Daily Endocrinology Research Analysis
Three standout studies advanced endocrinology this cycle: a 24‑protein, multi-center classifier that differentiates follicular thyroid adenoma from carcinoma and outperforms gene panels; a single‑cell and spatial atlas of pituitary neuroendocrine tumors revealing an invasive macrophage program (SPP1+ TAMs) and p53‑linked aggressive clusters; and a first‑in‑human stable‑isotope study mapping regional glucagon metabolism in type 1 diabetes vs. non‑diabetes. Together, they push precision diagnostic
Summary
Three standout studies advanced endocrinology this cycle: a 24‑protein, multi-center classifier that differentiates follicular thyroid adenoma from carcinoma and outperforms gene panels; a single‑cell and spatial atlas of pituitary neuroendocrine tumors revealing an invasive macrophage program (SPP1+ TAMs) and p53‑linked aggressive clusters; and a first‑in‑human stable‑isotope study mapping regional glucagon metabolism in type 1 diabetes vs. non‑diabetes. Together, they push precision diagnostics, tumor microenvironment–based targets, and mechanistic metabolism.
Research Themes
- Protein-based precision diagnostics in thyroid oncology
- Tumor microenvironment mapping in pituitary neuroendocrine tumors
- Human regional glucagon metabolism and implications for closed-loop therapy
Selected Articles
1. A protein-based classifier for differentiating follicular thyroid adenoma and carcinoma.
Across 24 centers and 1,568 patients, a targeted 24‑protein classifier robustly distinguished FTC from FTA, outperforming gene-panel classifiers and maintaining high performance in retrospective and prospective validations. Its 95.7% negative predictive value supports rule‑out of malignancy to reduce unnecessary surgery.
Impact: Delivers a clinically actionable, multi-center–validated proteomic test that surpasses gene-based approaches for a long-standing diagnostic gap in thyroid pathology.
Clinical Implications: A high‑NPV protein panel could be integrated into diagnostic pathways (e.g., surgical pathology or FNAB adjunct) to safely rule out FTC, decreasing diagnostic lobectomies and enabling more personalized management.
Key Findings
- Proteomics quantified 10,336 proteins and identified 187 dysregulated between FTA and FTC.
- Protein-based discovery model achieved AUROC 0.899 (95% CI 0.849–0.949), outperforming gene-based AUROC 0.670.
- A targeted 24‑protein classifier validated with AUROC 0.871 and 0.853 in retrospective cohorts and 0.781 in prospective biopsies.
- Negative predictive value reached 95.7% for ruling out malignancy.
Methodological Strengths
- Large, multi-center cohort with external and prospective validation
- Targeted mass spectrometry panel built upon discovery proteomics and machine learning
Limitations
- Clinical utility and cost-effectiveness not yet tested in prospective impact trials
- Potential pre-analytical variability across centers and sample types
Future Directions: Prospective clinical utility trials embedding the classifier into preoperative workflows (including FNAB) and health-economic evaluations; assay harmonization and regulatory validation.
2. Single-cell and spatial transcriptome analyses reveal tumor heterogeneity and immune remodeling involved in pituitary neuroendocrine tumor progression.
By integrating single‑cell and spatial transcriptomics across 57 PitNET samples, the study maps invasive programs, highlighting an aggressive p53‑high cluster and SPP1+ TAMs driving invasion via SPP1–ITGAV/ITGB1. These data nominate microenvironmental and tumor‑intrinsic targets for future therapies and stratification.
Impact: Provides a high-resolution atlas of PitNET progression linking immune remodeling (SPP1+ TAMs) and tumor programs to invasion, offering testable hypotheses for targeted therapy.
Clinical Implications: Suggests biomarkers (e.g., SPP1+ TAM signatures, p53‑mediated programs) for risk stratification and points to SPP1–integrin axis as a potential therapeutic target in invasive PitNET.
Key Findings
- Integrated single-cell RNA-seq and spatial transcriptomics across >177,000 cells and ~35,000 spots from 57 PitNET tissues.
- Identified an aggressive TPIT-lineage cluster with elevated p53-mediated proliferation and higher Trouillas classification.
- Invasive tumors were enriched for SPP1+ tumor-associated macrophages that promote invasion via SPP1–ITGAV/ITGB1 signaling.
- Resolved immune-stromal heterogeneity and TME reconfiguration along invasive trajectories.
Methodological Strengths
- Large-scale single-cell plus spatial transcriptomic integration
- Trajectory and microenvironmental interaction analyses linking cell states to invasion
Limitations
- Observational omics; functional and therapeutic validations are pending
- Sample representation may not capture all PitNET subtypes or treatment contexts
Future Directions: Preclinical testing of SPP1–integrin blockade and modulation of TAM phenotypes; development of prognostic assays using SPP1/p53/TME signatures; integration with imaging and clinical outcomes.
3. Splanchnic and Leg Glucagon Metabolism in Healthy Individuals and Those With Type 1 Diabetes: First-in-Human Study Using [13C9,15N1]Glucagon.
Using novel stable-isotope glucagon tracers with regional catheterization, investigators showed similar splanchnic extraction but altered leg extraction dynamics in T1D versus ND across physiological glucagon ranges. The study provides a translational framework to optimize dual-hormone closed-loop algorithms and evaluate incretin-based agonists’ effects on α‑cell function and glucagon clearance.
Impact: First-in-human application of dual-labeled glucagon tracers to quantify regional glucagon kinetics reveals tissue-specific alterations in T1D with direct implications for device algorithms and drug development.
Clinical Implications: Supports personalization of glucagon dosing in dual-hormone closed-loop systems and informs how GLP‑1/GIP/glucagon receptor agonists may modulate α‑cell secretion and glucagon clearance in different vascular beds.
Key Findings
- Baseline splanchnic glucagon extraction was similar in T1D vs ND (~31% vs ~29%), but leg extraction was lower in T1D (27.0% vs 40.6%).
- With rising glucagon, splanchnic extraction remained unchanged; leg extraction declined in ND (41→31→24%) but was unchanged in T1D.
- Net splanchnic glucagon production did not change with exogenous glucagon infusion across physiological ranges.
- Demonstrated feasibility of [13C9,15N1]-glucagon tracers to probe regional metabolism in humans.
Methodological Strengths
- First-in-human dual stable-isotope glucagon tracers with regional (splanchnic/leg) catheterization
- Controlled glucagon infusion across physiological ranges with paired measurements
Limitations
- Small sample size (n=14) and single fasting condition limit generalizability
- Short-term physiology study; no long-term clinical outcomes
Future Directions: Expand to larger and diverse cohorts, postprandial states, and integrate with closed-loop algorithm testing; examine modulation by GLP-1/GIP/glucagon co-agonists on regional glucagon kinetics.