Gene-therapy trials for children suffering from severe combined immunodeficiency disease (SCID) are set to resume in the United States, despite another cancer case in a French trial of the technique.

On 4 March, an advisory meeting in Rockville, Maryland, heard evidence that the cancer risk exposed by the French trial may be specific to one form of the disease, known as X-linked SCID.

The advisory panel recommended that a gene-therapy trial for another variant of the disease, called ADA-SCID, should be allowed to proceed.

The French trial has so far successfully treated six children with X-linked SCID, but three other patients have since developed cancer, and one of these has died. The third of the cancer cases was reported in January, and prompted the US Food and Drug Administration to suspend three US gene-therapy trials on SCID (see Nature 433, 561; 2005 10.1038/433561a).

The leader of the French trial, Alain Fischer of the Necker Hospital for Sick Children in Paris, has already said that he will not continue his trial until he has made changes to the methods he uses to deliver the therapeutic gene to children with X-SCID. But the US panel said that US trials on X-SCID could proceed on patients who have failed to respond to other treatments or are likely to do so.

And it said that, so far, the risk of cancer seems unique to X-SCID, and may not apply to ADA-SCID. At the meeting, oncologist Utpal Dave of the National Cancer Institute described unpublished research suggesting that some of the risk of cancer in the X-SCID trials was related to that form of the disease, and might not appear in other variants of SCID.

Dave's data linked the gene that is missing in children with X-SCID to cancer in mice. He theorized that the method used in the French trial to replace this gene could be increasing the children's risk of cancer. But he found no data to suggest a similar risk for ADA-SCID.

“Although not conclusive, these are the first data that suggest the gene may play a role in these cancers,” says Daniel Salomon, a molecular biologist at the Scripps Research Institute in La Jolla, California.

There is one planned US trial in ADA-SCID, led by Donald Kohn of the Childrens Hospital Los Angeles. “I'm glad they agreed that ADA clearly has a safer record to date,” Kohn says.