Manufacturing Opinion: What’s Not to Like About a Perfect Technology?

Here’s a provocative headline and subhead from OneZero, a publication by Medium, above an article by Emily Mullin published on August 17:

Men Are Twice as Likely to Support Genetically Engineered Babies Than Women
The public may be ready for gene-edited babies.

This is clearly an attention-grabber. It’s based on a terrible paper about public attitudes about heritable genome editing (also called human germline editing) by four University of Melbourne academics led by Christopher Gyngell. The article was published by the Journal of Human Genetics on July 31 with a rather less exciting title:

 Genetics experience impacts attitudes towards germline gene editing: a survey of over 1500 members of the public

Mullin’s article is much more careful than the headline (which was almost certainly written by someone else) and does a good job representing what the peer-reviewed article says. A better summary of the data, however, might be:

A survey by advocates of human germline gene editing finds some support for medical (but not enhancement) applications — except among women, the religious, and anyone with professional or personal knowledge of genetics or genomics. Parents and people with college degrees are more skeptical than those without children or degrees.

Bioethicist and legal scholar Hank Greely, a commentator who is by no means an anti-germline campaigner, told OneZero that he wasn’t surprised that “those who know less about genetics or parenthood would be more accepting of germline editing. … The less you know, I suspect the less you see some of the possible alternatives.” 

Greely is likely referring to the embryo screening technique known as pre-implantation genetic testing, which enables at-risk parents to have genetically related children who are unaffected by the condition they might transmit. He may also be thinking about somatic gene therapy, which in the future may be used post-birth to treat genetic diseases. Mullin mentions both of these. Also, of course, the more you know about the huge technical obstacles of germline genetic modification, the less inclined to it you are likely to be. The article continues:

Greely says he doesn’t find the level of support in the new study that surprising given the self-selecting nature of the survey. “I don’t think it reflects the full public’s view,” he says. The support for germline editing would likely shake out differently in a deliberative democratic process, he adds.

Greely is gentle in his comments, and Mullin is a professionally self-effacing journalist (who had the good sense to interview Greely). They are too kind. 

This paper is a classic case of spinning results you don’t want. Gyngell has been an advocate of germline gene editing (GGE) for years. In 2015, when he was billed as “a bioethicist from the University of Oxford” (at the Centre run by Julian Savulescu) he wrote a piece for The Guardian titled:

The case for genetically engineered babies

It’s obvious what results he hoped to find. Which presumably contributed to the choice of a thoroughly biased methodology:

• Participants were self-selected over the Internet — “social media, including via twitter and Facebook posts.”
   
• “A short educational video” was included, an explainer about CRISPR originally produced by STAT — two minutes and 13 seconds long, cute but superficial, and clearly on board with CRISPR’d babies.

The article mentions no other “educational” material provided to respondents. Which leads us to one more, crucial, issue:

• “Participants were asked to answer the questions on the basis that GGE was safe, accurate, and effective.”

Now there’s a heroic assumption.

Indeed, a month before the Gyngell article was published, Nature published a stunning (but not really surprising) summary article by Heidi Ledford with this headline and subhead:

CRISPR gene editing in human embryos wreaks chromosomal mayhem
Three studies showing large DNA deletions and reshuffling heighten safety concerns about heritable genome editing.

Mullin had already written about one of those studies:

Scientists Edited Human Embryos in the Lab, and It Was a Disaster
The experiment raises major safety concerns for gene-edited babies

Those particular papers were obviously not available when the survey was conducted. But concerns about off-target effects have been discussed since the technology appeared. The Gyngell paper actually includes this caveat, in the introduction:

While its potential therapeutic applications are compelling, gene-editing poses unprecedented ethical and scientific challenges. Firstly, the safety of CRISPR-cas9 is uncertain, with early research that targeted disease-causing genes in human embryos resulting in undesirable and unintended off-target effects and mosaicism.

And yet not only did they survey volunteers without mentioning these “unprecedented ethical and scientific challenges,” they specifically offered the hypothetical condition of safety, accuracy, and effectiveness. 

Well, if my grandmother had wheels, she’d be a bicycle. But I wouldn’t bet on her to win the Tour de France. 

The statistical analysis in the paper looks convincing, but the old computer adage applies: Garbage In, Garbage Out. Policymakers — and everyone else grappling with the ongoing controversy about heritable genome editing — should understand this paper for what it is: an attempt to influence discussion in one particular direction.