The study involved painstaking molecular analysis of blood samples taken annually from the patients, who participated in separate studies begun in 2000, 2002, and 2004.
"We were astonished that we could detect the modified cells for so long. It's a relatively small number of patients, but more than 500 years of patient data," said University of Pennsylvania pathologist Bruce Levine, a leader of the research. "But it's difficult to separate with certainty the effectiveness of this treatment from the antiretrovirals."
Gene therapy harnesses the insidious ability of viruses to slip their DNA into the cells they infect. By neutralizing a virus and then using it as a "vector" to insert DNA that is helpful rather than harmful, gene therapy can theoretically treat ailments ranging from arthritis to infections and cancer.
Levine, his Penn colleague Carl June, and their team have tested a variety of ways to outwit HIV with gene therapy. Their approach has focused on T cells, which are the big guns of the immune system but also the cells that HIV infects. The researchers took some of the patients' T cells and inserted a gene that makes them better at recognizing and killing HIV-infected cells. Then these super-T cells were multiplied using growth-stimulation technology and put back into the patient.
Over the years, many other research groups have tried using modified T cells, but the patient's immune system perceived them as invaders and wiped them out, sometimes within hours.
Another problem is finding precise ways to insert the modified virus and ensure that it will cause the desired genetic change. The chosen virus may put its genetic payload into a host's DNA in a way that triggers cancer. That happened in France when 3 of 10 children who underwent gene therapy and were cured of a severe combined immune deficiency disease - better known as the "Bubble Boy disease" - later developed leukemia.
The Penn analysis showed that its therapeutic gene entered varying spots in the DNA of the T cells, which then went on to divide and produce a new generation of cells. Yet annual monitoring of the patients "has not detected any suspected or documented" disorders that might be traced to the genetic manipulation.
The researchers speculated that mature T cells "are more resistant" to toxic changes the inserted genes may cause than other types of cells, such as the immature blood cells used to treat the Bubble Boy patients in France.
The T cell approach also avoids flooding the patient's body with viral vectors - the problem that killed Arizona teenager Jesse Gelsinger in a 1999 experiment at Penn. Trillions of vectors were infused directly into his liver, with the hopes of giving him a gene he lacked. Instead, his immune system perceived the neutralized viruses as a threat and went into lethal overdrive to destroy them.
The death, the first in a gene-therapy experiment, triggered a nationwide halt in human testing of such treatments after ethical and regulatory lapses were exposed. It cast a long shadow on Penn for years. "I think people have been scared of gene therapy for a long time," Levine said. "Now, maybe we are on the way to enlightenment."
June speculated that the T cell approach could be broadened to develop gene therapies for non-life-threatening diseases such as arthritis.
That may be a leap, said David Strayer, a professor of pathology at Thomas Jefferson University who has long studied gene therapy for HIV/AIDS.
"These excellent results are fairly specific for his system and may not mean a generalized breakthrough in gene therapy," Strayer said. "Both [the safety and durability] are significant. But it remains to be seen how broadly they apply to other disease settings."
And the road to a commercial therapy remains long. Two biotech companies that hoped to use the Penn team's technology to develop HIV gene therapies have run out of capital. Penn is now partnering with two other companies - Adaptimmune Ltd. and Sangamo Biosciences - to begin clinical studies of modified T cells for treating HIV.
Contact Marie McCullough at 215-854-2720 or email@example.com, or follow on Twitter @repopter.