By Grant Jacobs 15/02/2017

Just out today are the recommendations of a joint science advisory group of the USA National Academy of Science and the USA Academy of Medicine on gene editing.

As an alert to any that might be interested I’ve copied their summary of their recommendations below.This post is meant as a simple heads-up, rather than an analysis of this report, as I’ve yet to read the full report, but their summary recommends (very) cautious movement forward.

As expected, their summary of recommendations treat two different circumstances separately:

  1. where the gene editing only applies to the lifetime of the patient, it is not carried on to their children.
  2. where the edited genes are passed on to the children of those treated, so that the edit is inherited in the next generation(s).

For the first they recommend work proceed, but initially limiting clinical trials and therapies to those that prevent disease or disability.

For the second—applications that affect later generations—they recommend that work be limited to trials.

Both seem unsurprising to me.

Note that the second does not extend to therapies. I’m a bit confused here, as a number of early media reports read to me as if they think the recommendations do include therapies where the gene edit would be inherited, but the wording of the recommendation for heritable editing does not include ‘therapies’. It only refers to trials. (See 5.1 in block quote below.)

My reading of this, is that they recommend that work towards treatment of serious disability or disease be developed, limited to treatments that only affect the patient, not their children, and at this stage work might eventually result in treatments that would have heritable effects be limited to trials only for now.

Furthermore, their recommendations throw additional restrictions on the case of heritable changes, for example that they should be limited to cases where the parents have no other options for a healthy child, for example, and that substantial follow-on and testing be done.

They suggest ruling out applying gene editing for ‘enhancements’ at this time.

Readers might like to note the public engagement suggestions.

A few spare thoughts while I’m writing can be found after the summary of the recommendations,

Oversight and Use of Human Gene Editing

Summary of Recommendations

Global Principles for Research and Clinical Use

Consider and apply the global principles in governance of human genome editing (2.1)

  • Promoting well-being
  • Transparency
  • Due care
  • Responsible science
  • Respect for persons
  • Fairness
  • Transnational cooperation

Basic Laboratory Research

Use existing regulatory processes to oversee human genome editing laboratory research (3.1)

Somatic Genome Editing

Use existing regulatory processes for human gene therapy to oversee somatic human genome editing research and uses (4.1)

Limit clinical trials or therapies to treatment and prevention of disease or disability at this time (4.2)

Evaluate safety and efficacy in the context of risks and benefits of intended use (4.3)

Require broad public input prior to extending uses (4.4)

Germline (Heritable) Genome Editing

Permit clinical research trials only for compelling purposes of treating or preventing serious disease or disabilities, and only if there is a stringent oversight system able to limit uses to specified criteria (5.1)


Do not proceed at this time with human genome editing for purposes other than treatment or prevention of disease and disability (6.1)

Encourage public discussion and policy debate with respect to somatic human genome editing for uses other than treatment or prevention of disease and disability (6.2)

Public Engagement

Public input should precede any clinical trials for an extension of human genome editing beyond disease treatment and prevention (7.1)

Ongoing reassessment and public participation should precede any clinical trials of heritable germline editing (7.2)

Incorporate public participation into the human genome editing policy process about “enhancement” (7.3)

When funding genome editing research, consider including research on strategies to improve public engagement (7.4) and for long-term assessment of ethical, legal and social implications of human genome editing (7.5)

It’s worth remembering that applications for gene editing are more limited that many seem to think. All the fuss about gene editing in the media gives some the impression that almost anything can be done.* That’s not the case.

One of the reasons that rare single-gene disorders with strong effects are a focus is that they are caused by a simple well-defined genetic changes. You also know they’re very dominantly genetic essentially by definition. They’re also not common.

By contrast, most common conditions involve more than a few genes (often dozens or hundreds), and many (most) conditions are not ‘mainly genetic’ in a way that could be substantially altered using genetic changes.** Both make common conditions not suitable to be tackled using gene editing.

I think it may also be worth not confusing the applications with the techniques; interested readers are referred to the Footnotes.***

The full report is available as a PDF file from the National Academies of Science, Engineering and Medicine website, from their Human gene editing initiative page.


* As a rather silly example, I tried—seemingly in vain—to get a ‘futurist’ professor acknowledge that you can’t gene edit for things that are not dominantly genetic, a concept that I would have thought simple to grasp, but he just didn’t seem to want to understand. Over-playing of what gene editing includes some people who really ought to know better.

A related but different issue is that what works in mice may not necessarily work in humans. (There is a recent example of this, too, but I should probably leave that for another blog post. This is getting longer than the simple heads-up it’s meant to be… And like too many of my posts, it’s in danger of having more in the Footnotes than the piece itself…!)

** This is rather tricky, and I’m glossing over a large number of different issues in a short sentence to make it simple for readers. Bottom line for the point I’m trying to make is to not expect as many things to be tackled with gene editing as you might think – for now it’ll be mainly rare single-gene diseases, and perhaps specific types of cancer that are characterised by particular genetic changes in the tumour.

*** I need more time to think about this, but I wonder if the document might have been better titled something along the lines of ‘Recommendations for human gene therapy’ — focusing on the thing done, rather than using a title relating (mostly) on some ways (techniques) that it might be done. (This same issue affects discussions about GMOs.)

As a simple example, a recent study reported using genetically-modified bacteria to target human colon cancer tumors planted into mouse intestinal walls. A gene was introduced into Salmonella bacteria (with the Salmonella toxin disabled), that produced a protein that triggered an immune response, successfully encouraging the mice’s immune system to destroy the tumours. That case seems fairly straight-forward as you are not trying to “supplement” a faulty mouse gene, but it’s not difficult to imagine others cases that try affect gene editing-at-small-distance, with the gene in nearby bacteria rather than human cells. The point here is that if you try define things by specific techniques, you will always find discover approaches are not covered, particularly because science keeps developing new ways of doing things. (Ditto for GMOs.)

Featured image

Gene therapy using an Adenovirus to get the gene into the cells. Source: Wikipedia, public domain.