CRISPR-Cas9 gene editing has been around not even 4 years, and people are avidly discussing its promises and perils (see “The Public and the Gene Editing Revolution” in today’s New England Journal of Medicine). That’s great. But consider the historical backdrop.
April 1, the European Medicine Agency’s (EMA) Committee for Medicinal Products recommended a second gene therapy for marketing approval. “Strimvelis” treats adenosine deaminase severe combined immunodeficiency syndrome (ADA-SCID) and was developed at the San Raffaele Telethon Institute for Gene Therapy in Milan and GlaxoSmithKline (GSK). Regulatory approval is expected within a few months.
I wonder how many people realize, especially those fearful of how gene editing might be misused, that the gene therapy that is nearing approval actually entered clinical trials 26 years ago?
Yes, more than two decades to approve a gene therapy. The first, okayed in Europe in late 2012, was Glybera, to treat lipoprotein lipase deficiency. The first gene therapy approval in the US, my sources tell me, may happen this year, but more likely 2017.
ADA-SCID is NOT Bubble Boy Disease
The media, in reporting the April 1 EMA recommendation, widely dubbed ADA-SCID “bubble boy disease,” from the Wall Street Journal to Fortune and even to Medscape (although the article is accurate, the headline and photo not). ADA-SCID is NOT bubble boy disease.
The news releases from the EMA and GSK do not call ADA-SCID bubble boy disease, nor do technical publications, although scientists themselves sometimes do so in an effort to explain things. But oversimplification can ultimately be dangerous, especially when a technology is headed for the clinic. That’s why the “precision” is in precision medicine. The details matter.
The real bubble boy, depicted by John Travolta in a TV film and by Jerry Seinfeld in the 47th episode of his eponymous TV show, had a name, David Vetter. He had SCID-X1, which results from a mutation in the interleukin 2 receptor subunit gamma (IL2RG) gene. ADA-SCID is an enzyme deficiency and a more systemic illness.
Both ADA-SCID and SCID-X1 cripple T cells and therefore also B cells — that’s what the “combined” in the name means — but by entirely different mechanisms. ADA-SCID is autosomal recessive, SCID-X1 X-linked. The true bubble boy disease is the one that ran into trouble when a gene therapy vector pierced an oncogene, causing leukemia in some of the young boys in a clinical trial – a problem addressed by retooling vectors.
The true hero of the ADA-SCID story isn’t John Travolta or Jerry Seinfeld, but actually a heroine, Ashanthi (“Ashi”) DeSilva. She was 4 when she received the very first gene therapy of any kind – at the NIH in Bethesda, in 1990. I tell Ashi’s story in my book about gene therapy and excerpt it this week for Rare Disease Reports. She is alive and well today, but not keen on being interviewed. I talked to her briefly when she was 17.
Notes From a Gene and Cell Therapy Meeting
Several research groups have been working on a combination stem cell and gene therapy for ADA-SCID for many years. The approach is ex vivo (cells doctored outside the body) and autologous (patient receives her own cells).
The investigators graciously presented combined data at a news conference at the American Society of Gene & Cell Therapy annual meeting two years ago. At that time, 42 children had been treated, 18 from Milan, 8 from London, and 10 from Children’s Hospital Los Angeles/UCLA. The protocols used gamma retroviruses as the vector, with 100% survival and 73.8% disease-free survival.
Adrian Thrasher, PhD, from University College London, justified the new treatment for a disease so rare that it affects only 15 children in Europe a year. “Doing the math for ADA deficiency is easy. It costs $500,000/year for the enzyme replacement therapy, PEG-ADA. Autologous delivery is just the cost of a short hospital stay.” The children in the trials didn’t have matched bone marrow or umbilical cord stem cell donors. Many received PEG-ADA until the gene therapy took hold, as did Ashi in the initial trial, for safety.
Other speakers described a patient who contracted a cytomegalovirus infection after the gene therapy and fought it off so well that he didn’t even need acyclovir, and another whose liver enzymes and ADA levels were better than those of the researchers. And none of the kids showed a predominance of one white blood cell subtype that would indicate the feared leukemia.
At the news conference I asked about Ashi, because in researching my book I knew that some investigators questioned whether her gene therapy had really worked, since she was also receiving the enzyme. Other confounders, seen in retrospect, were that the gene transfer works better in babies, and that stem or progenitor cells are better targets than were Ashi’s mature T cells.
“The study used the best tools they had at the time. It was not enough to provide a clinical benefit, but there’s no question that by using the new methods, patients definitely have a clinical benefit. No one uses the word “cure,” but great clinical benefit,“ said Harry Malech, MD, Chief of the Genetic Immunotherapy Section at the National Institute of Allergy and Infectious Disease, who led the session.
History in a Citation List
So what is the new method for effectively delivering healing genes? (A gene therapy is called gene transfer until efficacy is shown.) A look at the citations from Don Kohn, MD, who leads the UCLA group, traces the evolution of ADA-SCID gene therapy, while also illustrating the incremental, non-breakthrough nature of medical research.
1986 Retroviruses deliver working ADA genes to human T cells in culture
1989 T cell lines are established from an ADA deficient patient
1995 Functional ADA genes are introduced via CD34+ cord blood stem cells into three newborns, one of whom graces the cover of Time magazine, with sustained gene expression.
1998 The three little boys are now three years old, and 1 to 10% of their T cells bear the correction.
2003 The boys still have corrected T cells, but not the explosive leukemia from SCID-X1 gene therapy that appeared in 2002. Just to be safe, further children with ADA-SCID treated in 2005 and later had genes delivered in retooled “self-inactivating” retroviruses, and also received a drug that makes more room in the bone marrow.
2006 Lentivirus delivers human ADA gene into knockout mice. Lentivirus (disabled HIV) is replacing gamma retroviruses in some gene therapy experiments.
2014 The lentivirus mouse model corrects the enzyme deficiency.
2016 “New approaches to gene therapy for ADA-deficient SCID, including the use of lentiviral and foamy viral vectors for ex vivo gene transfer to hematopoietic stem cells, direct in vivo ADA gene delivery and ADA gene correction using site-specific endonucleases to augment homologous recombination,” reads Dr. Kohn’s website. He’s moved on to gene editing, while still perfecting gene therapy.
Déjà vu All Over Again
Descriptions of the early ADA-SCID gene therapy experiments eerily parallel today’s concerns about CRISPR-Cas9. Consider “Human Gene Therapy”, a 1992 paper in Science by William French Anderson, MD, who treated Ashi. Here it’s important to distinguish gene therapy, which adds a gene but does not remove the mutant one, from gene/genome editing, which swaps in a gene or genes. Done on somatic cells the technologies could provide treatments; done on fertilized ova, the germline, they’d alter individuals.
In 1992 Dr. Anderson discussed human germline manipulation, but then ironically pointed out that an intervention that uses the general gene swapping phenomenon called homologous recombination would actually be safer, because it is much more precise. CRISPR-Cas9 does that. He wrote, “Until the time comes that it is possible to correct the defective gene itself by homologous recombination (rather than just inserting a normal copy of the gene elsewhere in the genome), the danger exists of producing a germline mutagenic event when the ‘normal’ gene is inserted. Therefore, considerable experience with germline manipulation in animals, as well as with somatic cell gene therapy in humans, should be obtained before considering human germline therapy.”
If that wasn’t prescient enough, Dr. Anderson then brought up:
• An infant’s right to an unmanipulated genome
• The impossibility of a fertilized ovum providing informed consent
• Playing God
• Genetic enhancement
Dr. Anderson can’t directly share his joy at the EMA announcement because he has been in prison since 2007. I’ve told some of his story here, and a more recent account is here. He told me last night, via his wife and her daily call to him, that he is “delighted that the approval in Europe is going forward.” When I can find more details of the EMA recommendation and get them to him, he’ll comment further for DNA Science. For a time he was known as the “father of gene therapy,” and read every line of my gene therapy book manuscript with pencil in hand.
Given the time it’s taken to get the first gene therapies into the clinic, I don’t think CRISPR-Cas9 DIY DTC kits will festoon the shelves at Walgreens anytime soon. But due to the relative ease of gene editing compared to gene therapy, it shouldn’t take two decades.
In the meantime, I hope that the media and public opinion polls make the effort to distinguish the two biotechnologies. The media are certainly getting better at this. But the Perspective in the NEJM published today points out that most public opinion polls lump gene therapy and gene editing together. Even more alarming, many polls refer to the subjects of germline manipulation – fertilized ova! – as “unborn babies.” With language like, equating microscopic cells to yowling babies, it’s little wonder that certain politicians are calling for punishment of women who end pregnancies. I don’t think the general public interested in genetic technology needs that level of dumbing down. The politicians might — how often do any of them mention any of this?
The Perspective authors call for “a level of public education that has not been achieved to date.” Amen. Guess I’ll keep on writing my textbooks.