Individualized Interventions for Rare Genetic Conditions and the Research-Treatment Spectrum: Stakeholder Perspectives

Abstract

Purpose

Advances in the study of ultra-rare genetic conditions are leading to the development of targeted interventions developed for single or very small numbers of patients. Due to the experimental but also highly individualized nature of these interventions, they are difficult to classify cleanly as either research or clinical care. Our goal was to understand how parents, IRB members, and clinical geneticists familiar with individualized genetic interventions conceptualize these activities and their implications for the relationship between research and clinical care.

Methods

We conducted qualitative, semi-structured interviews with 28 parents, IRB members, and clinical geneticists, and derived themes from those interviews through content analysis.

Results

Individuals described individualized interventions as blurring the lines between research and clinical care and focused on hopes for therapeutic benefit and expectations for generalizability of knowledge and benefit to future patients.

Conclusion

Individualized interventions aimed at one or few patients reveal the limitations of a binary framing of research and clinical care. As a hybrid set of activities, individualized interventions suggest the need for flexibility and new frameworks that acknowledge these activities across the spectrum of research and clinical care.

INTRODUCTION

Advances in methods to manipulate genes and gene expression are leading to the development of an array of targeted therapies for individual patients with rare genetic diseases. These “individualized interventions,” sometimes referred to as “individualized therapeutics,” closely resemble innovative clinical care in which treatment is tailored for an individual’s therapeutic benefit. This approach also resembles research where the administration of a novel intervention is carried out systematically in the hopes of producing knowledge that can benefit a broader patient population. An example is Milasen, a splice-modulating antisense oligonucleotide drug tailored to a particular patient’s rare genetic mutation and diagnosis of Batten’s disease that was funded through crowdsourcing of $3 million by a single patient’s family.^1,2 In addition to the cost of development, the expense incurred for administering the drug requires diverse financial models to make drugs like this possible for one patient.^3,4 Milasen is only one of several emerging individualized interventions, including Atipeksen which targets a specific ataxia telangiectasia mutation, and Jacifusen for a specific Amyotrophic Lateral Sclerosis causing mutation in the FUS gene.⁶ The lack of evidence on effectiveness and the significant cost raise questions about who should review and fund this sort of experimental approach.

Viewpoints in the literature vary and reflect little consensus over whether individualized interventions should be categorized as research or clinical care.^5,8,14 Similar questions are raised in the literature on single patient trials, often called N-of-1 trials, which typically compare alternative treatments, involve innovation to customize care, and implement multiple interventions in hopes of finding the best therapeutic option for the single patient.^4,7

Individualized therapy trials deviate from norms for systematic investigation as they involve one or very few patients and evaluate experimental treatments which are often unapproved medical products and/or are administered outside of randomized clinical trials. Some argue that interventions can be deemed research if the intention is to create generalizable knowledge and/or if studies share attributes of traditional clinical trials such as systematic data collection or measurement of endpoints⁸; however, they acknowledge that these practices will vary due to the unique context of these applications.^9,10 Chapman¹¹ highlights similar concerns regarding therapeutic outcomes for phase one trials, as the prospect of direct benefit is small and uncertain. Noting the views of “benefit enthusiasts” for whom these trials are motivated by “therapeutic intent” and scientific purpose, she raises ethical concerns about having individuals bear the brunt of risks while allowing society to gain the potential benefit derived from research knowledge. Justifications based on possible benefit for phase one trials identified by Chapman¹¹ resemble those frequently offered for individualized interventions.

The significant cost of individualized interventions raises additional questions about sources of funding and how cost should be distributed depending on the phase of administration. Furthermore, these costs of administering the intervention as well as potential unintended consequences raise the question of how payers will approach complications that arise and/or how this affects the initiation of treatment.^7,12,13 There is a common understanding that commercial and governmental sponsors may be disincentivized due to the extremely small number of people who can expect to benefit from the specific intervention, along with the usual financial risks of bringing a therapy or group of similar therapies to market.¹⁴ Others have proposed alternative funding models for novel interventions targeted at very small numbers of patients, including individualized genetically based therapies, that involve third parties such as government groups, manufacturers, academic groups, advocacy organizations and/or multi-stakeholder partnerships.^15–18

As part of a larger study of ethical issues related to individualized interventions for ultra-rare genetic conditions, we conducted an empirical investigation of stakeholder perspectives and considered the implications of designating these interventions as research or clinical care. To investigate these perspectives, we conducted an exploratory qualitative study with key stakeholder groups. Here we present a subset of those findings, focused on how these stakeholders understand the relationship between research and clinical care.

MATERIALS AND METHODS

In this exploratory, qualitative research, we took a purposive sampling approach to recruit participants.^19,20 We were interested in investigating stakeholder perspectives that did not rely only on a hypothetical case study and specifically recruited individuals who had experience with individualized interventions. This study investigated the perspectives of three stakeholder groups: parents of affected children, clinical geneticists, and Institutional Review Board (IRB) members. Individuals were selected for their familiarity with individualized interventions, with which many had direct experience: parents had explored individualized therapies for their children; clinical geneticists had direct involvement or knowledge of individualized genetic applications; and IRB members had reviewed individual intervention protocols.

Parents were primarily recruited from those whose children were seen at Columbia University Irving Medical Center in New York City and had considered an individualized intervention in the care of their child. Clinical geneticists were recruited among those known to the research team to have familiarity with individualized intervention approaches. Similarly, IRB members were recruited from institutions where cases of individualized interventions were known to have been reviewed and/or considered. All individuals who expressed interest in participating in the study were contacted via email and phone, eligibility was confirmed, written consent was obtained, and the interview was scheduled. In-depth, semi-structured 60-minute interviews were conducted via Zoom in English. Individuals received a $50 gift card upon completion of the interview.

Interviews followed a semi-structured, open-ended format, exploring the conceptual, societal, ethical and regulatory dimensions of individualized interventions. We presented a vignette describing a pediatric case involving an individualized intervention for an ultra-rare genetic condition (see Supplementary material). The vignette approach was chosen as a useful method for examining attitudes about a complex topic and to clarify perspectives across stakeholder groups by providing a singular scenario that could be used to clarify individuals’ responses. The vignette was particularly useful in allowing individuals, particularly parents, to openly express their perspectives without having to disclose information that they might not otherwise feel comfortable revealing.²¹ The interview guide and vignette were developed by team members with expertise and experience in molecular and clinical genetics, pediatrics, psychiatry, law, health policy, anthropology, bioethics, and philosophy. Individuals were asked how individualized interventions related to clinical and research care. What should impact the distribution of resources necessary for these approaches? How risks and benefits should be weighed and to whom they would accrue? What should the role of parents be in making decisions for children? And who should be charged with the responsibilities for oversight?

We enrolled 28 purposively recruited individuals, including 11 parents, 8 clinical geneticists and 9 IRB members. The clinical geneticists were affiliated with 7 academic centers and IRB members were recruited from 9 institutions; none were from our institution. Demographic information was obtained during interviews through self-report. (Table 1) Interviews were conducted by a team member (PA or SSL) via Zoom and were digitally recorded.

We transcribed all interview data verbatim and used the qualitative software program Dedoose²² (http://www.dedoose.com) to analyze the de-identified transcripts. Using a modified grounded theory approach, a subset of the research team developed a qualitative codebook based on a priori concepts from a systematic literature review, the interview guide and in vivo coding.^22,23 Investigators began by coding a subset of transcripts to test the initial codebook, followed by inductive coding to capture new and unexpected themes (Supplementary Box 2). To maximize intra- and inter-coder reliability, our team periodically jointly coded the same data, and discrepancies were discussed and reconciled. The coding team refined the codebook and coders achieved an inter-rater reliability kappa ≥ 0.8.²⁴ In data analysis,²² the research team wrote analytic memos on the codes and their corresponding data, and noted emerging themes that were discussed in regularly scheduled team meetings. The iterative process of data collection and analysis continued until saturation was achieved.²⁵

RESULTS

Our analysis shows that study participants 1) expressed uncertainty over categorizing the interventions as research or clinical care and did not categorize individualized interventions clearly. Despite recognizing their highly experimental nature, most respondents 2) expected generalizable knowledge to result from the development and application of each individualized intervention and held a capacious view of benefit as extending to unknown future patients. Participants also acknowledged the significant cost of these approaches and 3) identified the need for new models of collaborative funding and shared responsibility across stakeholders.

I. Uncertainty over Categorizing Individualized Interventions as Research or Clinical Care

Parents, clinical geneticists and IRB members struggled with definitively categorizing individualized interventions as exclusively either research or clinical care and indicated that disentangling these two aspects was challenging if not impossible. As one clinical geneticist stated, “they’re intertwined, there’s no separation” (Geneticist #2). Most IRB members recognized the intervention as not falling neatly in the realm of research or treatment. One IRB member suggested that activities can be identified along a continuum from research to clinical care. In this framing, he placed individualized interventions nearer to research, stating: “I’m definitely saying I don’t think this would just be therapy… I think it is the beginning of research, if itself is not research” (IRB #1).

Others identified characteristics that would put the individualized intervention in the category of research, citing its experimental nature and the lack of evidence of safety or efficacy. One IRB member expressed uncertainty: “I don’t know that I would feel comfortable with saying that it’s just clinical and it’s not research based on the fact that this is an agent that has never been used in another subject before, in another human” (IRB #5).

While most IRB members recognized research dimensions of individualized approaches, parents expressed more ambivalence. Many struggled to understand the difference between categories of research and care. Parents toggled back and forth, using the terms interchangeably, and grappled with the implications for categorization.

Well, honestly, I still don’t really quite understand the whole process of individualized treatment. I mean, I have a very general understanding of it. But what I imagine it means is that you’re thinking about giving a treatment to a person based on some understanding of the underlying pathophysiology of their illness, and I guess maybe there’s some bench research…I guess you’re pulling together certain strands of research or scientific concepts to make some best guess about this thing working in this condition. I don’t know how to think about that, I don’t know if you can explain a little bit more about what’s involved in developing an individualized therapy, exactly? Is it a more clinical process? A clinical decision? I don’t know. (Parents #6)

These parents seem to recognize the weak evidence base for an intervention but assumed that bench research” and “strands of research” had contributed to the development of the intervention. They admitted to not knowing the scientific requirements and process by which an intervention would become available and, like many of our parent-interviewees, were unable to distinguish individualized interventions as research vs. clinical care, or identify the impact of the categorization on an individual patient.

II. Individualized Interventions, Expectations for Scientific Benefit, and the Relationship between Research and Clinical Care

Whereas parents struggled to answer the general question of whether individualized interventions should be categorized as research or clinical care, several expressed more certainty about the potential benefit of the intervention beyond the therapeutic impact it may have on the individual patient or, specifically, their own child. Parents expressed this in terms of potential knowledge that could be used to treat future patients:

But, at the same time, sometimes you learn from these studies… things that may impact other research, as well. Benefits are not just for the patient… I mean, they are for the specific people at the time, but I’m saying there may be other tools, or skills, or things that are learned from the research that might be helpful. (Parent #6)

One parent, who founded a non-profit organization devoted to fundraising and creating awareness of her child’s rare genetic disease, suggested that individualized approaches were critical to building a pathway towards further research:

It’s a new modality that needs to be used on [my child’s] mutation. We know under 20 people in the world that have her mutation. So, if an N-of-1 works for her, it’s not going to work for [others] in our community. However, once [my child] can show proof of concept that it works, then it’s clinical care that those other nineteen kids get. But that first tip of the sword is clinical research. (Parent #1)

The question of potential benefit for future patients was addressed by most of our IRB members, who emphasized the need for clear disclosure about the limited benefit that might be derived from untested individualized interventions. However, several IRB members conveyed a capacious understanding of benefit and included the potential for adjacent learnings about method and technique that could contribute to a body of knowledge that would be built over time. As such, they framed the benefits in terms of research knowledge even if an intervention did not achieve desired clinical endpoints. In other words, they suggested that tertiary learnings could constitute generalizable knowledge:

The definition of clinical research and human subjects research is really generalizability, and when you’re talking about an N-of-1, which possibly might turn into an N-of-2 or 3, it’s hard to call it generalizable to a wider group. But at the same time, they’re going to learn about the side effects of giving this particular agent to this particular disorder or disease. (IRB #5)

Another IRB member described generalizable knowledge gained from individualized applications by invoking the metaphor of a platform that builds critical research infrastructure: I tend to think with this kind of research that there’s

…there’s a platform being developed. And so, while this may be incredibly rare, the questions I would ask the researcher are how much of this basic platform could be used in other areas. And by doing so, I’m trying to understand from an institutional standpoint are there learnings that could apply, even if this is literally a one case situation? And I think those learnings are not just on the science, but also on the application …We’re really trying to learn whether or not it’s working, and it may take years… building those capabilities into the way these studies get done is also part of that platform. I would ask the investigator to reflect on that and come back with what they see as those elements of generalizable knowledge. I think even the vector, the delivery method, there’s elements of that that potentially are generalizable. (IRB #2)

As the gold standard for determining whether a set of activities is research, generalizability in producing information that is relevant beyond an individual patient is critical to differentiate it from medical practice. The Belmont Report underscores that the fact that “a procedure is…new, untested or different does not automatically place it in the category of research.” Rather, it requires reasonable expectations that it will contribute to a knowledge base. IRB members make the case that individualized interventions should be considered research focused on the potential scientific benefits despite their being experimental approaches without evidence of efficacy.

IRB members recognized the novel features of individualized interventions that raised challenges for classifying them on the spectrum of research and clinical care. Some characterized them in terms of transformational catalysts, critical for the developmental pathway towards effective therapies. One IRB member drew an analogy to the history of chimeric antigen receptor T-cell (CAR-T) therapy, which genetically modifies a patient’s T cells to target antigens found on the cancer cells in the patient’s body.

At the beginning, children got very, or adults too, very ill and ended up in the ICU often. There were people who died from toxicity of cytokine release syndrome. Then people learned … now we know how to manage that better, but there were the first participants in CAR-T who could have given up and now people are trying to do similar therapy across all sorts of malignancies, not just T malignancies. So, I think that if their study team, the PI at the institution, demonstrated the rationale and that there is the potential for benefit for the individual patient and then this more nebulous benefit for future patients, I think that would weigh in the plus column for approval. (IRB #8)

By building the case of future benefits through research not specifically based on the efficacy of the therapeutic agent in a given trial, IRB members argued for a more expansive view of generalizability and to broaden an understanding of benefit to unknown future patients: “We push the boundary. Hopefully, we’ll benefit this child, but surely, it can benefit someone down the road.” (IRB #9)

Several IRB members acknowledged pushing beyond what would be normally considered in scope for IRB review by integrating a more expansive view of benefit. Others, who acknowledged the danger of this grey zone, worried about the slippage between expectations for clinical and scientific benefit; in particular, the potential risks to patients and parents who may not recognize when a set of activities administered in the context of care had tipped into research.

If you don’t designate it as research, and it turns out to be more systematic than care, systematic in the data collected … were the parents adequately informed about the experimental nature of it? The intent of it? If it ends up being more investigative rather than treatment and we’re really in a regimented way, following you in a protocol. If we’re going to end up following you for 10 years and we’re going to be collecting all this information above and beyond what the treatment would be but you’re doing it because you want to see how the treatment worked … that’s when it’s like, “Oh, did it evolve into something that is more research than treatment? (IRB #4)

IRB members struggled with how to address the ethical challenges of the grey zone in which activities could not clearly be defined on one or the other side of a research-care dichotomy.

III. Funding and Access to Individualized Interventions

Parents, IRB members and clinical geneticists in our study were sensitive to the financial burden of individualized interventions on patients and families. However, perspectives varied on how the costs should be distributed depending on individuals’ framing of the interventions as research or clinical treatment. As sources of potentially generalizable knowledge, several individuals pointed to universities, academic medical centers and government funders of research as being responsible for underwriting the cost of development. This parent draws an analogy to public investment in the U.S. space program:

When you think about the Apollo missions in the ‘60s, what did they get us besides a few moon rocks? What’s the big deal about going to the moon? Yeah, national pride, but it really was the technology that was advanced decades in a short period of time through that effort. I view these projects the same way. We may be only helping one child for one or tens of millions of dollars, but the technology is going to be generalizable. (Clin Gen #5)

Others, focused on the interventions as clinical care, assumed that health insurers would cover the cost. These individuals underscored the “economic payoff for the insurer” in avoiding the long-term financial burden of paying for a lifetime of healthcare for patients with rare genetic conditions, an argument that often assumed that the treatment would prove efficacious:

If there’s normal life span, we’re talking 50, 60, 70, maybe more years where health insurance has to cover all these therapies and support and maybe later, they’re going to need assisted living that Medicaid might have to cover versus paying up front some of that cost and maybe not have to deal with that for the rest of their lives. (Parent #3)

Several individuals suggested the need for new models that would fairly distribute financial burdens and risks to different stakeholders:

I think there’s a shared responsibility here…There is $1 million sunk cost but maybe at some point there’s profit for this pharmaceutical company because they’re leveraging it to other diseases or other mutations in the same gene, right? I just don’t think it’s fair whatsoever. Even in drugs for common diseases, that’s not getting passed on to one patient. That’s getting spread across many different people that are paying for that…Should the government pay? I do think the government should contribute. (Parent #11)

A few of our clinical geneticists expressed a similar perspective, suggesting as one articulated, that to “advance science and help families,” there needs to be “synergy between the government, state authorities, industry, academia…all these stakeholders.” (Clin Gen #4)

Our respondents wrestled with the research/clinical divide in their reflections on the challenges of funding. For some, government support was seen as necessary yet only justified if the interventions were considered research and tied to reasonable expectations for common benefit:

It really depends on whether the therapy is likely to be generalizable or not. From the standpoint of looking at federal dollars to go into something like this, I do feel strongly that a therapy for an N of 1 that would be funded with federal dollars that don’t have any generalizability would be problematic for me. As it goes from no generalizability to high level of generalizability, I would have increasing acceptance of that and increasing comfort and security that that’s an appropriate way to spend money. (Clin Gen #3)

Stakeholder perspectives on responsibilities for funding were often tied to whether individuals framed the interventions primarily in terms of research vs. clinical care. However, most interviewees could not classify the activities as clearly one or the other and conveyed the need for new models for funding that take into account the novel characteristics of these approaches that fail to fit existing models.

DISCUSSION

We present results of a qualitative investigation of multiple stakeholder perspectives on individualized interventions for ultra-rare genetic conditions, with a focus on how they understand the research and clinical care aspects of the interventions. See Supplementary materials for exemplar quotes that reflect stakeholder perspectives. Our data suggest the limitations of a binary framing of research versus clinical care to reflect and address the unique characteristics of emerging approaches to treat conditions that impact one or very few people. Our stakeholders emphasized that individualized interventions defied classification as wholly research or clinical care and may be best understood as a hybrid of the two, an observation that may have ethical and regulatory implications. Recognizing this hybridity may permit the creation of governance practices that take into consideration the unique features of individualized interventions that do not align with and are constrained by a research-care dichotomy.

Recognizing the Unique Features of Individualized Interventions

Effective regulatory governance of individualized interventions requires recognizing their unique features and challenge to norms based on conventional clinical trials. For example, the requirement that research protocols outline procedures for clearly defined clinical endpoints and stopping rules is often not possible to meet in the context of individualized interventions. Rather, a clinician must rely on clinical judgment in managing an intervention, based on real-time data from the patient. When an intervention is a treatment-research hybrid, she acts as both clinician and scientist, and her actions could be considered both research that adds to knowledge about a new intervention and clinical care aimed at improving the health of the patient in front of her.

However, adhering to a framework of a research-care divide can constrain ethical guidance by forcing a “fit” rather than exploring the dimensions of decision making that must be taken into account in this hybrid role of the physician-scientist and the hybridity of individualized interventions as both research and care. For example, many of the clinical cases that are considered for ultra-rare individualized interventions involve patients who have a very poor prognosis. With the high risk of fatality, sometimes in a short period of time, patients and their parents may believe they have few options and little to lose when assessing the risks and benefits of a highly experimental and untested intervention. Addressing the ethical issues that these approaches raise may require governance that recognizes their hybridity.

A Hybrid Framing Provides Space for Innovative Flexibility

In the context of research and clinical care, understanding individualized interventions as hybrids can provide conceptual space for needed flexibility in approaching governance and oversight. Recognizing the potential for individualized interventions to catalyze the development of new treatment pathways, a hybrid framing embraces aspects of both research and clinical treatment could take into account the unique and innovative features of individualized interventions. Our study results suggest that stakeholders either do not think of individualized treatments as being either research or treatment or understand that individualized interventions push the boundaries of conventional definitions of research by generating generalizable knowledge about methodologies, such as the development of a particular platform. This generous assessment of the generalizability of what is learned during development and administration of the intervention, if not the specific intervention itself, creates the possibility for a broader understanding of benefit that extends beyond the individual patient to others with whom similar methods might be applied, including future unknown patients who may be affected by rare conditions. To address the unique ethical challenges raised by individualized interventions requires going beyond characterizing these activities as both research and clinical care and thus, acknowledging that both governing regimes may raise wholly different considerations that warrant original approaches to oversight.²⁶

Whereas classification of activities as either research or clinical care implies specific modes of governance, recognizing individualized interventions as occupying a hybrid state offers an opportunity to draw on the best aspects of both areas of practice and policy. For example, to date, there has been little exploration into how individualized interventions should be funded at the earliest phase of development.²⁷ Funding decisions influence providers’ and individuals’ willingness and ability to participate. Without innovative approaches addressing who is responsible for funding these sorts of studies, issues of justice and equity will remain, as the ability to fund studies will often determine access to care.^8,14,28 Notably, individuals underscored the significant financial burdens placed on patients and their families, and the need for new conceptual and governance models not constrained by whether interventions were classified as exclusively research or care. For example, technology platforms could be funded in addition to treatments focused on specific diseases with the idea that there are many lessons that are generalizable across a new modality of treatment. Collaborative models that combine federal funds earmarked for research with funding from insurers and private foundations could be devoted to developing platform infrastructure. Developing new models that focus on equity through recognizing the hybrid nature of individualized interventions might include foregrounding shared responsibility involving partnerships across institutions. We are currently exploring these and other implications of recognizing hybridity.

LIMITATIONS

In our qualitative study, interviewees were recruited through a purposive sampling approach that relied on personal introductions. Our decision to recruit multiple groups of stakeholders who had experience with individualized interventions for rare genetic conditions may have limited the range and diversity of individuals enrolled in our study and the generalizability of our findings. For example, the lack of diversity across racial and ethnic identity, income and education in our sample of parents may have constrained discussion of the risks of individualized interventions in important ways that would not have occurred if our sample had included populations historically underrepresented in biomedical research.

FUTURE RESEARCH

Our findings suggest the need for future research that addresses a range of scientific, social and ethical issues stemming from the hybridity of individualized interventions. An important set of methodological challenges is identifying appropriate endpoints for administering individualized interventions. In contrast to typical clinical trials, the selection of endpoints is informed by an iterative process of clinical observation and adaptation over time that requires flexibility. Models of research that demand strict, pre-determined measures to evaluate the effect of an intervention are often not possible and do not serve the best interests of the patient. Exploring the hybrid nature of these interventions would allow for research on the methodological innovations, such as longitudinal observational studies that recognize the clinical realities on the ground and a more capacious understanding of what success means in the case of one or a very small number of individuals.

Related to the hybridity of these interventions and methodological challenges is the need to explore timelines for when outcomes can be determined. Clinical trial models that are funded within a set time frame to carry out “research” activities do not adequately assign responsibility for ongoing funding of the intervention should there be a positive outcome. Furthermore, how hybrid activities should be funded remains unclear. Categorizing different phases of development and administration of interventions as they relate to safety, efficacy and dosing adjustment as research vs. clinical care have downstream consequences for funding. Research could explore the potential for a hybrid framing to alleviate inequities, perhaps allowing for a greater array of potential funding sources, such as partnership with advocacy groups. An important area for further research is investigating whether the lack of diversity in our interview participants is reflected more generally among patients who might benefit from access to individualized interventions and, if so, the implications for funding models for equity.

CONCLUSION

Individualized interventions aimed at one or few patients reveal the limitations of a binary framing of research and clinical care. Our data suggest that addressing the complex and highly specific context of individualized interventions requires an approach to ethics and regulation that is not constrained by current conceptual boundaries. Framing emerging approaches in terms of hybridity could provide an important opportunity to address the different aspects of such interventions relevant for equitable and ethical regulation.

DATA AVAILABILITY

Because the data are qualitative and context specific, it is not possible to completely remove all identifiers. Data are not available for public use as guided by the Institutional Review Board (IRB). Qualitative data are available for individual researchers upon request.