Daily Endocrinology Research Analysis
Across endocrinology today, three studies advance fracture risk assessment, endocrine oncology safety, and perioperative aldosteronism care. A multi-faceted CT-based biomechanical test markedly outperforms DXA and FRAX for hip fracture prediction, ICIs double the risk of new-onset type 1 diabetes and quintuple DKA risk, and preoperative MRA in unilateral primary aldosteronism is safe and linked to superior long-term biochemical cure.
Summary
Across endocrinology today, three studies advance fracture risk assessment, endocrine oncology safety, and perioperative aldosteronism care. A multi-faceted CT-based biomechanical test markedly outperforms DXA and FRAX for hip fracture prediction, ICIs double the risk of new-onset type 1 diabetes and quintuple DKA risk, and preoperative MRA in unilateral primary aldosteronism is safe and linked to superior long-term biochemical cure.
Research Themes
- Advanced imaging-based fracture risk stratification
- Endocrine toxicities of immune checkpoint inhibitors
- Perioperative optimization in primary aldosteronism
Selected Articles
1. Improved Prediction of Hip Fracture Using Multi-Faceted Biomechanical Computed Tomography.
Using routine abdominal-pelvic CT scans, a new BCT Risk Score integrating bone, muscle, and soft tissue metrics outperformed DXA hip T-score and FRAX for predicting 5-year hip fracture. In women, c-statistic reached 0.89 and sensitivity 81.4% (vs 47.8% for DXA threshold and 75.9% for FRAX). Similar advantages were observed in men and for 2-year outcomes.
Impact: This test could leverage existing CT scans to identify high-risk patients with markedly improved sensitivity, addressing underuse and limited sensitivity of DXA. It offers an immediately scalable pathway to enhance fracture prevention.
Clinical Implications: Opportunistic BCT assessment on routine CT could be integrated into workflows to flag high-risk older adults for osteoporosis therapy and fall prevention, potentially increasing treatment rates and reducing hip fractures.
Key Findings
- In women, BCT achieved c-statistic 0.89 (95% CI 0.87-0.90) vs BMD 0.81 and FRAX 0.85 for 5-year hip fracture prediction.
- Using thresholds, sensitivity was higher for BCT (≥75: 81.4%) than BMD (T-score ≤ -2.5: 47.8%) and FRAX (hip risk ≥ 3.0%: 75.9%), with comparable PPVs (13.6% vs 15.3% vs 12.7%).
- Model inputs spanned age, femoral strength, trabecular/cortical BMD ratio, muscle area, intramuscular fat, femoral neck volume, hip width, and posterior fat thickness; performance advantages held in men and at 2 years.
Methodological Strengths
- Large healthcare system source population (n=341,364) with separate development (n=3,035) and geographically distinct validation cohort (n=2,124).
- Direct head-to-head comparison with DXA-equivalent BMD and FRAX; robust discrimination metrics and clinically actionable thresholds.
Limitations
- Retrospective design may introduce selection and confounding biases; calibration across different CT scanners/workflows may vary.
- FRAX was calculated without parental fracture history, potentially underestimating its performance.
Future Directions: Prospective implementation studies to test whether BCT-guided care increases osteoporosis treatment initiation and reduces incident hip fractures; cost-effectiveness and health equity assessments across diverse systems.
With the goal of preventing more hip fractures, a next generation of the VirtuOst® Biomechanical Computed Tomography (BCT) test was developed that integrates measurements from a clinical CT scan related to fall risk, impact force, and femoral strength, the three main determinants of hip fracture. Here, we introduce the test and validate it against bone mineral density (BMD) and FRAX®. Our source population from a large healthcare system comprised of 341,364 patients (≥ 65 years) with an abdominal-pelvic CT during care. Using data from 3,035 patients (1,790 with hip fracture), we developed a "BCT Risk Score" (range: 0-100) having input risk factors of age, femoral strength, ratio of trabecular/cortical BMD, muscle area, intramuscular fat, femoral neck volume, hip width, and posterior fat thickness. In a geographically distinct set of 2,124 patients (1,293 with hip fracture), we then compared the BCT Risk Score against a DXA-equivalent hip BMD T-score (lowest hip value, measured from the CT scan by VirtuOst) and FRAX hip fracture risk (with BMD but without parental fracture history) for predicting a first incident hip fracture within five years. For the women, the c-statistic for predicting fracture was higher for BCT (0.89, 95% confidence interval 0.87-0.90) than for BMD (0.81, 0.79-0.84) or FRAX (0.85, 0.83-0.87). Using binary thresholds to identify high-risk patients, sensitivity for BCT (Risk Score ≥ 75) was higher than for BMD (T-score ≤ -2.5) and FRAX (hip risk ≥ 3.0%): 81.4% vs. 47.8% vs. 75.9%, respectively; positive predictive values confirmed comparable high-risk status (BCT 13.6% vs. BMD 15.3% vs. FRAX 12.7%). Similar trends were observed for the men, two-year outcomes, and identifying very-high-risk patients. We conclude that, compared to both BMD and FRAX, the integrative BCT test better predicted hip fracture and its high sensitivity should improve fracture prevention. Although dual energy X-ray absorptiometry (DXA) testing to measure a bone mineral density “T-score” is the clinical standard for diagnosing osteoporosis, it is under-utilized and has only moderate sensitivity for identifying who will fracture and thus benefit from treatment. Addressing this care gap, here we introduce the latest generation of the biomechanical computed tomography (BCT) test. This diagnostic test integrates measurements of bone, muscle, and soft tissue from a routine clinical abdominal-pelvic computed tomography (CT or CAT) scan and in doing so accounts for fall risk, impact force from a fall, and bone strength, the three main determinants of hip fracture. The test was developed and validated using data from 341,364 patients (≥ 65 years) in the Kaiser Permanente healthcare system in Southern California. We found that hip fracture was better predicted by BCT than either bone mineral density T-score or the FRAX® fracture risk calculator. The sensitivity for predicting hip fracture using BCT exceeded 80% at five years, and 85% at two years. Millions of older patients already undergo an abdominal-pelvic CT each year during their clinical care. Use of this new BCT test in these patients, instead of relying on usual-care DXA, could substantially improve fracture prevention.
2. Immune checkpoint inhibitor therapy and risk of type 1 diabetes mellitus in metastatic cancer patients.
In a matched cohort of 50,926 metastatic cancer patients, ICIs doubled the risk of new-onset T1DM and increased DKA risk more than fivefold over a median 2-year follow-up. High-risk features included baseline HbA1c >6.0%, male sex, white race, and dual checkpoint blockade.
Impact: Quantifies rare but serious endocrine toxicities of ICIs at scale, enabling risk stratification and informing monitoring protocols across oncology and endocrinology.
Clinical Implications: Implement baseline and periodic glucose/HbA1c monitoring in ICI-treated patients, with heightened vigilance for DKA symptoms, particularly in high-risk subgroups and those on dual checkpoint blockade.
Key Findings
- After 1:1 matching (n=25,463 per group), ICIs increased T1DM risk (HR 2.35; 95% CI 1.81-3.04) and DKA risk (HR 10.58; 95% CI 4.21-26.59).
- Cumulative incidence: T1DM 0.75% vs 0.32% (RR 2.32), DKA 0.20% vs 0.04% (RR 5.00) for ICI vs non-ICI.
- High-risk subgroups included baseline HbA1c >6.0%, male sex, white race, and dual checkpoint blockade; median follow-up 764 vs 692 days.
Methodological Strengths
- Large, multi-system EHR dataset with robust 1:1 propensity score matching to balance baseline covariates.
- Consistent results across hazard ratios and cumulative incidence; subgroup analyses identify clinically actionable risk factors.
Limitations
- Outcome identification via ICD-10 codes may misclassify T1DM/DKA; lack of autoantibody or C-peptide confirmation.
- Residual confounding and channeling by indication cannot be fully excluded in observational design.
Future Directions: Develop prediction tools to identify high-risk patients pre-ICI; prospective monitoring protocols and mechanistic studies of β-cell autoimmunity under different ICI regimens.
AIMS: To assess the risk of new-onset type 1 diabetes mellitus (T1DM) and diabetic ketoacidosis (DKA) in metastatic cancer patients treated with immune checkpoint inhibitors (ICIs) compared to those receiving non-ICI therapies. METHOD: A retrospective cohort study using TriNetX global electronic health records (2014-2025) from multiple healthcare systems. Adult metastatic cancer patients initiating ICI or non-ICI therapy were included. Patients with preexisting diabetes within 6 months were excluded. After 1:1 propensity score matching, 25,463 patients remained in each group. Outcomes were identified by ICD-10 codes. RESULTS: Median follow-up was 764 days (ICI) vs. 692 days (non-ICI). ICI use was associated with a higher risk of T1DM (HR, 2.35; 95% CI, 1.81-3.04) and DKA (HR, 10.58; 95% CI, 4.21-26.59). Cumulative incidence analyses supported these findings, with ICIs showing higher risks of T1DM (0.75% vs. 0.32%; RR, 2.32 [95% CI, 1.79-3.00]) and DKA (0.20% vs. 0.04%; RR, 5.00 [95% CI, 2.54-9.86]). Subgroup analyses identified elevated baseline HbA1c (> 6.0%), male sex, white race, and dual checkpoint blockade as high-risk factors. CONCLUSION: ICIs significantly increase the risk of T1DM and DKA. These findings highlight the need for vigilant glycemic monitoring in cancer patients treated with ICIs, especially within identified high-risk subgroups.
3. Mineralocorticoid receptor antagonist pre-adrenalectomy in primary aldosteronism.
In 355 unilateral PA patients from a national registry, presurgical MRA (mostly spironolactone) did not increase immediate or short-term postoperative complications but was linked to superior long-term biochemical cure (81.7% vs 57.1%). Multivariable analysis supported an independent association.
Impact: Addresses a common, variably managed perioperative question in PA and provides safety and potential efficacy signals that could inform preoperative standardization.
Clinical Implications: Preoperative MRA can be used to optimize blood pressure and potassium control in unilateral PA prior to adrenalectomy without added postoperative risk and may improve long-term biochemical cure.
Key Findings
- Among 355 patients, 76.9% received preoperative MRA (spironolactone 64.5%, eplerenone 35.5%).
- No group differences in postoperative hyperkalemia, hypoaldosteronism, renal impairment, blood pressure changes, or short-term biochemical outcomes (≤90 days).
- Long-term complete biochemical response was higher with preoperative MRA (81.7%) vs non-pretreated (57.1%; p=0.004); MRA use independently associated with success.
Methodological Strengths
- Registry-based multicenter cohort with detailed clinical and hormonal data pre- and post-surgery.
- Multivariable analysis to adjust for baseline differences; multiple time horizons (immediate, short-term, long-term).
Limitations
- Non-randomized design with potential selection bias; MRA-pretreated patients had more severe baseline disease.
- Adrenal vein sampling guided surgery in only 33.5% overall, which may affect lateralization accuracy and outcomes.
Future Directions: Prospective trials to define optimal timing, agent, and dosing of MRA pre-adrenalectomy and to assess impacts on clinical (blood pressure remission) and biochemical outcomes.
BACKGROUND: Perioperative use of mineralocorticoid receptor antagonists (MRA) in patients with unilateral primary aldosteronism (PA) awaiting surgery is not well standardized. The aim of this study was to compare the risk of postoperative complications and surgical outcomes between PA patients treated with MRA prior to adrenalectomy and those non-pretreated. METHODS: Adrenalectomized patients for unilateral PA from the SPAIN-ALDO registry, with clinical, hormonal and treatment information before and after adrenalectomy, were analyzed. RESULTS: A total of 355 surgically-treated patients were included and 76.9% (n=273) received presurgical treatment with MRA (more commonly spironolactone [64.5%] than eplerenone [35.5%]). Adrenalectomy was guided by lateralization at AVS in 33.5% of the overall cohort, and by imaging in all the other cases. Patients pretreated with MRA had longer duration of hypertension, higher prevalence of hypokalemia and greater aldosterone concentrations than those not pretreated (n=82). No differences in the rate of postsurgical hyperkalemia, hypoaldosteronism, renal function impairment, blood pressure changes and biochemical outcomes were detected between groups in the immediate (≤30 days) and at short-term follow-up (≤90 days) following surgery. At long-term follow-up, patients pretreated with MRA exhibited better postsurgical biochemical outcomes (81.7% had complete biochemical response vs. 57.1% of non-pretreated patients; p=0.004). In the multivariable analysis, the use of MRA prior to adrenalectomy was independently associated with a successful postsurgical biochemical response. CONCLUSION: Preoperative MRA therapy can be safely introduced to control blood pressure and potassium levels in patients with PA awaiting surgery, without increasing the risk of postoperative hyperkalemia, hypoaldosteronism, renal impairment, or hypotension.