Introduction

At the end of an IVF cycle, a couple may have multiple embryos to select from. By using a genetic risk score (GRS), they can order these embryos by their estimated lifetime disease risk. Orchid offers GRSs for many common chronic diseases, therefore the embryo with the lowest risk of one disease may not have the lowest risk of another, so ordering embryos per disease typically is not enough to make a selection. Instead, the couple could also consider the magnitude of the risk – each embryo’s lifetime risk for each disease. However, considering GRS alone without family history may lead to an inaccurate estimation of this risk.

For example, If one of the parents has been diagnosed with a disease, their embryos’ absolute lifetime risk may increase many-fold. To get our most accurate estimations, the Family History Risk Calculator combines both GRS and family history into one estimation of absolute lifetime risk. Using this tool, couples can now confidently compare their embryos’ lifetime risks across multiple diseases. 

An Example

To better understand how to use this tool, let’s walk through an example. My mother passed away from breast cancer in her 60s, and my wife has a diagnosis of bipolar disorder. My wife and I want to better understand how this family history will impact our future child’s risk of these diseases. To do this, we use the Family History Risk Calculator.

We start by calculating each embryo’s risk of bipolar disorder. We start with the baseline population average of 2.8%. This is the risk of a child conceived at random – without any known disease history or genetic information.

Adding my wife’s diagnosis, we see our future child’s risk increases over 3.5x to about 10%.

Although all of our embryos share a large percentage of their DNA, each embryo has a unique combination and its therefore its own genetic predispositions.

Below, we see that Embryo 6 has 25% less risk than typical for an embryo with our family history, whereas Embryo 8 has a 31% greater risk.

Finally, we compare the lifetime risk of all embryos side by side.

Now let’s calculate each embryo’s risk of breast cancer. Following the same steps as before, we get the following results.

These results show that we can reduce our future child’s absolute risk of bipolar disorder slightly more than their risk of breast cancer. This is a very different result than if we hadn’t included our family history – in which case we would have expected significantly less absolute risk reduction of bipolar disorder (as you can see below).

Next Steps

• Want to better understand the risk reduction embryo selection can achieve? Check out our public risk calculator here.

• Ready to learn how to use Orchid for embryo screening? Just have questions? Get started with a free consultation with one of our board-certified genetic counselors.

How It Works

Most common diseases are the result of both environmental and genetic factors. For example, most of us know that a healthy diet and frequent exercise significantly decreases our risk of type 2 diabetes. But if two people follow the same eating and exercising routine, their underlying genetic predispositions may result in one developing type 2 diabetes and the other not.

Linking genes to increased risk for certain diseases is an active field of research. Many papers have already demonstrated millions of links (e.g. there are over a million alleles linked to breast cancer and nearly 300,000 to bipolar). However, there is still more research to be done, meaning we understand some but not all the genes that play a role in one’s predisposition to a disease. 

To estimate an embryo’s risk of a disease, we combine two methods: genetic risk scores and family history. On one hand, our genetic risk scores take the links researchers have discovered for a specific disease and combine them into one equation. We then take your embryo’s genomic data and plug it into this equation to get an estimated risk. On the other hand, family history helps us estimate the genetic risk due to genes that have not yet been linked to the disease. Together, these two methods provide our most accurate estimate of your embryo’s risk.

Going deeper

Whitepaper

These estimates were produced using data from the UK Biobank, an anonymous national database which links individual disease history and their genetic sequences.  You can read about how we modeled disease status and family histories in this whitepaper.

How do you validate your GRSs?

Our GRSs are built on top of large Genome Wide Association Studies (GWAS), like the UK Biobank. These GWASs include genomic and medical data on up to half a million individuals. To develop our GRS models, we follow machine learning best practices – training them on part of these GWAS datasets and validating their accuracy on another.

Still a natural question is: How can we be positive that our genetic risk scores are predictive without “running the experiment” – i.e., waiting until children born through embryo selection have grown up? This is a good question, and it usually arises from thinking about genetic risk scores the same way we think about pharmaceutical trials. But selecting an embryo is different from taking medication. With embryo selection, we can compare your embryo to individuals with the same genetic variants in large biobanks – where we have their full disease history. A new medication, however, is a novel intervention, so there’s no historical data to draw from. This means that our genetic risk scores are built on significantly more data than pharmaceutical trials.

‍