The inclusion criteria of this systematic review were as follows: articles in the adjacent fields of regenerative medicine, tissue-engineering, 3D bioprinting, cell-based therapy, organoid technology, synthetic biology, and bio-artificial organs describing ethical points to consider (issues, questions or challenges) for early-phase clinical trials. Letters to the editor, editorials and opinion articles were included as non-research manuscripts. Articles that only discussed pre-clinical research were excluded from our sample. For reasons of feasibility, articles discussing transplantation of non-biological medical devices instead of biological materials (e.g., pacemakers, blood glucose monitors, insulin pumps, or cardioverter defibrators) and articles discussing engineering of specific tissues for purposes other than organ transplantation (e.g., engineering of brains and reproductive organs for research purposes) were excluded. Finally, conference abstracts and articles were excluded (Table 1).

The method of qualitative content analysis was employed (13). Qualitative content analysis is an inductive (bottom-up) approach to categorize ethical considerations and to develop themes within a coding frame. One researcher (DJ) conducted the analyses. Firstly, codes were assigned to all the considerations mentioned in each publication. Secondly, themes (e.g., patient selection) were created out of these codes by DJ. Thirdly, DJ, EM and EB discussed whether the created words describing the themes were representative of the codes until agreement was reached. Finally, a coding framework was built out of the identified themes. The coding framework was used to systematically keep track of ethical considerations mentioned per article.

We did not conduct a quality appraisal procedure, as there are no suitable criteria for appraisal of the quality of the literature included. This is a well-documented limitation of systematic reviews of (bio) ethical literature (14, 15).

Six themes were identified: cell source, risk-benefit assessment, patient selection, trial design, informed consent, and oversight and accountability. The content of the article referring to the six identified ethical themes is summarized in Figure 2.

These six themes were found in seven different research fields (Table 2). The largest body of literature focusses on ethical considerations around early-phase trials in the field of cell-based therapy; 55 articles are published in this field, and the authoritative ISSCR guidelines are widely used (12, 16–26). There is less literature on ethical aspects of early-phase clinical trials in the field of 3D bioprinting, and organoid transplantation; seven articles were published on 3D bioprinting, three articles on bio-artificial organs, and two on organoid transplantation. Six empirical studies using questionnaires and interviews to investigate patients’ and professionals’ views on ethical considerations in early-phase clinical trials, were included. Seven papers were published in surgical journals.

53 out of 92 articles mention ethical considerations related to the sources of cells used to generate complex tissue-engineered products such as bio-artificial or 3D bio-printed organs for transplantation into humans (9, 12, 16–24, 26–68). There are four types of cell sources: 1) xenogeneic cells, 2) autologous, 3) allogeneic donor, and 4) highly manipulated or/and genetically modified cells in humans, each with their own sets of ethical considerations (Table 3).

Firstly, xenogeneic cells are associated with a risk of zoonosis (17, 20, 38, 47–49). For instance, issues related to the transmission of the infectious porcine retrovirus (PERV) from pig to human (69). Potential future patients could also reject the use of these cells to generate bio-artificial organs on religious grounds or for socio-cultural reasons (e.g., to protect animal rights/welfare) (33, 38, 48, 50, 52), even if their religious leaders take a more moderate stance (33). According to the literature, using these cells for transplantation into humans should be minimized as much as possible (12, 17, 38).

Secondly, the use of autologous cells (cells taken from the patient, who is both the donor and recipient) will make immunosuppressive therapy unnecessary (9, 16, 27–29, 33, 38–45, 68), and is perceived to carry fewer risks than the use of other cell types (33). However, challenges include the high production costs (29, 57, 70), extra surgical interventions for participants (50), the time required for their production (29, 40, 50, 57, 70), and the difficulty of standardizing manufacturing procedures (40 57, 70).

Thirdly, besides the medical risks of transplanting allogeneic donor cell (cells taken from another human being), for example developing immunological problems, use of these cells also raises relational issues (20, 27, 30, 38, 41, 43, 63, 71, 72). Relational issues include questions such as: Who is the owner of the human cells once it is separated from the body (30,38,41,43)?; Can cells from the human body be subjected to laws regarding property rights (38,43)?, and; To what extent can the donor’s privacy and confidentiality be ensured by adopting additional measures (e.g., pseudonymisation) (20, 27, 30, 38, 41, 43, 63, 71, 72). Removing the donor’s personal information is often not desirable, because subsequent research may necessitate ongoing access to the information about the cell donor’s health status requiring personal data of the donor (e.g., their name and/or address) (20,52). Further, some donors may not want their cells to become an integral, growing part of another person (12, 20, 32, 52, 73). In addition, in the course of donor cell (genetic) screening, researchers should develop a policy on whether and how incidental findings (e.g., genetic risk) will be returned to the donors and/or their relatives (12, 20, 52, 63). Donors might consider their privacy violated if scientists know their future susceptibility to genetic disorders (52). Researchers should obtain an adequate informed consent from donors to respect their autonomy (12, 20, 22, 27, 28, 34, 38, 43, 45, 52, 57, 63, 67, 72–76), and give them some degree of insight and perhaps control over the use of donated materials by informing them about the types of incidental findings they wish to receive, future commercial applications, individualized research and therapeutic uses (12, 20, 27, 38, 43, 52, 72, 76), for instance by maintaining an ongoing dialogue with the donors (76). Moreover, to safeguard the health of the recipient over the years, it may be necessary to periodically obtain records on the medical and family history of the cell donor to monitor potential health risks, such as long-term immunological or tumorigenic reactions (12, 19, 20, 22, 27, 28, 32, 34, 35, 39, 41, 49, 51–53).

Lastly, the use of highly manipulated cells (i.e., cells of which the biological nature or structural function has been altered during the manufacturing process) and/or genetically modified cells raises safety concerns, and requires more quality controls to avoid undesired events (9, 12, 18, 20–23, 27, 28, 33, 35, 40, 50, 61, 63). For instance, these cells could have an increased risk of being tumorigenic, genetically unstable or toxic (12, 18, 35). Therefore, some authors recommend avoiding the use of manipulated cells whenever possible (e.g., tumor formation, epigenetic or genetic instability) (9, 12, 18, 20, 22). However, cell manipulation and/or genetic modification might be useful and even necessary for the generation of a bio-artificial organ (e.g., to repair disease-causing mutations) (20). Cells used in tissue-engineered products are often differentiated in vitro prior to being combined with a scaffolding material, for example collagen, to form artificial tissue, therefore tissue-engineered products are mostly classified as more than minimally manipulated (18).

One of the conditions for ethically responsible clinical research is a favorable risk-benefit ratio (Table 4). This means that the risks and burdens of trial participation must be outweighed by the expected scientific or social value and the (potential) benefits for individual participants (12, 16, 21, 23, 24, 28, 29, 32, 34, 37, 45, 50, 53, 57, 64, 66–69, 77, 78) (Figure 3). The requirement of a favorable risk-benefit is difficult to meet in early-phase research, because the potential direct benefits to individual research participants in these trials are limited and uncertain (69). In the absence of direct medical benefit, justification of exposing individual research participants to potential harms in early-phase clinical trials is sought in expected scientific and/or social value (24, 30, 50, 66, 79). These include the benefits gained for science and society: generalizable knowledge and health gains for future patients (50). Knowledge of the working mechanism and the interaction of a regenerative medicine technology with the body, gathered in early-phase clinical trials, is necessary to move these technologies to the next clinical phase of clinical development (24, 30, 50, 66, 69). The anticipated social value of bio-artificial organs is potentially high, as they are intended as cures for patients with end-stage organ failure and might be more cost-effective than existing organ replacement therapies (66). At this stage, however, the social value is highly uncertain.

Transplanting regenerative medicine into human recipients requires an irreversible (innovative) surgical procedure, which is associated with risks of harms and complications. Once the regenerative product is implanted in the body, it may not be possible to completely remove it (50). For instance, surgical removal of the product will be impractical or associated with greater risks [i.e., infections or complications of anesthesia (33)], and there will be some irreversible changes, such as scarring (50, 70). In addition, unlike non-biological medical devices, the regenerative product will most likely interact and integrate with the rest of the body, which may have uncertain, possibly unforeseeable long-term adverse health events for the recipient (16, 18, 21, 23, 24, 27, 28, 31–34, 37–40, 48, 50, 58, 62, 66–70, 72, 73, 77, 79–86).

When researchers are dealing with uncertain but potentially high risks, they are advised, before undertaking an early-phase clinical trial, to provide preclinical evidence of high internal validity (e.g., through replication) and external validity (e.g. through careful study design) (12, 16, 23, 27–29, 31, 34–37, 43, 46, 49–51, 53, 57, 59, 61, 62, 64–69, 77, 79–81, 84, 85, 87–90). Some argue that large animals should be used, because these animals can better imitate the human anatomy and/or pathology than small animals (12 81). Others recommend to involve unbiased third parties to repeat some of the research (69). Even if robust preclinical evidence is available using these strategies, some unexpected risk will inevitably remain, such as unforeseeable long-term adverse health events for the recipient. Researchers should be aware that preclinical evidence from animal models may not correctly predict the duration, function and interaction that occur in a human body (16, 24, 27, 31, 34, 37, 39, 50, 65, 68, 79–82). In addition, the personalization of regenerative medicine makes the product variable, therefore, the quality control and safety requirements of mass manufacturing for external validity do not apply (32, 34, 35, 48). A major benefit of personalization, however, is that it may take away or reduce the need for the use of life-long immunosuppressive therapy for recipients, and avoid well-known side effects such as infections and nephropathy (45, 69).

To detect health risks associated with potential long-term adverse events, such as genetic instability, undirected or uncontrolled cell growth, research participants must be carefully monitored (16, 19, 21, 23, 24, 28, 29, 32, 34, 42, 46, 50, 58, 64, 67–70, 81–83, 85), with long-term follow-up (12, 19, 21, 23, 27–29, 32, 34, 35, 37, 38, 40, 46, 50, 51, 53, 62, 66–70, 73, 79, 81, 85, 87, 91, 92). On the one hand, intensive monitoring may be perceived as reassuring or beneficial by research participants (50, 83, 93). On the other hand, possible life-long follow-up could also be burdensome for participants (50). Given the complexity of tissue-engineered products, clinical teams conducting these studies should have experience with other regenerative medicine therapies (e.g., cell-based therapy) and with post-trial follow-up care (81).

In the patient selection procedure, a new kind of trade-off has to be made: against enormous benefits stand potentially large risks (e.g., tumour formation). Selection of patients in early-phase clinical trials is a major ethical theme in the literature (12, 27, 31–34, 37, 42, 43, 45, 48, 50, 66, 67, 69, 70, 77, 81, 82, 94). Potential target groups can be divided into 5 categories: healthy individuals, individuals at risk, children, patient with early-stage disease and patients with end-stage disease (Table 5). First, it is considered unacceptable to ask 1) healthy individuals for clinical studies of regenerative medicine applications, especially of tissue-engineered products which are designed to function in the body of the recipient, given the high risks (34) and lack of benefit (32,34). Also, when regenerative applications are personalized (i.e., composed, in part, of patient-derived material), the only eligible recipient will likely be the patient themselves (48). Second, the scholarly literature contains arguments in favour of the selection of 2) individuals at risks, with 3) early-stage disease (31, 37, 48, 50, 69, 77, 81, 94), and 4) children (37, 38, 48, 78). These individuals are relatively healthy, if a regenerative medicine application is used into one of these groups, it may help 1) to achieve more health benefit, and 2) to prevent (long-term) severe complications (31, 37, 48, 50, 69, 77, 81, 94). On the other hand, it is uncertain whether these individuals, who may not have developed or will develop symptoms at all, will indeed come to suffer from end-stage organ failure at all and be in need for a transplant. At the same time, as the procedure is novel, risky and invasive, their current physical condition could worsen significantly (50). Lastly, based on the literature, the most eligible patients for early-phase clinical trials are patients who have reached the 5) end-stage of their disease (12, 27, 31, 33, 34, 42, 43, 45, 48, 66, 69, 70, 81, 82, 94). These patients no (or no longer) have effective or suitable treatment options at the time of enrolment and may be facing limited life expectancy (12, 27, 31, 33, 34, 42, 43, 45, 48, 66, 69, 70, 81, 82, 94). When serious complications occur, they may have less to lose than healthy individuals or patients with stable disease (, 12, 32–34, 48, 50, 66, 67, 77, 94). Also, for patients who have reached the end-stage of their disease, a bio-artificial organ could potentially be associated with greater medical benefits.

The ethical requirements of clear informed consent is mentioned frequently in the literature (12, 16, 17, 20–25, 27, 29, 31–34, 37, 38, 43, 45, 50–52, 59, 60, 64–69, 75–77, 79, 81, 83, 85, 89, 90, 92, 93, 97, 98). Valid informed consent requires that participants must be adequately informed about relevant aspects of research participation, including the aim of the procedure, duration of the study, their right to withdraw, and the risks and benefits implications of the trial (Table 7). Less often mentioned as an essential component in informed consent is information on the specific composition of the regenerative medicine application, although some authors find it important (33, 81, 83). One survey showed that participants want to be especially informed about issues that could directly affect their health status, such as foreseeable risks, impact on quality of life and safety measures (83). Participants are worried about the risks associated with genetic manipulation of transplantable tissue and about commercialization of cells (33, 83).

Given the lack of evidence on the risks, however, it could be difficult for researchers to provide full disclosure. Rather, participants should be made aware of the uncertainties surrounding the risks and benefits of investigational regenerative medicine technologies (20, 21, 23, 24, 32–34, 65, 72, 81, 98). Participants should be given the opportunity to consult an independent expert (33, 98), and can be offered psychological support (81), or consult a patient advocates (81), to assist them in the decision-making process (33, 60, 81, 83, 84, 98). To minimize “the therapeutic misconception,” the (sometimes) mistaken belief among research participants that they will benefit from trial participation, measures should be taken to ensure that research participants are aware of the fact that research is conducted not with the goal of providing them medical treatment, but of obtaining generalizable information (12, 16, 17, 21, 24, 25, 29, 31, 33, 37, 50, 57, 60, 64, 67, 69, 81, 93, 97, 98). Researchers should avoid presenting the potential of the product in an overly optimistic light, overestimating the possible benefits, or giving unrealistic timelines for it to reach the clinic (30). Also, to strengthen comprehension, researchers are advised to present information about the trial not only in writing but also visually (33, 60, 68, 79), encourage patients to ask questions, and avoid scientific jargon by using only simple words or easily understood terminology during the informed consent process (20–22, 29, 31, 57, 69, 93, 98). Researchers may use the teach-back method (98) or even an “exam” or questionnaire (33) to ensure that participants understand the information and make an informed choice (33, 34, 81, 98, 99). Participants must also be aware that participating in a trial might diminish their chances of getting access to future treatment opportunities (21,48,50).

A widely endorsed norm in research ethics is that participants should always have the right to withdraw their consent without negative consequences for the health care they receive. However, for participants in early-phase clinical trials of regenerative medicine technologies, withdrawal may be complicated (34). While it may be possible to withdraw from follow-up, removal of bio-artificial organs (in their entirety) may not be possible. For this reason, the opportunities for withdrawal or lack thereof, and the implications of trial participation for the future health and safety of participants must be discussed beforehand, as part of the informed consent process (34). In particular, research participants should be aware of the need for a long-term follow-up and the possibility of (long-term) adverse events (32, 34, 81). Lastly, some authors suggest informing and asking participants to provide consent for a partial or complete autopsy after their death. Obtaining this information will improve the scientific value of the study and contribute to the safety of future research participants (12).

The literature suggests that researchers should be especially careful when communicating with patients, physicians, other stakeholders, and the general public about regenerative medicine applications, as overly optimistic expectations might easily arise (17, 21, 22, 25, 29, 46, 52, 57, 62, 64, 67, 69, 78, 80, 81, 86, 90, 93, 94, 100) (Table 8). The ways in which research is represented in the media affects societal perspectives and frames policy debates (17, 67, 86, 100). In frontier science, of which research on bio-artificial organ transplantation is an example, researchers might wish or feel compelled to attract media attention to obtain financial support (17). However, they should refrain from inaccurate or incomplete representation of research, as this could ultimately have negative consequences for the advancement of the field and the integrity. For instance, researchers should avoid sharing findings with the press before peer review (17, 62) or could follow the ISSCR guidelines with regard to the conduct, public engagement and accountability of clinical trials (12, 16). In addition, researchers should be open to (international) collaboration between scientists, ethicists and clinicians (18, 22, 23, 25, 28, 35, 36, 38, 39, 41, 45, 50, 54, 57, 63–65, 73, 77, 81, 84–86, 89, 96, 100–102) and the conduct of interdisciplinary dialogues, involving scientists, such as engineers and biologists, but also patients, clinicians, policy makers, industry partners, ethicists, and the general public (17, 24, 29, 35, 37, 38, 46, 55, 64, 73, 80, 81, 84, 86, 90, 93) to encourage responsible innovation, and build and maintain long-term trust in research and the development of regenerative medicine applications. Adopting a similar strategy around bio-artificial organ technologies is highly desirable.

All research involving clinical applications of regenerative medicine must be subjected to independent RECs for approval. The main task of these oversight bodies is to ensure ethical conduct of clinical research and to protect human research participants. However, it is uncertain whether existing RECs have sufficient specific technical and clinical expertise in the fields of both organ transplantation and regenerative medicine to be able to evaluate the risks associated with bio-artificial organ transplantation trials. Multiple authors have proposed to set up specialized RECs or advisory boards with experts from various backgrounds for the evaluation of clinical trials of regenerative medicine technologies (9, 16, 19, 20, 22, 24, 28, 29, 32, 45, 46, 62–65, 67, 69, 77, 78, 80, 85, 92). These experts could assist RECs in assessing the scientific underpinnings of the clinical trial protocols and the risks of abnormal product function and proliferation (16). According to some, such specialized RECs should ideally also include lay people (21, 80). Moreover, authors recommend providing education opportunities for surgeons, researchers, nurses and ethicist in training, on the ethical aspects related to ATMPs (9, 20–22, 29, 36, 40, 45, 64, 65, 69, 70, 73, 77, 87, 92, 93).

Researchers should pre-register clinical trials and publish understandable and complete data on each step along the research pathway regardless of whether the data is positive, negative or inconclusive (12, 16, 24, 28, 29, 69, 80, 81). Being transparent about data could also inspire other researchers to go into new research directions (69).