HOW CAN GETTING YOUR GENOME SEQUENCED IMPROVE YOUR HEALTH?
Although genomics can be useful for people with diseases, can it be useful for the healthy person? Many people believe that the answer is no. They feel it is too hard to predict disease risk from this information and raise concerns about the accuracy of both the technology and interpretation. They also worry that people will receive the information and worry excessively about their potential disease risk.
Others, however, have a different opinion. They feel that useful information can be extracted from a person’s genome, and that this can be used to help guide a person’s health care. A compelling argument is that family history is widely used in medical care. Based on family history, people are commonly placed on the alert for certain diseases, and diet and physical activity programs are often recommended. Shouldn’t a person’s genome sequence be better than family history?
The answer is certainly, “Yes,” even if we are not perfect at interpreting a person’s genome. We all have a multitude of variants in our DNA, many of which may cause disease—that is to say no one has the perfect genome. In fact, we all have at least 100 variants that cause a gene to be inactivated, although in the majority of these cases a good copy still exists. Methods have been established to analyze a person’s genome to find mutations that are predictive of disease. A version we use is summarized in Figure 1. By analyzing a person’s genome in detail we find several types of mutations:
- Those that are in a gene that previously has been shown to cause disease at very high occurrence. Furthermore, the actual mutation is very likely to interfere with the normal function of the gene and cause a disease. The genome is searched for all variants that cause disease at high Particular attention is devoted to variants in genes that may cause disease based on a person’s family history,
- Those that lie in a gene known to cause a disease, but the mutations are new or Computer algorithms suggest the mutation may be damaging, but it is unclear whether these mutations are disease- causing. These mutations are called “variants of unknown significance” (or VUSes).
- “Carrier mutations” in which the mutation itself is unlikely to cause disease but could be passed on to cause disease in a person’s children and grandchildren (e.g., recessive trait or lowly penetrant dominant one).
- Pharmacogenetic variants, which predict the individual’s response to the level of a drug or potential adverse side effect for a particular drug.
- Finally, it is possible to estimate the risk of complex diseases by aggregating the individual contributions of the many genetic loci that have been associated to the disease. This analysis forms the Risk-O- Gram described earlie.
Figure 1. Strategy for analyzing a healthy person’s genome. Variants identified by genome sequencing are analyzed for those that (a) reside in genes in which mutation in just one gene copy can cause disease; (b) are passed on to children and when combined with a mutation in the second gene copy may cause disease; (c) may affect response to certain drugs; and (d) have small effect but in aggregate with other changes can be associated with complex disease. For scenario a and b, sometimes the variants are “known” to be associated with disease and sometimes they occur in the same gene but are new and their significance in causing disease is not clear. These variants of unknown significance (VUS) can be difficult to interpret.
Examples of mutations that lead to disease with very high frequency are known disease-causing mutations in the BRCA1, BRCA2, and SHDB genes and the early Alzheimer’s genes (e.g., APP, presenilins). For BRCA1 and BRCA2, there are many known disease-causing mutations in these genes; women harboring these variants have a greater than 80% chance of developing breast or ovarian cancer. For mutations in SHDB, there is a high chance of developing paraganglioma. Although one might think that these mutations should be evident from a person’s family history, this is not always the case. In fact, in a recent study that we performed, a woman who did not have a family history of breast cancer was identified with an inactivating mutation in BRCA1. Although the reason for this is not clear; it is possible that she inherited the mutation from her father who is less susceptible to the disease or the mutation is “de novo.” This person learned about the mutation from her genome sequence and had surgery based on that information. She would not have known about this mutation had she not had her genome sequenced.
In addition to the BRCA, SHDB, and Alzheimer’s disease genes, many other mutations are highly predictive of diseases. The American College of Medical Genetics has identified 56 genes for which it is recommended that known disease-causing mutations from a sequencing study be reported back to the subject’s physician, who, in turn, can consult with the patients based on their wishes. This list can be easily expanded to include many more genes with known disease-causing mutations and will continue to grow as more disease-causing mutations are found.
“Variants of unknown significance” (VUS) is the second category of mutation. Our work indicates that people typically have one to three of these that might directly contribute to disease, and the number is even higher if carrier status is considered. It is difficult to know whether these mutations will cause disease. Their identification, however, can lead to follow-up tests that can determine whether the person might be likely to acquire this disease. For example, in the author’s genome, a variant in a TERT gene suggested that he may be susceptible to aplastic anemia, a loss of blood cells. The activity of the gene (which adds sequences to the ends of chromosomes) and levels of blood cells were analyzed in subsequent tests. The subject was found to have slightly shorter sequences on the ends of his chromosomes; however, his levels of blood cells have been normal and thus, do not currently manifest disease symptoms. For diseases that are associated with increased age, patients who have VUS can be on guard for acquisition of these diseases.
Finally, it is possible to infer risk of complex diseases from variants of low effect. In most cases, these may increase a person’s risk from a very low value, such as 0.1% to a higher, but still overall low value, such as 1%. Although this is a 10-fold increase, the person is still at very low risk for developing the disease. Presently, the accuracy of these tests is difficult to ascertain. However, they often match family history. Moreover, in the author’s case a prediction was made that proved to be accurate. My genome sequence predicted that I am at risk for type 2 diabetes, a condition not known to run in my family. I acquired the disease after a respiratory infection, and because of awareness of my genetic predisposition, detected it early and successfully managed it, at least initially (Figure 2). Like VUSes, complex disease risk can be a useful prognostic indicator of diseases for which to be on the alert, much like family history.
Figure 2. My genome was sequenced and I was found to be at risk for type 2 diabetes based on this Risk-O-Gram. My levels of glucose and glycosylated hemoglobin (HbA1c; an indicator of steady state glucose) were followed—elevated levels of glucose and HbA1c (not shown) were evident after a respiratory virus (RSV) infection. High glucose was controlled through diet and exercise initially, although elevated glucose and HbA1c levels returned two years later (not shown).
Genome sequencing for the healthy person already has been shown to have value in a number of cases: for predicting disease risk, catching disease early, and avoiding drugs that may lead to adverse side effects. As such, it can help facilitate a shift in the way medicine is currently practiced, from treatment after manifestation of the disease to taking a more preventative approach. Moreover, as our interpretation of the genome improves, the value of genome sequencing will continue to increase further.