The gene therapy trial during which a 36-year-old Illinois woman died in July was inappropriately designed, experts said at the 17 September meeting of the government panel investigating her death.

Researchers affiliated with Seattle-based Targeted Genetics injected a gene therapy treatment into Jolee Mohr's right knee first in February and again on 2 July. On 24 July, Mohr died.

The immediate cause of her death was massive bleeding behind the kidneys from unknown causes, which led to the collapse of multiple organs. Her body was also overwhelmed by a fungal infection known as histoplasmosis, which can kill people with deficient immune systems.

Mohr had also been taking Humira (adalimumab), an injectable protein that eases the symptoms of rheumatoid arthritis throughout the body by blocking a protein called TNF-alpha. The gene therapy product being tested also blocks TNF-alpha, but it was expected to do so only in the affected knee.

“The study was poorly designed. If you're trying to test safety of a product, then by definition the person shouldn't be allowed to be on [another medicine in] that same class of product,” Kyle Hogarth, a critical-care doctor at the University of Chicago, where Mohr died, told Nature Medicine after the Recombinant DNA Advisory Committee meeting. “What made her sick? Was it the Humira she was on or the gene [therapy] product? We'll never know.”

There is no test proven to distinguish between the two medicines, although a test made by California-based Amgen may be able to do so. Both drugs suppress the immune system, and if the gene therapy product leaked out of the joint it may have tipped a precarious balance, allowing the fungal infection to overwhelm Mohr.

“If you don't have an assay, don't do the trial—that's my sense,” Hildegund Ertl, who directs the Vaccine Center at The Wistar Institute in Philadelphia, said bluntly at the meeting.

These criticisms are off the mark, counters Stewart Parker, president and chief executive officer of Targeted Genetics. It is “good clinical practice” for people to be maintained on the medicines they are already taking when investigators introduce an experimental drug, Parker says. “If patients were denied current therapy, that would be considered unethical, as there is no guarantee in a safety study that benefit will be observed.”

The purpose of local injection of the gene therapy is to treat joints that are unresponsive to injected proteins that affect the whole body, Parker adds. To treat resistant joints, she notes, “One would have to give massive doses of systemic therapy, which would lead to serious, significant side effects from systemic immunosuppression.”