Dividing Diabetes by Cluster Instead of Types

SAID, MOD, MARD—Swedish researchers are proposing a new taxonomy for diabetes based on a cluster of factors.

Jan Greene

Take two patients with type 2 diabetes: one is a young woman whose pancreas is producing less insulin than it used to. The other is a middle-aged man whose pancreas is just fine, but whose belly fat is altering his metabolism to become very resistant to insulin. The two situations require completely different treatment—and yet they both have the same diagnosis, their doctors follow the same clinical guidelines, and what is happening inside these patients couldn’t be more different.

Diabetes is a generic term; it simply means high blood sugar. Of course, it has long been divided into type 1 diabetes, when the pancreas produces little, if any, insulin, and type 2, which may involve a shortage of insulin as well as insulin resistance. The young woman in our example is much closer on the spectrum of disease to type 1 diabetes and needs to replace the missing insulin. The heavy gentleman might be able to make his type 2 disease go away with the right combination of diet and exercise—and the discipline to stick with the program.

Clinicians faced with the wide range of types of diabetes need all the help they can get in responding appropriately to the patient in front of them. Breaking up type 2 diabetes into subtypes has been something that clinicians and researchers have discussed and debated for years. “There are many different elements that go into type 2 risk, and lumping under one umbrella a bunch of different conditions that have one thing in common—high blood glucose levels”—is not ideal, says Matt Peter­sen, vice president of medical information for the American Diabetes Association.

The imprecision has implications for treatment, he says: “We’ll probably find out some drugs work better in some people than in others, that certain changes in lifestyle have more of an effect in one person than another.”

“There is a need to understand” pathophysiologic subgroups in type 2 diabetes, says MGH’s Jose Florez, MD.

Some novel forms of diabetes have been identified in recent years. For instance, experts recently identified a new form of type 1 diabetes: latent autoimmune diabetes in adults (LADA). Another newcomer is maturity-onset diabetes of the young (MODY), which has a wide range of symptoms but affects a group of patients who have a genetic variation in common. There is also increasing evidence that type 2 diabetes should be treated quite a bit differently in elderly patients, partly because treatment drugs can more easily tip older people into hypoglycemia.

Diabetes ‘clusters’

Two years ago, the American Diabetes Association took a stab at creating some new type 2 categories but concluded that a new taxonomy will require a better understanding of the underlying beta cell dysfunction in the pancreas, along with staging of the disease by blood glucose status and complications. Petersen said the working group had envisioned that it would establish a new set of diabetes subgroups but in the end decided more information on genetics and other risk factors was needed.

Old and new way of categorizing diabetes

Researchers at Lund University in Sweden have proposed a new taxonomy that clusters patients into new categories

Source: Ahlquist E et al., Lancet Diabetes and Endocrinology, March 2018

In March in a paper published in Lancet Diabetes and Endocrinology, researchers at the Lund Institute in Sweden went ahead and proposed new, more nuanced categories of type 2 diabetes. Using data that captured information on insulin resistance, insulin secretion, blood sugar levels, and age of onset among 14,755 Swedes and Finns newly diagnosed with diabetes, they identified five distinct diabetes “clusters”:

Severe automimmune diabetes (SAID). Corresponds to type 1 diabetes and LADA. It is characterized by onset at a young age, poor metabolic control, impaired insulin production, and the presence of GADA antibodies.

Severe insulin-deficient diabetes (SIDD). Includes those with high HbA1c results, impaired insulin secretion, and moderate insulin resistance.

Severe insulin-resistant diabetes (SIRD). Characterized by obesity and severe insulin resistance.

Mild obesity-related diabetes (MOD). Includes obese patients who are diagnosed with diabetes at a relatively young age.

Mild age-related diabetes (MARD). Consists of older people and is the largest group (about 40%).

Jose Florez, MD, who runs a lab focused on precision therapies for diabetes at Massachusetts General Hospital, said the Lund Institute work will help with understanding the heterogeneity of type 2 diabetes. In an email responding to questions from Managed Care, he said, “There is a need to understand the various pathophysiologic subgroups in type 2 diabetes, and this is a good first step in that regard.”

More work is needed, he said, to reproduce their results in other ethnic groups and to show that patients identified with a particular cluster belong there.

Researchers need to investigate how a wide variety of factors related to diabetes interact over time, according to Christine Lee, MD, an NIDDK program director.

Along with clarifying the varieties of diabetes, researchers need to investigate how medical, physical, genetic, environmental, social, and other factors interact over time, Christine Lee, MD, program director in the National Institute of Diabetes and Digestive and Kidney Diseases, said in an email.

Researchers and clinicians had high hopes that genetics would produce simple gene-related variations that would make it easy to identify patients with subtypes of diabetes, says Petersen at the American Diabetes Association. “We know there are different genes involved and it’s been a little frustrating to pin them down,” he says. The challenges include significant variation among ethnicities—and with them. It’s not enough to look at Asians or Latinos, says Petersen. For example, diabetes prevalence is 4.3% among Chinese immigrants and their descendants in this country compared with 8.9% among Filipino immigrants and their descendants. Important information is lost when you lump those two groups together as Asians.

Long wait for genetic tests?

Classifying a patient using genetics rather than a biomarker would be handy, says Florez, because a gene variant is unchanging and needs to be identified just once in the patient’s lifetime. With diabetes, that simplified goal may not be possible for all patients, given the environmental and lifestyle factors that affect blood sugar metabolism. Still, Florez is hopeful that continuing work on genetics of diabetes will produce useful information for clinicians, particularly if patients’ genetic information is incorporated into their medical records.

A genetics-driven personalized medicine scheme for diabetes could be a few years away—or it could be decades, Petersen says. The trick will be coming up with treatment plans that fit individual patients based on more precise diagnostic categories while keeping them simple enough to use in everyday clinical practice. That balance will be difficult in personalized medicine—particularly in diabetes, which is seeing a flood of new treatments. Diabetes already has 12 approved classes of medications.

Clinicians are already subtyping patients in their heads, says Petersen. “But if you could do a simple array of genetic and metabolic tests for a $50 screening, everyone would welcome that.”

Jan Greene is an independent journalist based in the San Francisco Bay Area and a regular contributor to Managed Care.


Legislation and regulation
By Richard Mark Kirkner
Legislation and regulation
By Richard Mark Kirkner
Legislation and regulation
By Richard Mark Kirkner

A blueprint for high-volume, high-quality lung cancer screening that is detecting cancer earlier—and helping to save lives

Clinical Brief

Multiple Sclerosis: New Perspectives on the Patient Journey–2019 Update
Summary of an Actuarial Analysis and Report