BioWorld International Correspondent

LONDON A group of researchers in Germany is calling for additional stringent safety controls on gene therapy initiatives, having observed what they admit is a very rare instance of leukemia in mice transplanted with genetically manipulated bone marrow cells.

Christopher Baum, at the University of Hannover in Germany, and his colleagues published their findings in a short report in the April 19, 2002, issue of Science titled “Murine Leukemia Induced by Retroviral Gene Marking.”

Baum, who is a professor of experimental cell therapy in the department of hematology and oncology at Hannover Medical School, told BioWorld International: “It is important not to misinterpret this result. As it stands, this is just a case report. What we need now is to invest in research addressing the side effects of gene therapy, so that we can make gene therapy safe and predictable.” He proposes developing sensitive animal models that will make it possible to detect whether therapy with a particular gene is likely to cause cancer.

When they made their observation, Baum and his colleagues were working on evaluating the safety of a retroviral vector for gene delivery. The vector was unable to replicate. The gene in question encoded a molecule known as LNGFR, which is a marker gene whose product appears on the surface of cells that contain it, allowing researchers to track the survival of genetically manipulated cells using monoclonal antibodies.

LNGFR already is approved for clinical use in humans, for delivery to T lymphocytes. Baum’s group, however, was assessing its safety when delivered to murine bone marrow stem cells.

The researchers administered bone marrow stem cells, which had been exposed to the vector containing LNGFR, to five mice that had undergone lethal whole-body irradiation. The mice appeared to be hematologically normal for a period of 28 weeks. Samples of their bone marrow were then taken and pooled, and transplanted into 10 other mice, which had also previously undergone lethal whole-body irradiation.

All 10 of those animals, the team reported in Science, developed hematopoietic disorders within 22 weeks. Six of the animals had leukemia equivalent to a specific type of acute myeloid leukemia in humans. When the researchers transplanted bone marrow cells from some of these leukemic animals into eight other mice that had been given a dose of sublethal irradiation, all eight of the animals died from acute myeloid leukemia within four months.

Baum and his colleagues used the same retroviral vector to deliver other marker proteins to 70 control animals and found that none of the mice developed hematopoietic abnormalities. Further experiments confirmed that there had been no alteration in the transgene sequence and that the retroviruses used were not capable of replication. They also eliminated the possibility that endogenous retroviral sequences had been activated.

When the researchers carried out a genetic analysis of the mice that had fallen ill, they found that all had a single copy of the vector that had integrated itself into a gene called Evi1 (ecotropic viral integration site-1). The leukemic cells expressed RNA from Evi1.

“Activation of the oncogene Evi1 was definitely the first event,” Baum told BioWorld International, “but we then found preliminary evidence that the second event was signal interference by the co-transferred protein LNGFR.”

This is why, he added, he and his colleagues do not believe that their observation can be generalized to other gene therapy applications. Scientists already had estimated, he said, that the chance of a therapeutic gene inserting itself into the genome and activating an existing oncogene is about one in every 10⁷. But there always has to be a second event to bring about cancer in mice, and in humans, probably even more.

“Introducing a gene that does not play any part in hematopoietic differentiation makes it extremely unlikely that you will end up with a leukemia after gene therapy manipulations,” Baum said. “What we have to do now is to make sure that the genes we are introducing are themselves inert and are not likely to cause signal alterations in the cell.”

Genes should be screened, he said, by a process similar to that used in the experiments reported in Science, which passages the genetically manipulated cells through several generations of mice. “This type of so-called serial transplantation makes the cells proliferate more quickly, and forces the expansion of an aberrant clone,” Baum added.

Next, he and his colleagues are planning to investigate how the leukemic clone arose in this case, and the extent of the specificity of the interaction between the Evi1 oncogene and the transferred marker gene. They also will continue to take part in international studies examining data from ongoing human gene therapy trials and gene therapy trials on non-human primates, to find out whether vectors insert the genes they ferry into cells into the host genome at random or in front of specific genes.

“We also need to find out whether a similar event can occur in human cells because we know that murine cells are much easier to transform than human cells,” Baum said.