COMPARISON OF ANTI-HUMAN INSULIN-ANTIBODIES DETECTION BY COMMERCIAL ENZYME-LINKED-IMMUNOSORBENT-ASSAY KIT, DISPLACEMENT ENZYME-LINKED-IMMUNOSORBENT-ASSAY AND RADIOIMMUNOASSAY, IN THAI DIABETIC-PATIENTS
N. Tandhanandbanchuin et al., COMPARISON OF ANTI-HUMAN INSULIN-ANTIBODIES DETECTION BY COMMERCIAL ENZYME-LINKED-IMMUNOSORBENT-ASSAY KIT, DISPLACEMENT ENZYME-LINKED-IMMUNOSORBENT-ASSAY AND RADIOIMMUNOASSAY, IN THAI DIABETIC-PATIENTS, Diabetes research and clinical practice, 22(1), 1993, pp. 71-82
Antibodies specific to human insulin, in sera of Thai diabetic patient
s and normal healthy individuals, were detected by a commercial enzyme
-linked immunosorbent assay (ELISA) kit, displacement ELISA and radioi
mmunoassay (RIA). Among 36 insulin-treated patients, the antibodies we
re detected in 22, 23 and 20 individuals, by RIA, commercial ELISA kit
and displacement ELISA, respectively. Among those who had never previ
ously received insulin therapy, RIA showed positive results in 5/11 pa
tients with insulin-dependent diabetes mellitus (IDDM) and 1/2 patient
s with fibrocalculous pancreatic diabetes (FCPD); but were negative in
all 26 patients with non-insulin-dependent diabetes mellitus (NIDDM)
and 30 normal healthy individuals. The commercial ELISA kit could dete
ct such insulin autoantibodies (IAA) in 4/11 IDDM, 3/26 NIDDM, 2/2 FCP
D patients and 4/30 normal controls; while the displacement ELISA show
ed positive IAA detection in 5/11 IDDM, 2/26 NIDDM, 2/2 FCPD and 2/30
normal controls. By using RIA as the 'gold standard', the commercial E
LISA kit had 92.86% sensitivity, 87.01% specificity, 88.57% accuracy,
72.22% positive predictive value and 97.10% negative predictive value;
while these indices for the displacement ELISA were 85.71%, 90.01%, 8
9.52%, 77.42% and 94.59%, respectively.