Daily Endocrinology Research Analysis
Analyzed 28 papers and selected 3 impactful papers.
Summary
Analyzed 28 papers and selected 3 impactful articles.
Selected Articles
1. Efficacy and safety of HRS-7535, an oral small-molecule GLP-1 receptor agonist, in patients with diabetic kidney disease (SOLID-DKD): a randomised, double-blind, placebo-controlled, phase 2 trial.
In a 16-week, multicenter, randomized, double-blind, placebo-controlled phase 2 trial, the oral small‑molecule GLP‑1RA HRS‑7535 reduced UACR by 32–38% at 90 mg and 14–19% at 30 mg versus placebo, on top of widespread SGLT2 inhibitor and finerenone use. HbA1c decreased by 1.09% and body weight by 2.95% with a safety profile dominated by mild‑to‑moderate GI events and no deaths.
Impact: Demonstrates additive renoprotective effects of a novel oral small‑molecule GLP‑1RA in diabetic kidney disease under contemporary standard of care, supporting progression to phase 3 outcome trials.
Clinical Implications: If confirmed in outcomes trials, an oral small‑molecule GLP‑1RA could expand convenient, scalable reno‑ and cardiometabolic protection for patients with DKD, including those already on SGLT2 inhibitors and finerenone.
Key Findings
- At week 16, placebo‑corrected UACR reduction was −32% (ITT) and −38% (on‑treatment) with HRS‑7535 90 mg; −14% and −19% with 30 mg.
- Compared with placebo, HbA1c decreased by −1.09% and body weight by −2.95% with HRS‑7535 90 mg.
- Safety was consistent with GLP‑1RA class: mainly mild‑to‑moderate GI events; SAEs in 5.4% (30 mg), 4.3% (90 mg), 2.2% (placebo); no deaths.
- Effects were observed on top of intensive background therapy: 64.6% on SGLT2 inhibitors and 24.3% on finerenone.
Methodological Strengths
- Multicenter, randomized, double‑blind, placebo‑controlled phase 2 design with treatment‑policy (ITT) and efficacy (on‑treatment) estimands.
- Conducted on top of contemporary background therapy (SGLT2 inhibitors and finerenone), enhancing real‑world relevance.
Limitations
- Short duration (16 weeks) with surrogate albuminuria endpoint; no hard renal or cardiovascular outcomes.
- Single‑country study (China) and industry funding may limit generalizability and introduce potential bias.
Future Directions: Proceed to phase 3 trials powered for renal and cardiovascular outcomes, evaluate durability and safety, and assess interactions with SGLT2 inhibitors and finerenone across diverse populations.
BACKGROUND: The efficacy of non-peptide small-molecule glucagon-like peptide-1 (GLP-1) receptor agonists in diabetic kidney disease remains uncertain, particularly as add-on therapy to contemporary high-intensity treatments. We assessed HRS-7535, a novel oral small-molecule GLP-1 receptor agonist, in this population. METHODS: SOLID-DKD was a multicentre, randomised, double-blind, placebo-controlled, phase 2 trial conducted at 72 clinical sites in China. Adults with diabetic kidney disease (urinary albumin-to-creatinine ratio [UACR] 300-<3000 mg/g; estimated glomerular filtration rate [eGFR] ≥30 mL/min per 1.73 m FINDINGS: Between June 21, 2024, and March 30, 2025, 281 participants were randomised; 280 received at least one dose (30 mg: n = 93; 90 mg: n = 94; placebo: n = 93). Baseline median UACR was 763 mg/g; 181 (64.6%) participants were receiving SGLT2 inhibitors and 68 (24.3%) were receiving finerenone. At week 16, the placebo-corrected reduction in UACR with 90 mg was -32% (95% CI -43 to -18; treatment-policy estimand [intention-to-treat]) and -38% (-49 to -24; efficacy estimand [on-treatment analysis]); with 30 mg, -14% (-29 to 4) and -19% (-34 to -2), respectively. Compared with placebo, the mean difference in glycated haemoglobin with HRS-7535 90 mg was -1.09% (-1.32 to -0.86) and in body weight was -2.95% (-3.94 to -1.95). Adverse events were primarily mild-to-moderate gastrointestinal events during dose escalation. Serious adverse events occurred in 5 (5.4%), 4 (4.3%), and 2 (2.2%) participants in the 30 mg, 90 mg, and placebo groups. Discontinuation due to adverse events occurred in 3 (3.2%), 1 (1.1%), and 0 participants, respectively. No deaths occurred. INTERPRETATION: HRS-7535 dose-dependently reduced albuminuria in patients with diabetic kidney disease receiving intensive contemporary therapy, with a short-term safety profile consistent with the GLP-1 receptor agonist class, supporting its evaluation in phase 3 renal outcome trials. FUNDING: Jiangsu Hengrui Pharmaceuticals.
2. Machine learning-driven risk prediction for post-hospitalization diabetes case management: Integrating clinical and social determinants of health.
Using 178,227 encounters across derivation and temporal validation, an XGBoost model integrating clinical and SDoH features predicted 3‑month diabetes‑related ED visits with AUC 0.846 (0.842 temporal). Targeting the top 20% risk tier captured 64.1% of events with acceptable calibration and consistent performance across racial subgroups.
Impact: Demonstrates an operational, temporally validated prediction model that combines clinical and social risk to enable targeted, equity‑aware post‑discharge case management for diabetes.
Clinical Implications: Health systems can pilot daily risk‑stratified lists for diabetes case managers to proactively allocate resources, schedule follow‑ups, and address SDoH needs, while monitoring calibration, sensitivity, and fairness.
Key Findings
- XGBoost achieved AUC 0.846 (test) and 0.842 (temporal validation) for 3‑month diabetes‑related ED visits.
- Top 20% predicted risk tier captured 64.1% of all ED visits; calibration was clinically acceptable.
- Key predictors included prior ED visits, insulin prescriptions, age, and area‑level SDoH; individual SDoH (home safety, work disability) also contributed.
- Model discrimination was consistent across racial subgroups (AUC 0.843–0.851).
Methodological Strengths
- Very large cohort with temporal external validation, comprehensive calibration assessment, and subgroup fairness analysis.
- Compared multiple algorithms with cross‑validated hyperparameter tuning; integrated individual‑ and area‑level SDoH.
Limitations
- Single health system and retrospective design introduce potential selection and coding biases; generalizability needs testing.
- Sensitivity (0.296) indicates many events remain unflagged; complex models may challenge interpretability and workflow integration.
Future Directions: Prospective impact evaluations (e.g., stepped‑wedge trials), multi‑site external validation, continuous model updating, fairness audits, and ROI analyses of case management interventions.
OBJECTIVE: Patients hospitalized for diabetes-related conditions face elevated risks of emergency department (ED) visits post-discharge, driven by both clinical factors and social determinants of health (SDoH). This study aimed to develop and validate predictive models integrating clinical and SDoH data to identify high-risk patients for post-hospitalization diabetes case management. METHODS: We conducted a retrospective cohort study using electronic health record data from the University of Alabama at Birmingham Medical Center, including 162,063 inpatient encounters (January 2020-June 2024) for training and testing and 16,164 encounters (January-May 2025) for temporal validation. Patients were identified by diabetes-related ICD-10 codes or HbA1c ≥ 6.5 %, reflecting the scope of the institution's diabetes case management program. Predictors included demographics, diabetes-related comorbidities, surgical procedures, laboratory values, medications, and both area-level and individual-level SDoH. Logistic regression, decision trees, and XGBoost models were developed to predict diabetes-related ED visits within 3 months post-hospitalization. Hyperparameters for decision tree and XGBoost models were tuned via 10-fold cross-validation, and calibration was assessed using Brier scores and calibration plots. RESULTS: Among 162,063 hospitalizations, 6.2 % resulted in a diabetes-related ED visit. XGBoost achieved the best performance (area under the curve [AUC] 0.846, precision 0.420, sensitivity 0.296, specificity 0.972), maintained on temporal validation (AUC 0.842). Key predictors included past ED visit frequency, insulin prescriptions, age, and area-level SDoH indices. Individual-level SDoH factors, including home safety issues and work disability, also contributed to prediction. Targeting the top 20 % of predicted risk captured 64.1 % of all ED visits. Model discrimination was consistent across racial subgroups (AUC range: 0.843-0.851). Calibration was clinically acceptable across datasets. CONCLUSIONS: Integration of clinical and SDoH data achieved effective prediction of post-hospitalization ED visits. XGBoost provided excellent discrimination with temporal stability. Decision trees offered greater interpretability. A pilot implementation delivering daily risk-stratified patient lists to the diabetes case manager is underway, demonstrating a practical pathway from model development to clinical decision support.
3. Antenatal glycemic control in gestational diabetes and its association with postpartum glucose intolerance and weight retention.
In a prospective GDM cohort, lower late‑pregnancy time‑in‑target from SMBG independently predicted higher odds of postpartum dysglycemia at 6–12 weeks and greater 12‑month postpartum weight retention. The poorest control quartile had roughly double the risk of dysglycemia versus the best quartile.
Impact: Links an actionable antenatal glycemic metric (time‑in‑target) with postpartum metabolic outcomes, supporting risk‑stratified OGTT follow‑up and early lifestyle intervention in routine GDM care.
Clinical Implications: Incorporate late‑pregnancy TIT into discharge planning to prioritize postpartum OGTT and weight management for high‑risk women, and consider digital tools and coaching to sustain glycemic targets across the peripartum.
Key Findings
- Postpartum glucose intolerance occurred in 28.6% (82/287); each 10% lower late‑pregnancy TIT increased odds (aOR 1.24, 95% CI 1.10–1.39).
- Poorest versus best antenatal control quartiles had aOR 2.15 (95% CI 1.31–3.53) for postpartum dysglycemia.
- Each 10% lower TIT corresponded to +0.32 kg higher 12‑month PPWR (95% CI 0.12–0.52) and higher odds of substantial PPWR ≥5 kg (aOR 1.13, 95% CI 1.02–1.25); Q4 vs Q1 aOR 1.92 (95% CI 1.11–3.33).
- Follow‑up completion was high: 82.5% for postpartum OGTT and 76.7% for 12‑month weight.
Methodological Strengths
- Prospective cohort with prespecified confounder adjustment and multiple complementary outcomes.
- Use of continuous SMBG‑derived time‑in‑target enhances granularity over single HbA1c values.
Limitations
- Single‑center study with potential residual confounding; SMBG adherence and measurement bias possible.
- Generalizability to diverse populations and care settings requires external validation.
Future Directions: Externally validate TIT thresholds, and test TIT‑guided postpartum interventions (e.g., scheduling, digital support) in pragmatic trials for dysglycemia and weight outcomes.
BACKGROUND: Evidence linking antenatal glycemic control to postpartum dysglycemia and weight retention in real‑world GDM care is limited. The authors tested whether late‑pregnancy control predicts early postpartum glucose intolerance and 12‑month Postpartum Weight Retention (PPWR). METHODS: In this single‑center prospective cohort, women with GDM were enrolled from May 2021 to April 2024. Antenatal control (28-36 weeks) from routine SMBG was summarized as Time‑In‑Target (TIT). Mean fasting/postprandial glucose were recorded. The primary outcome was postpartum glucose intolerance by 75 g OGTT at 6-12 weeks, and secondary outcomes were PPWR and substantial PPWR (≥ 5 kg) at 12-months. Multivariable logistic and mixed‑effects models adjusted for prespecified confounders. RESULTS: Of 348 enrollees, 287 (82.5%) completed the postpartum OGTT and 267 (76.7%) had 12‑month weight. Postpartum glucose intolerance occurred in 82/287 (28.6%). Each 10% lower TIT was associated with higher odds of dysglycemia (adjusted OR [aOR = 1.24], 95% CI 1.10-1.39). The poorest‑ vs. best‑control quartiles had aOR = 2.15 (95% CI 1.31-3.53). Mean PPWR at 12-months was 3.4 ± 4.1 kg; each 10% lower TIT corresponded to +0.32 kg PPWR (95% CI 0.12-0.52) and increased odds of substantial PPWR ≥ 5 kg (aOR = 1.13, 95% CI 1.02-1.25), with Q4 vs. Q1 aOR = 1.92 (95% CI 1.11-3.33). CONCLUSIONS: Poorer antenatal control was independently associated with postpartum dysglycemia and greater 12‑month weight retention, supporting risk‑stratified postpartum OGTT follow‑up and early lifestyle support.