Precision healthcare and genomics: The delicate balance between what to ask and what to tell

Precision medicine aims to provide optimised treatments for disease based on more comprehensive and fine-grained information about the patient and their condition.

Such personalised treatment offers improved outcomes for patients while decreasing costs of care, by identifying targeted treatment, minimising unproductive procedures and avoiding adverse responses to therapy.

With the increasing success and accelerating adoption of precision medicine practices, there is growing interest in extending these principles more broadly in preventative healthcare.

This emerging domain, termed precision healthcare, aims to reduce the incidence and impact of disease through detecting conditions early, stratifying at-risk individuals, and providing individualised health advice and feedback.

A diversity of technologies is converging to enable precision medicine and precision healthcare, including decision-support networks, machine learning, and real-time biomonitoring.

However, the core technology and information base for personalisation is genomic. The capturing of a genome sequence – the genetic information used to to build and maintain a person’s body – effectively enables tens of thousands of individual tests to be conducted at once.

In turn, this permits this vast dataset to be interpreted to diagnose, or establish carrier status and predisposition of virtually every known genetic condition, as well as our response to disease (genetic or otherwise) and treatment.

In personalised medicine, a particular answer is sought to a clinical question, and the plethora of immediately accessible information from the genome has proved transformational in disease diagnosis and targeted cancer therapy.

However, in precision healthcare, the abundance of information presents the conundrum of how much information should be returned to the individual, and when.

On one hand, it seems intuitive to return as much information as possible on the premise that forewarned is forearmed, regardless of whether a pre-empted condition actually manifests.

As ever, the reality in genomics is more complicated. For the most part, there isn’t a straightforward correlation of cause and effect between a particular genetic variant and a given condition.

Outside of severe disorders involving variants in a single gene such as cystic fibrosis where particular variants consistently predict disease, the majority of changes in disease-associated genes results in variable degrees of effect.

In many cases, a gene change may not manifest in a clinically-observable disease. A large number of tests could be derived from a genome, as routinely seen and reported by both clinically-accredited and online genomic testing companies, so there is the possibility to return dozens of seemingly positive results that suggest elevated risk of various diseases.

At its extreme, this could lead to needless anxiety for the individual and their family, and countless unproductive follow-on tests. Ironically, this is exactly the situation that genome-informed medicine seeks to avoid.

As has been commented elsewhere, the consequences of over-diagnosis by genome testing could prove not only costly to the health system, but also cast doubt over the genuine benefits that widespread genomic testing could have for the population.

As genomics becomes increasingly accessible in the broader population, it will be critical to balance the overall clinical benefit with the cost of over-diagnosis.

Part of the solution rests in furthering the understanding of the factors that influence the way a disease manifests when a person has a variant in a disease-associated gene, which will better inform their life-long risk.

However, the greatest impact on minimising over-diagnosis is likely to result from a healthcare system that emphasises and enables the use of genomic information within the clinical context of the patient or their family.

Indeed, a future can be envisaged where genomic results are predominantly returned as part of any given clinical scenario; for example, results pertaining to cardiovascular conditions would be returned in the context of a cardiology clinic appointment or referral and similarly hereditary cancer results returned in a familial genetics clinic.

This approach maximises the utility of genomic information and simultaneously provides a precision care pathway with potential for both improved outcomes and increased efficiency across the health system.