The researchers have developed a biocompatible polymer that briefly changes from a liquid at 39 degrees Fahrenheit to a gel at body temperatures to block most gene-bearing viruses from being diverted through the blood stream to the wrong targets, the scientists reported in research journals.
"With this method we can reduce the misdirected virus dissemination by a factor of 100 to 1,000 times," said Fan Yuan, an associate biomedical engineering professor at Duke's Pratt School of Engineering who led the studies. "That's enough of a reduction to solve the problem."
The work was supported by the National Science Foundation.
According to Yuan, about 66 percent of the 918 gene therapy clinical trials conducted in 24 countries between 1989 and 2004 were aimed at treating cancer.
His interdisciplinary group from the Pratt School and the Duke Medical Center's radiation oncology department studied a preferred kind of anti-cancer gene therapy that uses relatively harmless adenoviruses to infect tumor cells. Once in a targeted tumor, genes in these genetically modified viruses are designed to express their modified genes to manufacture proteins that can either trigger tumor cell death or stimulate the immune system to attack the cancer.
Yuan said he and other Duke researchers have found that, because of the small size of pores in blood vessel walls and other access points, these adenoviruses cannot reach the majority of tumor cells by being injected into the blood stream but instead must be injected directly into tumors themselves.
However, direct intratumoral injection can cause the g ene-bearing viruses to escape their intended target and pose risks elsewhere, the researchers found in a study the researchers reported in the April 2005 issue of the British Journal of Cancer.
In that study, the researchers injected adenoviruses that produce glowing proteins in order to trace where the viral particles went during 24 hours after being injected into breast tumors implanted in experimental mice.
That investigation showed that 10 times more of the viral particles were transported to other organs than were retained within the tumors. And most of the errant viruses ended up in the liver, where the gene therapy protein products could cause cell death. Moreover, most viruses escaped from their intended tumor targets during the first 10 minutes.
Prompted by reports of animal deaths during past gene therapy trials with adenoviruses, this study provided the first direct research evidence that therapeutic gene-bearing viruses were being disseminated away from the tumor site via the blood stream, Yuan said. "In the past people wouldn't admit that," he said. "Nobody in the gene therapy field would believe this was happening."
The study also pinpointed the routes for the viral escapes. Those particles were escaping from the tumor area into the bloodstream through rifts in tiny tumor blood vessels created by damage from the injection needle.
Yuan said such viral escapes are inevitable since the needle's diameter is 30 times larger than the width of the microvessels that supply the tumor with blood. The needle is also wider than the spaces between each of those vessels, he added.
The first author of the British Journal of Cancer study was Yuan's graduate student Yong Wang. Other authors included Yuan, Yuan's research associate Ava Krol, associate research professor of radiation oncology Chuan-Yuan Li, and radiation oncology research associates Shanling Liu and Takashi Kon.
In a followup study in the September 2005 issue of the journal Cancer Research the researchers proposed a solution to the viral escape problem in the form of a polymer compound already employed for wound healing and for drug and gene delivery.
After being mixed with a gene-bearing virus, this compound -- called poloxamer 407 -- undergoes a temporary phase change as its stiffens from a liquid to a gel state when warmed to body temperature, that study reported.
This 1,000-fold viscosity change only lasts 10 to 90 minutes, long enough to clog the escape route for large numbers of viruses so those will link up with and infect tumor cells instead, Yuan said.
The Duke team found the polymer-virus mixture "could reduce virus dissemination by two orders of magnitude (100 times) and significantly increase transgene expression in solid tumors," their study reported.
In comparative tests on breast tumors implanted in live mice, the authors noted that two mice in the group injected with adenoviruses without poloxamer died within a half hour after administration. "But mice in the poloxamer group did not show any problem," they wrote.
Additional studies using the glowing protein to trace viral dissemination showed that "the poloxamer solution could significantly reduce the transgene expression in the liver and increase the transgene expression in the tumor," they reported.
Yuan said he has applied for a provisional patent covering this new use of poloxamer 407, which his group purchased from the BASF Corp.
In 2003, Yuan, Wang, Krol, Li and other Duke researchers previously reported that another chemical extracted from common brown algae can also block toxic leakages from tumor injections when mixed with the adenoviruses. However, the high viscosity of that jello-like mixture made it hard to inject.
"The bottom line is that when people try to do gene therapy on cancer they must inject the genes into the tumor," Yuan said. "One way would be to try and in ject the genes into the bloodstream. But that won't work because the virus particles are too large to get through to the tumor.
"Therefore, the only choice is to inject the viral particles directly into the tumor. But when you do that, we have demonstrated that dissemination elsewhere is inevitable. The result is that more will reach normal tissue than be left in the tumor. And that may cause all kinds of toxicity problems.
"In our first new paper we demonstrated the mechanism of this dissemination. And in the second paper we developed a solution to block it."