Daily Endocrinology Research Analysis
Policies and tools in endocrinology advanced on three fronts: health-economic modeling supports universal first-trimester TSH screening in pregnancy; brain-predicted age difference emerged as a potential biomarker reflecting reversible brain changes in anorexia nervosa; and a 23-year single-surgeon pituitary series provides robust real-world benchmarks for outcomes and recurrence risk. These studies collectively inform screening policy, biomarker development, and surgical quality counseling.
Summary
Policies and tools in endocrinology advanced on three fronts: health-economic modeling supports universal first-trimester TSH screening in pregnancy; brain-predicted age difference emerged as a potential biomarker reflecting reversible brain changes in anorexia nervosa; and a 23-year single-surgeon pituitary series provides robust real-world benchmarks for outcomes and recurrence risk. These studies collectively inform screening policy, biomarker development, and surgical quality counseling.
Research Themes
- Thyroid screening cost-effectiveness in pregnancy
- Neuroendocrine biomarkers in eating disorders
- Real-world outcomes in pituitary neuroendocrine tumors
Selected Articles
1. Cost-Effectiveness Analysis of Universal Screening for Subclinical Hypothyroidism in Pregnancy Based on the Czech Nation-Wide Pilot Screening Program in Light of Current Evidence.
Using a national pilot program and best-available evidence, universal first-trimester TSH screening in pregnancy was modeled as cost-effective (ICER ≈ €20,035/QALY) with modest QALY gains, prevention of miscarriages and preterm births, and low budget impact. Adding anti-TPO testing may further improve outcomes (lower ICER) but requires more resources and rests on lower-quality evidence.
Impact: This analysis directly informs national screening policy by quantifying trade-offs between health gains and resource use, grounded in a real-world pilot program. It provides decision-grade parameters to potentially standardize thyroid screening in pregnancy.
Clinical Implications: Health systems could adopt universal first-trimester TSH screening as a cost-effective policy, anticipating modest maternal QALY gains and prevention of miscarriages/preterm births with limited budget impact. Anti-TPO extension should be considered cautiously, balancing potential benefits against increased referrals and costs.
Key Findings
- Universal TSH screening modeled ICER: €20,035/QALY with 32.9 maternal QALYs gained per 100,000 pregnancies.
- Estimated prevention of 65 miscarriages and 52 preterm births annually under universal TSH screening.
- Anti-TPO extension improved ICER to €15,703/QALY with additional 111 QALYs but increased budget impact and referrals, based on lower-quality evidence.
- Budget impact for TSH screening estimated at ~€0.66 million with ~2,290 additional endocrinology referrals per 100,000 pregnancies.
Methodological Strengths
- Decision-analytic model parameterized with national pilot program data and best-available evidence.
- Probabilistic sensitivity analysis (3,000 simulations) supporting robustness of cost-effectiveness estimates.
Limitations
- Economic model depends on assumptions and external evidence; not a randomized trial.
- Anti-TPO extension relies on lower-quality evidence and may strain resources due to increased referrals.
Future Directions: Prospective implementation studies comparing universal versus targeted screening, with registry-based follow-up on maternal-fetal outcomes and healthcare utilization, and randomized evaluation of anti-TPO triage strategies.
OBJECTIVES: In 2023, the Czech Guideline by Endocrinology society introduced universal screening of thyrotropin (TSH) in the first trimester of pregnancy. We modeled the incremental cost-effectiveness ratio (ICER), budget impact (BI), and new endocrinology referrals (ER) for universal screening of (1) TSH for subclinical hypothyroidism compared with the current testing and (2) extension to anti-thyroid peroxidase antibodies (anti-TPO). METHODS: A decision tree was built assuming that levothyroxine reduces the risks of gestational hypertension, miscarriage, and preterm birth and anti-TPO test improves diagnosis of postpartum thyroiditis. Models were parametrized using the best available evidence, probabilities, and utilities from the national pilot screening program. ICER was compared with a willingness-to-pay threshold of 47 430 EUR per quality-adjusted life year (QALY). RESULTS: Expecting 100 000 pregnancies yearly, introduction of universal TSH screening will result in 32.9 (-40 to 91.7) maternal QALYs gained, 65 (45-90) miscarriages and 52 (30-77) preterm deliveries prevented, with an ICER of 20 035 EUR/QALY (2315/3000 simulations cost-effective), the BI of 659 756 (439 993-895 993) EUR, and 2290 (1883-2728) ER (randomized-controlled-trial-level evidence). The extension of the TSH screening with anti-TPO will bring additional 111 (-398 to 604) QALY gained, ICER of 15 703 EUR per QALY (1945/3000 simulations cost-effective), the BI of 1 746 486 (1 241 391-2 350540) EUR, and 6927 (6121-7765) ER (low-level evidence). CONCLUSIONS: Universal screening of TSH and extension with anti-TPO appear to be cost-effective. The population benefits of TSH screening are modest, but the BI is low. The model for anti-TPO is based on weak evidence and generates important BI.
2. Does restrictive anorexia nervosa impact brain aging? A machine learning approach to estimate age based on brain structure.
A multi-dataset ML model accurately predicted brain age and revealed overestimation of brain age in acute anorexia nervosa (+2.25 years), which normalized after weight restoration. The effect was stronger in younger patients and inversely related to BMI, supporting brain-PAD as a candidate biomarker reflecting reversible neurobiological impact of malnutrition.
Impact: Introduces an interpretable brain-age biomarker with evidence of reversibility tied to weight restoration, bridging neuroimaging, endocrinology, and clinical outcomes in anorexia nervosa.
Clinical Implications: Brain-PAD could aid risk stratification and monitoring of neurobiological recovery alongside nutritional and endocrine restoration in AN, especially in younger patients.
Key Findings
- ML brain-age model trained on 3,487 healthy females achieved MAE 1.93 years and r=0.88 in controls.
- Acute AN showed +2.25 years brain-PAD (advanced brain aging), while weight-restored AN normalized to HC levels (p=0.98 vs HC; p=0.0026 vs acute AN).
- Brain-age deviation was most pronounced in younger acute AN patients (group-by-age interaction p<0.001).
- Brain-PAD correlated negatively with BMI in acute AN (r ≈ -0.29).
Methodological Strengths
- Large, diverse training corpus (10 datasets) and external validation across two institutions.
- Robust model performance with clinically interpretable biomarker (brain-PAD) and subgroup analyses.
Limitations
- Cross-sectional design limits causal inference; longitudinal validation needed.
- Female-only sample aged 10–40 years limits generalizability to males and older adults.
Future Directions: Prospective longitudinal studies linking brain-PAD trajectories with endocrine recovery, cognitive outcomes, and relapse; testing intervention effects on brain-PAD as a surrogate outcome.
Anorexia nervosa (AN), a severe eating disorder marked by extreme weight loss and malnutrition, leads to significant alterations in brain structure. This study used machine learning (ML) to estimate brain age from structural MRI scans and investigated brain-predicted age difference (brain-PAD) as a potential biomarker in AN. Structural MRI scans were collected from female participants aged 10-40 years across two institutions (Boston, USA, and Jena, Germany), including acute AN (acAN; n=113), weight-restored AN (wrAN; n=35), and age-matched healthy controls (HC; n=90). The ML model was trained on 3487 healthy female participants (ages 5-45 years) from ten datasets, using 377 neuroanatomical features extracted from T1-weighted MRI scans. The model achieved strong performance with a mean absolute error (MAE) of 1.93 years and a correlation of r = 0.88 in HCs. In acAN patients, brain age was overestimated by an average of +2.25 years, suggesting advanced brain aging. In contrast, wrAN participants showed significantly lower brain-PAD than acAN (+0.26 years, p=0.0026) and did not differ from HC (p=0.98), suggesting normalization of brain age estimates following weight restoration. A significant group-by-age interaction effect on predicted brain age (p<0.001) indicated that brain age deviations were most pronounced in younger acAN participants. Brain-PAD in acAN was significantly negatively associated with BMI (r = -0.291, p
3. Pituitary surgery in New Zealand: A single neurosurgeon case series of 1,224 operations.
In a 23-year, single-surgeon series of 1,224 pituitary operations, transsphenoidal surgery showed low serious complication rates and endocrine remission rates in acromegaly and Cushing’s disease comparable to international standards. In non-functioning PitNETs, 20% recurred (median 3.7 years); cavernous sinus invasion strongly predicted recurrence, while older age was protective.
Impact: Provides comprehensive, real-world benchmarks for complication, remission, and recurrence across PitNET subtypes, informing surgical counseling, surveillance intensity, and quality metrics.
Clinical Implications: Counsel patients on expected remission and complication risks; consider closer surveillance in cases with cavernous sinus invasion; data support benchmarking for pituitary centers.
Key Findings
- Transsphenoidal surgery complications: CSF leak 7.3%, permanent diabetes insipidus 3.4%, meningitis 2.4%, visual deterioration 1.2%; carotid injury/stroke/mortality <1%.
- Non-functioning PitNETs: 20% recurrence at median 3.7 years; cavernous sinus invasion predicts recurrence (HR 2.9), older age protective (HR 0.98).
- Endocrine remission: acromegaly 76% (microadenomas) and 75% (macroadenomas); Cushing’s disease 82% (microadenomas) and 75% (macroadenomas).
Methodological Strengths
- Large single-surgeon cohort with consistent surgical technique over 23 years.
- Multivariate analyses and long-term follow-up enabling identification of recurrence predictors.
Limitations
- Single-surgeon, single-center retrospective series may limit generalizability.
- Potential selection and information biases inherent to retrospective design.
Future Directions: Prospective multicenter registries to validate recurrence predictors, integrate molecular markers, and harmonize outcome reporting across PitNET subtypes.
Pituitary neuroendocrine tumours (PitNETs) are among the most common tumours of the central nervous system, yet data regarding surgical outcomes is limited for New Zealand (NZ). This study presents a 23-year single-surgeon case series of pituitary surgery. A retrospective analysis of 1,224 pituitary operations performed between August 1999 and July 2023 was conducted. Patient demographics, tumour characteristics, complications, and long-term outcomes were analysed. Statistical analyses included univariate and multivariate models. Pituitary neuroendocrine tumours (PitNETs) accounted for 88 % of cases (n = 1,087), including 612 non-functioning, 175 GH-producing, and 112 ACTH-producing tumours. The complications of transsphenoidal surgery for PitNETs included postoperative cerebrospinal fluid (CSF) leaks (7.3 %), permanent diabetes insipidus (3.4 %), meningitis (2.4 %), and visual deterioration (1.2 %). Carotid injury, stroke, and death within 30 days each occurred in less than 1 % of cases. Among 554 non-functioning PitNETs with surveillance imaging, 20 % recurred after a median follow-up of 3.7 years (IQR 1.6-7.7), with 10 % requiring reoperation. Recurrence was detected in 60 % of cases by five years and 90 % by ten years, of those who recurred. Age was protective, with each additional year reducing risk (HR = 0.98, p = 0.005). Cavernous invasion strongly predicted recurrence (HR = 2.9, p < 0.001). No significant association between ethnicity and recurrence was observed, including among Māori and Pasifika patients. Surgical remission in acromegaly was achieved in 76 % of microadenomas (median follow-up: 4.8 years; IQR: 1.0-6.8) and 75 % of macroadenomas (median follow-up: 4.3 years; IQR: 1.5-8.5). In Cushing's disease, remission was achieved in 82 % of microadenomas (median follow-up: 3.1 years; IQR: 0.8-9.2) and 75 % of macroadenomas (median follow-up: 4.0 years; IQR: 0.6-8.7). These findings demonstrate postoperative outcomes, complication rates, hormonal improvement, and tumour control, comparable to international standards.