LONDON - A unique type of gene therapy that makes cancer cells commit suicide is safe and nontoxic, and caused tumors to shrink in some patients, a newly published study reported.

The targeted therapy, which exploits the genetic differences between cancer cells and normal cells, makes it possible to switch on an anticancer drug only in breast cancer cells.

The results set the foundation for a host of future clinical trials and suggest targeted gene therapies for many different types of cancer, including ovarian cancer and some types of breast cancer that are difficult to treat, could reach the clinic within a few years.

Topping the list are investigations into the treatment of liver metastases by infusing the liver, and ovarian cancer by an infusion into the peritoneum. In both cases, the infusion would deliver a gene therapy. Depending on the type of cancer and the genes chosen, the therapy could activate anticancer drugs at the tumor site, or trigger the production of cellular enzymes that push cells into the pathway leading to apoptosis (programmed cell death).

Nick Lemoine, director of the Imperial Cancer Research Fund's Molecular Oncology Unit at the Hammersmith Hospital in London, told BioWorld International, "The beauty of the system we are using is this targeting approach, which makes use of what we know about transcription in cancer cells compared to normal cells. In future studies, we hope to tackle more ambitious targets and produce a more dramatic therapeutic effect."

Lemoine and his colleagues, who are currently seeking partners to help develop their findings, reported their results in a paper in the July issue of the Journal of Clinical Oncology titled, "Genetic Prodrug Activation Therapy for Breast Cancer: a Phase I Clinical Trial of erbB-2-Directed Suicide Gene Expression."

The study is a culmination of a line of research, begun at Hammersmith in 1993, into the regulation in human breast cancer cells of the oncogene erbB-2. About 20 percent of women with breast cancer have tumors that overexpress the protein encoded by this gene. Those patients have a relatively poor prognosis and do not respond well to conventional chemotherapy or hormone therapy.

In 1995, the Hammersmith group discovered the promoter of erbB-2 is activated by a set of transcription factors that normally are expressed only during embryogenesis. In normal adult tissues, their production is switched off - but their reactivation results in cancer.

Lemoine and his colleagues decided to exploit this finding to drive the expression of genes that could cause suicide of tumor cells. They focused on an enzyme called cytosine deaminase (CD) found in fungi and bacteria that deaminates the base cytosine. CD also can deaminate 5-fluorocytosine (5-FC), an antifungal drug, turning it into 5-fluorouracil (5-FU), which is used to treat many cancers.

By fusing the gene for CD to the promoter of the erbB-2 gene, they hypothesized, CD would not be expressed in normal cells, but only in cancer cells that switch on erbB-2.

"If you then expose these cells to the harmless prodrug 5-FC," Lemoine explained, "they produce 5-FU and kill themselves because they poison their ability to make DNA."

The study reported in the journal was designed to evaluate safety and efficacy of the approach in patients with breast cancer. The 12 patients who took part had tumors that overexpressed erbB-2, and had failed to respond to conventional therapy. They all had recurrence of their breast cancers in the form of skin nodules, and all had at least three nodules of similar size and shape.

Each woman was first given injections of the genetic construct directly into the center of one of these nodules; 48 hours later, in eight of the patients, the prodrug was slowly infused into a vein. For purposes of comparison, a second nodule was injected with a genetic construct that did not include the CD gene, and a third nodule was left untreated.

Lemoine said, "In every patient we were able to find evidence of gene transfer and gene expression, and that it was specific to the cancer cells - there was no activation of the enzyme in normal tissues. In one case the lesion which received the gene transfer got substantially smaller." There was no evidence of toxicity in any of the patients, he said.

The significance of the study, Lemoine added, is that it showed targeted gene therapy based on the molecular profiles of tumors is a viable approach to cancer treatment. Other systems under investigation include the promoter for vascular endothelial growth factor (VEGF), which is overexpressed in many different types of cancer.

He also hopes to develop more effective suicide genes. "We want to exploit some of the natural mechanisms of cell death, including apoptosis," he said. "So, you might put under the control of your selected promoter - erbB-2, VEGF, or whatever - one of the caspase genes, which when expressed would send the cell into the apoptosis pathway." This could provide an alternative therapy for cancers that have developed resistance to several conventional agents, he said.

In future clinical trials, Lemoine and his colleagues hope to achieve more efficient gene delivery by using adenoviral vectors to deliver the gene constructs. They intend to try using the herpes simplex virus (HSV) thymidine kinase (TK) suicide gene system. TK is able to phosphorylate the prodrug ganciclovir (already used clinically to treat herpes infections) into toxic derivatives that, as with 5-FU, can poison nucleic acid synthesis.

High on the agenda are clinical trials that will use an adenoviral vector infused into the liver, to treat liver metastases, and use of the same vector (containing a suicide gene driven by the erbB-2 promoter) to treat ovarian cancer, via an infusion into the peritoneum.

Lemoine concluded, "There is an important lesson from this study, which shows that, as we develop more effective suicide genes, the ability to target their expression is going to become more and more useful."