ANTIBODIES IN TYPE I – PRACTICAL PROBLEMS IN INTERPRETATION

ANTIBODIES IN TYPE I – PRACTICAL PROBLEMS IN INTERPRETATION

            The best predictor of risk for Type IA diabetes is the expression of multiple is let auto-antibodies. The standard auto-antibody test for Type I A diabetes was the cytoplasmic ICA test, which used frozen sections of human pancreas. This test is difficult to standardize and measure. A combination testing for glutamic acid decarboxylase (GAD-65), ICA 512/IA-2 and insulin auto-antibodies, testing for cytoplasmic ICA is limited utility.

            More than 90% of individual with typical auto-immune insulin dependent diabetes  expression one or more of the three biochemical auto-antibodies. Among relatives at risk for Type I A diabetes, the highest risk is associated with expression of more than one of the bio-chemically defined auto-antibodies.

            Adding assay for cytoplasmic ICA to determination biochemical auto-antibodies runs the risk that a single biochemical auto-antibody (eg. GAD-65) may result in ICA positively, suggesting that a relative may have two antibodies (ICA and GAD-65) when he or she has only one. This form of ICA reacting only with GAD-65 has been termed restricted or selective ICA and is associated with a lower risk of progression to diabetes than is non restricted ICA, which is usually associated with expression of multiple auto-antibodies. The auto-antigen phogrin, or IA-2b is related to ICA 512/IA2 and most IA 2b positive sera are detectable ICA 512/IA-2 positive. The cation efflux transportes Zn T8 is an auto-antigen of Type I diabetes. Auto-antibodies against Zn T8 have been identified in 60-80% of patient with new onset Type I diabetes. These auto-antibodies were present in subject with diabetes and negative GAD-65, IA-2 and insulin auto-antibodies.

Caveats of Auto-antibody Testing:

1. Variation exists in sensitivity and specifity of assay. There is wide variation in different assays for cytoplasmic ICAs. Most assays produce labeled auto-antigens by transcribing and translating DNA of relevant clones. These assays are usually performed in semi-automated 96 well formats. Recent ELISA like assay developed where in auto-antibodies react with both “Labeled” fluid phase auto-antigen (eg. GAD-65) and antigens bound to plates.
2. Insulin autoantibodies cannot be measured after insulin therapy. Within weeks, certainly within a month of the institution of insulin therapy, insulin antibodies are produced.
3. Expression of insulin auto-antibodies is inversely correlated with the age at onset of Type IA diabetes. Thus, this assay is particularly sensitive for children developing Type IA diabetes before age 12 years. However, with excellent insulin auto-antibody assays, approximate 50% of adults developing diabetes also express such anti-bodies.
4. Auto-antibodies usually develop sequentially and the first auto-antibody often appears before 3 years of age. In some individual, the first or subsequent auto-antibodies can develop in late adulthood, hence, repetitive testing is necessary when children are followed prospectively to diabetes, one autoantibody may be developing, while another is disappearing and at diagnosis a small subject of children have lost auto-antibody expression.
5. The prognostic significance of even multiple auto-antibodies in presence of protective allele DQB1 0602 is unknown.
6. Although most individuals stably express auto-antibodies for years to decades before development of diabetes, as many as 10% of young children may transiently express a single auto-antibody.

            Thus the expression of a single persistent auto-antibody is often associated with a low risk (20%) for progression of Type IA diabetes compared to multiple auto-antibodies (>90%). Expression of no auto-antibodies can be reassuring, but auto-antibdies can appear later in life. The major benefit of identification of auto-antibodies should be the prevention of severe disability  or death at presentation in Ketoacidosis. This is estimated to occur in approximate 1 in 200 children presenting with diabetes in general population. The cons of such testing are potential anxiety and the potential for decreased insurability with “disease” detection. The best rationale for such screening is potential participation in a trial for the prevention of Type IA diabetes.

GAD-65 Antibodies:

1. GAD-65 enzyme present mainly in b cells, cerebellum and testis.
2. Increased in patient with Type I diabetes.
3. GAD-65 anti-bodies higher in females than male Type I diabetes patients.
4. GAD-65 anti-bodies are associated with slower rate of disease progression.
5. GAD-65 anti-body prevalence remains same with increasing age at onset of the disease.
6. GAD-65 anti-body positivity  in prediabetics marks impending autoimmune diabetes.
7. Associated more in Type I diabetes patients positive for DR3-DQ2.

Major b all auto-antigens associated with immune mediated Type I diabetes

Auto-antigen	Positive (%)	Comment
Insulin (ICA, IAA)	49-92	>90% children <5 years old auto-antibody positive
GAD	84	Low risk as single antibody
ICA 512/IA-2	74	Highly specific for Type I
Phogrin/IA-2b	61	Usually a subject IA-2

ICA 512 Anti-bodies

1. ICA 512 and IA-2 are refer to same autoantigen protein tyrosine phosphatase.
2. ICA 512 is increased in new onset Type I diabetes patients.
3. ICA 512 is expressed in pancreas and brain.
4. ICA 512 anti-bodies are associated with younger age at onset.
5. ICA 512 anti-bodies are in prediabetics are associated with rapid progression to Type I diabetes.
6. ICA 512 anti-bodies are associated more in Type I diabetes patients positive for DR4-DQ8.

Insulin Auto-antibodies (IAA)

1. Approximate 40-45% of new onset untreated patients are positive for IAA.
2. The IAA recognizes a conformational epitope of insulin molecule (Leu 32A).
3. IAA in presence of ICA in prediabetics enhances the risk to develop Type I diabetes.
4. IAA in younger patients is associated more with DR4.

Note:

Determination of biochemical auto-antibody can identify individuals at risk for Type I diabetes among family members and general population.

REFERENCE

            RSSDI Text Book of Diabetes Mellitus