Polygenic Risk and the Development and Course of Asthma

Lancet RM

LANCETRM Abstract

 

 

Article at THE LANCET Respiratory Medicine

Press Release from The Lancet

Press Release from Duke University

Press Release from University of Otago

 

FACT SHEET (scroll down for a detailed summary)

What the study was about:

Our study asked the question: What can new discoveries from genome-wide association studies (GWAS) tell us about the development and course of asthma?

What we did:

We looked to the largest-ever genome-wide association study of asthma (that study by the GABRIEL Consortium included more than 26,000 individuals) to identify genetic variants that could be used to construct a genetic profile of asthma risk. We then turned to The Dunedin Multidisciplinary Health and Development Study, a unique cohort of 1,000 individuals who have been followed from birth through their fourth decade of life with extensive measurements of asthma and related traits. We computed a “genetic risk score” for each person based on the variants identified in GWAS.  Then, we looked at who developed asthma, when they developed asthma, and what that asthma looked like in terms of allergic response and impaired lung function. 

What we found:

(1) People with higher genetic risk scores were more likely to develop asthma and they developed asthma earlier in life.

(2) Among children who developed asthma, the ones at higher genetic risk were more likely to have persistent asthma through midlife.

(3) Genetic risk was specifically associated with allergic asthma that resulted in chronic symptoms of impaired lung function.

(4) People with higher genetic risk score developed more severe cases of asthma. As compared to people with a lower genetic risk, they were more often absent from school and work because of asthma and they were more likely to be hospitalized for asthma.

(5) The genetic risk score provided new information about asthma risk that could not be obtained from a family history.

What it means:

Genetic risks can predict which individuals with childhood-onset asthma remit and which develop life-course-persistent asthma, although these predictions are not sufficiently sensitive or specific to support immediate translation into clinical practice.

GWAS discoveries for asthma are specifically related to a childhood-onset, life-course persistent form of asthma that is allergic in nature and that leads to chronic lung function problems.

Having high genetic risk can lead to severe asthma that prevents people from going to school and/or work and that can land them in the hospital.

Genetic information has the potential to enhance clinical assessments of asthma risk, but more work is needed to determine specific circumstances where genetic information can be useful.

Additional Details

Funders of the Dunedin study and data collection include the New Zealand Health Research Council, the Otago Medical Research Foundation, and the New Zealand Asthma Foundation. The research was supported by the US National Institute on Aging, the UK Medical Research Council, and the Jacobs Foundation.

You can find out more about where I work and my collaborators at the Moffitt Caspi Lab Website

For comment, contact me at dbelsky@duke.edu

 

 

DETAILED SUMMARY

What the study was about:

Our study asked the question: What can new discoveries from genome-wide association studies (GWAS) tell us about the development and course of asthma?

Asthma is a complex syndrome with many causes and multiple courses and levels of severity. In recent years, researchers have begun to describe development and biological characteristics of asthma that they hope will be able to distinguish important subtypes asthma with different causes and/or prognosis. At the same time these researchers were making progress by taking a more complex approach to measuring asthma, another set of researchers was making progress by doing the opposite. Genetic epidemiologists compiled samples of tens of thousands of asthma patients and comparison samples of people with no asthma in order to conduct genome-wide analyses to identify genetic variants  linked with asthma. Because very large samples were needed, analyses focused on single measurements of asthma, with the largest studies analyzing simply whether person had asthma or not.

We wanted to bring these two streams of research together to see if we could link the discoveries from GWAS with specific developmental and biological characteristics of asthma. Our study tested how newly discovered genetic risks were related to the timing of asthma onset, its persistence from childhood into adulthood, whether asthma was allergic in nature or not, and whether it caused symptoms of acute and chronic impairment to lung function.

How we did it:

We started by defining a measure of genetic risk for asthma. From the big-science world of genome-wide discovery consortia, we identified the largest-ever genome-wide association study of asthma (the study included more than 26,000 individuals). We examined the results and identified markers that could be used to construct a genetic profile of asthma risk. We then turned to a unique cohort of 1,000 individuals who have been followed from birth through their fourth decade of life (age 38) with extensive clinical and interview measurements of asthma and related traits. We used the DNA bank from this cohort to genotype 15 genetic variants called single-nucleotide polymorphisms (SNPs) that measured the genetic risk factors discovered in GWAS. With this information, we computed a “genetic risk score” for each person based on the number of risk variants in that person’s genome (you can see details on the score in the Supplement). Then, we looked at who developed asthma, when they developed asthma, and what that asthma looked like in terms of allergic response and impaired lung function. Our analyses tested whether people with higher genetic scores developed asthma earlier in life, whether their asthma persisted into adulthood, and whether their asthma was allergy-related and resulted in impaired lung function.

 

What we found:

(1) People with higher genetic risk scores were more likely to develop asthma and they developed asthma earlier in life.

The figure below shows “survival curves” that illustrate this finding. Survival curves plot the proportion of the sample remaining asthma-free. As time passes (moving left to right on the x-axis), more and more people develop asthma and the curve slopes downwards. Panel A shows “survival” to asthma onset for girls (red line) and boys (blue line). You can see that boys tend to develop asthma earlier in life than girls, but asthma rates in girls eventually catch up and even surpass the boys in adulthood. Panels B, C, and D show survival curves separated by genetic risk. The high genetic risk group is graphed in red and the low genetic risk group is graphed in green. The red lines slope downward more steeply than the green lines, showing that in both men and women, being at higher genetic risk predicts an earlier onset of asthma.

LANCETRM Survival

 

(2) Among children who developed asthma, the ones at higher genetic risk were more likely to have persistent asthma through midlife. This was true of the full sample we studied, and it was also true when we looked just at the subset who developed asthma in childhood. This results is especially important because many children who develop asthma recover naturally by young adulthood, whereas other suffer asthma symptoms their whole lives. Our result suggests the possibility that a genetic test could help to predict whether a child with asthma will recover naturally, or if their symptoms will persist into adulthood.

LANCETRM LCP Asthma

The figure at left graphs the probability a person will have what we called “life-course-persistent asthma” (y-axis) against genetic risk (x-axis). In these graphs, the genetic risk score is expressed in terms of standard-deviation units. A score of zero means the person had average genetic risk. Scores above/below zero correspond to the number of standard deviations above/below the sample average. Panel A shows the genetic association with life-course-persistent asthma in the full sample. Compared to the average person, someone with a genetic risk score of 1 would be 36% more likely to develop life-course-persistent asthma. Panel B shows the genetic association with life-course persistent asthma in just the people who developed asthma in childhood. Compared to someone in this group with  average genetic risk, a person with a genetic risk score of 1 would be 20% more likely to have persistent asthma through adulthood.

(3) Genetic risk was specifically associated with allergic asthma that resulted in acute and chronic symptoms of impaired lung function.

Asthma can be roughly divided into allergic and non-allergic categories (although many other subdivisions have been suggested). We wanted to test whether recently discovered genetic risks were associated with the allergic type. To do this, we looked at two different measurements. Skin tests  (conducted at ages 13, 21, and 32 years) involved giving participants a prick on the arm with a series of different allergens. Participants who exhibited an abnormal response–the skin around the prick became swollen and inflamed–were classified as “atopic” (allergic). We also tested levels of  Immunoglobulin E (IgE) in blood (at ages 11, 21, and 32 years). IgE is elevated in individuals with allergies. People with  higher genetic risk were more likely to manifest an abnormal response to the skin testing (to be “atopic”) and had higher levels of IgE in their blood.

Impaired lung function is a key indicator of severe asthma. Active airway inflammation, a hallmark of asthma, can make the airways hyper-responsive (quick to sieze up in response to a perceived pathogen), and ultimately result in chronic and irreversibly impaired lung function. We wanted to test whether genetic risk was associated with both acute hyper-responsiveness and long-term consequences of chronic airway inflammation.

We measured  airway hyper-responsiveness by comparing lung function before and after we administered two different types of drugs, methacholine (at ages 9, 11, 13, 15, and 21) and albuterol (salbutamol, at ages 18, 26, 32, and 38 years). In people with active airway inflammation, a small quantity of methacholine can impair lung function whereas a small quantity of albuterol can (in someone with asthma) make it better. People with asthma who manifested a big change in lung function in response to either drug (impairment following methacholine or improvement following albuterol) were classified as having “airway hyper-responsiveness.”Among people with asthma, the ones who had higher genetic risk scores were more likely manifest airway hyper-responsiveness to methacholine and/or albuterol.

To look at chronic symptoms, we wanted to measure long-lasting deleterious change to the airway resulting from chronic airway inflammation. To do this, we examined lung function of participants with asthma (at ages 18, 26, 32, and 38 years) after they had been dosed with albuterol. Albuterol acts to mitigate impairment to lung function resulting from current airway inflammation. But in people with a history of chronic airway inflammation, the structure of the airway is sometimes altered in ways that make lung function impairment more permanent. Clinically, the phenomenon is known as “airway remodeling.” People with asthma whose lung function did not improve above what we defined as the lower limit of normal function even after receiving albuterol treatment were classified as having “incompletely reversible airflow obstruction” (our study’s measure of airway remodeling-type pathology).

Graphs of genetic associations with atopy, airway hyper-responsiveness, and incompletely reversible airflow obstruction can be found in the supplement (page 6).

LANCETRM Overlap

The figure at left shows overlap among the three different biological characteristics of asthma (top) and average levels of genetic risk within groups of people with defined by different constellations of the three characteristics (bottom). The top part of the figure shows that atopy (allergic reaction to skin tests for various allergens) is a central feature of most cases of asthma that also manifested airway hyper-responsiveness (AHR) and/or incompletely reversible airflow obstruction (IRAO). In other words, most people who had asthma but who were not atopic did not develop impaired lung function–at least according to the measurements made in our study. The bottom part of the figure contains 2 important pieces of information: First,  high genetic risk was a clear predictor of the most severe asthma (atopy + airway hyper-responsiveness + incompletely reversible airflow obstruction, red bar). Second, genetic risk was predictive of impaired lung function only for people with asthma who were also atopic (in fact, people who had lung function problems but who were not atopic had the lowest genetic risk of any group in the sample).

(4) Genetic risk predicted asthma-related absenteeism from school and work and asthma-related hospitalization. People with asthma who had high levels of genetic risk not only exhibited biological characteristics of severe disease, they reported disruptions to their lives consistent with suffering severe asthma. Across follow-up from childhood (beginning at age 9) through midlife (age 38),  a higher genetic risk level was associated with a 36% increase in the rate of missed school/work due to asthma and a 38% increase in risk of being hospitalized for asthma.

(5) The genetic risk score provided information about asthma risk that could not be obtained from a family history. To get a sense of how the genetic risk score we were studying compared to other tools available for measuring inherited risk for asthma, we repeated our analyses of asthma using a family-history-based risk score. From interviews with study members and their parents, we identified first and second degree relatives who had asthma (siblings, parents, aunts and uncles, and grandparents). We calculated the family history score as the proportion of relatives who had asthma, weighted to account for the fact that people share twice as much of their genomes with first (parents/siblings) as compared to second (grandparents, aunts and uncles) degree relatives. In general, the family history score and the genetic risk score performed about equally well, although the genetic risk score was more informative about biological characteristics of asthma. A comparison of results from the two scores can be seen in this table.

What was more interesting was that the two scores were essentially uncorrelated. In other words, having a strong family history of asthma and having a high genetic risk score for asthma were not related to one another. We write in the paper

“This lack of concordance between genotypic and family-based risk assessments is not unique to asthma. We previously observed this phenomena in the cases of obesity and smoking problems, and others have reported similar results for some cancers. In studies of diabetes and cancer, taking family history into account did not change effect sizes estimated for GWAS-discovered genetic risks, suggesting low correlation between genotypic and family history risk measures. In turn, this lack of correlation suggests that family history captures more than the additive contributions of common variants. Replication is needed of the non-association between genetic risk and family history, using whole-genome scores. If replicated, future genetic-discovery research in asthma could focus on genetic features of family history not detectable in GWAS, including rare variants, interactions among genes (epistasis) and, because family histories capture shared environments as well as genes, interactions between genes and environmental factors.

 

What it means:

Genetic risks can predict which individuals with childhood-onset asthma remit and which develop life-course-persistent asthma, although these predictions are not sufficiently sensitive or specific to support immediate translation into clinical practice. (This turns out to be generally true of genetic discoveries for most health conditions–see a helpful review on this topic here) GWAS research had previously linked a number of the variants included in our genetic risk score with childhood onset asthma (see 1, 2, 3, 4). The contribution of our research was (1) to confirm this finding in a prospective study of a representative sample; and (2) to further refine the phenotype from childhood onset asthma to the subtype of childhood onset asthma that persists into adulthood. A large proportion of childhood asthma cases resolve by early adulthood. Although our results do not recommend that a genetic risk score be implemented as a screening tool, genotyping could guide the enrollment of subjects into prospective research studies seeking to understand the development of life-course-persistent asthma.

GWAS discoveries for asthma are specifically related to atopic asthma characterized by airway hyperresponsiveness and leading to incompletely reversible airflow obstruction. GWAS had previously linked some discovered SNPs with atopy phenotypes. Our results confirm this finding, for the first time linke GWAS discoveries with acute hyper-responsiveness of the airways and chronic impairment to lung function, and further show that genetic effects on lung function in asthma are specific to atopic asthma.

Collectively, our findings about the timing of onset, course, and biological characteristics of asthma help to describing a case profile that can be used by studies investigating the molecular mechanisms that link DNA variants discovered in GWAS with asthma. These studies may benefit from a focus on patients with childhood-onset, life-course-persistent asthma characterized by atopy and acute and chronic impairment to lung function.

GWAS-discovered genetic risks for asthma have real-life consequences: They lead  to illness that prevents people from going to school and/or work and that can land them in the hospital. Much has been made of the small effects of genetic variants discovered in GWAS and the limited specificity of predictions about individual outcomes that can be made using GWAS discoveries. And yet results like the ones we present clearly indicate that the morbidity attributable these discoveries is far from trivial. Developing prevention and treatment strategies to combat these genetic risks is a public health priority. One way we can get started is by looking for environments that protect high genetic risk individuals from developing asthma or that, conversely, act to amplify genetic risks.

 

Please let me know if you have questions or would like a copy of the full manuscript (dbelsky@duke.edu)

 

 

 

4 thoughts on “Polygenic Risk and the Development and Course of Asthma

  1. Pingback: Asthma Management » Asthma Test To Predict Which Children Will Grow Out Of It Steps Closer

  2. Pingback: Asthma Test To Predict Which Children Will Grow Out Of It Steps Closer | Bronchitis Advice

  3. Pingback: Asthma: Genetic Risks of Development and Course | Medical Research News and Interviews: MedicalResearch.com

  4. Pingback: Who Might Outgrow Childhood Asthma? | Emma's Children

Leave a Reply

Your email address will not be published. Required fields are marked *