The discovery that clustered regularly interspersed short palindromic repeats (CRISPR) and CRISPR-associated proteins found naturally in prokaryotic cells can be used to alter the genome of living organisms—including humans—has been one of the most exciting breakthroughs in biomedical science. Perhaps unsurprisingly, commentary on the ethical, social, and legal implications has focused primarily on concerns that, in the future, the technology could be used to heritably edit human germ cells, creating “designer babies”.

The current intense focus on germline genome editing—while it is understandable and necessary—diverts attention away from other pressing issues associated with clinical delivery in the non-germline context, where only the treated individual is affected. Given that attempts are already being made to use CRISPR-mediated somatic cell therapy to correct genetic mutations in diseases that are refractory to traditional therapies, somatic genome editing is an arguably more pressing issue [1].

Here we consider the regulatory conditions that need to be clarified for the medical community and the wider public to feel confident about moving somatic genome editing from bench to bedside. The regulatory parameters must be strong enough to engender confidence yet sufficiently flexible to respond to technological and social developments; a challenge that most jurisdictions are yet to fully address. We posit that these regulatory parameters should ideally be consistent globally, to provide greater certainty for the rapidly expanding industry and consistent safeguards for recipients. However, we recognize the almost insurmountable hurdles involved in achieving this end.

Responses to the regulatory uncertainties outlined in Table 1 need to be nuanced and flexible. The dangers of overly rigid regulatory responses to emerging technologies (e.g., through “command and control” legislation) have been demonstrated; for example, Canadian and Australian legislative responses to reproductive cloning, 15 years on, prevent scientists from undertaking work on some forms of mitochondrial transfer in those countries, irrespective of the potential benefits. Given the difficulty in future-proofing, mechanisms for periodic review need to be included to ensure responsiveness to evolving understanding of novel technologies and re-evaluation of their harms and benefits. This may help to prevent prematurely locking in regulatory parameters, which can otherwise remain untouched for generations, as has happened with the US Common Rule [5]. In the context of genome editing, and of any other emergent personalized therapy, regulation must be sufficiently flexible and technologically adaptive to address new technological advances and applications as they arise.

Quantifying the likely harms and benefits of novel, disruptive therapeutics such as genome editing will always involve considerable uncertainty. Within the permissible spectrum of CRISPR-mediated activities, rigorous regulatory approaches for risk and benefit assessment will be essential. This is not to assert that all regulatory insufficiencies require probing evaluations of their specific applicability and adequacy in respect of genome editing. Jurisdictions may legitimately set different thresholds for an acceptable risk–benefit ratio. Nonetheless, several features of any assessment process should be universally required. In particular, regulatory processes must be grounded in adequate scientific expertise, but take into account social as well as technical aspects of risk.

We argue that it will be challenging for clinicians or regulatory agencies to navigate the risks and benefits of translating genome editing into the clinic by relying on expert judgement alone, and that broader consultation is required. Trials of CRISPR-mediated somatic cell therapies present an opportunity to engage a wider range of stakeholders, especially patients and patient groups, in determining the acceptable thresholds for risk and benefit. Ideally, this means not only considering what magnitude of risk is acceptable, but discussing which outcomes are taken into consideration, and what counts as harm or benefit, and to whom [6].

Regulators have traditionally shown reluctance to engage widely in the course of their risk–benefit assessments. National regulators of medical products, for example, often operate in conditions of secrecy, with the European Medicines Agency, the Japanese Pharmaceuticals and Medical Devices Agency, and the Australian Therapeutic Good Administration typically only publishing information about product reviews and the evidence on which they rest after they have made made their decision. By contrast, US Food and Drug Administration Advisory Committee meetings are open to public participation.

Human research ethics committees, which take responsibility for approving and monitoring clinical trials in most countries, present a greater opportunity for broad engagement. They are generally required to have a diverse membership base, including members who are not trained in science. Nonetheless, medical and scientific members have consistently been assessed as having a greater influence on committee decision-making than others, even where community-wide ethical issues are present [7]. And astonishingly, some research ethics codes (e.g., the US Common Rule [5]) expressly prohibit considering the potential long-term social implications of a research project.

Some programmes, such as the European Patients’ Academy on Therapeutic Innovation and the US Professional Patient Advocate Institute provide training and support to enable patients to engage as advocates in the drug discovery process. But it remains unclear whether there will be longer-term institutional support for more direct involvement of patients and their representatives [8]. Incorporating improved public and stakeholder engagement mechanisms will assist in future-proofing regulation. More work needs to be done to develop appropriate mechanisms for open and inclusive multi-stakeholder dialogue.

Another complication in the further clinical development of CRISPR technology is the need to ensure that the financial benefits of clinical delivery outweigh the costs.

One way to reduce the large financial burdens associated with medical product reviews and approvals is to harmonize processes across countries. This means that if an approval is gained in one country, the same evidence can be used for approval in another (and perhaps even be automatically granted). Such harmonization would permit product development companies to reach economies of scale more quickly, reducing the cost of medical products and resulting in more favorable cost–benefit analyses. Although international harmonization is difficult to achieve, regulators should continue to work towards this goal, and considerable efforts have already been made in this regard.

Reimbursement is also a significant consideration. There are different national systems for evaluating whether a given intervention has adequate cost–benefit returns to make it worth paying for. This, when conducted by bodies like the UK National Institute for Health and Clinical Excellence, is sometimes regarded as the “fourth hurdle” of regulation. Glybera, the first gene therapy product to get European marketing authorization and approval, has recently been discontinued by its manufacturers following poor uptake by physicians resulting from cost recovery issues [9].

Patent rights add yet another layer of complexity. The development of any new technology generally prompts a raft of patent filings. In the case of genome editing this has been accompanied by a prolonged patent dispute, which has implications for both commercial and clinical applications. To date, this dispute has been resolved differently in Europe and the United States, although further judicial proceedings are likely in both jurisdictions.

Issues surrounding patent ownership and validity feed into clinical delivery. After all, a product will only be developed and marketed if there is some prospect of recouping research and development costs. The temporary exclusivity provided by patents can facilitate this. Patents can also be used as tools to minimize social harm and maximize social benefit if owners are able to negotiate ethical terms into patent license agreements. By contrast, delegation by owners to a small number of surrogate licensors—who may themselves be in competition with other developers—has the potential to create bottlenecks in rapidly developing fields [10]. In addition, patent licenses that incorporate significantly different terms for public and private end users may cause problems for academic–industry partnerships working to develop innovative CRISPR therapeutics. Ambiguities in the patent landscape surrounding CRISPR are likely to further confuse clinical translation pathways.