Ethical considerations about the collection of biological samples for genetic analysis in clinical trials

Abstract

Progress in precision medicine is being achieved through the design of clinical trials that use genetic biomarkers to guide stratification of patients and assignation to treatment or control groups. Genetic analysis of biomarkers is, therefore, essential to complete their objectives, and this involves the study of biological samples from donor patients that have been recruited according to criteria previously established in the design of the clinical trial. Nevertheless, it is becoming very common that, in the solicitation of biological samples, purposes that are beyond the objectives of the stated therapeutic trial research are introduced, like the development of ill-explained exploratory studies or the use in unspecified future research. In the digital era, the sequencing of patients’ DNA needs to be considered as a serious security matter, not only for the patients, but also for their relatives. Genetic information may be easily stored, even forever, in digital files. This engenders a permanent risk of being stolen or misused in many ways. Furthermore, re-identification of sample donors is technically feasible through their genetic data. For these reasons, genetic analysis of samples collected in clinical trials should be restricted to the accomplishment of their main objectives or well justified goals.

Randomized controlled trials are considered the gold standard research design for testing therapeutic interventions, assigning randomly individuals to treatment or control groups. Randomization helps to deal with the problem of bias by minimizing the probability that unknown factors are unevenly distributed between both groups. Participants do not respond equally, and a tested intervention may not work for a particular person due to inter-individual variability. The use of biomarkers allows for the implementation of precision medicine, the targeted treatment based on individual patient characteristics. Genetic biomarkers can aid in patient stratification (risk assessment) or as surrogate markers for treatment response identification (Mandrekar and Sargent, 2009). New clinical trial designs have been necessary to contend with biomarker-driven therapies. For example, the enrichment design whereby only biomarker-positive persons are randomly distributed between treatment and control groups; or the randomized-all design in which biomarker-negative persons are also considered (Janiaud et al., 2019). For these clinical trials biological samples must be collected to enable grouping the participants depending on biomarkers characterized by genomics, transcriptomics, or proteomics. Genomics is the most important source of genetic information, which is obtained from samples such as tissues, blood-derived circulating tumor DNA, circulating tumor cells, or exosomes (Tsimberidou et al., 2020).

Genetic data is so singular as medical information that the term “genetic exceptionalism” has been coined to deal with the potential implications it could have in socio-cultural spheres. Its singularity arises from its predictiveness, historical misuse, immutability, or hereditability (Mannette, 2021). Because the mutation rate of a genome is very low, once it is sequenced and stored in a digital file it might remain accessible indefinitely. Security breaches might occur several decades later if this information falls into the wrong hands. Furthermore, the hereditary character of the genetic information of a person implies that it is shared by relatives, and sequencing compromises their genomic privacy. One consequence of the decreasing costs of DNA sequencing is the large number of publicly available human genome sequences. This body of information may allow the inference of genomes from unsequenced individuals (Backes et al., 2018) and deductions about their phenotype (Bonomi et al., 2020). The main precaution taken to address problems such as these and to protect the privacy of trial participants is anonymization. The intention is to dissociate personal information from health data. Nevertheless, this procedure is not feasible for DNA sequences because it is possible to re-identify the individual through a partial genomic information (Bonomi et al., 2020).

There is much interest in accumulating health related records. A list kept by the US Department of Health and Human Services documents privacy breaches in the form of medical information that has been lost or stolen, usually from unencrypted computers in clinics and hospitals (U.S. Department of Health and Human Services, 2022). Although these breaches have not occurred frequently in research institutions, the increasing availability of large repositories of DNA sequences may become attractive targets for individuals with illegitimate interests. Unintended and unforeseen uses of health data collections have already been acknowledged such as forensic investigations for the identification of criminals by comparing the suspect’s DNA to sequences available in national biobanks or paternity claims in civil lawsuits (O’Doherty et al., 2016). The report National and Transnational Security Implications of Big Data in the Life Sciences, by the American Association for the Advancement of Science (AAAS), the Federal Bureau of Investigation (FBI), and the United Nations Interregional Crime and Justice Research Institute (UNICRI) warns that “sensitive, patient information could be exploited to support other activities with national security and criminal concerns such as the design of potent, custom narcotics based on specific genetic information about dopamine receptors or subversion of patient, clinical trial, and/or drug study data in furtherance of pharmaceutical espionage” (Berger and Roderick, 2014: 39). Surreptitious genetic testing, in which genetic testing is performed on a biological sample from persons accessed without their knowledge or consent, has raised many concerns. For instance, testing of samples from a romantic partner could be used unknowingly to evaluate the risk of passing problematic traits on to future children (Strand, 2016). Genetic information obtained within the context of a research study might be used for a purpose not specified at the time of collection, but which the individual finds offensive; or for the discovery by insurance companies of an increased risk of diseases like early-onset Alzheimer’s (Caulfield et al., 2008). Numerous other examples may be imagined like the analysis of genetic material from a work associate to glean information about disease risk determination that could have an impact upon employment decisions. The massive sequencing of complete genomes and the storage of the genetic information in public or private databases would render the isolation of DNA from a personal item or biological sample unnecessary.

The solicitation of biological samples from therapeutic clinical trials participants is becoming very common (Olympios et al., 2021; Perez-Gracia et al., 2017). Consequently, we are confronted with the dilemma as to how to gather the required genetic information for clinical research without exposing the participants to serious risks of inappropriate use of their genetic data. A strict approach to this problem requires that only the testing necessary for precision medicine should be allowed. Analyzed genes would enable the stratification of participants in the corresponding trial arms, but this narrow point of view would prevent other studies that are not directly related to the clinical trial, but may be useful for other diseases that share molecular pathways. Additional applications of gene sequencing include the comparison of genetic alterations identified in hereditary cancers with other neoplasia, such as mammary, ovary, or colon. Or testing the known response to pharmacological treatments developed from this genetic information, for example immunotherapies, in carriers of the mutations in other diseases. Such studies, in which validated biomarkers are analyzed (Mandrekar and Sargent, 2009), must be based on scientific knowledge or reasonable hypotheses, always related to the objectives of the clinical trial, implying that the results will possibly benefit the participants or the population that they represent.

Access to a large biorepository has many advantages, and may prompt researchers to ask for biological samples for use in exploratory studies. Frequently, a long list of genes that could be analyzed is provided to justify exploratory genetic studies, with the argument that it may indicate whether they are involved in the pathogenesis of the disease or could be used as prognostic factors. But these sorts of investigations are not the objective of clinical trials, and they should be performed as academic work. The design and implementation of studies to discover or validate biomarkers is very complex; most of them have no impact upon the routine clinical care (Ioannidis and Bossuyt, 2017). Biomarker analysis requires improvements in methodology and reporting, together with large collaborative studies, external independent validation, and pivotal randomized trials (Ioannidis and Bossuyt, 2017). Several initiatives are being developed in different regions of the world to meet these demands, such as the Nordic Society for Human Genetics and Precision Medicine, in which many actors, including universities, hospitals, patient representatives, or governments, join efforts to accelerate discovery of disease susceptibility genes and genes which afford protection from disease, among other objectives (Njølstad et al., 2019).

Therapeutic clinical trials are not conceived for these purposes, and it is doubtful that they may be accomplished with the collected samples. There are various reasons why. Results from exploratory studies are underreported, they might have no impact upon future drug development and consequently be of no benefit to current or future patients. Strong scientific rationale for exploratory studies is often lacking, they commonly fail to identify robust candidates for prediction of efficacy and toxicity, and solid biomarker identification programs are rare (Olympios et al., 2021; Perez-Gracia et al., 2017). The aforementioned risks associated with sequencing the whole genome may not be offset by the uncertain benefits of exploratory analysis. When studies are designed for the analysis and discovery of genetic biomarkers, participants know that the goal is not therapeutic and that the aim is the acquisition of knowledge about their disease. They can decide whether or not to give a sample because they understand that it may help to design new therapeutic strategies in the long term. The reasons for providing a sample are very different in therapeutic clinical trials and academic studies, and they should not be mixed.

Although it might be expected that people are willing to help with biomedical research, a significant number may prefer to keep their genetic information private. For instance, after the widescale genome sequencing of Icelanders, some demanded that their genetic information was removed from databases (Hauksson, 1999), or the return of samples collected from members of the Havasupai Indian tribe for diabetes studies when it was found that they had been analyzed for studies unrelated to diabetes (Mello and Wolf, 2010).

In conclusion, whole genome sequencing should not be performed in clinical trials for the future purpose of exploratory studies, undefined investigations or any other unjustified purposes. Institutional ethics review committees and boards, whose role is to protect the rights and welfare of participants (Anderson et al., 2023), should be especially aware of the dangers to genetic information privacy when evaluating studies that request use beyond the main objectives of the clinical trials. Important participant information, such as the right to either know or not know results of the analysis of their genome that may have implications for their health/health behaviors (Davies, 2020), may need to be supplemented by clear information about how it is not possible to guarantee the anonymity of the collected genetic information, and that this affects not only them, but also their relatives.