Q&A with Pilar Ossorio: Is whole genome sequencing an effective diagnostic tool?

As the cost of whole genome sequencing (WGS) approaches $1,000, the possibility of using it to help diagnose patients becomes economically feasible. But is it the right tool for doctors?

At an international workshop in Montreal, more than 30 scientists, bioethicists and doctors presented and discussed the challenges and potential consequences of the implementation of clinical WGS. The effort resulted in an influential November 2013 Public Library of Science (PLoS) Bioethics paper, “Reflections on the Cost of ‘Low Cost’ Whole Genome Sequencing: Framing the Health Policy Debate.”

Pilar Ossorio, UW–Madison professor of law, Morgridge Institute Ethicist in Residence and a co-author of the paper, discusses some of their conclusions.

Imagine you were talking to one of your law school colleagues. How would you describe your recent paper?

My elevator speech would be that we urged caution in the kinds of claims made about the potential benefits of WGS as a clinical tool.

We have a lot of reasons to believe that just screening everyone for everything is a bad idea. There is a whole world of science in understanding what actually works as a clinical application. Rather than saying ‘We should just do this,’ we need to study how to do it and when to do it.

What would be an example where knowing a person’s gene sequence would not be useful?

There are variants of two genes that strongly influence how people metabolize Warfarin (a drug used to prevent blood clotting). Warfarin dosing is a tricky thing. Overdose people and they end up having major bleeding events, they could even die. Under-dose them and they don’t get the benefit that they need.

The people who discovered the genes that influence Warfarin response are just baffled by why doctors who prescribe Warfarin aren’t ordering up genetic tests.

There are a couple of reasons. A lot of doctors don’t feel comfortable because they don’t know the genetics well enough.

The other thing is doctors saying, “I don’t need to know this patient’s variants because I know how to dose this drug.” To a genome scientist, that sounds arrogant and misguided.

But, recently they published the biggest, truly randomized, controlled, clinical trial on Warfarin dosing. For one arm of the trial the docs used the best clinical algorithms without any additional genetic data, and in the other arm those same algorithms with the genetic information. Having the genetic information didn’t improve outcomes. Turns out, doctors do know how to dose Warfarin.

In the lab, these variants are important to understanding the biology of drug metabolism. But they don’t necessarily confer benefit in a clinical setting.

Can you give an example of a clinical application where WGS would be beneficial?

An example described in our paper is tumor sequencing. A lot of times, the way we categorize cancers might not be the best way for deciding how to treat them.

A pancreatic cancer and a breast cancer may have similar biochemical pathways that have gone awry. You might be able to learn that by doing transcriptomics (sequencing those genes actively expressed in a given cell type), or some other ‘-omic’ analysis on the cells. Ultimately, if it’s not WGS, some other ‘-omics’ will be a great clinical tool in understanding how to treat people with cancers.

Some people view genomic sequencing as a way to unlock secrets about themselves. Do you feel there is more good or harm in allowing an individual to sequence themselves?

I don’t think that’s the right question. There isn’t anyone who’s saying you shouldn’t go into the lab and sequence yourself tomorrow if you want to.

The question you might have been getting at is empowerment — the idea that more data is always better. That isn’t necessarily true. Not all information is really knowledge, or wisdom. For most people, if you gave them a readout of their sequence — that’s data — they wouldn’t know how to interpret it, because there would be no information.

In the kinds of interpretation we have right now, there’s not a huge amount of valuable knowledge. But I don’t think it’s bad if someone has that. Some people might misinterpret it, but there would be others who might use it wisely.

What effect do you see the PLoS paper having on the implementation of WGS as a clinical tool?

My hope is it will give science funders and the hospitals who are feeling pressure to start offering these services the ammunition to say, ‘Wait a minute, I don’t have to be offering everyone WGS. We can study this and see where it makes sense to offer it.’