The Limitations of VEGF Inhibition in Colorectal Cancer

Leonard Saltz, MD

The idea of using antiangiogenic therapy to fight cancer was first proposed in the early 1970s by Judah Folkman, MD, a professor at Harvard and director of the vascular biology program at Boston Children’s Hospital. Folkman, who died in 2004, believed that by creating an agent that attacked the blood supply instead of the tumor itself, cancer cells would be deprived of the nutrients they needed to grow and thereby unable to sustain themselves.

Folkman’s theory eventually led to the approval of VEGF inhibitors bevacizumab (Avastin), ziv-aflibercept (Zaltrap), and ramucirumab (Cyramza) in colorectal cancer.

However, these drugs are not true antiangiogenic therapies, and because of that they can only advance the treatment of colorectal cancer so far, says Leonard Saltz, MD, chief of gastrointestinal oncology and head of colorectal oncology at Memorial Sloan Kettering.

“My politically incorrect opinion is that we haven’t come out with any real antiangiogenic therapies,” says Saltz. “We need to be honest with ourselves and each other in admitting what these drugs are and what they are not. They are small steps forward, much smaller than any of us wanted them to be, and we need to keep that in perspective, especially when we look at the cost of them, which is very significant.”

OncLive: Why don’t you consider VEGF inhibitors like bevacizumab, aflibercept, and ramucirumab to be true antiangiogenic agents?

Below, Saltz explains his definition of an antiangiogenic agent, how VEGF inhibitors can best be optimized in colorectal cancer, and what their limitations are, via a series of questions and answers.Dr Saltz: The work of Folkman was based on the idea that we would no longer be attacking the tumor, but rather the blood supply. It was thought that by now we’d look back with disbelief that we used to give people agents that had toxicities like nausea, neutropenia, hair loss, and so on. It has not worked out that way. What we clearly have is agents that are anti-VEGF. If something was truly antiangiogenic, we would expect that it would work as a single agent—it wouldn’t need to be given with cytotoxics, and it might even interfere with cytotoxics by blocking the blood supply in order to decrease delivery of the drug to the tumor.

Are true antiangiogenic agents possible?

Are there VEGF inhibitors that you feel have shown the most patient benefit?

What we see with VEGF inhibitors is just the opposite. They don’t work as single agents, and they modestly enhance the effectiveness of cytotoxic regimens. We don’t spare patients the toxicity of those regimens, we just add the toxicities of the VEGF inhibitors on top of it. What happens when you give an anti-VEGF agent with chemotherapy is that it appears to improve the delivery of that chemotherapy to the tumor. That is its major benefit. I think that antiangiogenic is a somewhat wishful marketing term, and it is more accurate to define these drugs as anti-VEGF.I don’t believe we have succeeded in creating clinically available antiangiogenic agents, certainty not in the context in which the word was initially used in the discussion of the development of these drugs. No one ever set out to develop bevacizumab, aflibercept, ramucirumab, or any of these agents, for the purpose of combining them with chemotherapy to get a little bit of benefit. The hope was that these would replace chemotherapy, and they have not been able to do that.Bevacizumab has been on the market for more than a decade now and it has become our standard anti-VEGF agent. Everyone hoped that the two agents that came after it, aflibercept and ramucirumab, would be better. Unfortunately, the data suggest that these drugs are not better. The problem that we’ve had with both of them is that they’ve cost more than double what bevacizumab costs and without adding any apparent benefit. It is hard to see a role for them.

As a result, perhaps of the awareness of the cost differences, aflibercept dropped its price by more than half and now is priced very comparably to bevacizumab. There are no data supporting aflibercept as a first-line agent, and there are no data supporting aflibercept without FOLFIRI. In first-line, if one is going to use a VEGF agent, bevacizumab is the only one with data that proves it to be effective.

In second-line, the data suggest that either continuing bevacizumab is best, or switching to aflibercept or ramucirumab if you are also switching chemotherapies. These agents are likely to have similar effectiveness, but there is no reason that one is better than the other and no reason that one would rescue the other. There is no rational for example, to use FOLFIRI and bevacizumab and then try FOLFIRI and aflibercept or FOLFIRI and ramucirumab after that. It would be an inappropriate exposure of toxicity for the patient with no real chance of benefit.

What are the options for patients who progress on bevacizumab?

Are there settings or patient populations within colorectal cancer where VEGF inhibitors should not be used?

Are there any options for combining anti-VEGF agents with other agents besides chemotherapy?

What’s on the horizon in VEGF inhibition?

If one uses FOLFOX with or without bevacizumab first-line, it is acceptable to use FOLFIRI with any of these three agents as second-line, but I would strongly advocate that bevacizumab is the preferred agent. This is based on the fact that we simply have more experience with it and we better understand it and it’s less than half the price of ramucirumab.Unfortunately, there are not a lot of options. What we know is that when a patient progresses on a first-line bevacizumab-containing regimen, there is a modest benefit to continuing anti-VEGF therapy. However, the results are virtually the same, whether you continue bevacizumab or switch to either of the other two.They don’t work in the adjuvant setting. That was a disappointment to all of us. We hope when we find a drug that we can add it to the treatment of metastatic disease and then further use it in earlier stage-disease to increase the cure rate in the adjuvant setting. Two extremely large studies were done adding bevacizumab to chemotherapy in stage II colon cancer and both were negative studies. One actually trended toward inferior results with bevacizumab. In the absence of data, I think it is fair to say all three should not be used in the adjuvant setting.That is an experimental question. People have investigated the question of combining bevacizumab with VEGF tyrosine kinase inhibitors and that turned out to be rather surprisingly toxic. That approach has been largely abandoned. Whether these agents will have any type of beneficial effect with immunotherapies is an interesting question that others are exploring. However, since none of the immunotherapies have show benefit thus far in colorectal cancer, it is probably a question that is going to be asked in other disease first.I think we’ve mined VEGF as far as we are going to be able to. Part of the problem in oncology is that people like to say we have a lot of drugs in development and there is a robust pipeline of new agents in clinical trials, but there is a big difference between a new drug that is a novel mechanism, first-in-class drug, versus a next-in-class drug that is just another VEGF inhibitor or another EGFR inhibitor. The likelihood that a next-in-class drug is going make a dramatic beneficial leap is unfortunately very small. I think that we are likely going to have to look in other areas to make meaningful advances.