Daily Endocrinology Research Analysis
A multicenter randomized trial in NEJM shows automated insulin delivery significantly improves glycemic control in insulin-treated type 2 diabetes. A Science Translational Medicine study introduces BASTA, a simple whole-blood assay to detect beta cell antigen–specific CD4 T cells, potentially enabling broader immune monitoring in type 1 diabetes. An EJE study uses circulating microRNAs and machine learning to classify endocrine hypertension subtypes from primary hypertension with high accuracy.
Summary
A multicenter randomized trial in NEJM shows automated insulin delivery significantly improves glycemic control in insulin-treated type 2 diabetes. A Science Translational Medicine study introduces BASTA, a simple whole-blood assay to detect beta cell antigen–specific CD4 T cells, potentially enabling broader immune monitoring in type 1 diabetes. An EJE study uses circulating microRNAs and machine learning to classify endocrine hypertension subtypes from primary hypertension with high accuracy.
Research Themes
- Automated insulin delivery for insulin-treated type 2 diabetes
- Simplified immune diagnostics for type 1 diabetes
- MicroRNA-based machine learning to classify endocrine hypertension
Selected Articles
1. A Randomized Trial of Automated Insulin Delivery in Type 2 Diabetes.
In 319 insulin-treated adults with type 2 diabetes, automated insulin delivery reduced HbA1c by 0.6 percentage points more than control over 13 weeks and increased time-in-range by 14 percentage points, with low hypoglycemia risk. All CGM hyperglycemia metrics favored AID.
Impact: This multicenter RCT in NEJM provides high-level evidence that AID benefits insulin-treated type 2 diabetes, a population historically excluded from closed-loop trials.
Clinical Implications: Clinicians can consider AID to improve glycemic control and time-in-range in insulin-treated type 2 diabetes, with minimal increased hypoglycemia risk over 13 weeks.
Key Findings
- HbA1c decreased by 0.9% with AID vs 0.3% with control; adjusted difference −0.6% (95% CI −0.8 to −0.4; P<0.001).
- Time-in-range (70–180 mg/dL) increased from 48% to 64% with AID vs 51% to 52% with control; mean difference 14 percentage points (P<0.001).
- CGM hyperglycemia metrics were significantly better with AID; hypoglycemia was low in both groups with one severe event in AID.
Methodological Strengths
- Randomized, multicenter controlled design with continuous glucose monitoring.
- Clear primary endpoint and multiplicity-controlled CGM outcomes.
Limitations
- Short follow-up duration (13 weeks) limits assessment of long-term safety and durability.
- Potential lack of blinding and single-industry sponsorship may introduce bias.
Future Directions: Longer-term, diverse-population trials assessing durability, cost-effectiveness, quality of life, and rare adverse events; head-to-head comparisons among AID algorithms in type 2 diabetes.
BACKGROUND: Automated insulin delivery (AID) systems have been shown to be beneficial for patients with type 1 diabetes, but data are needed from randomized, controlled trials regarding their role in the management of insulin-treated type 2 diabetes. METHODS: In this 13-week, multicenter trial, adults with insulin-treated type 2 diabetes were randomly assigned in a 2:1 ratio to receive AID or to continue their pretrial insulin-delivery method (control group); both groups received continuous glucose monitoring (CGM). The primary outcome was the glycated hemoglobin level at 13 weeks. RESULTS: A total of 319 patients underwent randomization. Glycated hemoglobin levels decreased by 0.9 percentage points (from 8.2±1.4% at baseline to 7.3±0.9% at week 13) in the AID group and by 0.3 percentage points (from 8.1±1.2% to 7.7±1.1%) in the control group (mean adjusted difference, -0.6 percentage points; 95% confidence interval [CI], -0.8 to -0.4; P<0.001). The mean percentage of time that patients were in the target glucose range of 70 to 180 mg per deciliter increased from 48±24% to 64±16% in the AID group and from 51±21% to 52±21% in the control group (mean difference, 14 percentage points; 95% CI, 11 to 17; P<0.001). All other multiplicity-controlled CGM outcomes reflective of hyperglycemia that were measured were significantly better in the AID group than in the control group. The frequency of CGM-measured hypoglycemia was low in both groups. A severe hypoglycemia event occurred in one patient in the AID group. CONCLUSIONS: In this 13-week, randomized, controlled trial involving adults with insulin-treated type 2 diabetes, AID was associated with a greater reduction in glycated hemoglobin levels than CGM alone. (Funded by Tandem Diabetes Care; 2IQP ClinicalTrials.gov number, NCT05785832.).
2. BASTA, a simple whole-blood assay for measuring β cell antigen-specific CD4
The authors present BASTA, a simplified whole-blood assay that detects human CD4 T-cell responses to beta cell antigens. By reducing assay complexity and blood volume requirements, BASTA could enable broader clinical immune monitoring beyond autoantibodies in type 1 diabetes.
Impact: Enabling routine T-cell monitoring for T1D would be a major advance over current autoantibody-only approaches, with implications for early detection and response monitoring in prevention trials.
Clinical Implications: If validated in clinical cohorts, BASTA could complement autoantibody testing to risk-stratify individuals, track immunologic responses in prevention/teplizumab-like therapies, and monitor disease activity with minimal blood volume.
Key Findings
- Introduces BASTA, a whole-blood assay detecting human CD4 T cells specific for beta cell antigens.
- Addresses long-standing barriers of assay complexity and large blood volume requirements that have limited T-cell monitoring to research settings.
- Positions T-cell measurements to potentially complement autoantibody testing in T1D.
Methodological Strengths
- Translational assay design using whole blood to reduce sample handling and volume.
- Focus on disease-causal T-cell responses rather than surrogate autoantibodies.
Limitations
- Abstract does not detail cohort size, analytical performance, or multi-center validation.
- Clinical utility and cutoff definitions require prospective validation across diverse populations.
Future Directions: Prospective, multi-center validation comparing BASTA with standard autoantibodies and tetramer/ELISPOT assays; integration into T1D screening and prevention trials; assay standardization and quality control.
Type 1 diabetes (T1D) is an autoimmune disease where T cells mediate the destruction of the insulin-producing β cells found within the islets of Langerhans in the pancreas. Autoantibodies to β cell antigens are the only tests available to detect β cell autoimmunity. T cell responses to β cell antigens, which are known to cause T1D, can only be measured in research settings because of the complexity of assays and the large blood volumes required. Here, we describe the β cell antigen-specific T cell assay (BASTA). BASTA is a simple whole-blood assay that can detect human CD4
3. Identification of hypertension subtypes using microRNA profiles and machine learning.
Using circulating miRNAs and eight supervised ML algorithms, the authors classified endocrine hypertension subtypes from primary hypertension with AUC up to 0.9 (PPGL, CS, EHT) and 0.8 for PA. Key discriminators included hsa‑miR‑15a‑5p and hsa‑miR‑32‑5p.
Impact: Noninvasive miRNA biomarkers with ML could streamline early diagnosis of endocrine hypertension, reducing delays to curative surgeries or targeted therapies.
Clinical Implications: If validated prospectively, a plasma miRNA panel could triage patients with suspected secondary hypertension toward targeted endocrine workup (e.g., PA, CS, PPGL), improving diagnostic yield and care pathways.
Key Findings
- Eight supervised ML methods on circulating miRNAs achieved AUC ~0.9 for PPGL, CS, and overall EHT vs primary hypertension, and AUC 0.8 for PA vs primary hypertension.
- Top discriminative features included hsa‑miR‑15a‑5p and hsa‑miR‑32‑5p across disease comparisons.
- Demonstrates feasibility of plasma miRNAs as diagnostic biomarkers to classify endocrine hypertension subtypes.
Methodological Strengths
- Comparative evaluation across eight supervised machine learning methods.
- Feature attribution identifies specific miRNAs driving classification.
Limitations
- Cohort size, demographics, and external validation details are not provided in the abstract.
- Risk of overfitting and need for prospective, multi-center validation with clinical endpoints.
Future Directions: Prospective multi-center studies to validate performance across ancestries; integration with clinical/biochemical screening algorithms; development of standardized miRNA panels and thresholds.
OBJECTIVE: Hypertension is a major cardiovascular risk factor affecting about 1 in 3 adults. Although the majority of hypertension cases (∼90%) are classified as "primary hypertension" (PHT), endocrine hypertension (EHT) accounts for ∼10% of cases and is caused by underlying conditions such as primary aldosteronism (PA), Cushing's syndrome (CS), pheochromocytoma or paraganglioma (PPGL). EHT is often misdiagnosed as PHT leading to delays in treatment for the underlying condition, reduced quality of life and costly, often ineffective, antihypertensive treatment. MicroRNA (miRNA) circulating in the plasma is emerging as an attractive potential biomarker for various clinical conditions due to its ease of sampling, the accuracy of its measurement and the correlation of particular disease states with circulating levels of specific miRNAs. METHODS: This study systematically presents the most discriminating circulating miRNA features responsible for classifying and distinguishing EHT and its subtypes (PA, PPGL, and CS) from PHT using 8 different supervised machine learning (ML) methods for the prediction. RESULTS: The trained models successfully classified PPGL, CS, and EHT from PHT with area under the curve (AUC) of 0.9 and PA from PHT with AUC 0.8 from the test set. The most prominent circulating miRNA features for hypertension identification of different disease combinations were hsa-miR-15a-5p and hsa-miR-32-5p. CONCLUSIONS: This study confirms the potential of circulating miRNAs to serve as diagnostic biomarkers for EHT and the viability of ML as a tool for identifying the most informative miRNA species.