Another Antibody Test for Type 1 Diabetes

The FDA just approved a new antibody test to help diagnose type 1 diabetes. It will be a laboratory-test, not a use-at-home test, and will be available for labs to order and use in the near future.

The new test, which measures whether or not a person's immune system is producing an antibody (AKA an autoantibody) called ZnT8Ab (zinc transporter 8 autoantibody). This antibody is pretty specific for T1D; it's rarely present in people with other conditions such as T2D or gestational diabetes.

The company presented data to the FDA from a clinical study of 569 blood samples: 323 from patients with diagnosed T1D and 246 samples from patients diagnosed with other kinds of diabetes, other autoimmune diseases, and other clinical conditions. The results showed that their test "was able to detect the ZnT8 autoantibody in 65 percent of the samples from patients with diagnosed type 1 diabetes and incorrectly gave a positive result" in less than 2% of the samples from patients who had other diseases. (Of course, with 35% of of samples from people with T1D being negative, a negative result from the test does not exclude a possible diagnosis of type 1 diabetes.)

As a representative of the manufacturer pointed out to me [personal communication], there are already several autoantibodies on the market for assistance in making a diagnosis T1D. They are already marketing three that have been around for a while: Glutamic Acid Decarboxylase (GAD65) Antibodies, IA-2 Antibodies, and Insulin Autoantibodies.

Where does this new test fit in, in the diagnosis and/or management of people with diabetes?

The new ZnT8Ab test should complement the other three, so in theory someone with T1D might have all four tests positive -- although there have been cases where only one or two of these tests are positive. Whether any of these antibody tests is positive or not apparently depends on a multiple factors. One study indicated that higher levels of ZnT8 were associated with higher levels of stimulated C-peptide. Another showed a correlation between variants of the SLC30A8 gene and the type of ZnT8Ab that develops. Duration of diabetes also seems to make a difference: after many years of T1D, the antibodies wane and disappear.

Clearly, a high-risk individual who doesn't show glucose intolerance but has other risk factors for T1D (such as autoimmune thyroid disease or multiple family members with T1D) might want to know if they are at risk: their physician could order a panel of these four tests from a commercial laboratory and if 4 of 4 came back positive, or if none of the four tests came back positive, then the patient and the physician could draw inferences about the likelihood of future T1D.

And in some cases of recently-diagnosed diabetes where it's unclear, for instance, if the patient has T2D or T1D or LADA, a panel of the four diabetes autoantibodies could help clarify which variant of diabetes exists.

Finally, as might be obvious, but the FDA none-the-less states, "The test should not be used to monitor the stage of disease or the response to treatment." It's a diagnostic tool to sort out whether T1D might occur in the future, or if a case of diabetes is or is not antibody-related.