Genetics offers new possibilities for diabetes management

As advances in genomic research propel the field of diabetes toward novel approaches for prevention and management, endocrinologists and researchers are speculating about the effect that genetic information will have on the diabetes epidemic.

“We are still in the early phase of using genomics to guide the future of preventing and optimally treating diabetes,” Eric J. Topol, MD, chief academic officer for Scripps Health and professor of genomics at the Scripps Research Institute, said in an interview. “But we are also in a rapid, accelerated phase of research.”

Genetic risk testing, for example, has already been pegged as an important area of exploration, but the implications of genomic research extend beyond predicting disease risk. Health care providers interviewed by Endocrine Today hope results of these studies may help individualize therapy in terms of medications and prevention strategies.

Endocrinologists who spoke with Endocrine Today said they are well aware that it is too early to tell if and when these tools will be available, and any change in treatment paradigms will be implemented slowly. Even so, many see the benefits as worth the wait.

“I am very optimistic about the future of this field,” Dawn Belt Davis, MD, PhD, assistant professor of endocrinology at the University of Wisconsin-Madison and Endocrine Today Editorial Board member, said in an interview. “We have a long way to go, but in the end, our goal is to have the ability to make much more educated therapeutic decisions for patients with diabetes.”

To date, researchers have identified approximately 40 loci on the genome that influence type 1 diabetes and another 40 that influence type 2 diabetes, according to Stephen S. Rich, PhD, director of the Center for Public Health Genomics at the University of Virginia. Despite these discoveries, researchers are far from having a thorough understanding of the role that genes play in the development of diabetes.

The loci, for instance, contain seven genes, on average, but this number can vary from zero to 25, Rich told Endocrine Today. Moreover, none of the loci identified for either type 1 or type 2 diabetes overlap, meaning that different pathways are likely involved in the diseases and their progression.

Stephen S. Rich

“The real predictive ability [for diabetes and its type] is limited to whether you think you have the right gene or not. Then there is always the question if you have the right gene, have you identified the appropriate variant or variants within the gene that contribute to risk?” Rich said.

The next few years hold more promise, Rich said. Currently, there are studies that may determine whether structural variations, including insertions, deletions, duplications and rearrangements of DNA, affect diabetes disease pathways. The identification of common structural variants would help researchers design tests to determine whether a person has these variants and the subsequent risk for disease.

The ability to code regions of the genome is also a big breakthrough, according to Rich. Compared with whole-genome sequencing, exome sequencing is quick and cost-effective.

“With exomes — the coding regions — if you find the variant that causes an amino acid change, then it may have some deleterious predictive function,” he said. “This can at least be incorporated into a plan, whereas with whole-genome sequencing, you may find variants that are not in coding regions that may not even be in genes. It is hard to interpret and to determine causality at that level.”

Predictive screening potential

Although precise predictive screening of diabetes is currently beyond reach, researchers are already conducting studies to gauge patients’ and physicians’ interest in such tests and to evaluate hypothetical reactions to the results. According to Topol, interest is high on both ends.

“When I’ve given lectures with physicians and people outside the health care field in attendance and I ask if they would like to have genetic testing available, the response has been highly enthusiastic,” he said.

Dawn Belt Davis

Some physicians and researchers, however, question whether accurate genetic risk testing for diabetes is possible because the disease is so dependent on external factors, including diet and exercise.

“This is the problem with any multifactorial disease such as diabetes,” Davis said. “We know that genetics play a role, but it is a relatively small role and there are multiple genes involved. So, in any one individual, it may always be somewhat difficult to highlight genetic risk above and beyond environmental influences.”

Similarly, Rich said genetic information may not strengthen predictive or diagnostic capabilities for diabetes. Rather, these tests would have a greater effect on prevention and management.

Daniel Einhorn, MD, medical director of the Scripps Whittier Diabetes Institute and chancellor of the American College of Endocrinology, said physicians may be able to detect a patient’s genetic predisposition to diabetes-related complications, such as kidney disease, and approach diabetes treatment accordingly. More importantly, however, screening for diabetes would allow physicians to identify high-risk patients before they develop diabetes or even prediabetes.

“That is the point where inexpensive, safe and rather modest interventions can make the largest impact,” Einhorn told Endocrine Today. “If you don’t catch people at that stage, you can never achieve as good an outcome as you would have before they developed the disease.”

Real-life ramifications

Despite physicians’ hopes, patient reactions to predictive screening for diabetes appear mixed, according to findings in recent studies.

In a study published in Diabetologia in 2009, Richard W. Grant, MD, MPH, assistant professor at Massachusetts General Hospital and Harvard Medical School, and colleagues polled a random sample of primary care patients who did not have diabetes but were at high risk for the disease about their views on genetic testing for diabetes. They found that 79% of physicians and 80% of patients at risk for diabetes supported genetic risk testing. Further, 71% of patients surveyed believed that high-risk results would trigger positive changes in diet and exercise.

In a later study, published in Diabetes Care in February 2011, Grant delved deeper into this issue. He and researchers polled 22 overweight participants with phenotypic risk factors for type 2 diabetes about how genetic screening would hypothetically affect behavior. Some participants reported that a high risk score would increase their motivation to make healthy lifestyle changes. Others, however, adopted a defeatist attitude, believing that they were likely to develop diabetes, regardless of improvements in diet and exercise.

Eric J. Topol

Topol and colleagues performed a similar study, which was published in The New England Journal of Medicine in February 2011. They followed 2,037 participants who purchased a genome-wide susceptibility test to evaluate diabetes risk. Results suggested that knowledge of genetic risk did little to inspire lifestyle changes. Data also demonstrated no significant association between a person’s risk and improvements in diet, physical activity or psychological health.

Nevertheless, recent research has yielded some positive findings. In Grant’s second Diabetes Care study, the researchers said patients inspired to make lifestyle changes reported that genetic risk testing would boost their motivation to improve diet and exercise more than the presence of traditional risk factors, such as obesity and family history.

“One of the themes that came out of [our study] was that genetic information is unchangeable,” Grant said in an interview. He noted that one participant believed that other factors, such as weight or abnormal glucose levels, can be lowered and then the risk for diabetes would disappear. “This is the contrast between the changeability of some of these risk factors and the permanence of a genetic result.”

Communication of test results

Genetic information may hold the key to unlocking effective treatment and prevention strategies, but it is complex and remains difficult for physicians to explain and patients to understand. Even so, how a genetic test’s results are communicated may make a difference, according to experts Endocrine Today interviewed.

According to Topol, patients in his study who reviewed their test results with physicians were more likely to implement lifestyle changes. Yet, many health care providers “aren’t comfortable with genomic data, so they minimize or neglect it,” he said.

Moreover, in the case of diabetes, genetic risk can also often be misconstrued because the disease differs greatly from other genetic disorders.

“People are confused by the difference between complex risk and single-gene disorders, like cystic fibrosis, in which case, if you have the gene, you’re going to get the disease. In the case of type 2 diabetes, you may simply have twice the risk of someone else,”

Michael F. Christman, PhD, president and CEO of the Coriell Institute for Medical Research, told Endocrine Today. Therefore, genetic counseling is vital, he said.

Jessica Waxler

Jessica Waxler, MS, a certified genetic counselor at Massachusetts General Hospital, said genetic counselors receive specialized training and have a broad depth of genetic knowledge.

“Genetic counselors also have the communication skills to help people break down information to an understandable level and then help them cope with that information,” she said in an interview.

Genetic counselors primarily work with patients who have or are at risk for conditions more strongly influenced by genetic factors, such as certain types of cancer or Down syndrome. They less commonly see those who want to discuss diabetes risk, according to Waxler. Nevertheless, the approach to counseling remains similar in those cases. Assessing risk by taking a detailed family history is key; then, there would be a discussion of known risk to the patient along with prevention and management strategies, she said.

Rebecca Nagy

Unfortunately, genetic counselors are not easily accessible for everyone. Rebecca Nagy, MS, a certified genetic counselor and assistant professor of clinical internal medicine at The Ohio State University, said although some states have genetic counselors in at least every ZIP code, other states have none.

Nagy urged physicians to visit the National Society of Genetic Counselors’ website to find counselors in their area if they have questions about patients who may have a genetic predisposition to or a family history of disease.

“Making that connection is important for future patients because everybody has patients in their practice with a genetic disease. And if you delve into the more common diseases, like diabetes, genetic testing and genomic testing is going to become more and more commonplace,” she said.

Direct-to-consumer dilemmas

Although many physicians believe that genetic risk testing lacks precision, several companies, such as 23andMe and Navigenics, already cater to customers’ curiosity about their risk for an assortment of diseases, including diabetes, with direct-to-consumer tests, which usually cost about $200 and can be ordered online. Customers register with the website, send a saliva sample via mail and wait several weeks for their results.

“These are primarily candidate gene tests,” Rich said. “The interesting thing is the way they come up with their risk estimates.”

Testing for every genome on the locus would be impossible, according to Rich. However, one test, for example, only screens variants on six of the approximately 40 loci associated with diabetes. Thus, such a test may have considerably limited accuracy in detecting diabetes.

Topol said the direct-to-consumer tests “are not up-to-date, not complete, not comprehensive.” Most of the pharmacogenomic data relevant to diabetes management is not included. For instance, manufacturers only used data from earlier studies on TCF7L2 when developing these tests. In addition, they do not take into account the recently discovered, much rarer variants in other genes, such as HHEX, KCNJ11 and the melatonin receptor MTNR4 — all of which carry high risk for diabetes, Topol said.

Physicians interviewed by Endocrine Today said they also took issue with how direct-to-consumer test results are delivered. Occasionally, these companies have genetic counselors available to discuss concerns via telephone, but it is not required or standard practice.

“We’ve had some patients come in with their test results in hand from a Web-based company, not knowing what to make of the results,” Nagy said.

Despite these difficulties, now that information on one’s personal genome is so easily available, interest will only continue to grow, according to Rich.

“The genie is out of the bottle, and the best thing is to make sure the tests are as good and as interpretable as possible,” he said.

Beyond risk screening

Risk testing is not the only type of genetic screening that is important, according to Toni I. Pollin, PhD, a genetic counselor and associate professor in the departments of medicine and epidemiology and public health at the University of Maryland School of Medicine, Baltimore.

“Testing for monogenic forms of diabetes, such as maturity onset diabetes of the young (MODY) and neonatal diabetes, is useful because if you identify a gene mutation, you identify the specific etiology of that diabetes. That can actually provide an opportunity to get more effective treatment and also more precisely assess risk in families,” Pollin told Endocrine Today.

She cited infants diagnosed with what looks like type 1 diabetes before age 6 months as prime examples. Many cases, Pollin said, are a type of permanent neonatal diabetes in which patients can produce insulin, but a genetic mutation prevents them from secreting it. Once the diagnosis is confirmed through genetic testing, these children no longer need insulin therapy and can manage their diabetes using high-dose sulfonylureas.

Davis said genetic tests facilitate discussion with patients; for example, MODY can be difficult to describe because the disease is neither type 1 nor type 2 diabetes.

“When I explain that to patients, it is hard for them to understand because it gets lost in the linguistics. [Genetic testing] makes it easier for the patient to say, ‘Now I clearly see this is what’s wrong with me,’ and I can better explain why they developed diabetes,” she said.

Promise of personalized medicine

Advances in genomic research are likely to help health care providers after the diagnosis is made, too. Data generated by these studies may eventually allow researchers to further dissect already identified types of diabetes into subclassifications. This breakthrough would help physicians tailor treatment to individual patients based on subtype, according to Topol.

“It’s this concept of taxonomy, of reclassifying diseases [such as diabetes and obesity] based on their molecular story,” he said. “These diseases lend themselves well to a total reclassification, not a simplistic, reductionist diagnosis of diabetes or obesity. Rather, we can tell patients, ‘This is the specific pathway that is implicated in your diabetes and this is exactly how we plan to work on that pathway.’”

Pollin, who is a member of the Diabetes Prevention Program (DPP) genetics working group, said she and colleagues found interesting genetic differences between patients who do and do not respond to the intensive lifestyle or metformin interventions during the course of the DPP study.

Moreover, genomic information may lead to optimization of existing treatments. Topol said several large, randomized trials suggest that various medications reduce the risk for diabetes by 70% to 80%, but the drugs are not used for this purpose. Likewise, many physicians are unsure of what treatment to try after one therapy has failed.

“Much of using any medication is a bit of trial and error,” Einhorn said. “You try and see if you get the results you want, and if that doesn’t work, you try a different drug class. Wouldn’t it be wonderful to know which medications are more or less likely to work before you start treatment?”

Moving forward

As the field of genetics evolves, physicians continue to target new ways to apply this knowledge.

Researchers at Massachusetts General Hospital and Harvard Medical School, for instance, discovered an association between five amino acids — isoleucine, leucine, valine, tyrosine and phenylalanine — and the development of diabetes. “The point is to be able to identify how an interaction of genes and environment affect risk for diabetes,” Einhorn said.

Grant and colleagues are also pushing their studies beyond the hypothetical realm. In an ongoing, randomized trial, they performed genetic risk testing, sent study participants to genetic counseling and evaluated how actual test results influenced their adherence to a diabetes prevention program.

Others are focusing on how best to implement changes in clinical practice affected by genomic discoveries. Christman, who is also head of the Coriell Personalized Medicine Collaborative, said medical schools within their network will emphasize the importance of genetics so that awareness spreads and it becomes part of physicians’ rhetoric in the future.

This integration will not happen quickly, he said, adding that 50 years ago, discussing LDL cholesterol or triglycerides was also not customary in a primary care physician’s office.

“It is not an easy task, and it will probably take a generation for the medical community to fully embrace and integrate it, but genetic advances are going to have a profound effect on medicine,” Christman said. – by Melissa Foster

For more information:

• Bloss CS. N Engl J Med. 2011;364:524-534.
• Grant RW. Diabetologia. 2009;52:2299-2305.
• Markowitz SM. Diabetes Care. 2011;34:568-573.
• Wang TJ. Nat Med. 2011;17:448-453.

Disclosure: Drs. Christman, Davis, Einhorn, Pollin, Rich and Topol, and Ms. Nagy and Ms. Waxler report no relevant financial disclosures.

At what age would it be appropriate to perform genetic testing for diabetes?

As early as possible

Gerald Bernstein

Given that one in 300 to 400 births in the United States results in a patient with type 1 diabetes, genetic risk screening for diabetes should be conducted at birth, which we do for other, much rarer conditions.

For type 1 diabetes, accurate genetic testing in children would allow parents to look for early symptoms of the disease, such as excessive thirst and frequent urination. Some parents may become unnecessarily concerned, yet often the disease is not identified until children enter an altered state of consciousness due to dehydration and high blood sugar. Therefore, patients and parents should have access to a diabetes educator so that they recognize symptoms of diabetes but understand that they need not panic if their child takes an extra glass of water or wets the bed.

Genetic risk testing for type 1 diabetes may also be helpful in the young adult population, especially in people who are in their 20s and not regularly receiving physicals from physicians. Because these patients generally have larger bodies than children, they can absorb some of the abnormalities associated with the disease and may attribute symptoms such as excessive thirst to external factors. Knowing they are at a higher risk for type 1 diabetes may help them look for signs.

With type 2 diabetes, no one knows when the disease process begins. Certain factors, such as poor diet and lack of physical activity, can accelerate this process, but the fact is that type 2 diabetes probably has a multifaceted series of molecular abnormalities, meaning that any two individuals who develop the disease are most likely very different. Currently, physicians can reasonably predict a patient’s risk based on certain factors, such as family history or ethnicity, and genetic testing at this point may not add much to that equation. Once we know more about the disease, then we can determine if screening in young children will be beneficial.

Gerald Bernstein, MD, is director of the diabetes management program at the Friedman Diabetes Institute at Beth Israel Medical Center in New York.

Disclosure: Dr. Bernstein acts as a consultant for Generex.

May be problematic in pediatrics

Leonid Poretsky

Even if we had a genetic test that could accurately predict risk for developing diabetes, screening for type 1 diabetes may not be as useful as screening for type 2 diabetes in pediatric patients.

In children, it is important to distinguish between type 1 and type 2 diabetes because they are very different. Currently, we have fairly strong knowledge about genetic influences — for example, HLA haplotypes — that are associated with type 1 diabetes. The problem, however, is that we do not currently have effective prevention strategies for people at high risk for the disease. Researchers are conducting trials and programs examining the efficacy of various approaches to prevent the development of type 1 diabetes, such as using oral or injectable insulin, but at present the issue remains unresolved. Therefore, how is knowing one’s risk beneficial?

Furthermore, rather than helping patients and their parents, results indicating high risk for type 1 diabetes may cause unnecessary anxiety. For instance, a child may not be able to completely comprehend the information provided by his or her physician. Parents, on the other hand, would hopefully understand, but might become overly cautious and experience stress, even though their child, while at high risk, may not develop the disease.

Genetic testing for type 2 diabetes is more complicated because, unlike type 1 diabetes, many more genes appear to be involved, but the influence of any one of them appears to be very small. Nevertheless, a high risk result may prompt patients and parents to implement proven interventions — primarily involving weight maintenance and exercise — to prevent development of the disease. Once we develop more accurate genetic tests for both types of diabetes, find preventive measures for type 1 diabetes and learn more about the genetic influences on type 2 diabetes, then genetic screening in children may be both feasible and clinically meaningful.

Leonid Poretsky, MD, is director of the Friedman Diabetes Institute and chief of the division of endocrinology at Beth Israel Medical Center in New York. He is also professor of medicine at Albert Einstein College of Medicine.

Disclosure: Dr. Poretsky reports no relevant financial disclosures.