Shih Ling Kao

Primary diabetes care and diabetic retinopathy (DR) screening persist as major public health challenges due to a shortage of trained primary care physicians (PCPs), particularly in low-resource settings. Here, to bridge the gaps, we developed an integrated image-language system (DeepDR-LLM), combining a large language model (LLM module) and image-based deep learning (DeepDR-Transformer), to provide individualized diabetes management recommendations to PCPs. In a retrospective evaluation, the LLM module demonstrated comparable performance to PCPs and endocrinology residents when tested in English and outperformed PCPs and had comparable performance to endocrinology residents in Chinese. For identifying referable DR, the average PCP's accuracy was 81.0% unassisted and 92.3% assisted by DeepDR-Transformer. Furthermore, we performed a single-center real-world prospective study, deploying DeepDR-LLM. We compared diabetes management adherence of patients under the unassisted PCP arm (n = 397) with those under the PCP+DeepDR-LLM arm (n = 372). Patients with newly diagnosed diabetes in the PCP+DeepDR-LLM arm showed better self-management behaviors throughout follow-up (P < 0.05). For patients with referral DR, those in the PCP+DeepDR-LLM arm were more likely to adhere to DR referrals (P < 0.01). Additionally, DeepDR-LLM deployment improved the quality and empathy level of management recommendations. Given its multifaceted performance, DeepDR-LLM holds promise as a digital solution for enhancing primary diabetes care and DR screening.

CONTEXT: Isolated elevation of TSH in the absence of thyroid symptoms can be very rarely caused by a macromolecule formed between TSH and Ig (macro-TSH), confounding the interpretation of thyroid function test results. OBJECTIVE: We described the use of several commonly available laboratory-based approaches to investigate an isolated TSH elevation [232 mIU/liter; free T(4), 10 pmol/liter (reference interval, 10.0-23.0 pmol/liter), Vitros platform] in a clinically euthyroid elderly gentleman, which led to the diagnosis of macro-TSH. METHODS AND RESULTS: TSH concentration of the patient was significantly lower (122 mIU/liter) when measured on the Advia Centaur platform. Serial dilution of the patient's sample showed a nonlinear increase in TSH recovery at increasing dilution (nonlinearity). Polyethylene glycol precipitation and mixing the patient's sample with a hypothyroid patient sample showed reduced TSH recovery, suggesting the presence of a high molecular weight interfering substance and excess TSH binding capacity, respectively. Heterophile blocking tube studies and rheumatoid factors were negative. Gel filtration chromatography demonstrated a TSH peak fraction that approximated the molecular size of IgG; together with the excess TSH binding capacity, this indicated the presence of TSH-IgG macro-TSH. A review of 12 macro-TSH case reports showed that samples with macro-TSH produce over-recovery with dilution, return negative results on anti-animal and anti-heterophile blocking studies, and commonly have recovery of less than 20% when subjected to polyethylene glycol precipitation. CONCLUSION: Macro-TSH is an underrecognized laboratory interference. Routine laboratory techniques described above can help diagnose this rare entity. A close dialogue between the physician and the laboratory is important in approaching such cases.

10.1056/NEJMc0807678

AIMS: To study whether HbA(1c) , and its relationship with fasting plasma glucose, was significantly different among Chinese, Malays and Indians in Singapore. METHODS: A sample of 3895 individuals without known diabetes underwent detailed interview and health examination, including anthropometric and biochemical evaluation, between 2004 and 2007. Pearson's correlation, analysis of variance and multiple linear regression analyses were used to examine the influence of ethnicity on HbA(1c) . RESULTS: As fasting plasma glucose increased, HbA(1c) increased more in Malays and Indians compared with Chinese after adjustment for age, gender, waist circumference, serum cholesterol, serum triglyceride and homeostasis model assessment of insulin resistance (P-interaction < 0.001). This translates to an HbA(1c) difference of 1.1 mmol/mol (0.1%, Indians vs. Chinese), and 0.9 mmol/mol (0.08%, Malays vs. Chinese) at fasting plasma glucose 5.6 mmol/l (the American Diabetes Association criterion for impaired fasting glycaemia); and 2.1 mmol/mol (0.19%, Indians vs. Chinese) and 2.6 mmol/mol (0.24%, Malays vs. Chinese) at fasting plasma glucose 7.0 mmol/l, the diagnostic criterion for diabetes mellitus. CONCLUSIONS: Using HbA(1c) in place of fasting plasma glucose will reclassify different proportions of the population in different ethnic groups. This may have implications in interpretation of HbA(1c) results across ethnic groups and the use of HbA(1c) for diagnosing diabetes mellitus.

The coding variant (p.Arg192His) in the transcription factor PAX4 is associated with an altered risk for type 2 diabetes (T2D) in East Asian populations. In mice, Pax4 is essential for beta cell formation but its role on human beta cell development and/or function is unknown. Participants carrying the PAX4 p.His192 allele exhibited decreased pancreatic beta cell function compared to homozygotes for the p.192Arg allele in a cross-sectional study in which we carried out an intravenous glucose tolerance test and an oral glucose tolerance test. In a pedigree of a patient with young onset diabetes, several members carry a newly identified p.Tyr186X allele. In the human beta cell model, EndoC-βH1, PAX4 knockdown led to impaired insulin secretion, reduced total insulin content, and altered hormone gene expression. Deletion of PAX4 in human induced pluripotent stem cell (hiPSC)-derived islet-like cells resulted in derepression of alpha cell gene expression. In vitro differentiation of hiPSCs carrying PAX4 p.His192 and p.X186 risk alleles exhibited increased polyhormonal endocrine cell formation and reduced insulin content that can be reversed with gene correction. Together, we demonstrate the role of PAX4 in human endocrine cell development, beta cell function, and its contribution to T2D-risk.