If we could get the specificity up high enough so that we can get inside the tumor in its totality, I think we could insert any of a dozen apoptotic genes that can drive sensitivity to such cytokines as TRAIL, for example. So infect, allow integration time, pour in the TRAIL, and nail the tumor. Hopefully, the reasoning of these last few paragraphs works and the patient doesn't dissolve into a pile liquefying apoptotic goo.
An idealized therapy along these lines could also be constructed to infect the tumor with a highly visible MHC-class antigen to expose on the cells surface that would make the tumor exceptionally foreign. Then you use apheresis to remove soluble inhibitors of the immune system exposing the tumor to attack. Repeat until eradication. Devil in the details of course, but I see no show stoppers. Nothing a $50 million effort couldn't run to the ground. The basic research would go for about $5 million to prove each independent piece.
Alternatively, one could go for 1st-degree relative cross-transfusions of immune cells that are loosely compatible with the host. The resulting donor-vs-host flavor of graft-vs-host disease will be relatively stronger against the tumor. The Israeli's have done work in this area with surprising results. Enhancing the visibility of the tumor would drive these therapies into high gear.
Any cancer drug can cause potential heart damage, even death, and many doctors do not adequately monitor their patients or manage their care to minimize the health risk. It's hard to tell a doctor (and patient) to ratchet back on the anti-VEGF drug they're using when the disease setting is stage IV lung, ovarian or pancreatic cancer.
Patients and doctors may not be aware of the spectrum of heart problems that can arise from cancer treatment. Heart problems can occur during treatment or months and even years after treatment. So yes, even the newest "targeted" therapies, designed to attack only cancer cells or microvascular cells, can cause cardiotoxicity.
Some clinical scientists have found a profile of cardiotoxicity for the most often anticancer drugs, but it is important to know that every patient has different risk factors that will determine how their hearts handle the treatment. Monitoring and mangement is key to surviving cancer with a good and lasting heart.
In "curable" diseases, therapy-related, late onset sequelae are becoming a real problem. Many of these new "targeted" therapies often get a pass on toxicities because they are just so darn cool (Herceptin and CHF in the adjuvant setting is another example). This is perhaps why Herceptin causes heart problems. Among other things, it's an antiangiogenic. It's all risk/benefit ratio.
The problem is that few drugs work the way oncologists think and few of them take the time to think through what it is they are using them for.