BOSTON -- Gene therapy is rapidly migrating westward from itsroots at the National Institutes of Health (NIH). A fortnight agoKenneth Culver, one of the original trio that injected the firstgene into the first patient three years ago next month, movedto Iowa Methodist Medical Center in Des Moines as executivedirector of its new Human Gene Therapy Research Institute.

Earlier this month Culver gave a plenary address here atInnovation Science, a conference sponsored by the AmericanAssociation for the Advancement of Science. He recalled that onSept. 14, 1990, he and his two associates at NIH's NationalCancer Institute, W. French Anderson and R. Michael Blaese,introduced a gene construct encoding adenosine deaminase(ADA) into the bloodstream of a 4-year-old girl with SCID(severe combined immunodeficiency disease), then treated a 9-year-old some months later. This Sept. 14, a celebration at theCapitol Building in Washington, D.C., will mark the thirdanniversary of this gene-therapy milestone (see adjoiningsidebar).

Since then, the most ambitious gene therapy protocolsapproved by the Recombinant DNA Advisory Committee (RAC)to NIH have broadened from inherited diseases such as SCID,cystic fibrosis and Gaucher's to embrace cancer. Here, the portof entry is -- paradoxically -- the most inaccessible and deadlymalignancy, brain tumors. Culver explained that because braintissues are immunologically privileged, their tumors can beattacked with gene strategies without concern for immunerejection.

His experimental gene therapy game plan against malignantglioblastomas was used to treat the first of eight patients lastDecember. RAC has authorized trying this maneuver in a totalof 20 patients, half with primary brain tumors and half withmalignancies metastasized from elsewhere in the body.

Victims of glioma have an average life expectancy of ninemonths after their disease is diagnosed. Conventional treatmentconsists of surgery to remove the bulk of the tumor followedby irradiation to ablate the remaining cancerous cells.

Culver's one-time therapy for killing malignant cells whilesparing healthy ones consists of injecting two deliverypackages a week apart directly into the cranial tumor bed via ahole drilled in the skull. The packages contain a series ofcomponents crafted to confront tumor cells with the seeds oftheir own self-destruction.

-- The first package contains a congeries of up to 1 billionrecombinant mouse fibroblast cells transformed by anemasculated retrovirus packing a herpes simplex virus geneencoding thymidine kinase (HSVtk).

-- A second package follows into the patient's brain a weeklater. It consists of the potent anti-viral drug gancyclovir,which the genetically expressed thymidine kinase (tk)unleashes to wipe out all rapidly dividing cells and viruses inits purview, namely the glioma cells, and, incidentally, thenow-redundant retroviral vector particles.

Culver, who perfected this ploy jointly with Edward Oldfield atNIH, told BioWorld how its components work.

The recombinant mouse host cells harboring the viral tk geneact as a slow-release system for at least seven days, spreadingover the tumor. They express two pieces of DNA. The first re-creates the retroviral vector, starting with the gag envelopegene that makes the viral coat, and then enzymes responsiblefor a functioning virus. The second DNA sequence expresses theHSVtk gene that the vector transfers to the target tumor cell.

The retroviral vector has been engineered so that it cannotreplicate in situ after delivering its tk DNA sequence. The tkgene is one of 174 in the herpesvirus genome and the only oneappropriated from this infectious pathogen.

The thymidine kinase enzyme expressed by that viral gene hasthe unique function of phosphorylating, and thus turning loosethe gancyclovir anti-viral drug.

"The herpesvirus thymidine kinase gene we've taken out of thevirus has a much higher affinity for phosphorylatinggancyclovir than our own human tk gene," Culver explained."That's why we can have selective toxicity zapping the tumor."

Beyond that he invokes the "bystander effect," by whichmalignant cells disseminate gancyclovir's death-dealingmessage to neighboring tumor cells untouched by the actualgene therapy package.

Results of the treatment given to those initial eight brain-tumor patients at NIH will be published this fall. Meanwhile,plans are afoot to extend these trials functionally as well asgeographically. Clinicians have developed a fine tube or shuntthat can dwell in the skull so treatment may be repeated atwill.

Treatment of the remaining 12 patients in the initial RAC-approved protocol will be completed at NIH by neurosurgeonsEdward Oldfield and Zvi Ram.

Culver will be collaborating with the University of Iowa andthe University of California, San Francisco, for a newmulticenter brain tumor trial of up to 30 patients at his HumanGene Therapy Research Institute.

Separately, he plans a brain-tumor gene therapy program atLos Angeles Childrens Hospital to treat youngsters ages 2 to 18afflicted with recurrent astrocytomas rather than gliomas.

Culver plans to initiate these trials in early November, bywhich time Genetic Therapies Inc. of Gaithersburg, Md., willhave formulated "a big new batch of murine producer cells" tosupply for use in these and other trials.

Nelson Wivel, who directs the NIH's Office of Recombinant DNAActivities, sees two important differences between the initialNIH brain-tumor trial and Culver's proposed Iowa protocol.

"One of the principal differences," Wivel told BioWorld, "is thatin this new approach, they want to put in a little tube or shuntthat leads from the interior of the brain to the outside. Whatthat does is open up the option for multiple treatment, whichwas not a feature of the original protocol. I think that does givesome flexibility in dealing with these patients."

However, Wivel sees a perceived problem that he said the Iowaprotocol will unveil.

"Clearly," he said, "the major delivery vehicle here is mousecells, which contain tk genes -- up to 10 to the ninth powercells. That's a lot of foreign protein. One could conceivablydeliver the first treatment, but what if the patient's coursedictates a second treatment? There's a question now of animmune reaction or a sensitization to this mouse protein. That'ssomething that remains to be determined as we move on tomultiple treatments of one patient."

He added that whether the patients have primary brain tumorsor metastatic brain tumors can be a significant point because"the blood-brain barrier may be breached in cases ofmetastatic cancer. In that setting, one might be a little moreconcerned about the immune response. If there is an immuneresponse, then the mouse cells could be destroyed before thatseven-day period, spread of the virus vector would necessarilybe limited, and you would be denied the opportunity for themaximum therapeutic effect."

As Culver plans to move beyond the brain's immune sanctuaryinto gene therapy of solid tumors throughout the body, he isacutely aware of the immune-response complication. "Now thatwe're moving out of the brain," he asked rhetorically, "are wegoing to have to immune-suppress somehow so that theproducer cells will live long enough to transfer the genes?Because once you leave the brain you lose that privileged site."

Culver has animal studies and protocols on the Iowa drawingboard for ovarian, breast, metastatic colorectal and primaryliver cancers.


Of the 40 known SCID (severe combined immunodeficiency)children in the world today, Kenneth Culver reported, nine areon T-cell-based gene therapy, one without the chemical cushionof direct injections of PEG-ADA, the missing enzyme. Both ofthe first two patients that were treated, Culver said, "hadimprovement in clinical health beyond what they got on PEGtherapy."

On May 13, the older girl, now 12, received another "first": herown stem cells primed to express the ADA gene. To harveststem cells from her peripheral blood, Culver and fellowresearchers R. Michael Blaese and W. French Anderson gaveher six subcutaneous injections of granulocyte colony-stimulating factor (G-CSF). These caused demargination, a shiftof stem cells from bone marrow to blood. They took these stemcells out, and ex vivo in the lab used a retroviral vector totransfer ADA genes, which they then gave back by threeintravenous infusions.

It's too soon to assess how that stem-cell therapy is working.Data on the initial T-cell treatments will be published soon,Culver said, adding, "both girls are healthy kids." But he added,"We do believe that their immune systems are not 100 percentnormal. They have vulnerabilities, even though they're doingmuch better against common pathogens.

"We're really forging ahead to stem-cell gene therapy becausewe believe that will provide as complete an immunereconstitution as we can give them. And hopefully, it will bedurable, lasting -- hopefully forever."

All of the gene therapy was all done at NIH, paid for by thetaxpayers. There was no charge to the family, Culver said.

If the stem-cell therapy succeeds and the two girls need it onlyonce or rarely, he declared: "I don't know what it's going tocost, but not a quarter-million dollars. PEG-ADA treatment,with injections once or twice a week, comes to around$200,000-$250,000 a year, paid for by insurance companies. Sostem cell treatment, even if expensive, would be a tremendoussaving overall." -- David N. Leff

-- David N. Leff Science Editor

(c) 1997 American Health Consultants. All rights reserved.