American Association for Cancer Research

Anti-Angiogenesis Drugs


Tumor Blood Vessels


What do tumor blood vessels look like and why does their structure affect chemotherapy? Think of a solid tumor as a city. In this city, the roads, which are akin to a tumor’s blood vessels, are highly disorganized. If all the roads were perfectly made and you wanted to bring milk to everybody in the city, you would have no trouble delivering it. But since the roads are very chaotic, it's hard to deliver the milk to every house. In the same way, a tumor’s blood vessels are very disorganized. And when we inject a chemotherapeutic drug into the patient’s circulation, the drug would have trouble reaching all cells of the tumor because the tumor’s supply system is disorganized.

In some clinical trials, oncologists gave higher and higher doses of chemotherapy. But the outcome did not improve. Based on our research, we suggest another approach. We suggest that instead of giving higher doses of toxic chemotherapeutic drugs, we should fix the blood vessels, so we can distribute the drugs more effectively throughout the tumor.

Over the last 25 years, my colleagues and I have been trying to figure out why tumor blood vessels are abnormal and how we can fix them. We discovered quite accidentally that one way to fix these blood vessels is to use anti-angiogenic drugs.

When tumors grow in humans and animals, they have many incoming and outgoing vessels that make up the blood supply of tumors. Instead of having one door to enter and one door to exit, the tumors have thousands of entries and thousands of exits, which make them difficult to study.

In 1974, when I was working on my doctorate in chemical engineering, I met a scientist named Pietro M. Gullino, who was the chief of laboratory pathophysiology at the National Cancer Institute (NCI), in Bethesda, Md. He developed a tumor in rats with a single input (a single artery) and a single output (a single vein). With such a tumor preparation, Gullino could keep an eye on what was going into and what was coming out of the tumor.

I went to work in his laboratory and asked the following question: If we were to inject a drug into a tumor’s artery, how much drug would leave by the vein and how much would be retained by the tumor? And I found that when I injected a drug into the tumor’s artery, most of the drug bypassed the tumor like a beltway around a city. Some regions of the tumor accrued enough drug whereas others had very little to no drug.

That observation in 1975 haunted me and I realized that no matter how potent the newly developed drugs are, we needed to figure out how to deliver them throughout the tumor in effective quantities. Otherwise, we would just kill some cancer cells, and the tumor would eventually regrow from the cells that survived.

Being an engineer by training, I wanted to visualize what was going on inside the tumor. If we wish to find out what everybody in this room is doing, we could replace all the walls of this room with glass windows and put video cameras outside and then observe and record the activities in this room. In the same way, we implanted transparent windows around tumors in mice so that we could scan the growing tumors and record the events taking place in the various regions of the tumors. Initially, we focused on the tumors’ blood vessels, to see how they grew, what they looked like and how well they functioned.

We discovered that as a tumor grows, the blood vessels become disorganized and begin to leak fluid and even red blood cells. That's why tumors appear bloody sometimes. On the other hand, blood vessels in a normal tissue are organized in a way that every cell gets adequate oxygen and nutrients.

We also discovered that the blood flow is chaotic in tumors. So at a given time, blood flow might be brisk in some vessels while it  might shut down in nearby vessels. We have videos where we can see the flow reversal: Blood will flow in one direction for a while and then in the opposite direction. What is even more remarkable is that flow may resume in blood vessels that were stagnant at another time. Thus one big plus of giving drugs continuously over the long term is that the flow might resume and reach some of the cells that were not exposed to the drugs when the flow in that region was stagnant. Giving high doses of drugs continuously, however, could be toxic to normal tissues. On the other hand, giving lower doses of drugs over a longer period of time might improve drug delivery.

So, one of the obvious questions lingering in the back of my mind was: This is all observed in mice—what about human beings?  So we conducted a study not exactly like this, but very close to this, in cancer patients.

I knew from the scientific literature that some colorectal tumors in patients have an isolated blood supply similar to what Dr. Gullino developed in rodents—the tumors are connected to the patient’s circulation with a single artery and a single vein. So in collaboration with Dr. Mitch Posner, a surgical oncologist at the University of Pittsburgh at that time, my post-doctoral fellow Dr. Joanne Less and I made vascular casts of human colon cancers. After surgical resection, of tumors from six patients—with their artery and vein intact—we injected into the artery a blue-colored polymer (a liquid plastic that hardened inside the tumor’s blood vessels). Then, for one month, we bathed these tumors in potassium hydroxide, which ate away the tumor and left behind an intact cast of the blood vessels.

We then carefully analyzed the structure of the blood vessels in tumors from the six patients. And indeed, these tumor vessels were highly abnormal—similar to what we saw in our mouse studies. So what has been haunting me for 20 or 25 years is this: How do we straighten out these abnormal vessels and fix them so that we can deliver  chemotherapy to all regions of tumors in effective quantities? And, I came up with a strategy that went against the traditional thinking.