Ethical Considerations in Clinical Proteomics

Proteins exert most biological functions in the body – so it’s no surprise that their levels vary over time and between environments, adjusting to the needs of the organism. Proteomics studies these fluctuations to determine which proteins are responsible for specific functions. Why? The main role of clinical proteomics is to identify proteins correlating with health and disease states in humans.

These proteins are so called biomarkers, and can streamline disease diagnosis and personalize treatments. To this end, state-of-the-art workflows aim to assess all proteins in a sample in an unbiased way – a difficult task, given the sheer number of proteins and their vast ranges of abundance in samples. Realizing the promise of clinical proteomics is one of the great challenges we face in analytical science.

Bringing the benefits of proteomics into the clinic requires us to attend to issues that extend beyond technology. As with most activities with implications for health and wellbeing, the broad inferential powers of proteomics come with related ethical responsibilities. Many of these, including the importance of informed consent and data security, are familiar from other biomedical and human subject-related fields. However, the capacity of proteomic profiles to broadly reflect an individual´s biological state refashions old questions relating to the types of information that can be derived – regardless of the purpose for which the information was originally collected.

We set out to illustrate this ethical dilemma by re-analyzing a previously conducted weight loss study from MS-based proteomics of human plasma (1). Protein levels varied widely between people, but were stable between sampling periods in individuals. So we were able to identify individuals in our cohort by their characteristic protein levels and alleles. We were also able to measure levels of proteins that vary between different ethnicities and genders, as well as proteins associated with pregnancy. Clearly, this is all potentially sensitive information.

We also found that proteomic profiles contain “incidental findings” – clinically relevant information on disease states other than the one for which the sample was taken. For example, glycation patterns and levels of apolipoproteins could indicate somebody’s risk of developing diabetes or Alzheimer’s disease.

The ease with which we obtained this information from existing datasets convinced us of the need to discuss the ethical implications of proteomics. To kickstart this conversation, we applied systematic review methods adapted for qualitative data to capture nascent discussions of these ethical issues in the literature (2). Of the 16 relevant articles we identified, most were out of date or contained only cursory references to ethical topics. 

We also looked to the literature on clinical genomics. How did that field deal with similar issues? First, there appears to be consensus in clinical genomics that actionable information relevant to a person’s health should – or must – be returned to them. Moreover, worries about how well sample genomes represent global human diversity appear to be just as prevalent in genomics as in proteomics.

But there are also clear differences between the fields that prevent us from adopting genomics’ guidelines wholesale. Most important in our eyes: the potential for proteomic profiles, in contrast to genomic data, to capture phenotypic information in samples and track changes over time. This offers clear opportunities, such as periodic proteomic profile collection with healthcare providers for diagnostic, treatment, and prevention purposes. Such preventative measures give us a chance to maximize the benefits of our science (for example by empowering us to enhance disease resilience) – another ethical consideration we must bear in mind.

There are clear potentials and pitfalls when we think about proteomics from an ethical perspective. Our main contribution, we hope, is stimulating an early and comprehensive conversation on the topic. The experiences of clinical genomics and other biomedical fields show that legislation- and guideline-based regulation will eventually come. And, though this may come from external sources, we argue that clinical proteomics would benefit from self-regulation through open discussions and dissemination of consensus-based professional guidelines. This is the wisest way by which we can fulfil the promise of clinical proteomics while protecting patients from ethical issues in the long term.