Aims: To improve sensitivity and specificity of the diabetes risk assessment of the population-based genetic screening used in the Finnish Diabetes Prediction and Prevention (DIPP) trial.
Methods: One thousand consecutive newborns enrolled in the DIPP were compared with 316 samples from children with Type 1 diabetes mellitus. A modification of the previously described technique based on hybridization of relevant PCR products with five lanthanide-labelled probes detected by time-resolved fluorometry (TRF) was used. A new probe was designed and allowed discrimination between DQB1*0602 and 0603 alleles, in addition to DQB1*02, *0301 or *0302, each of which required specific probes. A new, added screening strategy was developed for individuals carrying low-risk genotypes through specific typing of DQA1 *05 and *0201 alleles in DQB1*02 positive, and DRB1 typing for DR4 subtypes in DQB1*0302 positive subjects, with a new specifically designed high-resolution TRF-based DR4 subtyping technique.
Results: This two-step screening approach enhanced the sensitivity of the detection of genetic risk for Type 1 diabetes mellitus in this cohort up to 85.4%. In the general population cohort, 24.4% were identified for prospective follow-up, 2.6% of these are expected to develop Type 1 diabetes mellitus before the age of 15 years. Exclusive typing for HLA-DQB1 locus as an alternative screening strategy had sensitivities of 26.3-77.2% with general population cohorts of 2.3-23.1% identified for follow-up.
Conclusions: The described strategy for genetic prediction of Type 1 diabetes mellitus relies on the convenient genotyping procedure and could be applied in large scale screening projects such as DIPP.