What are the morally relevant features of stem-cell-based embryo models?

We recently collaborated with colleagues around the world on a white paper (see BioNews 1335) setting out a foundational framework for research, technological development and regulation in relation to stem-cell-based embryo models (SCBEMs). One of the steps recommended in our white paper was the identification of features of SCBEMs that are of particular ethical concern.

What follows is our initial attempt to fulfil that recommendation. These proposals are set out in greater detail in a paper that forms part of a SCBEM-themed article collection in the journal Nature Cell Biology.

Complexity and proportionality

Ideas of 'complexity' and 'proportionality' play an important role in guidance and governance for research involving SCBEMs. Consider the following:

• The UK's Code of Practice for the Generation and Use of Human Stem-Cell-Based Embryo Models, published jointly by Cambridge Reproduction and PET (the Progress Educational Trust), proposes a 'degree of oversight' for SCBEM-related research that is 'proportionate to the complexity of the SCBEM and proportionate for each individual project' (see BioNews 1246a, 1246b and 1247).
• The latest Guidelines published by the International Society for Stem Cell Research state that 'review of research... should be proportionate to the degree of complexity of the model' (see BioNews 1289, 1293 and 1302).

However, it is not entirely clear how complexity and proportionality are to be assessed or measured in this context (see BioNews 1309).

To help address this gap, we propose three broad categories of ethical value and five categories of biological feature that – we argue – are of particular relevance to research involving SCBEMs. Our focus here is on human SCBEMs, but it is possible that (some) nonhuman SCBEMs may involve their own distinct ethical considerations.

Broad categories of ethical value

There are three broad categories of ethical value that we might want to consider in relation to both human embryos and human SCBEMs.

• Intrinsic value concerns the value, or moral status, of the entity (the embryo or SCBEM) in and of itself. This might be based on features such as the entity's capacity to feel pain, or its capacity for other forms of awareness.
• Extrinsic value concerns the value of an entity when it is considered in relation to something else. For instance, some may attribute value to human embryos because of the intention behind their creation. For others, embryos may hold symbolic value because of what they are thought to represent.
• Public trust in science is relevant in this context, because governance of research involving human embryos and related materials can be understood as a form of 'social contract' that must balance wide-ranging points of view, objectives and perceptions (see BioNews 1245).

Morally relevant features

Next, we propose five categories of biological features that are of moral relevance in relation to human SCBEMs.

• Developmental stage.
• Integration of organ systems.
• Fetal potential.
• Capacity to form functional neural circuits.
• Features of public concern.

This list – explored in greater detail below – is not exhaustive, but we hope that it might serve as an initial basis for devising more detailed oversight procedures for SCBEM-related research, and that it might help to support consistent decision-making.

These morally relevant features are not intended to serve as simplistic binary cutoff points (ie, what is allowed versus what it not allowed) or to imply a single linear scale (as though it is always apparent that one SCBEM is of greater moral concern than another). Rather, these features are intended to enable multifactorial and multidimensional evaluation.

Developmental stage

Human development involves dynamic transitions along a well-defined continuum, and progression along this trajectory is a critical feature of both human embryos and human SCBEMs. However, some SCBEMs do not adhere to canonical developmental timelines, and this makes simplistic chronological regulation impractical.

Nonetheless, we maintain that the explicit purpose of SCBEMs is to provide insight into embryonic processes. Researchers are therefore likely to be aware of which 'developmental window' they intend their SCBEM to replicate in vitro.

We propose that researchers should set out to investigate within a defined developmental window with a defined endpoint, and no further. If researchers then wish to proceed beyond the defined endpoint, they should seek additional review, with an updated developmental window proposed and associated scientific justification.

Such developmental windows need not be straightforwardly time-based (that is, defined in terms of the number of days or weeks in culture). Rather, they might be linked to the emergence of certain features or cell types, and/or to equivalent embryonic stages.

Integration of organ systems

Comparisons are often drawn between SCBEMs and organoids. Organoids model a single organ or tissue (see BioNews 1300, 1341 and 1347) whereas SCBEMs co-develop multiple tissues and replicate the developmental dynamics of embryos (see BioNews 1321a, 1321b and 1339). The distinction can become blurred, as SCBEMs transition beyond early stages and into organogenesis, and as organoids become increasingly complex or are grown as assembloids (see BioNews 1105) or 'multi-organoids' (see BioNews 1083).

Nonetheless, we maintain that multilineage SCBEMs – which comprise several organ systems, integrated into a single entity – warrant greater scrutiny than disconnected single organoids. One reason for this is that biological systems often show emergent features when their components are integrated.

Furthermore, SCBEMs that only partially model the fetal body plan (such as somitoids, axioloids or trunk-like structures) are likely to be less morally concerning than SCBEMs that recapitulate the entirety of the fetal body plan – for example, if researchers were to succeed in developing human equivalents of certain mouse SCBEMs that have fairly complete body plans (see BioNews 1157).

We therefore propose that researchers should explain which tissues/organs might – and might not – be expected to emerge during research involving SCBEMs. Researchers should also state whether any measures will be taken to minimise, or prevent, the integration and interaction of any organ systems that are present.

Fetal potential

Various arguments have been made for and against the ethical evaluation of biological systems according to developmental potential. There are many aspects to this discussion, not least because the idea of 'developmental potential' can be understood in several different ways (for example as a matter of possibility, of probability or of predisposition).

For the purposes of evaluating SCBEM-related research, we propose that it could be important to distinguish between entities that do – and do not – have the potential to develop to fetal stages.

Any such distinction should take account of those features of a SCBEM that might enable it to develop into a fetal-stage entity, including features we have already discussed above (advanced developmental stage and integration of multiple organ systems). We propose that research involving simplistic models, cultured for short periods of time, requires less stringent oversight than models which have the potential to reach highly advanced stages.

At the same time, account should be taken of external conditions that might enable prolonged development, such as a particular experimental design or bioengineering approach. This should help to ensure that in instances where SCBEMs do not seem to possess inherent fetal potential, and yet researchers still intend to supply (via external conditions) components that could enable development to fetal stages, oversight of the research will be appropriately stringent.

Capacity to form functional neural circuits

There are different scientific arguments, covering a wide range of weeks, regarding the point at which a fetus might be capable of feeling pain or possessing awareness.

For many, a key indicator is the presence of the cortex and connection of the peripheral nervous system through the spinal cord and thalamus, likely to occur at 22-24 weeks of fetal development. Others argue that because the first neural projections from the thalamus to the subcortical plate happen at 12-18 weeks, this could be a suitable cutoff point for possible pain perception based on neurological morphology. (More basic touch sensitivity is already present after around eight weeks of embryonic development).

We argue for precaution. Any experimental design involving new SCBEMs should – unless very strong reasoning is provided that justifies doing otherwise – take active steps to ensure that these SCBEMs stop short of constituting an entity that could be said to be sentient. This might mean considering whether the SCBEMs will develop central neural processing capacity or mature states of neuromuscular circuitry.

We propose that researchers should state whether and why they expect (or do not expect) their research to lead to entities which have the capacity to form neural circuits. This should include means by which the researchers will actively seek to ensure that experiments stop well short of pain perception.

Features of public concern

Public engagement and dialogue exercises in the UK (see BioNews 1234b) and in the Netherlands have found that although there is considerable excitement about SCBEM-related research and associated opportunities, many would like to see clearer regulation and increased public involvement in governance discussions.

Opinions are likely to vary substantially within and across different segments of society, but evidence so far suggests that public concern may arise if SCBEMs develop defining and identifying features of human anatomy – for example beating cardiac structures, appendages (hands/feet and digits), or the rudiments of facial structure. Public responses elicited by the appearance of such features deserve to be considered and discussed, regardless of whether or not we consider the features in question to be – in themselves – morally significant.

We propose that wherever possible, researchers who work with SCBEMs should engage in dialogue, events and exercises that include the broader public, and we propose that funding for this sort of public-facing work should be made available. Researchers should also consider how to communicate responsibly in public, and should discuss potential concerns about SCBEMs in public contexts and in a responsive and dynamic two-way manner.

Conclusion

We propose that researchers justify the generation and use of particular SCBEMs according to the presence (or absence) of the features described above, and also according to the anticipated impact of their investigations (the possible benefits and the possible risks).

We hope that our proposals might act as a launchpad for wider discussions of ethical and regulatory aspects of SCBEM-related research. As this research progresses, governance of the research must also progress, in order to ensure adequate oversight and safeguard public trust.

Just as international coordination between researchers is important for scientific progress, so too is international coordination with regard to governance. Exchanges of decision criteria and relevant expert knowledge, between oversight committees and relevant bodies in different countries, should help to promote more equitable and transparent decision-making in relation to SCBEMs.