**4. Consent**

This section considers three issues concerning consent in data-intensive medicine for rare disease. The first concern is that it is impossible to fully specify all the potential users and uses of patient data at the time of collection. A common solution when seeking consent to research and sharing of samples and data is seeking broad consent to future not-fully-specified uses accompanied with ongoing governance. A second issue is re-use or sharing of legacy collections of samples and data that have significant scientific and societal value but where the original consent is absent or is silent about key matters. A third issue is consent in the pediatric context, which raises special concerns about capacity, protection from harm and exploitation, inclusion, and shared decision-making. Arguably, the tension between promoting science and respecting individual autonomy is greater in the rare disease context.

#### **4.1 Broad consent**

The International Rare Diseases Research Consortium (IRDiRC) in collaboration with the GA4GH has shared template consent clauses for rare disease research [32]. These clauses emphasize some of the special characteristics of rare disease research, such as the collection of photos and videos of patient phenotypes, the participation of and feedback of health findings to family members, as well as the imperative of international data sharing to support both research as well as to match patients to inform diagnoses. This initiative also demonstrates the importance of engaging patients in the development of research governance and consents.

Biobanking and data sharing aim to make samples and data available for research that cannot be fully specified at the time of recruitment and collection. This present risks to patient autonomy: how much information can and should be provided at the time of consent? What kinds of meaningful choices can and should be offered to patients about who can access their samples and data? As samples and data are typically stored for long periods of time, can and should patients be able to withdraw consent or change their preferences over time? Broad consent—consent to notfully-specified research uses coupled with ongoing governance—has been adopted in many research contexts internationally, is now expressly permitted under the US Common Rule and recognized under the EU GDPR (rec 33) [33].

Especially where samples and data are collected in clinical care contexts, there is concern that sharing those samples and data for research may be done coercively, or that patients may have limited knowledge or comprehension. Moreover, rare disease patients may see data sharing as a necessity for receiving a diagnosis, or to advance research on a cure, and thus may feel compelled to forgo their privacy. Where consents cover a broad set of purposes, this can be seen as coercively tying purposes together, unless patients are given granular choices. But these kinds of arguments can result in inefficient sharing and use of data that precludes effective care and research, and that contradicts the wishes of many rare disease patients. Broad consent may be especially important in rare disease, given the scarcity of data and the risk of losing that data if every subsequent use is subject to re-consent. There is also an argument that rare disease data, again considering its scarcity, should be made available for a wide range of purposes, including diagnostic matchmaking, research to discover new biomarkers, natural history studies to better understand the nature

of a rare disease, and the recruitment of individuals into precision medicine clinical trials. Generating multiple siloed resources for multiple different purposes is simply not feasible.

#### **4.2 Legacy collections**

Legacy collections of samples and data are those collected without consent or without consent covering core consent elements required to conduct research or data sharing. These collections present an ethical dilemma: they often continue to have great scientific and societal value if shared and used for research, but the consent to do so is missing or insufficient. This is particularly a problem in biobanking, where samples are often collected many years before they are able to be distributed, aggregated, or analyzed. As the years pass, scientific and data sharing practices can change, and regulatory and ethical frameworks can evolve. As a result, the existing consents may become insufficient. One could argue the ethical dilemma is even more pressing in rare disease contexts, because of the high unmet need for research into novel diagnostics and therapies, as well as the associated practical difficulties of recruiting geographically disperse patients and collecting samples.

Solutions have been developed for legacy collections that aim to strike an appropriate balance between making them available for research, while also making best efforts to communicate with and respect the expectations of patients. When seeking to study or share legacy collections, an important starting point is to assess the existing consent materials (if applicable) [34]. If core elements of consent are met, then the research may be able to proceed. If the consent is silent on the desired research or data sharing, then patients should be recontacted to renew their consent. If permitted by applicable norms, it may also be sufficient to notify the patients and provide them with an opportunity to opt-out. In many cases, however, re-consent or re-contact will be impossible if patients can no longer be found. In such cases, some jurisdictions allow research ethics committees (RECs) to alter or waive consent requirements, as long as certain conditions are met. In Canada for example, a consent alteration/waiver is available where research is minimal risk, consent is impracticable, the alteration/waiver will not adversely harm individuals, there are appropriate safeguards in place, and there has been no clear refusal by the individual (art 3.7B (samples); art 5.5A (identifiable information)) [35].

Where consent is silent, an ethics waiver may be more easily justified. Where consent makes a specific commitment (e.g., guarantees data will be kept confidential, or will only be used for a specific research project), it may be harder to justify a waiver. Because consent is often take-it-or-leave-it, however, it is not necessarily clear if the commitment was determinant to the patient's decision, or what the patient would have preferred. An active refusal by the patient to participate in research or data sharing is a more clear-cut case [36]. For example, the patient may have been offered the option to participate in research or to share their data and may have refused. Indeed, the revised US Common Rule for research on human subjects in the US prohibits use of an ethics waiver when a patient has rejected a broad consent (§46.116) [37]. Practically, however, given the limits of tracking systems, it is unclear how these refusals can be tracked and respected over time.

Typically, ethics waivers are used to approve a specific research project using legacy collections, but in some cases, they have also been used to approve the deposit of data into international databases for onward sharing. The GA4GH recommends ethics waivers under certain conditions for international sharing of genomic and health-related data [34].

A final consideration offering some additional flexibility in rare disease contexts is patient involvement. Patient groups or representatives may lend moral support

to a particular interpretation of an existing consent, or to a particular decision to re-use or share legacy collections with an ethics waiver. This can help to alleviate uncertainty over what patients would have wanted in cases of uncertainty. If core consent elements are legally required, however, such modifications will not be possible.
