Gene therapy 'caused T-cell leukemia'

Insertional mutagenesis pinpointed as cause of T-cell leukemia in X-SCID gene therapy trial | By Jo Lyford

An unanticipated complication of gene therapy has been confirmed as the cause of T-cell leukemia in two boys receiving the pioneering treatment for X-linked severe combined immunodeficiency (X-SCID). In the October 17 Science, an international research team confirms that inappropriate insertion of the retroviral vector near the proto-oncogene LMO2 promoter led to uncontrolled clonal proliferation of mature T cells (Science, 302:415-419, October 17, 2003).

“This is the nightmare scenario,” said Terry Rabbitts of the Medical Research Council's (MRC) Laboratory of Molecular Biology in Cambridge, whose previous work on LMO2 played a key role in the investigation. “It's time to step back and give ourselves some breathing space.”

The affected children were the youngest of 10 boys with X-SCID enrolled in the gene therapy trial at Hôpital Necker Enfants Malades under the direction of Alain Fischer. Inherited X-SCID, the so-called “baby in the bubble” disease, is characterized by a lack of both T and B cells. The only certain cure at present is a bone marrow transplant from a histocompatible sibling. For the vast majority of children, however, an allogenic transplant is the best option, which carries a 5-year success rate of just 65%.

Building on the discovery that X-SCID was due to a defect in the common γ (γ c) chain of the interleukin 2 receptor on chromosome Xq13, Fischer and colleagues at the Necker Hospital developed an alternative therapy based on the ex vivo transfer of the γc gene into autologous hematopoietic precursor cells, employing a vector derived from a defective Moloney murine leukemia virus.

Fourteen children with X-SCID—10 in France and four at Great Ormond Street Hospital in London—were subsequently treated and showed an impressive and sustained clinical response. “These children are coping with lots of infections, have a fully functioning immune system, and are effectively cured,” said Professor Rabbits.

In October of last year, the first serious side effect in a trial subject who exhibited a “leukemia-like syndrome” 30 months after therapy was reported. In January, the development of similar complications in a second participant in the French trial triggered a moratorium and frantic investigations aimed at discovering what had gone wrong.

The Science paper, a collaboration among Fischer's team, the MRC, and experts in the United States and Australia, confirms that insertional mutagenesis—originally perceived as a possible but remote complication—caused both children's leukemias. Using molecular techniques, Hacein-Bey-Abina et al. observed that clones from the patients showed retrovirus insertion in proximity to the promoter of LMO2, which encodes a transcription factor required for hematopoiesis.

“Until this report, retroviral insertion in the context of gene therapy has been considered an untargeted and largely random event,” comment David A. Williams and Christopher Baum in an accompanying Perspectives article. Indeed, such insertional leukemogenesis without widespread viral replication was reported in just one of thousands of preclinical experiments.

What remains unclear is why the viral vector inserted near the LMO2 promoter locus in just two of 14 children treated with the same technique. Fischer believes that disease- and protocol-specific issues may have played a part.

“These patients were the youngest to be treated—just 1 and 3 months—while the others were all at least 6 months old,” Fischer told The Scientist. “They also received the highest number of corrected cells. So, statistically speaking, the more you inject, the higher the probability.”

Professor Hidde Haisma, from the Department of Therapeutic Gene Modulation at the University of Gronigen, agreed that the patients' age may be significant. “Very young infants have a different subset of stem cells, with a higher proportion of precursor T cells, so they may be at higher risk of insertional mutagenesis,” he said. The Necker team is already testing the “age” hypothesis in mice.

Another worrying aspect of the study is the apparent preference of the retrovirus vector to insert in or near active genes. Professor Christine Kinnon, head of the Centre for Gene Therapy of Childhood Disease at Great Ormond Street Hospital for Children, commented that gene therapy in X-SCID children involves introducing a strong selective advantage, which would inevitably carry a theoretical risk of leukemia.

“The insertional activation of the LMO2 locus in two independent leukemias suggests an essential event that is required for the initiation of a malignant cascade,” add Williams and Baum. “Given the lack of similar side effects in previous studies, a combinatorial process seems to be the likely culprit.”

In addition to identifying patient-related risk factors, attempts are also underway to develop a vector with a reduced propensity for insertional genotoxicity. “Retroviruses were the logical choice, as they had a good track record and can relatively easily infect bone marrow cells,” said Haisma. “But we could use lentiviruses or adenoassociated viruses, or even artificial systems in which DNA is combined with proteins to do the same job.”

Fischer said his team is already making progress toward the next generation of highly selective vectors. “I expect that by next year, we will be in a position where we have a safer therapeutic approach and can start treating patients once again,” he said.

Kinnon agrees that vector modifications hold the key to success. In the meantime, the Great Ormond Street trial is continuing to recruit X-SCID children for gene therapy, albeit on a case-by-case basis and only when authorized by the Gene Therapy Advisory Committee.

“I still believe gene therapy has a relatively good risk–benefit ratio,” Kinnon said, noting that all 14 children treated to date are alive and well, and the two boys who developed leukemia responded well to chemotherapy and are in complete clinical remission. “With a mismatched bone marrow transplant, at least one third would have died by now—that's four or five children versus none—so it's still an huge improvement on current options. If it were my child, I would give them this treatment.”