Background: Marijuana use is rising globally amid expanding legalization. Although prior studies suggest potential epigenetic effects, its relationship with systemic biological aging remains unclear. This study aimed to investigate the association between marijuana use and accelerated biological aging, and to preliminarily explore the mediating role of metal exposure.
Methods: We analyzed data from 12,806 U.S. adults aged 20–59 years from the 2005–2018 NHANES. Marijuana use was self-reported via structured interviews. Biological age (BA) was estimated using two validated algorithms, PhenoAge and KD-BioAge. Biological aging acceleration was then defined as the difference between BA and chronological age (CA), with positive values indicating accelerated biological aging (e.g., PhenoAge delta = PhenoAge − CA). Survey-weighted linear regression models, subgroup and joint exposure analyses, and mediation models assessing serum and urinary metals were performed.
Results: Current marijuana use was significantly associated with accelerated biological aging for both PhenoAge delta (β = 0.72, P < 0.001) and KD-BioAge delta (β = 0.36, P = 0.002), independent of demographics, lifestyle, and clinical factors. Subgroup and joint exposure analyses supported consistent associations, with additive effects observed among concurrent marijuana and tobacco users. Mediation analysis identified blood cadmium as a partial mediator, accounting for 15.6% and 8.3% of the total effects on PhenoAge and KD-BioAge acceleration, respectively.
Conclusions: This study provides robust epidemiological evidence linking marijuana use to accelerated biological aging, with cadmium exposure as a potential mechanistic contributor. These findings underscore the need for greater public health attention to the long-term physiological impacts of marijuana consumption.
Keywords: Marijuana; Biological aging; Serum cadmium; Mediating effect

Introduction:
In recent decades, there has been a significant increase in drug-related problem (DRP)-associated hospital admissions, primarily due to undesirable outcomes such as frequent outpatient visits, prolonged hospital stays, and increased healthcare costs. DRPs are a major concern as it negatively impact patient health outcomes. Clinical pharmacists can help ensure safer and more effective treatment plan and play a significant role among the healthcare team.

Aim:
To provide safe and effective treatment by identifying and managing drug-related problems in patients.

Methods:
Concurrent hospital-based cohort study was conducted at JSS Hospital, Mysuru during August 2024 to January 2025. Total of 122 patients of all ages and genders diagnosed with refractory epilepsy were included. Data were collected from case sheets, treatment charts, and patient or caregiver interviews. Hepler and Strand’s classification, World Health Organisation - Uppsala Monitoring Centre Causality Assessment, Naranjo ADR Probability Scale, Modified Hartwig and Siegel ADR Severity Assessment, and criteria for determining ADR predictability and preventability (Modified Shumock and Thornton) tools were used for the study.

Results:
Total of 320 DRPs were identified: 190 (59.3%) were drug-drug interactions (DDIs) and 108 (33.75%) were adverse drug reactions (ADRs). Of the DDIs, 80.4% were accepted, 19.4% were not; 50% required a change in frequency, and 82.6% were of moderate significance. Among the ADRs, 60.19% occurred in males, 28.7% were due to levetiracetam, and 62.96% were of moderate severity (Level 3). Based on assessments, 90.74% were predictable, 62.04% were not preventable, 66.66% were classified as "possible" by WHO-UMC, and 51.85% by Naranjo. Overall, 53.13% of DRPs were resolved.

Conclusion:
This study highlights the role of clinical pharmacists in optimizing therapy and reducing irrational prescribing. Although personalized care can reduce prescription volume, improper DRP management contributes to longer hospital stays and poorer health outcomes.

Keywords: Anti-seizure drugs, Drug-drug interaction, Drug-related problems, Refractory epilepsy

Background: Gabapentinoids, such as gabapentin and pregabalin, are widely recognized as first line therapies for managing neuropathic pain. Despite their benefits, these medications pose safety concerns, including sedation, dizziness, and cognitive impairment.
Aim: To evaluate the prescribing trends for gabapentin and pregabalin among patients taking opioids for their pain management.
Method: This cross-sectional study was conducted from 2010 to 2020 using the prescription databases of a tertiary hospital in Malaysia. All prescriptions for gabapentinoids that were issued to patients treated with opioids for their pain management during the study period were included. Annual number of prescriptions and prescriptions per patient were measured in repeat cross-sectional estimates. Descriptive statistics and linear trend analysis were performed using Stata version 15. Ethical approval for this study was secured from the Medical Research Ethical Committee, Ministry of Health Malaysia (NMRR-16-2135-33068).
Results: A total of 6.7% (n=2363/35237) of opioid patients were prescribed with gabapentinoids during the study period. These patients were issued with 10676 gabapentinoid prescriptions (gabapentin: 89.2%, n=9527/10676 and pregabalin 10.8%, n=1149/10676). Gabapentinoid prescriptions increased from 479 in 2010 to 1367 in 2020, representing a 185.3% increase (P<0.001). A similar trend was observed for gabapentin alone (from 468 in 2010 to 1115 in 2020, 135.5% increase, P<0.001) and pregabalin alone (from 11 in 2010 to 252 in 2020, 2190.9% increase, P<0.001).
Conclusion: Gabapentinoid prescriptions among opioid users in Malaysia significantly increased from 2010 to 2020, with gabapentin remaining the most prescribed medication, while pregabalin use rose sharply. These trends highlight the need for careful monitoring and call for action to ensure safe prescribing practices.

Introduction
Preclinical evidence indicates that glucagon-like peptide 1 receptor agonists (GLP-1 RAs) may reduce alcohol and drug abuse. However, large scale real world evidence is lacking, highlighting the need to estimate treatment effects specifically among those receiving GLP 1 RAs.

Aims
To evaluate the effect of alcohol and drug abuse in patients with diabetes receiving GLP-1 RAs.

Methods
We emulated a target trial emulation from the National Health Insurance Research Database (NHIRD). We included patients diagnosed with diabetes receiving GLP-1 RAs or dipeptidyl peptidase 4 (DPP-4) inhibitors between 2013 and 2021. We assigned them at the date of their first prescription, which was defined as the index date. Propensity scores with fine stratification weights were used to generate similar probability of treatment assignment between groups. Specifically, we calculated the average treatment effect on the GLP-1 RA weights. The primary outcomes were alcohol or drug abuse. Additionally, the secondary outcome was bipolar disorder. Cox proportional hazards models were used to estimate the effect of GLP 1 RAs, generating hazard ratios (HRs) and 95% confidence intervals (CIs).

Results
Of 493,788 patients receiving GLP-1 RAs or DPP-4 inhibitors, the mean age was 46.11 ± 12.84 years and 55.4% male. Use of GLP 1 RAs was associated with a trend toward reduced risk of alcohol abuse (HR 0.60; 95% CI 0.36–1.02) and drug abuse (HR 0.34; 95% CI 0.06–3.77). Furthermore, GLP-1 RAs was associated with a reduced risk of bipolar disorder (HR 0.82; 95% CI 0.48–1.38). These results remained consistent across subgroup and sensitivity analyses, further strengthening the possible causation.

Conclusions
Our findings suggest that treatment with GLP 1 RAs may be associated with a reduced risk of alcohol and drug abuse, as well as bipolar disorder, compared with DPP 4 inhibitors.

Keywords GLP-1 RAs, alcohol and drug abuse, bipolar disorder

Introduction: Dementia is a neurodegenerative condition that impairs daily functioning. Antipsychotics (APs) are frequently prescribed to manage behavioral symptoms in dementia; however, their potential role in increasing fall and fracture risk remains unclear due to inconsistent evidence.
Aims: To investigate the risk of falls and fractures associated with AP use in dementia people.
Methods: PubMed, Scopus, Cochrane Library, and ClinicalTrials.gov were searched up to January 23, 2025. Randomized controlled trials (RCTs) and observational studies reporting fall or fracture risk in dementia people using APs versus placebo or non-use were included. Pooled risk ratios (RRs) and 95% confidence intervals (CIs) were calculated using a random-effects model, with RCTs and observational studies analyzed separately.
Results: Twenty-four studies (10 RCTs and 14 observational studies) met the inclusion criteria. Findings from the RCTs indicated that AP use did not significantly increase the risk of falls compared to placebo (pooled RR=1.01, 95% CI: 0.84-1.21, p=0.952, I²=0.0%), which was consistent with the results from observational studies (pooled RR=1.10, 95% CI: 0.79-1.51, p=0.575, I²=80.5%). However, subgroup analysis revealed that typical APs were significantly associated with increased risk of falls (pooled RR=2.08, 95% CI: 1.32-3.26, p<0.001, I²=0.0%). For fractures, RCTs showed no significant increase in risk (pooled RR=0.88, 95% CI: 0.25-3.09, p=0.847, I²=46.6%). In contrast, observational studies demonstrated a significant association between AP use and increased fracture risk (pooled RR=1.44, 95% CI: 1.27-1.64, p<0.001, I²=96.6%), including both typical (pooled RR=1.57, 95% CI: 1.03-2.41, p=0.04, I²=26.7%) and atypical APs (pooled RR=1.23, 95% CI: 1.07-1.43, p<0.001, I²=0.0%)
Conclusions: Overall, APs were not associated with increased fall risk, except typical APs. Observational data suggested a higher fracture risk with both typical and atypical APs. These findings underscore the need for cautious prescribing of typical APs and further studies to assess long-term safety of APs in dementia people.

Keywords: Dementia, Antipsychotics, Fractures

Introduction
Exploratory post-hoc subgroup analyses in the AEGIS-II trial suggested that the efficacy of CSL112 – human apolipoprotein A-I (apoA-I) intravenously infused in disc-shaped high-density lipoprotein (HDL) particles – may be influenced by baseline low-density lipoprotein cholesterol (LDL-C) levels.

Aims
To genetically evaluate the potential role of LDL-C in apoA-I and cardiovascular outcome associations.

Methods
We performed drug-target Mendelian randomization analyses in 339,390 European-ancestry participants from the UK Biobank. We leveraged the cis-acting variant (rs12225230) within APOA1 to mimic the pharmacological effects of apoA-I infusion (CSL112). Outcomes included ischemic heart disease, myocardial infarction, stroke, and their composite. To address potential horizontal pleiotropy and quantitative inference limitations, we applied the Katan approach. Four analyses were performed: 1) Primary associations of rs12225230 (apoA-I) with cardiovascular outcomes; 2) LDL-C stratification (≥ 100 mg/dL vs < 100 mg/dL), mirroring the AEGIS-II trial post-hoc design; 3) Evaluation of LDL-C effect modification via an interaction term (rs1222530 × LDL-C), adjusting for potential collider bias and power preservation; and 4) Statin use stratification to address residual confounding.

Results
The instrument (rs12225230) demonstrated significantly increased circulating apoA-I and HDL-C levels, but no concomitant association with any cardiovascular outcome, including the LDL-C stratified analyses (≥ 100 mg/dL vs < 100 mg/dL). These genetic results directly contradict the AEGIS-II’s post-hoc signal of reverse association. Consistent null associations were observed in interaction models (rs1222530 × LDL-C), and statin use strata. The comprehensive genetic evidence therefore coherently indicates that the effects of apoA-I are independent of LDL-C.

Conclusions
These analyses provide consistent genetic evidence that circulating apoA-I concentrations do not associate with any of the cardiovascular outcomes, even when baseline LDL-C ≥ 100 mg/dL. The protective effects attributed to CSL112 infusion in AEGIS-II’s post-hoc subgroup analysis are likely reflecting residual confounding rather than causal LDL-C modulation.

Introduction: Choice of biologic therapy for plaque psoriasis (PsO) requires simultaneous consideration of various patient characteristics. However, their complexity and diversity render understanding patient characteristics and making informed decisions on biologic choice challenging.
Aims: To identify PsO patient groups based on similarities in characteristics using hierarchical clustering.
Methods: This study utilized data from Medical Data Vision, Japan, an administrative database, spanning 2015.01.01 to 2022.12.31. Adult patients with PsO (ICD-10=L40.0) initiating biologics (TNF-α, IL-12/23, IL-17, and IL-23 inhibitors) were included. Patient characteristics encompassed demographics (age, sex), treatment history (topicals, systemic agents, phototherapy) and comorbidities. Hierarchical clustering (Gower distance, Ward’s criterion) based on patient characteristics was performed. Pearson’s Chi-square test was used to compare initial biologic therapies across clusters.
Results: Overall, 3787 eligible patients (mean age 56.4 years, 68.2% male) were grouped into four clusters, comprising 19.5%, 43.6%, 30.2% and 6.7% of patients, respectively. Clusters 1 and 2 include younger patients (mean age 52.3 and 55.3 years), while those in Clusters 3 and 4 were older (59.0 and 63.1 years). The proportion of male patients was lowest in Cluster 2 (54.6%) and higher in others (>70%). Treatment history varied, with Cluster 1 receiving minimal (no topical; 15.2% systemic), Cluster 4 moderate (76.7% topical; 31.2% systemic), and Clusters 2 and 3 substantial topical treatment (>95% topical; 23.1% and 87.0% systemic, respectively). Comorbidities were most prevalent in Cluster 4 (75.5%) and less prevalent in others (6.1%-31.9%). Initial biologic therapies varied significantly (p=0.026), with Cluster 1 preferred IL-17 inhibitors (39.4%), whereas others favored IL-23 inhibitors (39.9%-41.8%).
Conclusions: We identified four distinct clusters based on demographics, treatment history, and comorbidities. Hierarchical clustering consolidated multiple patient characteristics, thus simplifying their diversity and offering an innovative, holistic perspective for understanding biologic therapy decision making for PsO patients in a real-world setting.
Keywords: Hierarchical clustering, Biologics, Plaque psoriasis

Introduction: Volume-based procurement (VBP) policy reshapes drug markets, yet its differential impact across hospital levels is unclear. Understanding VBP's effects on lipid-lowering drug procurement patterns is critical for healthcare policy refinement. This study addresses variations in policy implementation efficiency across healthcare settings in Jiangsu Province.
Aims: We evaluated VBP-induced changes in procurement expenditures, utilization, and costs across hospital tiers. Our objective was to inform drug policy optimization.
Methods: Procurement data (October 2019–September 2023) were analyzed using defined daily doses (DDDs) and defined daily cost (DDC). Mixed-effects models compared pre- and post-VBP changes, adjusting for covariates.
Results: Statins dominated the market in Jiangsu Province (rosuvastatin DDDs: 748 million). Post-VBP, expenditures and DDC decreased by 53.9% and 35.4%, respectively, with primary hospitals showing the largest expenditure reduction (61.6%) and secondary hospitals the greatest DDC decline (53.9%). DDDs increased significantly in primary care settings (e.g., pitavastatin +239.8% in secondary hospitals) but decreased for some drugs in tertiary hospitals (amlodipine/atorvastatin -7.3%). Mixed-effects models confirmed that VBP significantly reduced expenditures (OR = -1.07, p < 0.001) and DDC (OR = -2.70, p < 0.001). After covariate adjustment, expenditure reductions for rosuvastatin and atorvastatin narrowed, ezetimibe expenditures increased (OR = 0.13, p = 0.002), and pitavastatin usage declined ( OR = -0.10, < 0.001). Changes in amlodipine/atorvastatin and ezetimibe lacked statistical significance due to short VBP implementation periods. Tertiary hospitals demonstrated the strictest policy adherence with the largest reductions (p < 0.001).
Conclusion: Jiangsu's lipid-lowering drug structure aligns with guidelines (statin-based, moderate-intensity preference). VBP reduced costs with tier-specific variations where tertiary hospitals prioritized generics while primary hospitals showed cost-driven shifts. Continuous tier-specific optimization is recommended to provide insights for healthcare reform.
Keywords Lipid-lowering drugs, Drug utilization, Defined daily dose system, Defined daily cost, Mixed effects model

Introduction: Effective management of Type 2 Diabetes Mellitus(T2DM) depends on self-care practices, including glucose monitoring, dietary management, physical activity, and emotional well-being. However, sociodemographic, therapeutic, and psychological factors often hinder glycemic control and reduce quality of life.

Aims: This study evaluates self-management practices using the Diabetes Self-Management Questionnaire (DSMQ) and emotional distress associated with diabetes using the Problem Areas In Diabetes scale (PAID-5). It also aims to identify influencing factors, and to provide targeted counselling to patients with low DSMQ scores.

Methods: A prospective observational study was conducted using DSMQ questionnaire. Sociodemographic and therapeutic data were collected through interviews and medical records. DSMQ scores categorized as adequate (≥6) or inadequate (<6). Emotional distress was evaluated using the PAID-5, with scores ≥8 indicating distress. Patients with low DSMQ scores received physician-supervised counselling and were followed up via telephone after three months to reassess self-management.

Results: The analysis of 140 T2DM patients revealed significant associations between self-management practices and several factors, including age group 40–59 years (p = 0.032), urban residence (p = 0.029), private-sector employment (p = 0.041), middle socioeconomic status (p = 0.018), regular exercise (p = 0.001), and absence of comorbidities (p = 0.047), which were linked to higher DSMQ scores. No statistically significant associations were found for gender, BMI, education, dietary habits, smoking, alcohol, or type of diabetes medication. Targeted counselling led to significant improvement in DSMQ scores at three months.

Conclusions: This study highlights the need to assess sociodemographic, therapeutic, and emotional factors to improve self-management in T2DM patients. Targeted counseling significantly improved self-care and glycemic control in patients with low DSMQ scores. Limitations include its single-center design and small size, limiting generalizability. Future studies could include longer follow up and integrate digital health tools to monitor self-management behaviors.

Keywords: Type 2 diabetes mellitus, DSMQ, PAID-5, Self-management

Introduction: CDK4/6 inhibitors have significantly improved survival in patients with hormone receptor-positive (HR+) advanced breast cancer. As their use expands and treatment durations increase, concerns about potential cardiotoxicity have emerged. It not only impairs quality of life but may also offset survival gains by increasing the risk of cardiovascular-related mortality. However, real-world studies assessing the cardiotoxicity of CDK4/6 inhibitors remain limited.

Aim: This study aims to estimate and compare the risk of cardiotoxicity associated with different CDK4/6 inhibitors.

Methods: This retrospective cohort study was conducted at a tertiary medical center in Taipei, Taiwan, utilizing electronic medical records from July 1, 2017, to September 30, 2024. Women with HR+ breast cancer who initiated CDK4/6 inhibitors were included. Cardiotoxicity was defined as any occurrence of cardiomyopathy, takotsubo syndrome, pericardial disease, heart failure, arrhythmias, QT prolongation, acute coronary syndrome, or venous thromboembolism. An as-treated analysis was performed, and a multivariable Cox proportional hazards model was used to estimate the risk.

Results: The study included 175 patients treated with palbociclib, 175 with ribociclib, and 74 with abemaciclib. The incidence rates of cardiotoxicity were 6.83, 15.15, and 6.44 per 100 person-years, with median follow-ups of 312, 336, and 186 days, respectively. Compared with palbociclib, ribociclib was associated with a significantly higher risk of cardiotoxicity (adjusted hazard ratio [aHR] 2.20; 95% confidence interval [CI], 1.14–4.25), whereas abemaciclib did not show a significant difference (aHR 0.66; 95% CI, 0.21–2.10).

Conclusion: Ribociclib was associated with a higher risk of cardiotoxicity, particularly cardiac rhythm abnormalities, consistent with patterns observed in prior trials. As CDK4/6 inhibitors expand into early-stage breast cancer treatment, these findings underscore the need for cardiovascular monitoring to ensure patients’ long-term safety and treatment success. Given the single-center nature of this study, further multi-center research is warranted to confirm these observations.

Keywords: breast cancer, CDK4/6 inhibitors, cardiotoxicity

Introduction: Acute Kidney Injury (AKI) affects approximately 26% of hospitalized children globally, contributing to prolonged hospitalization and increased mortality. Dexamethasone, a potent corticosteroid widely used for inflammatory conditions, exhibits conflicting renal effects: preclinical studies suggest it worsens renal injury via tubular epithelial cell injury, while clinical evidence links it to renal protection in some cohorts. This discrepancy necessitates real-world evidence in pediatric cohorts.
Aims: To examine the association between dexamethasone and AKI risk in hospitalized children, and identify clinical risk factors and high-risk subgroups for AKI post-dexamethasone exposure.
Methods: This retrospective cohort included hospitalized children (1–18 years) with ≥2 hospital days in 2022. AKI was defined according to the KDIGO criteria (2012). Follow-up spanned admission to AKI onset/discharge/death (median 15 days, IQR 9–24). Patients without available serum creatinine (Scr) tests or with severe renal impairment on admission were excluded. Baseline characteristics were balanced using inverse probability weighting (IPW) based on propensity scores. Multifactorial logistic regression was used to assess the association between the use of dexamethasone and the risk of AKI. Subgroup analyses and propensity score matching-based sensitivity analysis were conducted using all covariates. Variables with ≤5% missing data were multiply imputed, while those exceeding 5% missingness were excluded.
Results: Among 1,978 patients (685 dexamethasone-exposed, 1,282 non-exposed), AKI incidence was 14.0% (96/685) vs. 6.6% (84/1,282). Dexamethasone was an independent protective factor (IPW-adjusted OR 0.47, 95% CI 0.32–0.69, P<0.01). Subgroup analysis identified male sex, younger age, low baseline Scr, diuretic use, and absence of fever/pulmonary infection as factors associated with reduced AKI risk (P<0.05). Sensitivity analysis confirmed protection (Adjusted OR 0.54, 95% CI 0.29–0.99, P=0.05).
Conclusions: Dexamethasone may reduce AKI risk in hospitalized children, challenging preclinical safety assumptions. Large-scale prospective studies are warranted to validate these findings.
Keywords: Dexamethasone; Acute Kidney Injury; Hospitalized Children; Inverse Probability Weighting

Introduction:
Non-alcoholic fatty liver disease (NAFLD) is the most prevalent chronic liver disorder worldwide and is increasingly recognized as a significant public health challenge. While metabolic risk factors are well established, association of calcium (Ca), magnesium (Mg), phosphorus (P), and selenium (Se) with NAFLD risk remains insufficiently characterized. Understanding these associations could inform targeted nutritional strategies for NAFLD management.
Aim:
To systematically evaluate and quantify the association between circulatory and dietary Ca, Mg, P, and Se with risk of NAFLD
Methods:
A comprehensive literature search was conducted in PubMed, Embase, and Web of Science from inception through June 2025 to identify observational studies (cross-sectional/case-control) published in English, examining associations between micronutrient levels and NAFLD reported multivariable-adjusted odds ratios (ORs). Study quality was assessed using the Newcastle-Ottawa Scale. Random-effects meta-analyses were performed using R. Sensitivity analyses were conducted using leave-one-out methods, and potential publication bias was assessed using funnel plots and Egger’s test.
Results:
17 studies comprising 103,242 participants were included from 3303 screened records. The overall pooled OR was 1.05 (95% CI: 0.86–1.28), with substantial heterogeneity (I² = 88.7%). Subgroup analyses indicated that elevated blood calcium levels were associated with an increased risk of NAFLD (pooled OR = 1.32; 95% CI: 1.00–1.75), while higher dietary magnesium intake was inversely associated with NAFLD risk (pooled OR = 0.59; 95% CI: 0.39–0.91). A borderline inverse association was observed for lower serum calcium. No significant associations were identified for phosphorus, selenium, or the Ca/Mg ratio. Sensitivity analyses supported the robustness of the findings, and no significant publication bias was detected (Egger’s test: t = 1.71, p = 0.0992).
Conclusion:
Elevated blood calcium appears to confer increased risk, while dietary magnesium may offer protective benefits. These findings may inform future dietary recommendations and preventative strategies in NAFLD management.
Keywords: NAFLD, Micronutrients, Risk assessment

Introduction: Tacrolimus is a first-line immunosuppressant used in renal transplant patients but exhibits a narrow therapeutic index and high inter-individual variability.
While therapeutic drug monitoring using trough concentrations (C0) is common, it may not consistently reflect true drug exposure. Area under the concentration-time curve (AUC) is a more realistic metric, and Bayesian estimation methods often improved precision for AUC estimation using priori pharmacokinetic knowledge.

Aim: To focus on the reliability in using Bayesian approaches over other traditional therapeutic drug monitoring approaches for the estimation of AUC and thereby recommending the Tacrolimus dosage adjustment in renal transplant patients.

Methods: We conducted systematic literature search on PubMed, Scopus & Embase from inception to August 2024, for studies utilizing Bayesian approaches for estimating Tacrolimus exposure in renal transplant patients. Studies were included if they employed Bayesian techniques for AUC estimation or dose adjustment using population pharmacokinetic (PopPK) models or limited sampling strategies. Data was extracted, appraised using the CACPK tool, and summarised.

Results: A total of 30 articles were included. The majority reported that Bayesian methods, particularly Maximum a Posteriori Bayesian Estimation (MAP-BE), provided superior AUC prediction compared to traditional methods. Tools such as ISBA and software like NONMEM were frequently used. Most studies supported the inclusion of trough and post-dose samples >2 hours for accurate predictions.

Conclusion: Bayesian approaches demonstrate enhanced accuracy and efficiency in tacrolimus dose adjustment for renal transplant patients, outperforming conventional TDM strategies and showing potential for broader clinical application.

Keywords: Bayesian Approach, Tacrolimus, Renal Transplantation, MAP-BE

Introduction:
Opioids are potent analgesics commonly used for acute and cancer pain. However, they are being increasingly used for long-term non-cancer pain, which is associated with adverse outcomes.
Aims:
To determine the association between patients’ beliefs and persistent opioid use (POU: opioid use for & >90 days).
Methods:
An anonymous online survey was administered between June 2024 and November 2024, to recent opioid users (within 12 months of study participation) in New Zealand, without a prior diagnosis of opioid misuse or dependence. All participants were asked for information related to demographics, patterns of opioid use, the Beliefs about Medicines Questionnaire (BMQ-specific), and agreement on experiences from opioid therapy. The BMQ-specific was analysed via the necessity-concern differential (NCD, -4 to 4).
Results:
Of 777 responses, 686 responses were included in the analysis, with 61.0% (n=419) considered to be POU. We found that for every point increase on the NCD, there was an increased association of POU (aOR= 1.25; 95% CI: 1.01-1.55, p=0.036). A total of 619 participants completed the section on experiences from opioid therapy, nearly half of the participants agreed that they have experienced a side-effect, with around one in four agreeing that opioids impaired their ability to drive or operate machinery. In addition, approximately one in ten participants agreed that opioids may have resulted in taking time off work or school.
Conclusion:
This is the first NZ study to demonstrate a significant association between patient beliefs and POU. Overall, the study highlights a reduction in treatment concerns with ongoing opioid use, which may drive future addiction and dependence on opioids. Our study further described how patients’ adverse experiences related to opioid use may impact areas outside of health.

Keywords: opioid, medication beliefs

Introduction: While several studies have been conducted to investigate the effects of chimeric antigen receptor (CAR) T-cells, the overall impact of these remained unclear.

Aims: This study investigates the effects of CAR T-cells in patients with relapsed or refractory (R/R) acute myeloid leukemia (AML).

Methods: A comprehensive search of PubMed, Cochrane CENTRAL, EMBASE, ClinicalTrial.gov, and Scopus from inception to January 31, 2025 identified relevant studies. Evaluated outcomes included complete response (CR), partial response (PR), overall response rate (ORR), relapse, overall survival (OS), stable disease (SD), progressive disease (PD), persistence, post-hematopoietic stem cell transplantation (post-HSCT), myelosuppression, cytokine release syndrome (CRS), and immune effector cell-associated neurotoxicity syndrome (ICANS). A random-effects model meta-analysis was performed.

Results: Twenty-three studies were included, and the meta-analysis revealed that 68% (0.68 [95% CI: 0.52, 0.84]) of patients achieved CR, while none (0.00 [95% CI: 0.00, 0.08]) showed PR. The ORR was 73% (0.73 [95% CI: 0.57, 0.86]); however, approximately 41% (0.41 [95% CI: 0.25, 0.58]) of patients experienced relapse, and the OS was 59% (0.59 [95% CI: 0.41, 0.77]). Additionally, 31% (0.31 [95% CI: 0.20, 0.42]) of patients exhibited PD, while 27% (0.27 [95% CI: 0.11, 0.45]) demonstrated SD. CAR T-cell persistence was reported in nearly all patients (97%) (0.97 [95% CI: 0.79, 1.00]), and 40% (0.40 [95% CI: 0.23, 0.58]) received post-HSCT. Myelosuppression was observed in most patients (95% [0.95 [95% CI: 0.75, 1.00]), CRS occurred in 91% (0.91 [95% CI: 0.73, 1.00]), and ICANS was detected in 5% (0.05 [95% CI: 0.00, 0.13]).

Conclusions: Although CAR T-cell therapy demonstrated measurable anti-leukemic efficacy in R/R AML, the high incidence of treatment-related toxicities underscored the imperative for further clinical trials.

Keywords: Acute myeloid leukemia; Chimeric antigen receptor; Systematic review

Introduction: Hypertension and diabetes are increasingly prevalent chronic diseases. Although prior studies examined the effects of oral hypoglycemic agents (OHAs) on blood pressure, their impact on antihypertensive drug use in real-world settings remains unclear.
Aim: To evaluate the changes of antihypertensive drugs, use before and after the initiation of second-line OHAs —including sulfonylureas (SUs), dipeptidyl peptidase-4 inhibitors (DPP-4is), thiazolidinediones (TZDs), and sodium-glucose cotransporter-2 inhibitors (SGLT2is).
Methods: The retrospective cohort study was conducted using the Korean Health Insurance Review and Assessment database. Patients aged ≥50 years with newly diagnosed type 2 diabetes mellitus in 2019 who initiated a second-line OHA within six months of diagnosis were included. Eligibility required ≥ 90 days of OHA use and ≥180 days of follow-up. Changes in the number of antihypertensive drugs were assessed by comparing the proportions of patients with increased versus decreased use across OHA classes. Initiation and discontinuation rates of specific antihypertensive drug classes, with 95% confidence intervals (CIs), were estimated.
Results: Among 2,566,086 eligible patients, 8,085 initiated a second-line OHA in 2019. The cohort included 5,604 DPP-4i users, 3,014 SU users, 1,045 SGLT2i users, and 775 TZD users. Following initiation of second-line OHAs, the number of antihypertensive drugs increased in the DPP-4i (22.6% increase vs. 20.4% decrease) and SU groups (23.6% vs. 20.3%), decreased in the SGLT2i group (24.3% vs. 16.3%), and showed no net change in the TZD group (20.8%). Among SGLT2i users, significant discontinuation rates were observed for ARB/ACE inhibitors (OR: 1.45, 95% CI:1.12–1.87), CCB (OR: 1.57, 95% CI: 1.19–2.07) and thiazide diuretics (OR: 1.76, 95% CI: 1.17–2.63).
Conclusions: Using real-world data from Korea, this study demonstrated that the previously reported
blood pressure-lowering effects of SGLT2 inhibitors significantly influenced the discontinuation and
adjustment of antihypertensive prescriptions, differentiating them from other second-line OHAs.
Keywords: SGLT2 inhibitors, Real-world evidence, Cohort

Introduction:
Emerging evidence for benefits beyond glycaemic control for new glucose-lowering drugs (GLDs), such as Sodium-Glucose Cotransporter-2 inhibitors (SGLT2i) and GLP-1 receptor agonists (GLP-1RA), has transformed the landscape of GLD use. However, the trends and patterns of GLD use in dementia and Parkinson’s disease (PD) populations remains underexplored.
Aims:
To examine the trends and patterns of GLD use in Australia (2015-2024), focusing on individuals with dementia and PD.
Methods:
We analysed prescription claims for people aged 50–99 using a 10% sample of the Australian Pharmaceutical Benefits Scheme data. Annual prevalence and incidence were calculated for each GLD class across the general population, and within the dementia and PD cohorts, defined using prior dispensing of dementia- and PD-specific medications, respectively. Log-binomial regressions were used to assess predictors of GLD initiation by cohort in 2024.
Results:
Between 2015 and 2024, we identified 196,573 prevalent and 162,539 incident GLD users. The prevalence and incidence of SGLT2i use increased markedly (incidence; general population: 4.6→16.1/1,000 people; dementia: 3.5→15.9/1,000; PD: 12.5→ 25.2/1,000), with similar trend for GLP‑1RA (general population: 1.0→6.9/1,000; dementia: 0.9 →4.7/1,000; PD: 3.0→9.7/1,000). In contrast, sulfonylurea and thiazolidinedione use declined across all groups. In 2024, compared to the general cohort, individuals in the PD cohort were more likely to initiate DPP‑4 inhibitors (RR = 1.2, 95% CI: 1.1–1.4) and GLP‑1RA (RR = 1.2, 95% CI: 1.1–1.4); those in the dementia cohort were less likely to initiate GLP‑1RA (RR = 0.8, 95% CI: 0.6–0.9) and SGLT2i (RR = 0.9, 95% CI: 0.8–1.0).
Conclusion:
Overall, GLD use has shifted toward newer GLDs, but the pattern of GLD use in people with dementia or PD differed from the general population. These findings highlight the urgent need for examining drivers of these differences and potential guideline implementation tailored to neurodegenerative populations.
Keywords:
Glucose-lowering drugs; Dementia; Parkinson’s disease

Introduction
Gram-negative infections in children are increasingly associated with longer hospital stays, high morbidity, and mortality. Pediatric studies that relate bacterial etiology, clinical presentations, and outcomes are few.
Aims
To study clinical characteristics, antibiotic utilization, risk factors, and outcomes of pediatric patients with Gram-negative infections.
Methods
A retrospective study was conducted in patients aged between one month and 18 years with culture-confirmed Gram-negative infections from January 2022 to December 2023 in a tertiary care teaching hospital. Demographics, clinical characteristics, bacterial etiology, and antibiotic utilization were collected from medical records. Data were analysed using SPSS v.20, and logistic regression was used to identify the risk factors.
Results
There were 227 gram-negative isolates from 209 patients, including 114 males and 95 females with a mean age of 6.34±5.97. The common conditions recorded were urinary tract infection (59.81%), sepsis/septic shock (24.40%), bloodstream infections (10.53%), and pneumonia (5.26%). The common isolated pathogens were E.coli (46.89%), Klebsiella (25.84%), and Acinetobacter (9.57%). Among these, multidrug-resistant (MDR) (16.27%), extensively drug-resistant (XDR) (5.26%), and pandrug-resistant (PDR) (5.74%) pathogens were observed. Aminoglycosides (19.88%), 3rd generation Cephalosporins/β-lactamase inhibitors (19.13%), and Sulphonamides (11.90%) were the commonly prescribed antibiotics. Furthermore, there was an increase in the average length of stay and overall mortality rates with the resistant isolates as compared to the sensitive isolates (18.25±13.07 Vs 9.04±6.80 (p<0.01), and 13 (23.63%) Vs 5(3.24%) (p<0.01), respectively). Age (OR=1.629, p = 0.001), mechanical ventilation (OR=15.61, p=0.037), central venous catheter (OR=22.48, p<0.01), malignancy (OR=17.36, p<0.01), and autoimmune diseases (OR=32.31, p=0.010) were identified as risk factors for the development of gram-negative infections.
Conclusion
Gram-negative pathogens cause severe infections, causing increased mortality and hospital stay in children. Understanding the epidemiology of Gram-negative infections at institutional and regional levels is essential for improving management decisions and treatment outcomes.

Key Words
Gram-negative infections, pediatrics, antibiotic usage

Introduction: Chemotherapy toxicities are prevalent in older adults with cancer and often compromise treatment tolerance. The Cancer and Aging Research Group Chemotherapy Toxicity Tool (CARG-TT), developed in 2011, is recommended to assess risks of severe (G3–5) toxicities, but its applicability in older adults in mainland China under current treatment practices is limited.
Aims: This study examined associations between CARG-TT scores and chemotherapy toxicity and intolerance in Chinese patients aged ≥70 years with cancer to provide preliminary evidence for its predictive value.
Methods: In this ongoing prospective cohort study, we enrolled 238 patients aged 70 years or older who initiated new chemotherapy regimens at a single cancer center in Shanghai from December 2024 to June 2025, with a planned 6-month follow-up. Baseline CARG-TT scores were obtained through face-to-face interviews prior to chemotherapy. Outcomes included G3–5 toxicities and chemotherapy intolerance, defined as a composite of discontinuation, dose reduction, delay, or missing doses. Associations between CARG-TT risk groups and outcomes were analyzed using chi-squared tests.
Results: The median age was 74 years (range 70–88); 30.7% were female, 63.4% had gastrointestinal cancer, and 62.2% had stage IV disease. Overall, 78.2% received reduced-intensity chemotherapy; 34.0% received immunotherapy, and 36.6% targeted therapy. Based on CARG-TT, 20.2% were low risk, 55.0% medium, and 24.8% high risk. Overall, 46.2% experienced G3–5 toxicities (including 8 deaths), and 26.1% did not complete planned treatment due to intolerance. G3–5 toxicities occurred in 39.6%, 48.1%, and 47.5% of low-, medium-, and high-risk patients, respectively (P=0.5852). Chemotherapy intolerance occurred in 25.0%, 29.8%, and 18.6%, respectively (P=0.2660).
Conclusions: Patients classified as medium or high risk by CARG-TT appeared more likely to develop severe toxicities, while statistical significance was not observed. Further follow-up is needed to assess the predictive ability of CARG-TT.
Keywords: Geriatric oncology, Chemotherapy toxicity, Tolerance

Introduction: Difficult-to-treat rheumatoid arthritis (D2T RA) is defined by poor disease control despite standard therapies and poses a growing public health challenge. In aging populations like Japan, with rising multimorbidity, understanding the clinical profile and burden of D2T RA is essential. While the European Alliance of Associations for Rheumatology (EULAR) definition of D2T RA has gained international consensus, data on DT2 RA in Asian populations remains scarce.
Aims: This study aimed to compare the clinical characteristics, functional status, and comorbidity burden of D2T and non-D2T RA patients in Japan.
Methods: We conducted a retrospective observational study using electronic health records from a Japanese tertiary care center (2022–2025). RA patients with ≥12 months of follow-up were classified as D2T or non-D2T based on EULAR criteria. Data on demographics, laboratory measurements, disease activity, radiographic and functional status, and comorbidities were analyzed. Statistical comparisons were made using Welch’s t-tests and Fisher’s exact tests.
Results: Of 1,581 RA patients (D2T: n=102; non-D2T: n=1,479), D2T patients were more often female (90% vs. 78%, p=0.002) and diagnosed at a younger age (48 vs. 53 years, p=0.005). On average they exhibited higher disease activity (DAS28-CRP: 3.11 vs. 2.01, p<0.001; clinical disease activity index (CDAI): 13.0 vs. 5.2, p<0.001) and greater functional impairment (HAQ-DI: 0.88 vs. 0.49, p<0.001). Inflammatory markers (CRP, RF), renal function (eGFR), advanced radiographic stage (IV: 42% vs. 15%, p<0.001), and severe functional class (III–IV: 16.9% vs. 5.3%, p<0.001) tended to be worse in D2T patients. History of comorbid lymphoproliferative disorders (16% vs. 5.9%, p=0.007) was more common.
Conclusions: D2T RA patients in Japan are more likely to have younger disease onset, higher inflammatory activity, worse functional status, and greater comorbidity burden. These findings emphasize the need for early identification and tailored intervention strategies for RA patients.
Keywords: rheumatoid arthritis, difficult-to-treat RA, Japan

Introduction: Psychotropics are commonly used in people with dementia. Health service utilisation is important in managing both dementia-related symptoms and multimorbidity. However, patterns and clusters of psychotropic and health service utilisation among people with dementia remain poorly understood in the Australian context.
Aims: To identify factors associated with different clusters of psychotropic and health service use, and assess their association with mortality in Australians with dementia.
Methods: This cohort study utilised linked 2021 Census, death registration, Pharmaceutical Benefits Scheme (PBS) and Medicare Benefits Scheme (MBS) data. People with dementia aged ≥65 years in Australia were included in the study in 2021. Latent class analysis involving 16 variables, including different classes of psychotropics (antipsychotics, opioids, antidepressants, antiepileptics, benzodiazepines and Z-drugs), chronic health service use (such as chronic disease management plans and medication review), mental and physical healthcare (such as allied health and psychiatrist visits). Binary logistic regression was used to identify factors associated with class assignment. Association with 12-month mortality was assessed using Kaplan Meier curves.
Results: Overall, 177,809 people with dementia were included. A five-class model was selected with groups described as (1) low psychotropic use, high chronic disease management (35.4%), (2) low psychotropic use, low health services use (33.9%), (3) high psychotropic use, high chronic disease management (13.4%), (4) high psychotropic use, low health service use (14.5%) and (5) high psychotropic use, high mental health service use (2.8%). People with dementia residing in remote/regional areas were less likely to belong to classes with high health service utilisation. Despite high psychotropic use, the class with high mental health service use was associated with the lowest mortality (p <0.001).
Conclusions: Health services have the potential to improve mortality outcomes in individuals with dementia; however, significant disparities in access to these services persist across Australia.

Keywords: Psychotropics, People with dementia, healthcare utilisation

Introduction:Preclinical studies have suggested that colchicine, a medication widely used for gout, can inhibit tumor cell proliferation, migration, and metastasis across various cancer models. Nevertheless, evidence regarding the reduced cancer risk of colchicine in a real-world setting remains limited.
Aims:To investigate the association between colchicine use and the risk of cancer in patients with gout compared to non-steroidal anti-inflammatory drug (NSAID) use.
Methods:We conducted a target trial emulation study using the Korean National Health Insurance Service-National Sample Cohort (2002–2019). We compared new users of colchicine and NSAIDs for gout in an active comparator, new-user design to emulate random assignment. Follow-up began one year after treatment and continued until cancer diagnosis, all-cause death, or the end of the study (December 31, 2019). The outcome was overall cancer, and analyses were based on an intention-to-treat approach. We used propensity score fine stratification weighting to control confounding. Incidence rates per 1,000 person-years and 95% confidence intervals (CIs) were estimated using Poisson regression, and hazard ratios (HRs) with 95% CIs were estimated using Cox proportional hazards models.
Results:We identified a total of 5,255 patients with gout, including 1,189 colchicine users and 4,066 NSAIDs users. Colchicine users had a mean age of 50.54 years and 78.97% were male, while NSAID users had a mean age of 52.04 years and 68.25% were male. Over a mean follow-up of 12 years, the weighted IRs for overall cancer were 9.36 (95% CI, 7.79–11.14) per 1,000 person-years among colchicine users and 9.50 (95% CI, 8.66–10.40) among NSAIDs users. The crude and adjusted HRs were 0.95 (95% CI, 0.78–1.16) and 0.99 (95% CI, 0.81–1.20), respectively.
Conclusions:Despite preclinical findings suggesting anticancer potential of colchicine, our real-world analysis over a 12-year follow-up period did not support an association between colchicine use and cancer risk in gout patients.
Keywords: Gout, Cancer, Colchicine

Introduction: The use of older antibiotics like colistin has been revisited to combat multidrug-resistant Gram-negative bacterial (MDR-GNB) infections. However, the comparative efficacy of colistin monotherapy versus combination therapy remains unclear in pediatric care.
Aims: To evaluate the clinical efficacy and safety of colistin monotherapy compared to combination therapy in pediatric patients with MDR-GNB infections.
Methods: This retrospective cohort study included pediatric patients under 18 years diagnosed with MDR-GNB infections and treated with intravenous colistin at Chiang Mai University Hospital between 2014 and 2024. Inverse probability weighting (IPW) was used to balance baseline characteristics between monotherapy and combination therapy groups. The primary outcome was 30-day mortality. Secondary outcomes included clinical response, end-of-treatment mortality, and microbiologic response. Univariate and multivariable logistic regression analyses were performed after IPW adjustment.
Results: A preliminary analysis of 128 pediatric patients (45 on monotherapy, 83 on combination therapy) was conducted. The median ages were 3.65 years (range, 0.71–13.19) in the monotherapy group and 2.12 years (range, 0.57–7.37) in the combination therapy group. After IPW adjustment, baseline characteristics were balanced. No significant differences were observed in 30-day mortality (adjusted odds ratio [aOR], 0.22; 95% confidence interval [CI], 0.01–4.07; p = 0.311), end-of-treatment mortality (aOR, 0.49; 95% CI, 0.06–4.35; p = 0.525), clinical response (aOR, 2.02; 95% CI, 0.23–17.85; p = 0.525), or nephrotoxicity (aOR, 1.18; 95% CI, 0.39–3.56; p = 0.764) between groups. However, microbiologic response was significantly higher with combination therapy (aOR, 31.39; 95% CI, 1.34–737.20; p = 0.032).
Conclusions: Colistin-based combination therapy improved microbiologic outcomes but showed similar mortality, clinical response, and nephrotoxicity rates compared to monotherapy. These findings suggest that colistin monotherapy may be as effective as combination therapy. However, future studies with larger sample sizes or prospective randomized controlled trials are needed to determine optimal treatment strategies in pediatric patients.
Keywords: colistin, pediatrics, MDR-GNB

Introduction:
Allergic conjunctivitis (AC) is a common ocular inflammatory condition triggered by IgE-mediated hypersensitivity to environmental allergens. Second-generation H1 antihistamines, such as Bepotastine and Olopatadine, are widely used. However, comparative efficacy data are limited. This meta-analysis evaluates their effectiveness in managing AC symptoms, including itching, hyperemia, and ocular discomfort.

Aims:
To compare the efficacy of Bepotastine besilate 1.5% and Olopatadine HCl 0.1% in relieving AC symptoms using standardized clinical outcomes.

Methods:
A systematic search of PubMed, Embase, Web of Science, and Google Scholar (January–March 2025) identified randomized controlled trials (RCTs) comparing the two drugs in patients with allergic conjunctivitis. Eligible studies were RCTs involving patients with allergic conjunctivitis, Studies using validated symptom scores, and providing data suitable for meta-analysis. Six RCTs (N = 320) met the inclusion criteria.
Efficacy outcomes included the Total Ocular Symptom Score (TOSS), a composite score assessing itching, redness, tearing, and foreign body sensation, as well as individual scores for conjunctival hyperemia, itching, watering, and discomfort. Meta-analysis was performed using a random-effects model. Heterogeneity was assessed using the I² statistic.

Results:
Bepotastine demonstrated significantly greater improvement in TOSS at Day 14 (Mean Difference [MD]: –0.35; 95% CI: –0.53 to –0.18; I² = 0%; p < 0.001), Itching at Week 8 (MD: –0.40; 95% CI: –0.60 to –0.20; I² = 12%; p < 0.001), and ocular discomfort at Week 8 (MD: –0.28; 95% CI: –0.45 to –0.12; I² = 0%; p = 0.001). Hyperemia and watering outcomes favored Bepotastine but were not statistically significant.

Conclusion:
This meta-analysis suggests that Bepotastine offers superior relief of itching and discomfort compared to Olopatadine in AC. These findings are specific to RCTs and should not be generalized to other study designs. Further large-scale trials are warranted.

Keywords: Allergic conjunctivitis, Bepotastine, Olopatadine, meta-analysis, antihistamines

Introduction: Longitudinal analyses of medication-use trajectories are critical for understanding adherence, switching, and discontinuation patterns in pharmacoepidemiology. Group-Based Trajectory Analysis (GBTA) is widely used for its intuitive group-level summaries. However, emerging machine learning approaches, such as mixture Hidden Markov Models (HMMs) or process mining, offer alternative frameworks for capturing complex, real-world prescribing behaviours.
Aims: We aim to conduct a comparative evaluation of three longitudinal clustering approaches to evaluate its use in exploring medication use patterns.
Methods: We conducted a comparative evaluation of three longitudinal clustering approaches using general practice electronic medical records and administrative dispensing data to classify patient medication-use phenotypes. GBTA modelled continuous or count-based outcomes (days covered, switch counts) as polynomial trajectories, assigning individuals to latent adherence groups. Mixture HMMs represented medication use as transitions between discrete states (e.g., on/off medication), estimating subgroup-specific transition matrices and covariate effects on both group membership and transition probabilities. Process mining reconstructed prescribing pathways from time-stamped medication events, enabling visualisation of treatment sequences (first, second, and third-line medications). Each method was assessed for its ability to capture adherence, switching, and discontinuation behaviours, as well as interpretability and suitability for covariate modelling.
Results: GBTA effectively summarised adherence patterns and identified discontinuation timepoints, with the added benefit of incorporating risk factors via multinomial logistic sub-models. Mixture HMMs offered a unified probabilistic framework, capturing dynamic transitions, switching patterns, dwell times, and covariate effects. Process mining revealed rich, real-world pathway variants, including non-linear switching, but required separate clustering and lacked built-in uncertainty quantification or parametric inference.
Conclusions: While GBTA remains foundational for adherence and discontinuation analysis, mixture HMMs provide a robust framework for modelling switching dynamics and covariate effects. Process mining offers valuable exploratory insights into real-world treatment pathways. Method selection should align with study objectives, data structure, and desired interpretability.
Keywords: Time-series clustering, Medication Patterns

INTRODUCTION
Abiraterone and enzalutamide are widely used agents for advanced prostate cancer (PC). Compared to abiraterone, enzalutamide is known to have higher blood-brain barrier permeability, raising concerns about potential neurocognitive risks. However, comparative real-world evidence remains limited.
AIMS
To compare the risk of dementia and PD between patients treated with abiraterone versus enzalutamide.
METHODS
We retrospectively analyzed South Korean nationwide claims data using a new-user, active-comparator design. Individuals aged 40-80 years with PC, who newly initiated either abiraterone or enzalutamide (but not both) between 2018 and 2020 were included. Patients with prior dementia, PD, or cerebrovascular disease from one year before to six months after the index date (first prescription date) were excluded. Propensity score matching (1:1) was employed to balance baseline characteristics, using the following covariates: age, type of National Health Security program enrollment, Charlson comorbidity index, medical history (hypertension, diabetes, dyslipidemia, cardiovascular diseases, peripheral vascular disease, chronic obstructive pulmonary disease, asthma and liver diseases) and concomitant medications (statins, antiplatelet agents, and anticoagulants). Cox proportional hazards models were employed to estimate the risk of dementia and PD, with mortality considered as a competing risk. Models were adjusted for use of anticholinergics, antidepressants, antipsychotics, and chemotherapy.
RESULTS
The matched cohort included 2,692 patients (1,346 per group) with well-balanced baseline characteristics. Adjusted hazard ratios (aHRs) for abiraterone versus enzalutamide (reference) were: all-cause dementia, aHR 0.91 (95% confidence interval [CI]: 0.54-1.54); Alzheimer’s disease, 0.83 (0.45-1.53); vascular dementia, 1.41 (0.08-25.39); other dementias, 0.82 (0.32-2.14); and PD, 2.83 (0.45-17.93).
CONCLUSIONS
No statistically significant differences in dementia or PD risk were observed between abiraterone and enzalutamide. Further large-scale studies with longer follow-up are needed to validate these findings.
This research was supported by a grant from Korea Institute of Drug Safety and Risk Management in 2023.
Keywords: Prostatic Neoplasm; Abiraterone; Enzalutamide; Dementia; Parkinson’s Disease;

Introduction The use of attention-deficit/hyperactivity disorder (ADHD) medications is on the rise globally. However, there is a lack of evidence concerning the specific drug classes, geographic distribution, economic levels, hospital types, and the impact of COVID-19 on medication consumption in the Chinese market.
Aims To evaluate the consumption pattern of ADHD medication in mainland China.
Methods This is an observational study utilizing aggregated medication sales data from the China Medical Economic Information (CMEI) database, covering the period from 2015 to mid-2023. The impact of the COVID-19 pandemic on medication consumption was examined using interrupted time series analysis.
Results This study investigated ADHD medication consumption in mainland China. Monthly DDDs/TID exhibited both seasonal fluctuations and a consistent upward trend (changing rate from the linear regression model: +2.54%, 95% confidence interval [CI]: +2.36% to +2.71%). Methylphenidate demonstrated higher consumption and growth rates compared to atomoxetine. Tertiary hospitals had higher consumption values, while secondary hospitals showed a greater monthly change rate. Consumption rates corresponded to economic levels, with high-income areas having the highest DDDs/TID but the second-lowest growth rate. During the initial eight months of the COVID-19 pandemic, ADHD medication consumption was lower than expected but gradually recovered thereafter.
Conclusions This research firstly uncovers the increasing patterns of ADHD medication consumption in mainland China, the correlation between economic status and medication consumption rates, and the short-term impact of the COVID-19 pandemic on medication consumption. It emphasizes the importance of raising awareness and providing support for ADHD, particularly in regions with lower income levels. Furthermore, it offers valuable insights for further management of ADHD during pandemics.
Keywords attention-deficit/hyperactivity disorder, medication, consumption, COVID-19, China

INTRODUCTION: Data transportability is a metric of external validity involving extrapolating results from a source population to a distinct target population. By incorporating real-world evidence (RWE), transportability methods can allow randomised controlled trial (RCT) findings (e.g. from Europe/US) to be applied to broader/external populations across geographies (e.g. Asia, often under-represented in pivotal RCTs), to reduce research burden and accelerate healthcare decision-making.
AIM: To review the application of data transportability in studies using Asian populations as the source and/or target.
METHODS: An update to a published global review of transportability studies was conducted via MEDLINE and Embase searches in June 2025 to identify studies specifically using data from patients in Asia. Data were extracted on key study objectives and findings.
RESULTS: A total of 10 studies were identified. Most (8/10) were proofs-of-concept, with the primary objective to demonstrate/test novel statistical methods. RWE was used most often to characterise covariates in the target population. Nine studies used RCTs as source data. Seven aimed to transport treatment effect (TE) estimates of interventions from RCTs to real-world (RW) target populations, including two studies transporting US findings to Asia (Japan, Thailand), another generalising findings from a Chinese RCT to the RW China population, and one using a Chinese source trial to predict results in European populations. The remaining three studies transported findings from multinational RCTs with trial sites in Asia to France/US RW populations. Except one methods-only paper, all noted successful transport/generalisation processes, while highlighting caution in the interpretation of transported outcomes, as some modelling assumptions may not be valid in the RW.
CONCLUSIONS: Data transportability analyses have been successfully applied using various methods, and appear particularly promising for transporting TE estimates. As there were limited examples where data transportability methods were used to bridge data gaps in Asia, future studies should investigate this application.

Introduction:
The FDA’s Accelerated Approval (AA) program grants early access to therapies based on surrogate endpoints, requiring confirmatory trials to verify clinical benefit. A Boxed Warning (BW) is the FDA’s strongest safety alert, highlighting the risk of serious adverse effects. Drugs labeled with a BW are more likely to fail confirmatory trials, potentially delaying regulatory decisions and AA withdrawal.

Objective:
To compare the period from AA approval to the final regulatory decision (withdrawal vs traditional approval) and assess the impact of a Boxed Warning on regulatory outcomes.

Methods:
Oncology drugs granted AA between 1992 and 2024 were identified using the FDA’s Drugs@FDA database. Information on Boxed Warnings was obtained through the FDA Label Search. Drugs were categorized into two groups based on regulatory outcome: (1) AA Withdrawn and (2) converted to traditional approval. Duration was defined as the interval between AA approval and final regulatory decision. Non-parametric statistical tests were applied due to non-normal data distribution.

Results:
A total of 144 oncology drugs were analyzed: 30 were withdrawn, and 114 received traditional approval after confirmation of clinical benefit. Among the withdrawn, 10 had a Boxed Warning and 20 did not. The median duration to regulatory decision was longer in the withdrawn group (1,378 vs. 1,196 days; p = 0.059). Cumulative incidence analysis showed a trend toward delayed decision in the withdrawn group (p = 0.099). Within this group, drugs with a Boxed Warning had the longest median duration (1,953 days), indicating prolonged market presence without confirmed clinical benefit.

Conclusion:
Oncology drugs granted Accelerated Approval remained on the market for extended periods without confirmed clinical benefit, despite being labeled with a Boxed Warning.
The delayed regulatory action raises concerns about prolonged patient exposure and highlights the need for timely oversight in the post-approval setting.

Keywords: Accelerated Approval, Boxed Warning, Regulatory Withdrawal

Introduction: Growing concern exists that blood–brain barrier (BBB)-crossing angiotensin-converting enzyme inhibitors (ACEis) may impair cognitive function by inhibiting brain ACE, an enzyme that degrades amyloid-β peptides, thereby potentially increasing the risk of dementia. However, current evidence on this association remains inconclusive.
Aims: To investigate the association between BBB-crossing ACEis and dementia risk compared to non-BBB-crossing ACEis.
Methods: We conducted a retrospective cohort study using the National Health Insurance Service–Health Screening Cohort (2002–2019) of South Korea. We identified individuals aged ≥ 40 years who newly started BBB-crossing or non-BBB-crossing ACE inhibitors during the period from January 1, 2003, to December 31, 2018, and maintained therapy for over 6 months were included. The outcome was incident dementia. Follow-up continued until outcome occurrence, death, or December 31, 2019, whichever came first. Propensity score (ps) 1:1 matching was used to balance the covariates between treatment groups. Incidence rates (IRs) per 1,000 person-years and their 95% confidence intervals (CIs) were estimated. Cox proportional hazards models were applied to estimate hazard ratios (HRs) and their 95% CIs among matched populations.
Results: 9,524 BBB-crossing (mean age 62.9 years; 66.5% male) and 10,747 non-BBB-crossing ACEis users (mean age 62.0 years; 59.4% male) were included. After 1:1 PS matching, 6,705 patients remained in each group, and all covariates were well balanced between groups. During a median 11.7-year follow-up, the IRs of dementia were 11.14 (95% CI, 10.40–11.94) and 10.74 (95% CI, 10.02–11.51) per 1,000 person-years, corresponding to an HR of 1.04 (95% CI, 0.95–1.15).
Conclusions: In this retrospective cohort study, BBB-crossing ACEis use was not associated with an increased risk of dementia compared to non-BBB-crossing ACEis use. Despite prior concerns, our population-based cohort study found no elevated risk of dementia associated with BBB-crossing ACEis.
Keywords: ACEis, blood–brain barrier, dementia

Introduction: Prediabetes elevates cardiometabolic morbidity and mortality risk. Early identification of high-risk individuals is crucial for preventing disease progression; however, long-term prediction models in this population are scarce. This study addressed the need for accurate 10-year risk prediction models for cardiometabolic outcomes in prediabetic patients.

Aims: To develop and validate 10-year risk prediction models for five outcomes: cardiovascular disease (CVD), hypertension, type 2 diabetes mellitus (T2DM), chronic kidney disease (CKD), and all-cause mortality in prediabetic adults defined by glycated hemoglobin (HbA1c) or fasting plasma glucose (FPG).

Methods: This retrospective, population-based cohort study utilized anonymized data from the Clinical Management System in Hong Kong. Two prediabetic cohorts (HbA1c 5.7-6.4% [N = 66,378] or FPG 5.6-6.9 mmol/L [N = 237,717]) were identified (2008-2012), with the first prediabetic diagnosis as the index date and follow-up until 2022. Patients with prevalent hypertension, T2DM, or CKD were excluded from analyses of each relevant outcome. Stepwise Cox regression was applied to a derivation cohort (2/3 of data), using routinely available demographic, physical, laboratory, and medication use variables. Model performance was assessed in the remaining validation cohort.

Results: HbA1c and FPG cohorts had 10-year cumulative incidence of CVD (20.0%; 19.5%), hypertension (26.6%; 32.3%), T2DM (22.1%; 29.0%), CKD (21.9%; 17.1%), and mortality (15.9%; 15.8%). Significant predictors included age, gender, smoking status, Charlson comorbidity index (CCI), and body mass index (BMI). We observed non-linear and interaction effects for CCI (hypertension/T2DM) and BMI (mortality). Adding laboratory results significantly improved the predictive performance for T2DM. All models demonstrated good discrimination (Harrell’s C-statistic >0.7 except T2DM) and calibration.

Conclusions: We developed and validated robust 10-year risk models for cardiometabolic outcomes in prediabetic populations. These models, based on routinely available variables, can aid in the early risk stratification, enabling individualized preventive strategies for high-risk patients.

Keywords: prediabetes, risk prediction, cardiometabolic outcomes

Background
Evaluating the knowledge, attitude, and practices (KAP) of tuberculosis with type 2 diabetes mellitus (TB-T2DM) patients is vital to reducing the burden of the disease, especially for advancing the end TB strategy for 2030. To our knowledge, no reliable and validated tool is currently designed to assess the KAP among patients with concurrent TB-T2DM.
Objectives
To develop and validate a questionnaire evaluating KAP for TB-T2DM co-morbid patients.
Methodology
This quantitative study was performed in Karnataka, India, between January to July 2024. The inclusion criteria were patients with TB-T2DM condition. In first phase, we developed the questionnaire through a comprehensive literature review and a semi-structured interview with five specialists. In second phase, we validated the questionnaire by content validity, face validity, and reliability. Content validity was performed by taking expert opinion from a panel of 6 experts, and then item content validity index (ICVI) and scale content validity index (SCVI) were calculated. The impact score estimated face validity. Reliability was performed by IBM SPSS 22.0. by internal consistency from Cronbach’s alpha value.
Results
In the development phase, initially, 27 items (questions) were developed, comprising 15 knowledge, 5 attitude, and 7 practice items. The ICVI score for seven questions was 0.83. For one question, it was 0.5, and the remaining questions were 1. So, the question with an ICVI score of 0.5 from the knowledge section of the questionnaire was removed. Therefore, the total number of questions after content validation were 26. However, the SCVI score obtained was 0.95. Reliability was determined by calculating the internal consistency by Cronbach’s alpha values, obtained as 0.765, 0.700, 0.643, and 0.870 for knowledge, attitude, practice, and the overall questionnaire, respectively, which were acceptable.
Conclusion
The validated questionnaire provides a reliable tool for future research to assess the KAP, improve treatment outcomes in TB-T2DM patients.

Introduction: People living with dementia and Type 2 Diabetes (T2D) are predisposed to an increased risk of adverse drug events (ADEs) from T2D medication. Deprescribing or de-intensifying T2D medication regimens in this population could reduce the risk of ADEs and is often recommended in those approaching end-of- life.

Aims: To investigate the prevalence of T2D medication use in the last year-of-life in people living with dementia and the patient-level factors associated with discontinuation in the 6-months prior to death.

Methods: We used linked hospital and administrative data for individuals in the United Kingdom (UK) and Victoria, Australia. The cohort included individuals with data available for at least 12-months prior to death between July 2012 and June 2018. Logistic regression estimated the association of co-variates such as age, sex, cardiovascular disease, chronic kidney disease (CKD) and cancer with discontinuation of T2D medications.

Results: 10,016 individuals from the UK and 6,436 from Australia were identified. Those aged 85+ made up 55.8% of the UK and 37.7% of the Australia cohort. In both datasets there was little reduction in use of T2D medications, the largest change was observed in prevalence of insulin supply with a 25.6% (UK) and 43.5% (Australia) reduction in use in the last 3-months-of-life compared to use at 12-months. CKD was associated with higher odds of discontinuing metformin in the UK (OR 1.05, 95% CI 1.03–1.07) and Australia (OR 1.60, 95% CI 1.15–2.12). Those aged 85+ had an increased likelihood of discontinuation for sulfonylureas in Australia (OR 1.67 95% CI 1.16–2.40).

Conclusions: Despite observing some discontinuation of T2D medications, over half of individuals were supplied one or more of these mediations in the final 3-months-of-life. To increase comfort and minimise potential ADEs, strategies to facilitate deprescribing of T2D medications near end-of-life need to be explored.

Keywords: multi-database, dementia, type-2-diabetes

Introduction: While surrogate endpoints have expedited the evaluation of anti-cancer drugs, their correlation with overall survival (OS) remains uncertain. In real-world settings, pragmatic endpoints—time to next treatment (TTNT), time to treatment discontinuation (TTD)—are increasingly utilized as alternative indicators of OS.
Aims: To investigate the association between surrogate endpoints and OS in patients receiving biologics using real-world data.
Methods: A retrospective cohort study utilized the Korea-Clinical Data Utilization Network for Research Excellence registry database (2013-2021). Patients with advanced breast (BC), colorectal (CRC), and gastric cancer (GC) who received biologics (trastuzumab, cetuximab, or bevacizumab) were identified. Candidate surrogate endpoints included TTD/TTNT were defined respectively as the time from treatment initiation to discontinuation or death, and to the commencement of the next line of therapy or death. The primary outcome was patient-level correlation between TTD/TTNT and OS.
Results: A total of 550 BC, 490 GC, and 1902 CRC patients were identified. Median OS was longest in BC at 56.0 months (49.0-64.0), followed by CRC (bevacizumab: 21.0 months, 21.0-22.0; cetuximab: 26.0 months, 25.0-28.0), and GC (13.0 months, 11.0-14.0). TTNT generally presented higher correlation than TTD, though both showed weaker correlations in cancers with longer survival durations for both TTNT/TTD. Despite using the same biologic (trastuzumab), the correlation between surrogate endpoints and OS was weak in BC (TTNT Spearman ρ = 0.230, 0.143-0.315) and relatively strong in GC (ρ = 0.772, 0.716-0.820). In CRC, moderate correlations were observed irrespective of biologic types (bevacizumab: ρ = 0.456, 0.427-0.489; cetuximab: ρ = 0.637, 0.595-0.680).
Conclusions: The correlation between surrogate endpoints and overall survival appeared to be driven more by disease characteristics than specific treatment. Our findings suggest that the surrogacy reliability may be limited in cancer with longer survival, highlighting the importance of reassessment of drugs approved based on surrogate endpoints.
Keywords: surrogate endpoints, real-world data, oncology

Introduction: Attention-deficit hyperactivity disorder (ADHD) is a prevalent neurodevelopmental disorder in children contributing to significant socioeconomic burden. However, novel drug targets for ADHD remain limited. Given its early onset age, elucidating how protein perturbation impacts ADHD risk can help identify putative modifiable targets for drug development in disease prevention.
Aims: We aim to identify specific proteins as potential drug targets for ADHD prevention.
Methods: We conducted a two-sample Mendelian Randomization (MR) study leveraging protein quantitative trait loci (pQTLs) associated with 443 childhood plasma proteins (N = 2,147; ages 5 - 20 years, in standard deviation [SD]) as genetic instruments. Summary statistics from an ADHD genome-wide association study (Ncase = 38,691, Ncontrol = 186,843) were used as outcome datasource. Inverse-variance weighted method was used as the main analysis for estimating possible causal effects, with false discovery rate to account for multiple comparison. Colocalization analysis was performed to assess whether identified associations were driven by shared causal variants.
Results: Amongst 443 proteins, we identified five potential protein drug targets for ADHD. Genetically predicted higher coagulation factor XIII B chain (OR: 1.04; 95%CI: 1.01-1.06) and extracellular matrix protein 1 (OR: 1.03; 95%CI: 1.01-1.04) were associated with increased ADHD risk, and higher haptoglobin (OR: 0.96; 95%CI: 0.93-0.98), vitamin D-binding protein (OR: 0.94; 95%CI: 0.92-0.97), and sonic hedgehog protein (OR: 0.96; 95%CI: 0.94-0.98) were inversely associated with ADHD risk. However, evidence for genetic colocalization was only evident for extracellular matrix protein 1 (coloc.susie-PPH4 = 0.998).
Conclusion: Our study suggested extracellular matrix protein 1 is a promising candidate for preventing ADHD during childhood. Whether existing medications which reduce extracellular matrix protein 1 (e.g. captopril) can potentially be repurposed for ADHD prevention should be further explored using randomized controlled trials or real world evidence from electronic health records.
Keywords: attention-deficit hyperactivity disorder; mendelian randomization; childhood proteomics.

INTRODUCTION
Psychiatric disorders rank as the fourth leading cause of disability-adjusted life years, affecting 2–3% of India's population and around 10% of adults globally. Drug Utilisation Evaluation (DUE) studies help us assess how medications are prescribed and used, as well as their medical and social impact.

AIM
The present study was conducted to assess the antipsychotic drug use patterns at a tertiary care teaching hospital.

METHODS
This prospective study was conducted over 12 months at the outpatient department of Psychiatry, following Institutional Ethics Committee approval. A total of 150 psychiatric patients receiving antipsychotic treatment were included, and their prescriptions were assessed against the standard therapeutic guidelines. Descriptive analysis was performed using Microsoft Excel, and the results were expressed in numerical values and percentages.

RESULTS
A total of 428 drugs were prescribed across 150 prescriptions, including 302 (70.6%) antipsychotics and 126 (29.4%) concomitant medications. Female patients made up 54%, and males 46%. The most common diagnosis was schizophrenia (38%), followed by bipolar affective disorder (20%), alcohol dependence psychosis (14%), and others at 28%. The frequently used atypical antipsychotics were risperidone (31.1%), olanzapine (23.2%), amisulpride (20.2%), quetiapine (11.6%), aripiprazole (8.6%), and clozapine (5.3%). About 77.5% of patients received polytherapy, while 22.5% received monotherapy. The average number of antipsychotics per prescription was 2.01. Prescribing by generic name was at 100%. Use of drugs from the WHO and National Essential Medicines Lists was 68.2%. No injectables or fixed-dose combinations were prescribed.

Conclusion
This study's findings conclude that the atypical antipsychotics are the most prescribed antipsychotics. Most prescriptions were complete and followed rational prescribing principles, highlighting the need for a drug utilisation evaluation (DUE) to improve patient outcomes.

Keywords: Antipsychotics, Drug Utilisation Evaluation, Atypical Antipsychotics

Introduction
The potential for clinically relevant drug-drug interactions between direct oral anticoagulants (DOACs) and statins in atrial fibrillation management remains uncertain.

Aims
This study aimed to evaluate the risks of clinical outcomes associated with DOACs plus statins in patients with atrial fibrillation.

Methods
We conducted a cohort study and a six-parameter model case-crossover study, using electronic health records from the Hospital Authority of Hong Kong between 1 Jan 2011 and 31 Dec 2023. The cohort study compared hazards of hospitalised and emergency adverse outcomes (major bleeding, cardiovascular diseases, all-cause mortality) in DOACs plus CYP3A4/P-glycoprotein-inhibiting statins (interacting) users with DOACs plus non-interacting statins users. The case-crossover study compared the odds of exposure to different drug initiation patterns associated with adverse outcomes in the hazard period with the referent period.

Results
Of 145,394 people with atrial fibrillation, 31,466 co-prescribed DOACs and interacting statins and 2,822 co-prescribed DOACs and non-interacting statins. In cohort design, we observed no difference in the hazard of bleeding and cardiovascular diseases but a higher hazard of all-cause mortality in DOACs plus interacting statins users (hazard ratio with 99% confidence interval: 1.36 (1.02-1.82), compared to DOACs plus non-interacting statins users. In the case-crossover design, odds ratios for all outcomes in drug-interaction-related initiation patterns were not higher than initiating DOACs-monotherapy or statins-monotherapy patterns.

Conclusions
We found no evidence of higher risks of bleeding and cardiovascular diseases associated with DOACs plus interacting statins versus DOACs plus non-interacting statins. The results of the case-crossover study suggested that residual between-people confounding is likely to explain the elevated hazard of all-cause mortality observed in the cohort design.

Keywords: direct oral anticoagulants, statins, drug-drug interactions

Introduction: The FOURIER trial demonstrated that evolocumab reduces cardiovascular events. Chinese dyslipidemia management guidelines advocate initiating therapy with statins and stepwise intensification, adding a cholesterol-absorption inhibitor and/or a PCSK9 inhibitor when needed.

Aims: This study aimed to evaluate, among patients with atherosclerotic cardiovascular disease (ASCVD), the association between use of PCSK9 inhibitors and the risk of three-point major adverse cardiovascular events (3P-MACE), providing real-world evidence to inform clinical practice.

Methods: Using the Ningbo Regional Health Information Platform and a target trial emulation framework, we applied an active-comparator new-user design with ezetimibe (EZE) initiators as the comparator. We included adult ASCVD patients who initiated a PCSK9 inhibitor or EZE between January 1, 2019 and June 30, 2025 while concurrently receiving statin therapy. Propensity scores were constructed from demographics, lifestyle factors, comorbidities, and medication history, followed by 1:1 matching to emulate randomization; Cox proportional hazards models were then fitted in the matched cohort to estimate the hazard ratio (HR) and 95% confidence interval (CI) for the association between PCSK9 inhibitor use and 3P-MACE.

Results: The final cohort included 4,591 new users of PCSK9 inhibitors and 4,591 new users of EZE. Over a median follow-up of 1.93 (1.89 - 1.97) years, 437 and 684 events occurred in the PCSK9 inhibitor and EZE groups, respectively. In the matched cohort, multivariable Cox models showed a significantly lower risk of 3P-MACE in the PCSK9 inhibitor group [HR (95% CI): 0.85 (0.75 - 0.96)]. Subgroup and sensitivity analyses supported the primary findings.

Conclusions: Among patients with ASCVD, initiation of PCSK9 inhibitors was associated with a reduced risk of 3P-MACE, supporting their clinical benefit in routine practice.

Keywords: PCSK9 inhibitors; target trial emulation; ASCVD; MACE; regional healthcare big data.

Introduction:
Though many studies support sodium intake among athletes for various reasons, evidence of its effectiveness remains inconclusive. Sodium plays a vital role in fluid balance and thermoregulation, and its depletion during intense or prolonged exercise can lead to serious outcomes such as exercise-associated hyponatremia. This systematic review evaluates current evidence on sodium intake to better understand its role in preventing performance decline and heat-related illness in athletic populations.

Aim:
To clarify the role of sodium products among athletes.

Methods:
A comprehensive literature search was conducted in PubMed, Scopus, Embase, and Cochrane from inception to January 2025. Interventional and observational studies assessing the effect of sodium or its products in any type of athlete were included. Improved athletic performance and reduced heat-related illness were primary outcomes, while adverse effects were secondary. The Cochrane Risk of Bias Tool-2 and Newcastle-Ottawa Scale were used for quality assessment of RCTs and observational studies. Findings were analyzed by sport type to identify potential variations in efficacy.

Results:
A total of 19 out of 792 screened studies met inclusion criteria, comprising 12 randomized controlled trials and 7 observational studies. The majority investigated endurance sports (cycling, running), where sodium supplementation consistently improved performance outcomes like race completion times and cycling velocity. In contrast, studies of other sports (swimming, wrestling) showed minimal benefits. Among endurance athletes, sodium intake demonstrated protective effects against heat-related illness, though certain formulations (particularly sodium bicarbonate) were associated with increased gastrointestinal complaints. The observational studies generally supported these findings but with less consistent effect sizes compared to RCTs.

Conclusions:
Sodium supplementation benefits endurance sports most clearly, enhancing performance and reducing heat risks. Effects are limited in other sports. Sport-specific approaches considering individual tolerance are recommended. Further research is needed to refine guidelines on optimal sodium use across different sports.

Keywords:
athlete, sodium, sports

Introduction: Emerging evidence suggests vitamin D (VD) may influence glucose metabolism and insulin sensitivity, offering potential benefits in preventing the development of diabetes and its complications.
Aims: To evaluate the effects of VD supplementation on multiple health outcomes in diabetes or prediabetes.
Methods: We conducted this review strictly following the PRISMA guidelines. A comprehensive search of PubMed, Cochrane, and Embase was conducted for systematic reviews with meta-analyses of RCTs published from inception up to January 2024. Studies were included if they evaluated the effect of VD interventions on any health outcomes in prediabetes, type 2 diabetes mellitus (T2DM), or diabetic nephropathy (DN), with placebo as the control. For meta-analyses, summary effect estimates with 95%CIs were extracted. Methodological quality and certainty of evidence were assessed using AMSTAR2 and GRADE respectively.
Results: A total of 2,614 articles were retrieved and finally 34 publications with 60 meta-analyses of RCTs were included in this review. Methodological quality was high to moderate in 15% and low to very low in 85% of the included publications. The certainty of evidence was rated as moderate for 28%, low for 20%, and very low for 52% outcomes. Results from meta-analyses indicated that in T2DM, VD supplementation significantly reduced fasting blood glucose levels (MD-5.02 mg/dL, 95%CI -6.75, −3.28) and insulin levels (SMD-0.49, 95%CI -0.68, -0.31). In prediabetic populations, VD supplementation was associated with a reduced risk of progression to T2DM (RR0.89, 95%CI 0.80, 0.99). Among patients with DN, VD supplementation significantly reduced the urinary albumin to creatinine ratio (SMD-0.49, 95%CI -0.90, -0.08) and hs-CRP levels (MD-0.69 mg/L, 95%CI -0.86, -0.53). Additionally, VD showed potential benefits in lowering blood pressure, blood lipids, and inflammatory markers.
Conclusions: VD may benefit patients with prediabetes, T2DM, or DN, though the overall certainty of evidence remains low.
Keywords: vitamin D, diabetes, umbrella review

Introduction: Immune checkpoint inhibitors (ICIs) have revolutionized first-line therapy for advanced non-small cell lung cancer (NSCLC). With emerging PD-1/PD-L1 agents, we excluded bevacizumab—an anti-VEGF confounder—to present the latest network meta-analysis (NMA) focused solely on immunotherapy efficacy and safety, aiming to guide personalized treatment decisions across all PD-L1 levels.
Aim: To compare the efficacy and safety of first-line ICI-based treatments for advanced NSCLC across all PD-L1 expression levels.
Methods: This systematic review and frequentist network meta-analysis (NMA) were conducted following PRISMA guidelines and registered in PROSPERO (CRD42024616245). Phase II/III trials were identified through PubMed and Embase. Primary outcomes were overall survival (OS) and progression-free survival (PFS), while secondary outcomes include the objective response rate (ORR) and adverse events (AEs). Treatments were ranked by Surface Under the Cumulative Ranking (SUCRA) and validated by a league table. Bias assessment done via funnel plot, Egger’s test, and Risk of Bias 2 (ROB2) tool.
Results: Thirty-one studies involving 15,172 patients were included, assessing 21 ICI-based regimens regardless of PD-L1 expression. In OS and PFS analyses, Prolgolimab+chemotherapy (hazard ratio [HR]: 0.51, 95%CI: 0.35–0.74; SUCRA:0.91) and Penpulimab+chemotherapy (HR: 0.43, 95%CI: 0.32–0.58; SUCRA:0.91) showed the best efficacy outcomes, respectively, across non-selective PD-L1 expression levels. Compared to chemotherapy, Pembrolizumab+chemotherapy (odds ratio [OR]: 3.10, 95%CI: 2.18–4.41; SUCRA:0.86) showed superior ORR. In OS subgroup analysis, Prolgolimab+chemotherapy ranked best for PD-L1 expression ≥50% and ≤1%, while in PFS, Penpulimab+chemotherapy and Serplulimab+chemotherapy ranked highest for PD-L1 ≥50% and ≤1%, respectively. However, Atezolizumab monotherapy was the safest option regarding any AE and grade ≥ 3 AEs.
Conclusions: Our comprehensive analysis suggests that prolgolimab and penpulimab regimens offer superior survival benefits compared to chemotherapy alone, although the combination therapies have higher AEs. The optimal treatment strategy depends entirely on patient characteristics and PD-L1 expression.
Keywords: Immune checkpoint inhibitors, non-small cell lung cancer, network meta-analysis

Introduction:
Real-world data (RWD) is essential for generating robust real-world evidence (RWE) supporting clinical and regulatory decisions. Diverse RWD sources are available in China, such as electronic health records, claims data, and disease-specific cohorts. However, choosing the optimal data source remains challenging and impacts study quality. We address how to optimize RWD source selection to enhance RWE generation within the Chinese healthcare setting.

Aims:
We aim to systematically assess China’s real-world data sources and develop practical selection guidance, enhancing the reliability and applicability of real-world evidence for informed healthcare decision-making.

Methods:
We summarized key elements for evaluating databases and proposed a framework to guide fit-for-purpose data source selection, highlighting critical considerations throughout. Two case studies demonstrate practical application based on specific research questions and study requirements.

Results:
We catalogued an overview of the RWD landscape in Chinese Mainland and Hong Kong by identifying and characterizing widely used sources. Our evaluation covered data coverage, longitudinality, data types, information flow, and key elements vital for addressing diverse research questions.

We developed a tailored data assessment workflow to guide fit-for-purpose RWD source selection based on specific study objectives, enabling researchers to assess cases individually and identify data best suited for each research question’s unique requirements.

We presented two case examples illustrating the application of this framework in selecting appropriate data sources based on study design and target population. The first example features a retrospective study utilizing regional electronic health record (EHR) database from Chinese Mainland, while the second leverages data from a Hong Kong database.

Conclusion:
Our proposed framework offers a structured way to navigate China’s complex RWD landscape, helping researchers align data selection with study objectives and speed feasibility assessments. Demonstrated through case studies, this approach enhances high-quality, relevant real-world evidence for healthcare.

Keywords:
Real-world data, Data source selection, China healthcare

Introduction: Drug hypersensitivity reactions (DHRs) are unpredictable adverse drug reactions that range from mild skin rashes to severe anaphylaxis. Accurate diagnosis is crucial but challenging due to lack of standardized testing for many commonly used drugs beyond penicillin. Skin prick testing (SPT) is a reliable tool for diagnosing immediate, IgE-mediated drug allergies.

Aim: To develop and clinically validate an in-house SPT method for identifying drug allergies to selected drugs: Paracetamol, Diclofenac, Aceclofenac, and Levofloxacin.

Methodology: A five-year prospective study was conducted at a 1200-bed teaching hospital in Southern India. Healthy volunteers (n=25) and atopic individuals (n=10) were enrolled to identify non-irritating SPT concentrations for the selected drugs. Serial dilutions were prepared from pure active compounds, avoiding commercial excipients. Positive (histamine) and negative (50% glycerin) controls were used. Wheal reactions were observed every 5 minutes for 45 minutes. The validated non-irritating concentrations were tested in 10 patients with a known history of drug allergy. ROC analysis was used to evaluate diagnostic performance.

Results: The identified non-irritating concentrations were: Paracetamol (0.5 mg/mL), Diclofenac (0.2 mg/mL), Aceclofenac (0.2 mg/mL), and Levofloxacin (0.1 mg/mL). Optimal reaction times were between 15–25 minutes. ROC analysis showed moderate diagnostic accuracy (AUC ~0.70). Validation in patients showed 70% concordance with clinical history. One negative case was found to be non-IgE mediated.

Discussion: The study successfully established standardized, safe SPT concentrations for selected drugs. Use of pure compounds minimized false positives. This method enhances diagnostic accuracy for immediate-type drug allergies and may reduce mislabeling, unnecessary drug avoidance, and inappropriate substitutions in clinical practice.

Conclusion: This validated SPT protocol offers a reliable, safe, and cost-effective method for diagnosing immediate drug allergies and can improve clinical decision-making in routine practice.

Keywords: Drug hypersensitivity, skin prick test, drug allergy, paracetamol, diclofenac, levofloxacin, aceclofenac, ROC analysis, IgE-mediated reaction.

Background: Comparability of two treatment assignments can be evaluated using negative control outcomes (NCOs); however, differing criteria for defining NCOs may lead to misinterpretation of comparability.

Objective: We aim to evaluate comparability using NCOs, applying different assessment criteria.

Method: We selected 50 negative control outcomes (NCOs) based on previous studies and clinical expertise and applied two criteria to assess comparability: (1) fewer than 5% of NCOs should be non‑null, with null defined as an effect estimate crossing 1.0; and (2) fewer than 5% of NCOs should be non‑null, with null defined as an effect estimate crossing the broader range of 0.75–1.3. Only NCOs with at least 100 events were included to ensure sufficient statistical power. We conducted a new user cohort study in the TriNetX US Collaborative Network comparing vertebral fracture risk between denosumab and bisphosphonates, including patients with DEXA T scores ≤ –2.5 and incorporating DEXA results into 1:1 propensity score matching to address confounding by indication. Moreover, we compared our findings with a trial under a similar scenario to evaluate whether the less stringent criterion could ensure comparability. We hypothesized that, even with this less stringent criterion, our results would remain consistent with the RCT while improving comparability.

Results: After 1:1 propensity score matching, 2,195 patients were included in each group with all balanced covariates. Denosumab was associated with a lower risk of vertebral fracture (HR 0.60, 95% CI 0.43–0.85), similar to the RCT (HR 0.42, 95% CI 0.18–0.96). Using the first criterion, 2 of 16 (12.5%) NCOs were non-null, indicating lack of comparability. With the second criterion, 0 of 16 NCOs were non-null, suggesting improved comparability.

Conclusions: Applying multiple NCOs with less stringent criteria may reduce the risk of misinterpreting treatment comparability and, at the same time, enhance data usability.

Introduction
Observational studies have shown association between recent elevated anticholinergic burden and a higher risk of falls and fractures, while this association may not be directly attributable to the anticholinergic burden itself, but the underlying indication for which the medications were prescribed.
Aims
To evaluate the association between a transient increase in anticholinergic burden and risk of falls and fractures.
Methods
We conducted a case-case-time-control study using National Health Insurance Research Database in Taiwan. We included individuals over 65 who were hospitalized or admitted to emergency room due to falls and fractures, using admission date as index date. Future cases were those occurring within 60-150 days after matched index date. We defined hazard period as 60 days before index date, with reference period randomly selected from 121 to 300 days before index date. We categorized the sum of Anticholinergic Cognitive Burden Scale (ACB) into three groups: 0, 1-2, 3+ points, using conditional logistic regression to estimate effect of anticholinergic burden on risk of falls and fractures. We also conducted subgroup analysis by osteoporosis, dementia, Parkinson disease and comedications related to fall and fracture, including alpha blocker, benzodiazepines and antiacids.
Results
We included 434,322 patients, with a mean age of 77 and 36% male. The case-case-time-control analysis suggested that an ACB score of 1-2 points (OR 1.29, 95% CI 1.24-1.33) and 3+ points (OR 1.40 95%, CI 1.37-1.44) were associated with an increased risk of falls and fractures. Results of subgroup analysis were consistent with the main findings.
Conclusions
Our findings suggest an association between recently elevated anticholinergic burden and an increased risk of falls and fractures. We recommend that clinicians monitor cumulative effects from multiple drugs with anticholinergic properties to prevent acute risk of falls and fractures.

Keywords: anticholinergic burden, fall and fracture

Introduction: Coronary artery disease (CAD) remains a leading cause of death and loss of Disability Adjusted Life Years in India and globally. Despite therapeutic advances, patients often experience symptoms that affect their quality of life. Traditional clinical markers do not fully capture patient experiences, making patient reported outcome measures (PROMs) essential for understanding treatment outcomes from the patient’s perspective. To our knowledge, this study is the first to assess the percentage improvement in angina symptoms after medication using PROMs.
Aims: To identify factors associated with improvement in PROMs among patients with CAD.
Methods: A prospective cohort study was conducted from November 2022 to April 2023 in hospitalised patients with CAD. The 19-item Seattle Angina Questionnaire (SAQ) was administered through face-to-face interviews at baseline and via telephone at one-month follow-up. Patients who did not provide consent were excluded. The SAQ scores were compared before and after therapy to assess changes. A total of 117 pre-specified variables—including socio-demographic, clinical, medication-related factors, potential drug-drug interactions, and drug-related problems—were selected based on literature and clinical relevance. These were analysed using chi-square tests and multivariable linear regression to identify predictors of PROMs improvement.
Results: Of the 61 patients, 98.24% showed improvement in SAQ scores, with a mean (SD) increase of 30.17 (13.9)%. Regression analysis identified occupational status [r²=0.126, p<.01], use of beta blockers [r²=0.11, p<.05], and use of gastrointestinal medications [r²=0.12, p<.05] as significant predictors. Together, these explained 30.4% of the variance in PROMs improvement [r²=0.30, p<.01].
Conclusions: Occupational status and medication use, particularly beta blockers and GI medications, significantly influenced PROMs in patients with CAD. These findings underscore the need to consider socio-clinical factors in treatment evaluations. Further studies should explore causal mechanisms underlying these associations.
Keywords: coronary artery disease, patient reported outcome measures, Seattle Angina Questionnaire.

Introduction
Multimorbidity and polypharmacy are nowadays prevalent in many countries. Deprescribing based on medical advice could be a countermeasure for polypharmacy, but identification of deprescribing from the healthcare database is often challenging. In 2020, Japan introduced a financial incentive for deprescribing focusing on hospitalized patients who lived with polypharmacy, with its real-world application understudied.
Aims
To characterize the profiles of patients assigned a reimbursement code for deprescribing in the claims-based database.
Methods
We analyzed an employee-based health insurance claims database (> 10 million individuals) in Japan from 2020 to 2022, in which prescriptions were traceable across all medical institutions. We identified patients assigned a reimbursement code for deprescribing at hospital discharge, and summarized medication usage at the 4-digit ATC level. Only oral medications prescribed at outpatient visits were included for analysis. In this study, regular or chronic medication use was operationally defined as a total of ≥28 days within 6 months prior to deprescription.
Results
A total of 73 patients were identified. Among them, 59% were male, and the median age was 58 years (range: 13–74 years). We unexpectedly found that only 8 patients (11%) were regularly prescribed ≥5 medication classes. As post-hoc analyses, we varied the ATC classification level and changed the definition of regular use, but the results remained largely unchanged. Finally, when the threshold for regular use was lowered to “a total of ≥7 days”, 61 patients (84%) had prescriptions from ≥5 medication classes.
Conclusions
In our study population with a limited sample size, deprescribing under Japan’s national health insurance system was mostly applied to medications with relatively short-term use. These findings suggest that there may be limited opportunities to study chronic polypharmacy through the current payment scheme for deprescribing.

Introduction:
Treatment-resistant depression (TRD), defined as the failure to respond to at least two adequate antidepressant trials, presents a significant burden in psychiatric care. Poor medication adherence, misconceptions about antidepressant therapy, and limited awareness of emerging tools like pharmacogenomic testing may contribute to treatment failure. Understanding patients’ knowledge, attitudes, and practices (KAP) can inform targeted interventions to improve treatment outcomes in Major Depressive Disorder (MDD).

Aims:
This study aimed to assess the KAP surrounding antidepressant therapy and pharmacogenomic testing among individuals with MDD and to identify behavioral and perceptual domains that influence treatment resistance.

Methods:
A cross-sectional study was conducted between January and March 2025 involving 144 adults diagnosed with MDD attending outpatient psychiatric clinics in South India. Participants completed a validated KAP questionnaire covering antidepressant knowledge, adherence practices, and perceptions of genetic testing. Exploratory Factor Analysis (EFA) using Principal Component Analysis (PCA) with Varimax rotation was performed to extract latent constructs. Internal consistency was measured using Cronbach’s alpha, and regression analysis examined associations between factor scores and self-reported treatment resistance.

Results:
The mean participant age was 36.7 years (SD = 10.9), with 59% female. Only 28% demonstrated high knowledge about antidepressant use, and 41% reported full adherence. Awareness of pharmacogenomic testing was low (14%), but 62% showed interest in its future use. Four latent factors were identified, explaining 68.4% of the total variance: (1) antidepressant knowledge, (2) adherence behaviors/barriers, (3) attitudes toward pharmacogenomics, and (4) trust in healthcare providers. The KAP scale showed strong internal consistency (Cronbach’s alpha = 0.87). Higher knowledge and attitude scores significantly predicted better adherence (β = 0.39, p < 0.001) and reduced likelihood of TRD (β = -0.27, p = 0.008).

Conclusions:
This study emphasizes the need for education on genetic testing to address behavioral barriers and promote pharmacogenomic testing awareness for precision therapy.

Introduction:
Tyrosine kinase inhibitors (TKIs) are targeted therapies that block aberrant signaling pathways often activated in cancer cells. Although TKIs provide clinical benefits in cancers like chronic myeloid leukemia and non-small cell lung cancer, they are also linked to adverse drug reactions, particularly hepatotoxicity. Identifying genetic factors in TKI-induced liver injury is key to personalized treatment.

Aims:
This study aimed to identify genetic variants associated with TKI-induced hepatotoxicity using large-scale genomic data.

Methods:
We utilized data from the All of Us Research Program. Participants with available whole genome sequencing data, documented TKI exposure, and recorded liver enzyme levels were included. The observation window was defined from the first to the last TKI exposure date. A phenotype file was created, and genome-wide association analyses were performed using PLINK, adjusting for age, sex, and genetic principal components. Significant associations were visualized using Manhattan and QQ plots.

Results:
Among 805 individuals with available liver enzyme data within 60 days after TKI initiation, we identified 98 cases with hepatotoxicity and 707 controls without. Genome-wide association analysis was conducted based on this case-control classification. Eight single nucleotide polymorphisms (SNPs) exceeded the suggestive threshold of P < 1×10⁻⁵. The lead SNP, rs72734306, showed an odds ratio (OR) of 3.73 (P=4.18×10⁻⁷). Other significant associations included rs71425326, associated with EML5, a gene involved in microtubule organization and immune function (OR=9.86); rs35928059, associated with CFH, a key regulator of the complement system (OR=4.43); and rs190933696, located between ERLEC1 and GPR75-ASB3, genes related to endoplasmic reticulum stress and immune response (OR=9.46).

Conclusion:
These findings highlight immune regulation and ER stress–related genes, particularly CFH and ERLEC1, as potential contributors to TKI-induced hepatotoxicity. This study offers new insights into genetic susceptibility and supports personalized approaches to managing TKI-related adverse effects.

Keyword: Tyrosine kinase inhibitor, Hepatotoxicity, Genome-Wide Association Study

Introduction: Myoclonic epilepsy of Infancy (MEI), also known as benign myoclonic epilepsy in infancy, is a rare, self-limited epilepsy syndrome, marked by brief myoclonic seizures in otherwise healthy and developmentally normal children. The Epidemiology of MEI has not been quantified in the literature to date.
Aims: To estimate the worldwide incidence and prevalence of MEI.
Methods: A systematic search on Medline, Embase, Scopus, and Web of Science was conducted from inception to March 2025 to identify studies reporting prevalence or incidence of MEI. Eligible studies were selected using pre-defined criteria. Meta-analysis was performed using the Freeman-Tukey double arcsine transformation. Pooled estimates with corresponding 95% confidence intervals (CIs) were reported. Subgroup analyses explored variations across countries, WHO regions, the influence of sample size, and the year of publication. Sensitivity analysis was performed to assess the robustness of the findings.
Results: A total of 17 studies were included, of which 13 reported incidence and 4 reported prevalence, encompassing a combined population of 1,491,596, with 27 cases of MEI. The pooled prevalence was 1.69 per 100,000 individuals (95% CI: 0.00-5.51, I2= 80.6%), and the pooled incidence was 0.3 per 100,000 live births (95% CI: 0.0-1.0; I2=21%).
Subgroup Analysis showed that incidence ranged from 0.0 per 100,000 individuals in the United States to 13.4 per 100,000 in France. Across WHO regions, incidence varied from 0.0 per 100,000 in the Region of the Americas to 0.8 in Europe. Incidence was also higher in studies with >100000 participants and in those published after 2015, suggesting a potential improvement in case identification over time.
Conclusion: MEI has a low global burden but notable regional variation, particularly in Europe. Country-specific estimates can guide national planning and policy decisions. More research from under-represented regions, such as Asia, is needed.
Keywords: Myoclonic infancy in epilepsy, self-limited epilepsy, burden, prevalence, incidence

Introduction: Use of glucagon-like peptide-1 receptor agonists (GLP-1 RAs) for diabetes and/or obesity also disseminates into pediatric populations, notwithstanding their poorly established safety profile compared to better characterization in adults. Understanding spontaneous adverse event (AE) patterns appears critical for guiding clinical practice.
Aims: We aim to describe potential safety signals of GLP-1 RAs in FDA Adverse Event Reporting System (FAERS) in children.
Methods: We included AE reports where GLP-1 RA was the sole suspected active ingredient in children aged 6 to 17 years, submitted up to 31 March 2025 to the FAERS. We evaluated demographic characteristics, drug distribution, reporter qualifications, and the seriousness of AEs. These were classified according to the MedDRA version 28.0 hierarchy at the System Organ Class and Preferred Term levels.
Results: We identified 159 pediatric cases where liraglutide (37.1%) and semaglutide (28.3%) were the most frequently reported agents. Female patients constituted 71.5% of cases, and the median age was 15 years (IQR:14–17). Around half of the reports (52.2%) were submitted by healthcare professionals, and 55.3% involved serious AEs; 1.9% of cases included fatal outcomes. Semaglutide had the highest rate of serious outcomes (75.6%), which was significantly higher compared to non-semaglutide GLP-1 agonists (p=0.001). Among the 159 cases, a total of 414 distinct adverse events (AEs) were reported. Gastrointestinal disorders were identified in 37.7% of the cases, while the most frequently reported individual AE, off-label use, was observed in 18.9%. Suicidal and self-injurious behaviors were reported in 6.3% of all cases, with the highest proportion observed in semaglutide-related reports (11.1%), followed by liraglutide (5.1%).
Conclusion: A substantial portion of reported AE in children using GLP-1 RAs were serious, with semaglutide showing a notably higher rate. These results emphasize the need for cautious use and monitoring in pediatric populations.
Key words: FAERS, Glucagon-like peptide-1 (GLP-1) receptor agonists, pediatrics

Objective: To evaluate the impact of self-monitoring blood glucose (SMBG), continuous glucose monitoring (CGM), and flash glucose monitoring (FGM) alongside non-insulin glucose-lowering medications on glycaemic control in adults with non–insulin-dependent T2DM.

Methods: An umbrella review was conducted following the Cochrane Handbook and PRISMA 2020 guidelines (PROSPERO: CRD420250628416). Searches were performed across MEDLINE, Cochrane Library, Embase, and Scopus. Reviews were eligible if they included adults with non–insulin-dependent T2DM using glucose monitoring and reported HbA1c outcomes. Two reviewers independently conducted study selection, extraction and quality appraisal using AMSTAR 2. A random-effects meta-analysis was undertaken. Publication bias and study overlap were assessed, and subgroup analyses explored potential effect modifiers.

Results: Sixteen systematic reviews comprising 81 unique studies and 23,657 participants were included. SMBG versus no monitoring showed significant HbA1c reduction (–0.23%; 95% CI: –0.29 to –0.18; I² = 20.9%). CGM demonstrated superior benefit over SMBG (–0.30%; 95% CI: –0.35 to –0.25; I² = 0%). FGM showed no statistically significant advantage over SMBG (–0.26%; 95% CI: –0.57 to 0.05) with substantial heterogeneity (I² = 73.9%). Egger’s and Begg’s tests indicated no publication bias (p = 0.80; p = 0.685). Overlap analysis identified a total of 205 citations and a mean of 2.53 citations per study, indicating moderate overlap.

Conclusion: Glucose monitoring interventions consistently demonstrated statistically significant improvements in glycaemic control among adults with non–insulin dependent T2DM. While SMBG offers measurable benefits over no monitoring, CGM provides superior HbA1c reductions compared to SMBG, supported by robust evidence and minimal heterogeneity. The moderate overlap in included studies strengthens the reliability of this synthesis. These results underscore the importance of structured and evidence-based monitoring in personalised diabetes care.Integrating these strategies into routine practice may significantly improve long-term outcomes and enable proactive, data-driven approaches to chronic disease management.

Keywords:
Type 2 diabetes, glucose monitoring, glycaemic control

Introduction:
Steatotic liver disease is the most prevalent chronic liver condition, affecting one-third of the adult population. Habitual aspirin use is common among older adults.
Aims: To investigate the relationship between aspirin use and liver fat accumulation and inflammation.
Methods:
We used data from the UK Biobank, including participants who underwent liver MRI. Liver fat percentage was assessed using proton density fat fraction (PDFF), and liver inflammation was measured using corrected T1 (cT1). Aspirin use was self-reported and participants were categorized as habitual users or others. Covariates included age, sex, socioeconomic status, lifestyle, and cardiometabolic risk factors. We used multivariable linear regression to examine associations between aspirin use and PDFF and cT1 values. Logistic regression was used to assess the association with steatotic liver disease (PDFF > 5%).
Results:
A total of 36,404 participants (mean age 64.5 ± 7.6 years; 51.4% female) were included, of whom 3,530 (9.7%) reported habitual aspirin use. Compared with non-users, aspirin users were more likely to be male, older, consume more alcohol, and have higher BMI, waist circumference, blood pressure, diabetes/pre-diabetes, high triglycerides, and low HDL. They were less likely to be physically active or highly educated. Aspirin users had higher unadjusted liver fat (5.4% vs. 4.9%) and cT1 (715 ms vs. 699 ms). However, after adjusting for covariates, aspirin use was associated with lower liver fat (β (95%CI): –0.41 (–0.55, –0.27); p < 0.01) and higher cT1 (β (95% CI): 5.9 (4.0, 7.8); p < 0.01). Aspirin use was also associated with lower odds of steatotic liver disease (OR (95% CI): 0.77 (0.71, 0.84), p < 0.01).
Conclusion:
Habitual aspirin use was associated with reduced liver fat but higher liver inflammation. Further research is needed to explore potential prospective dose–response relationships and underlying mechanisms.
Keywords: aspirin; liver fat; liver inflammation

Introduction:
In China, most psychotropic drugs are not strictly regulated and thus allowed in primary healthcare institutions (PHIs). Yet, uneven resource allocation has led to persistent gaps in drug availability and service capacity between PHIs and non-Primary Healthcare Institutions (nPHIs). With rising mental health needs, evaluating primary-level drug access is vital for evidence-based policy planning.
Aims:
This study investigates disparities in psychotropic drug availability between PHIs and nPHIs in Shandong Province, aiming to assess whether access has shifted downward and whether PHIs are equipped for basic pharmacological interventions, thereby informing policies on service integration.
Methods
Procurement data from all public healthcare institutions in Shandong Province (2016–2024) were analyzed, covering four psychotropic drug classes: antipsychotics, anxiolytics, hypnotics/sedatives, and antidepressants. Coverage was defined as the proportion of institutions procuring each drug within the corresponding facility level. Temporal trends were assessed using statistical methods.
Results:
Between 2016 and 2024, psychotropic drug coverage in nPHIs increased significantly from 56% to 92% ( 5.15% per year, P < 0.001), while PHIs rise from 50% to 52% (0.46% per year, P < 0.001). The largest disparity occurred in antipsychotic drug coverage (2024: 78% in nPHIs vs. 16% in PHIs), while anxiolytics showed the narrowest gap (85% vs. 41%). Most psychotropic drugs had coverage rates below 10% in PHIs, except benzodiazepines such as diazepam (31%) and alprazolam (34%). In contrast, coverage in nPHIs exceeded 50% for most drugs, including chlorpromazine (53%), olanzapine (62%), diazepam (80%), and midazolam (55%).
Conclusions:
Over half of PHIs lack essential pharmacological capacity, falling short of the WHO’s 75% coverage benchmark. Drug accessibility has improved mainly in nPHIs, while PHIs remain constrained by structural limitations—offering primarily anxiolytics, in contrast to the broader all psychotropic spectrum available in nPHIs. Strengthening primary-level capacity may require enhanced provider training and improved distribution incentives.

Introduction: While both empagliflozin and dapagliflozin are known to improve hepatic steatosis and liver enzyme levels, robust Phase Ⅲ trial evidence for improving steatohepatitis and fibrosis is currently exclusive to dapagliflozin. Therefore, a head-to-head comparison between two agents is warranted.
Aims: To compare the hepatic effectiveness of empagliflozin versus dapagliflozin in metabolic dysfunction-associated steatotic liver disease (MASLD).
Methods: Using nationwide Korean claims data, we constructed a new-user cohort of adults with MASLD—defined as having T2D and a fatty liver index (FLI) ≥ 60—who initiated empagliflozin or dapagliflozin between May 2016 and December 2023. The index date was the first prescription of empagliflozin or dapagliflozin, with follow-up based on an on-treatment exposure definition. Co-primary outcomes were MASLD improvement (achieving FLI < 30) and decompensated hepatic events (ascites, variceal bleeding, hepatic failure, liver transplantation). After 1:1 propensity score (PS) matching, hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated using Cox models. Analyses stratified by low (empagliflozin 10 mg; dapagliflozin 5 mg) and high (empagliflozin 25 mg; dapagliflozin 10 mg) doses were further conducted.
Results: Among 265,098 new users of empagliflozin and dapagliflozin, 101,495 pairs were 1:1 PS matched (mean age 54.3 years; male 71.2%). Compared with dapagliflozin, empagliflozin was associated with a higher likelihood of MASLD improvement (HR 1.05, 95% CI 1.00-1.11; p=0.0557), although the result was not statistically significant. The risk of decompensated hepatic events was comparable (HR 1.01, 95% CI 0.96-1.06). Dose-stratified analyses yielded consistent findings, with no significant differences observed in either MASLD improvement (low dose: HR 1.02, 95% CI 0.76-1.80; high dose: 1.06, 0.94-1.20) and decompensated hepatic events (low dose: 1.02, 0.79-1.31; high dose: 1.09, 0.98-1.21).
Conclusion: Use of empagliflozin and dapagliflozin showed comparable effectiveness in MASLD improvement and risk of decompensated hepatic events.

Keywords: Sodium-glucose cotransporter 2 inhibitors; metabolic dysfunction-associated steatotic liver disease; hepatic decompensation.

Introduction Evidence on the heterogeneity in the efficacy of statins for the primary prevention of cardiovascular disease (CVD) remains limited.
Aims To identify the sources of heterogeneity in the efficacy of statins for the primary prevention of CVD.
Methods Studies were systematically identified from a previous systematic review and searched in Medline, Embase, and Cochrane. Randomized controlled trials (RCTs) in adults without a history of CVD that investigated statin treatment were included. A meta-analysis was conducted to pool risk ratios (RR) and 95% confidence intervals (CI) for major adverse cardiovascular events [MACE], myocardial infarction [MI], stroke, and cardiovascular death. Univariate meta-regression analyses were conducted to assess the impact of trial-level covariates on the effect sizes ,using the ratio of risk ratios (RRR). Subgroup analyses were conducted by pooling ratio of effect sizes.
Results 24 RCTs were included. Compared to no statin treatment, statin treatment was associated with a reduced risk of MACE (RR, 0.73[95% CI, 0.67, 0.80]), MI (RR, 0.68[95% CI, 0.60, 0.77]), stroke (RR, 0.76[95% CI, 0.65, 0.89]) and cardiovascular death (RR, 0.81[95% CI, 0.71, 0.95]). Univariate meta-regression indicated that higher baseline of total cholesterol (RRR, 1.29 [95% CI, 1.07, 1.57]), higher low-density lipoprotein cholesterol (LDL-C) (RRR, 1.31 [95% CI, 1.07, 1.62]), higher triglycerides (RRR, 1.81 [95% CI, 1.17, 2.81]) and lower statin intensity (RRR, 0.74 [95% CI, 0.59, 0.94]) were significantly associated with reduced efficacy of statins in stroke prevention at trial-level. While the statistical significance was borderline, the subgroup differences indicated a potentially greater benefit of statin therapy in males and participants without metabolic syndrome.
Conclusions Statin therapy exhibited consistent efficacy in preventing CVD across diverse populations, but for individuals with elevated levels of total cholesterol, LDL-C, or triglycerides, more intensive treatment beyond statins may be necessary for effective stroke prevention.
Keywords: statins, cardiovascular disease, meta-analysis

Introduction: Elderly patients with acute myeloid leukaemia (AML), ineligible for chemotherapy, usually receive hypomethylating agents (azacitidine or decitabine). The combination of venetoclax and azacitidine was observed with a survival advantage over azacitidine alone among patients with AML in a pivotal randomised controlled trial. However, the small sample size in the Asian subgroup means we cannot confirm a comparable survival benefit in this population.

Objective: We aimed to compare overall survival and healthcare resource utilisation in elderly patients newly diagnosed with AML receiving combination therapy (venetoclax with hypomethylating agents) versus monotherapy (hypomethylating agents alone).

Methods: Elderly patients (≥60 years old) newly diagnosed with AML during 2018 and 2023 were identified from territory-wide electronic medical records in Hong Kong. Patients were excluded if they had certain previous prescriptions, e.g., chemotherapy, or diseases, e.g., malignant neoplasm. A target trial emulation was conducted with 1:1 propensity-score matching, balancing the baseline characteristics. Kaplan-Meier curve and Cox regression compared overall survival, and negative binomial regression assessed healthcare resource utilisation with time offset. Subgroup analysis was performed based on the age of 75 years.

Results: The matched cohort included 159 patients per arm with a median follow-up duration of around 12 months. The combination therapy significantly improved overall survival compared to monotherapy [hazard ratio with 95% confidence interval (HR)=0.72 (0.57, 0.93), p=0.01], with pronounced benefit in patients ≥75 years [HR=0.68 (0.47, 0.98), p=0.04]. The combination therapy arm required fewer packed cell transfusions [26.98 vs. 35.20 units/person-year; incidence risk ratio with 95% confidence interval (IRR)=0.70 (0.56, 0.87), p=0.001], and also in patients ≥75 years [IRR=0.72 (0.53, 0.98), p=0.04].

Conclusion: Compared to hypomethylating agent alone, venetoclax plus hypomethylating agent benefits elderly patients with AML on overall survival and healthcare resource utilisation, especially among patients aged ≥75 years old.

Keywords: Acute myeloid leukaemia, Venetoclax plus hypomethylating agent, Asian patients

Next-generation sequencing (NGS) tools facilitate high-resolution genotyping of Human Leukocyte Antigen (HLA) genes in clinical settings, contributing to precise results and the advancement of personalized medicine. Pharmacogenetic variability, particularly in the HLA class I and II genes, affects how individuals respond to medication. The objectives of our study were to perform high-resolution sequencing of HLA class I and II genes for 537 samples at a Thai Pharmacogenomic Personalized Medicine Centre using Third-Generation Sequencing (NGS) technology at a 3-field resolution. This approach aimed to reduce ambiguities in achieving accurate typing and exploring the lengths of reads, subsequently comparing the frequencies of these variants with those in other populations. The study sequenced 537 samples using NGS with the Pacific Biosciences (PacBio) platform, identifying 233 distinct alleles out of 6444 detected variants. The highly frequent HLA alleles in six genes were HLA-A*11:01:01 (24.9%; 95% CI: 22.3-27.5), HLA-B*46:01:01 (9.9%; 95% CI: 8.1-11.7), HLA-C*01:02:01 (17.2%; 95% CI: 14.9-19.5), HLA-DRB1*12:02:01 (17.4%; 95% CI: 15.3-19.8), HLA-DQB1*03:01:01 (20.9%; 95% CI: 18.6-23.5), and HLA-DPB1*05:01:01 (23.6%; 95% CI: 21.1-26.2), The prevalence of these variants was found to differ statistically among populations (p < 0.05). Remarkably, the detection of some variants with distinctions in exon synonymous mutation regions, for class I variations such as A*11:01:01/43, B*51:02:01/02, B*51:01:01/02 and C*03:04:01/04 as well as for class II genes DPB1*02:01:01/02 and DRB1*15:02:01/02 based on 3-field call reads in high resolution was considerable. The study successfully sequenced HLA alleles in a 3-field high resolution to assess the diversity of HLA alleles within the Thai population. The results accentuate the importance of enhancing HLA typing screening, which could significantly improve patient clinical outcomes. Additionally, the findings provide valuable data for studies on human genetics in populations, HLA matching, transplantation techniques, HLA-associated disorders, and implications for personalized medicine.
Keywords: Human Leukocyte Antigen (HLA) genes; Next-generation sequencing; Pharmacogenomics

Introduction: The balance between rejection prevention and immunosuppressants (IS) related adverse drug events (ADEs) is crucial, as most transplant patients require IS throughout their entire life.
Aims: To investigate the IS use and the occurrence of ADEs, especially infections, during the first six months using IS in organ transplant patients.
Methods: We performed a cross-sectional study on patients who received organ transplantation and were subsequently followed up at University Medical Center Ho Chi Minh City from 01/2018 to 04/2024. Data was collected from medical records including patient demographics, IS use, acute rejection and ADEs especially infection. Factors associated with infections were identified using univariable logistic regression analysis.
Results: There were 77 patients - 66 adults, 11 children - in the study, including 41 kidney and 36 liver transplant recipients. At the end of follow-up period, most kidney recipients maintained a steroid-tacrolimus-mycophenolate regimen (87.8%). In liver transplantation, tacrolimus-mycophenolate was predominant in adults (52.0%) while tacrolimus monotherapy was more common in children (63.6%). The tacrolimus concentration achieved target range most frequently in the first month (63.6-80.0%) and decreased thereafter. The most common ADEs in kidney recipients, adult and pediatric liver recipients were endocrine-metabolic disorders (65.8%), nervous system disorders (80.0%) and infection, respectively. Infection incidence was 34.1% in kidney; 48.0% in adult liver and 90.9% in pediatric liver recipients. Liver transplantation (OR=3.385; 95%CI: 1.323-8.659; p=0.011), age ≥ 18 (OR=0.061; 95%CI: 0.07-0.505; p=0.010), time in intesive care unit (OR=1.216; 95%CI: 1.045-1.415; p=0.012) and time in hospital (OR=1.048; 95%CI: 1.014-1.083; p=0.006) were significantly associated with infection.
Conclusion: These results emphasize the role of closely monitoring patients post-transplant to manage ADEs and intervene promptly. Optimizing IS therapy, particularly in the therapeutic drug monitoring, is also essential.
Keywords: Organ transplantation, immunosuppressants, infection

Introduction:
Globally, cardiovascular disease remains the leading cause of mortality, with dyslipidemia as a major modifiable risk factor. While pharmacist-led interventions have demonstrated significant clinical benefits in developed nations, evidence from developing countries like India remains scarce and underutilized.

Aim:
This study aimed to evaluate whether structured pharmacist-led care could significantly improve lipid outcomes among dyslipidemia patients in India, addressing a critical evidence gap.

Methods:
This randomized controlled trial was conducted over 18 months (July 2021 – December 2022) at a tertiary care teaching hospital in India. The study was prospectively registered with the Clinical Trial Registry of India (CTRI). Eligible patients (N=378) were randomized to an intervention group receiving individualized counseling, recommendations per ATP III guidelines, and culturally adapted patient information leaflets, or to a control group receiving standard care. Monthly follow-ups were performed in-person and via telephone. The primary outcome was mean change in Low Density Lippoprotein-Cholesterol (LDL-C) at 12 months.

Results:
A total of 378 patients were enrolled in the study, with 186 assigned to the intervention group and 192 to the control group. The cohort comprised 56% males, while the remaining were females. The mean low-density lipoprotein cholesterol was found to be significantly different after twelve months (P=0.001), with an average of 105.22±13.26 mg/dl in the control group and 101.02±12.44 mg/dl in the intervention group. After twelve months, the intervention group demonstrated a mean LDL-C reduction of 25.31% compared to 21.3% in the control group (P<0.001).

Conclusions:
Structured pharmacist-led interventions significantly improve lipid profiles and adherence among dyslipidemia patients in India. This evidence supports expanding clinical pharmacy services to enhance cardiovascular outcomes in resource-limited settings.

Keywords: Pharmacist-care, Interventions, LMIC

INTRODUCTION: Urinary tract infections (UTIs) are common bacterial infections that affect millions worldwide, posing challenges in timely diagnosis. Although they are common, achieving prompt and precise diagnosis remains a significant hurdle. This study explores the potential of artificial intelligence (AI) and machine learning (ML) to enhance diagnostic accuracy for better patient outcomes.
OBJECTIVE: To develop and validate machine learning algorithms for early UTI diagnosis using clinical symptoms and laboratory parameters, and evaluate their performance for potential integration into clinical decision-making systems.
METHODS: A mixed retrospective-prospective observational study was conducted in a tertiary-care teaching hospital with 330 adult patients (18–80 years) with suspected UTIs. Patients were selected using systematic sampling from multiple departments. Clinical data, including demographics, symptoms, comorbidities, and laboratory parameters, were collected. An external-validation dataset comprising 80,000 UTI cases was utilized. Predictor variables were identified using logistic regression and Chi-square tests. Random Forest and Decision Tree algorithms were evaluated using standard performance metrics.
RESULTS: Among 330 patients, females comprised 62.7% (p = 0.0025), and patients aged ≥60 had significantly longer hospital stays (p = 0.010). Key clinical symptoms included fever (59.7%), malaise (35.15%), and chills (33.94%). Prevalent comorbidities were diabetes (56.97%), hypertension (53.03%), and chronic kidney disease (18.78%). E. coli was the predominant pathogen (27.27%). Specific drugs associated with UTI included SGLT2 inhibitors, anticholinergics, calcium channel blockers, and furosemide. Random Forest outperformed Decision Tree with 82% accuracy, 86% precision, 90% recall, and 88% F1-score. External validation yielded 50% sensitivity and 70% specificity. Clinical feedback indicated 73.3% clinician support, highlighting the need to improve sensitivity.
CONCLUSION: ML algorithms, particularly Random Forest, demonstrate potential in enhancing UTI diagnostic accuracy. Integrating clinical symptoms and laboratory parameters provides a promising framework for early diagnosis, though optimization of sensitivity remains crucial for widespread clinical adoption.
KEYWORDS: Urinary tract infection, Machine learning, Clinical symptoms

Introduction:
Attention-deficit/hyperactivity disorder (ADHD) is one of the most common neurodevelopmental and neurobehavioral disorders among children and adolescents. Pharmacological treatment is considered a first-line therapy in many countries, including China. However, real-world study on medication treatment patterns and persistence in the Chinese pediatric population remain limited.

Aims:
To examine the pharmacological treatment patterns, one-year medication persistence, and factors associated with treatment discontinuation among children and adolescents with ADHD in China.

Methods:
This retrospective cohort study utilized electronic health records from the Jiangsu Provincial Health Information Platform. Patients aged 6–17 years with a diagnosis of ADHD who initiated pharmacological treatment between January 2020 and November 2023 were included. Treatment discontinuation was defined as a ≥180-day gap without medication during the one-year follow-up period. Descriptive statistics characterized demographic and treatment variables. Multivariable logistic regression was used to identify factors associated with treatment discontinuation.

Results:
A total of 11,152 patients met the inclusion criteria, with a mean age of 9.02 years; 1,803 (16.17%) were female, and 429 (3.85%) had baseline psychiatric comorbidities. The most commonly prescribed initial medications were methylphenidate (54.38%) and atomoxetine (44.39%). During the one-year follow-up, 5,166 patients (46.32%) maintained pharmacological treatment, while 5,986 (53.68%) discontinued. Patients initiated on atomoxetine had a significantly higher risk of treatment discontinuation compared to those initiated on methylphenidate (adjusted odds ratio [aOR], 1.94; 95% CI, 1.79–2.11). Older age was also associated with a greater likelihood of discontinuation (aOR, 1.08; 95% CI, 1.06–1.10).

Conclusions:
In this large real-world cohort of children and adolescents with ADHD in China, methylphenidate and atomoxetine were the predominant treatment choices. However, overall treatment persistence was low. Atomoxetine initiation and older age were significant predictors of treatment discontinuation. These findings highlight the need for targeted interventions to improve long-term adherence and optimize treatment outcomes in this population.

Keywords: ADHD, medication persistence, China

Introduction:
Early-stage risk communication and knowledge adoption by target populations are essential for evaluating the effectiveness of risk minimization measures (RMM) within 12-24 months of regulatory implementation, enabling timely adjustments in healthcare practices. Knowledge-based surveys, which provide quantitative measurements of dissemination methods, perceptions of RMM, and intended actions for risk minimization, are frequently used to identify dissemination barriers and knowledge gaps. However, there is little research on recently conducted post-authorization safety studies (PASS) that focus on general principles and methodologies related to RMMs assessment through knowledge surveys.

Aims:
This review aims to describe the study characteristics of recent PASS that evaluate effectiveness of RMMs through knowledge-based surveys for healthcare professionals (HCPs).

Methods:
PASS evaluating the effectiveness of RMMs conducted in 2022-2024 were identified through the HMA-EMA Catalogue. Titles and full-texts were screened and study characteristics were extracted.

Results:
Seventeen studies out of 720 PASS in 2022-2024 were eligible for this review. The majority of PASS were category 3 (required in the Risk Management Plan to investigate safety concerns or evaluate the effectiveness of risk minimization activities; 12/17, 70.6%), with the remaining (5/17, 29.4%) requested by regulators outside the European Union (EU). Six studies (35.3%) evaluated the effectiveness of RMMs within the GVP Module XVI recommended time window (12-24 months) following the launch or regulatory implementation of RMMs. In addition to assessing knowledge outcome (included in all studies), 11 (64.7%) studies evaluated HCPs’ awareness or receipt of RMMs, 6 (35.3%) assessed RMMs utilizations, and 4 (23.5%) evaluated self-reported behaviors related to risk minimization. Among seven studies were finalized, five (71.4%) drew conclusions based on pre-defined thresholds for determining RMM effectiveness.

Conclusions:
Clear understanding of methodological and data source limitations, and sources of biases is warranted to design survey studies for assessing RMMs dissemination and risk knowledge.

Keywords:
PASS, RMM, survey

Introduction
Current recommendations for eGFR monitoring intervals in type 2 diabetes mellitus (T2DM) patients are largely consensus-based, with limited empirical evidence.
Aims
To investigate the appropriate eGFR monitoring intervals for T2DM patients at various CKD progression risk levels.
Methods
180,984 T2DM adults (2009–2015) were stratified into low, moderate, and high CKD progression risk groups based on ADA and KDIGO guidelines. A target trial was emulated in each group to assess the impact of different eGFR monitoring intervals (2–8, 9–15, and 16–24 months for low and moderate risk; 2–4, 5–8, 9–15, and 16-24 months for high risk) on renal function decline and ESRD. Hazard ratios (HRs) and 95% confidence intervals (CIs) were estimated using pooled logistic regression. Follow-up continued until the earliest occurrence of an outcome event, death, or December 31, 2021.
Results
In the low-risk group, extending the eGFR monitoring interval from 2–8 to 16–24 months did not significantly increase the risk of renal function decline (HR [95% CI]: 1.021 [0.771, 1.351]) or ESRD (HR [95% CI]: 1.048 [0.863, 1.272]). In the moderately increased risk group, a 16–24-month interval (vs 2-8 months) was significantly associated with a higher risk of renal function decline (HR [95% CI]: 1.03 [1.002, 1.06]), but not ESRD (HR [95% CI]: 1.092 [0.921, 1.296]). For the high-risk group, extending the interval from 2–4 to 9–15 months increased the risk of renal function decline (HR [95% CI]: 1.044 [1.005, 1.085]). Monitoring every 16–24 months (vs 2–4 months) was also associated with a higher risk of ESRD (HR [95% CI]: 1.269 [1.085, 1.484]).
Conclusions
For low-risk T2DM patients, eGFR monitoring intervals can be safely extended to 16–24 months. Annual assessments are recommended for those with moderately increased or high risk of CKD progression.
Keywords
Long-term care based on risk stratification, eGFR monitoring, type 2 diabetes mellitus

Introduction: Epilepsy is a chronic neurological disorder affecting 23 million people in Asia, with 30-40% having drug-resistant epilepsy.

Objective: To study clinical characteristics and therapeutic outcomes in refractory epilepsy using self structured seizure tracking chart.

Methods: Prospective cohort study at JSS Hospital, Mysuru (August 2024-January 2025) using a Seizure Tracking Sheet.

Results: A total of 122 patients with refractory epilepsy were enrolled, including 68 males (55.73%) and 54 females (44.26%). The largest age group was young adults (19–44 years), accounting for 36.06% of cases (n=44).
Seizure timing showed a marked morning predominance (53.28%), followed by evening (25.41%), afternoon (24.59%), and night (20.49%). Seizure durations varied: 20.49% experienced 1-minute seizures, 31.15% had 2-minute episodes notably among11% pediatric and 4% geriatric patients on valproate and levetiracetam, while 37.70% had seizures exceeding 5 minutes—these cases frequently involved pediatric (17%) and geriatric (5%) populations treated with fosphenytoin and midazolam.
The most common clinical manifestations were body tremors (68.8%), blank stares (34.4%), falls (33.6%), tongue biting (14.7%), up-rolling eyes (13.1%), and frothing (11.4%). Leading seizure triggers included sleep deprivation (50.8%), psychological stress (47.5%), missed medications (31.9%), and sensory stimuli such as flashing lights (8.2%). Global developmental delay was reported in 8.2% of cases.
Post-ictal symptoms were predominantly drowsiness (69.7%) and confusion (60.7%), along with sensory changes (29.5%), speech difficulty (14.8%), and behavioral alterations (13.9%).
Refractory periods varied widely: 26.2% of patients were seizure-free for 6–12 months, 24.6% for 1–3 months, 22.1% for over a year, 10.7% for 3–6 months, while 16.4% experienced recurrence within 48 hours.
Levetiracetam was the most prescribed anti-seizure medication (72.95%), followed by sodium valproate (36.88%) and phenytoin (34.42%). Clobazam was commonly used as an adjunct therapy, particularly in cases with prolonged episodes.

Conclusion: Study validates structured tracking for personalized anti-seizure medication optimization in refractory epilepsy management

keywords: Refractory Epilepsy, Drug-resistant epilepsy.

Introduction: Type 2 diabetes (T2DM) and chronic obstructive pulmonary disease (COPD) frequently coexist, increasing morbidity and mortality. Antidiabetic agents may modulate respiratory outcomes via pleiotropic anti-inflammatory effects relevant to COPD pathogenesis. Their comparative impact on exacerbations remains unclear. This network meta-analysis (NMA) addresses optimal management of coexisting T2DM/COPD by evaluating oral glucose-lowering therapies' effects on exacerbation risk.
Aims: To compare the effects of oral antidiabetic agents on overall and severe COPD exacerbations using network meta-analysis, based on their potential mechanistic links to inflammation and respiratory outcomes. To rank treatments by efficacy for exacerbation prevention.
Methods: Following PRISMA-NMA guidelines, we searched PubMed, Embase, Cochrane Library, and Web of Science until March 2025 for observational studies and randomized controlled trials (RCTs) in adults with confirmed T2DM and COPD. Outcomes were overall and severe COPD exacerbations. Risk ratios (RRs) with 95% CIs were pooled using random-effects NMA models. Treatments were ranked using the Surface Under the Cumulative Ranking curve (SUCRA). Heterogeneity was assessed via I² statistics.
Results: Eleven studies (n=482,314 participants; all observational studies) were included. Compared to sulfonylureas, SGLT-2 inhibitors significantly reduced overall exacerbations (RR 0.71, 95% CI 0.66-0.77), as did GLP-1 receptor agonists (RR 0.77, 95% CI 0.71-0.84). DPP-4 inhibitors (RR 1.05, 95% CI 0.97-1.13) and meglitinides (RR 1.13, 95% CI 0.93-1.36) showed non-significant increases. For severe exacerbations, SGLT-2 inhibitors (RR 0.42, 95% CI 0.38-0.46) and GLP-1 receptor agonists (RR 0.46, 95% CI 0.42-0.51) demonstrated the strongest reductions.
Conclusions: SGLT-2 inhibitors and GLP-1 receptor agonists significantly lower COPD exacerbation risk in patients with T2DM and COPD compared to sulfonylureas. These findings support considering these agents for managing patients with coexisting T2DM and COPD. Future randomized trials are needed to validate these associations.
Keywords: Hypoglycemic Agents, COPD exacerbations, Type 2 Diabetes Mellitus, Network Meta-Analysis.

Introduction: Appropriate antibiotic prescribing is crucial to prevent both, adverse effects and the development of antimicrobial resistance.
Aims: We aimed to assess the quantity and quality of outpatient antibiotic prescribing in children aged 0 to 2 years in Switzerland.
Methods: We conducted a drug utilization study using data from a Swiss health insurance (Helsana) with approximately 16% population coverage. The study population comprised children born between 2014 and 2021. We calculated the median time from start of insurance period to a child’s first outpatient antibiotic prescription, the prescription prevalence and the rate of antibiotic treatment episodes in the first two years of life. Furthermore, we assessed the quality of antibiotic prescribing, considering the first antibiotic dispensed.
Results: We included 106‘036 children. At the beginning of the study period, the median time to the first outpatient antibiotic stayed at around 33 to 34 months before rising sharply to 44 months (in birth-cohort from 2018), followed by a subsequent decline to 37 months (2020). The prescription prevalence ranged from 43% (2014) to 32% (2019). The rate of antibiotic treatment episodes in the first two years of life declined from 427 to 277 episodes per 1000 person-years. Amoxicillin was the most frequent first-ever dispensed antibiotic throughout the study period and the proportion of the World Health Organization’s Access group (i.e. first-choice antibiotics) increased from 86% to 93%.
Conclusions: In our sample of Swiss claims data, children received antibiotics markedly later and less frequently than in a previous study with data from Germany and Denmark. Quality indicators were slightly poorer in Switzerland compared to Denmark, but more favourable than in Germany. These discrepancies may be partially explained by different national guidelines. Temporal trends indicate an improvement in the prescribing quality.
Keywords: antibiotics, young children, drug utilization research

Background:
Radiation dermatitis is a common adverse effect in breast cancer patients undergoing chest wall radiotherapy post-mastectomy, significantly impacting quality of life.
Aim: This study evaluates the efficacy of prophylactic topical corticosteroids—Betamethasone 0.3% cream—in mitigating radiation-induced dermatitis using Functional Assessment of Cancer Therapy–Breast (FACT-B) scores as patient-reported outcomes.
Methods:
A randomized controlled trial was conducted with 12 post-mastectomy breast cancer patients undergoing chest wall radiotherapy. Participants were randomized into two groups: intervention (Betamethasone 0.3% cream, n=6) and control (no topical corticosteroid, n=6). FACT-B scores were recorded at baseline and weekly for 5 weeks.
Results:
At baseline, the intervention group had a slightly higher mean FACT-B score (45.83 ± 20.69) compared to the control group (40.67 ± 12.66). The intervention group showed a significant increase in mean score during the 1st week (72.50 ± 21.20), indicating early improvement in patient-reported quality of life. However, the benefit declined over the following weeks, with scores dropping to 45.67 ± 17.64 by the 5th week. In contrast, the control group demonstrated a gradual but consistent increase in scores, peaking at 68.67 ± 0.52 in the 5th week. Variability in response was higher in the intervention group, especially between the 2nd and 4th weeks, suggesting inconsistent patient response or potential rebound effects.
Conclusion:
Prophylactic use of Betamethasone 0.3% cream resulted in early symptomatic relief as evidenced by improved FACT-B scores in the first week. However, the sustained benefit was not observed over five weeks, and a higher variability was noted. While initial outcomes were promising, the long-term impact appears limited. A larger sample size and extended follow-up are warranted to validate the clinical significance of prophylactic corticosteroid use in managing radiation dermatitis

Background: ALK inhibitors are standard treatment for ALK-rearranged NSCLC, but their efficacy in patients with concurrent PD-L1 expression remains uncertain. This study compared outcomes of first-line alectinib in ALK-rearranged NSCLC patients with variable PD-L1 expression levels.
Methods: This retrospective study conducted target trial emulation to compared Time to Treatment Failure (TTF) and overall survival (OS) in patients had ALK–rearranged NSCLC with variable PD-L1 expression levels receiving alectinib, based on the Flatiron Health database (2011-2025). Propensity Score Matching (PSM) balanced baseline characteristics. Time-to-first-event analysis was performed using Cox proportional hazards and Kaplan–Meier survival analysis, with the hazard ratio and 95% confidence interval calculated.
Results: 292 ALK–rearranged NSCLC patients were enrolled, with 171 in PD-L1–negative and PD-L1-low group (PD-L1 expression < 50%) and 121 in PD-L1-high group (PD-L1 expression ≥50%). After PSM, 101 patients remained in each group. Median 5-year TTF was not reached in PD-L1–negative and PD-L1-low group versus 24.1 months in PD-L1-high group. Median 5-year OS was 46.9 months (PD-L1–negative and PD-L1-low versus 43.4 months (PD-L1-high). PD-L1–negative and PD-L1-low showed a significant trend toward improved TTF (HR 0.39, 95% CI 0.23-0.67, p=0.274) and showed no significant difference in OS between groups (HR 0.91, 95% CI 0.59-1.42, p=0.68).
Conclusion: For ALK-rearranged NSCLC patients, high PD-L1 expression is associated with worse TTF among those receiving alectinib as first-line treatment.

Introduction: Peripheral edema is a common adverse reaction to amlodipine and celecoxib, but whether the risk increases when both drugs are used simultaneously is unknown.
Aims: To estimate the effect on edema of amlodipine and celecoxib, odds ratios (ORs) and relative excess risk due to interaction (RERI) were obtained by the case-crossover method.
Methods: We used a Japanese claims database (JMDC 2005-2021). Cases used amlodipine and celecoxib at least once during a 180-day period before newly starting furosemide (used as a surrogate for peripheral edema) after >365 days of non-use of furosemide. The odds ratios (ORs) for edema from amlodipine only (OR1), celecoxib only (OR2), and both amlodipine and celecoxib (OR3) compared to no amlodipine/celecoxib, and RERI and their 95% confidence interval (CI) were estimated by standard conditional logistic regression (SCL) and a recently proposed weighting method for multiple exposures (ICPE 2025 Washington DC). Exposure status was determined on the last day of each period.
Results: We identified 1014 cases. When the time window for control periods=120 days and 1 period=10 days, OR1=1.9 (95% CI: 1.5, 2.4), OR2=2.7 (1.9, 3.8), and OR3=4.7 (3.6, 6.2), and RERI=1.2 (0.0, 2.4) by the SCL method, while OR1=1.5 (1.2, 1.9), OR2=2.2 (1.6, 3.2), and OR3=3.2 (2.5 4.2), and RERI=0.5 (-0.9, 1.8) by the weighting method. When the length of 1 period was varied between 2 to 60 days, point estimates were OR1=1.6 to 2.0, OR2=2.0 to 3.1, OR3=3.4 to 5.8, and RERI=0.8 to 1.7 by SCL, while OR1=1.5 to 1.5, OR2=1.9 to 2.4, OR3=3.1 to 3.2, and RERI=0.4 to 0.7 by the weighting method.
Conclusion: Estimates of the risk of edema due to amlodipine and celecoxib using SCL were higher and more unstable than the weighting method. Significant RERI was suggested by SCL but not by the weighting method.
Keywords: Peripheral edema, Case-crossover study

Introduction: Elderly patients with heart failure (HF) often present with multiple comorbidities and complex medication regimens. However, research on polypharmacy and potentially inappropriate medication (PIM) in Asian population with HF remains insufficient.

Aims: To provide scientific evidence for improving clinical medication safety in elderly patients with heart failure.

Methods: The research studied elderly patients with HF (≥65 years) from the Ningbo Yinzhou District medical database from 2011 to 2022. We categorized them into four groups based on the number of drug types recorded during the 90-day period following their diagnosis: non polypharmacy (<5 drugs), polypharmacy (5-9 drugs), hyperpolypharmacy (10-14 drugs), and super hyperpolypharmacy (≥15 drugs). PIM was identified based on the 2019 Beers Criteria from inpatient and outpatient records. The primary outcome was a composite of cardiovascular death or HF hospitalization, while secondary outcomes included HF exacerbation, all-cause mortality, and hospitalizations.

Results: The research included 7,361 elderly patients with HF, 82.3% were prescribed five or more medications, and 53.2% had PIM. No significant differences were found between the polypharmacy and control groups, but the hyperpolypharmacy and super hyperpolypharmacy groups showed a 22% (OR, 1.22; 95% CI, 1.01-1.48) and 53% (OR, 1.53; 95% CI, 1.26-1.87) increased risks on a composite of cardiovascular death or HF hospitalization, respectively. For secondary outcomes, the super hyperpolypharmacy group had a 26% increased risk of HF exacerbation (OR, 1.26; 95% CI, 1.01-1.56), a 64% increased risk of recurrent HF hospitalizations (OR, 1.64; 95% CI, 1.27-2.12), and a 25% increased risk of recurrent all-cause hospitalizations (OR, 1.25; 95% CI, 1.09-1.44).

Conclusions: Polypharmacy and PIM use are prevalent among elderly HF patients in China. Hyperpolypharmacy and super hyperpolypharmacy are associated with worse cardiovascular and hospitalization outcomes. Interventions to optimize medication regimens—beyond eliminating PIMs—are urgently needed to improve safety in vulnerable population.

Keywords: heart failure, polypharmacy, potentially inappropriate medication

Introduction:
Immune checkpoint inhibitors (ICIs) have become integral in cancer therapy but are associated with immune-related adverse events such as hepatotoxicity. The genetic factors underlying ICI-induced hepatotoxicity are not well understood. Identifying these factors could improve risk prediction and patient management.

Aims:
This study aimed to identify genetic variants associated with ICI-induced hepatotoxicity using a large-scale genomic dataset. The objective was to uncover candidate loci for further investigation.

Methods:
Patients receiving ICIs were identified from the All of Us Research Program, a large-scale, population-based genomic and health database designed to reflect diverse ancestries. Hepatotoxicity was assessed using laboratory data (ALT, AST, ALP, total bilirubin) and classified according to the Common Terminology Criteria for Adverse Events (CTCAE). Cases were defined as patients with grade ≥2 hepatotoxicity, and controls were defined as patients without clinically significant hepatotoxicity. Genome-wide association analyses were performed on array genotype data using PLINK, with quality control and adjustment for population stratification.

Results:
A total of 553 patients were analyzed at the 60-day observation window following ICI initiation, including 74 cases and 479 controls. No variants reached genome-wide significance (p<5E-08). However, several loci demonstrated suggestive associations (p<1E-05). Notably, rs16957038 in CLYBL was associated with hepatotoxicity (OR=9.1, p=1.90E-06). Additional loci included rs1257300 (OR=0.36, p=3.68E-06) and rs76857861 (OR=6.8, p=6.03E-06).

Conclusions:
This exploratory GWAS identified potential genetic signals for hepatotoxicity in ICI-treated patients. Larger cohorts and functional validation are needed to confirm these findings and elucidate their biological relevance.

Keywords:
Immune checkpoint inhibitors, hepatotoxicity, GWAS

Introduction: Ischemic stroke (IS) is a major complication in atrial fibrillation (AF) patients, but current risk models often ignore complex drug-protein-disease interactions, reducing their accuracy. We developed ABioSPath, an interpretable deep learning model, to predict one-year IS risk in AF patients by integrating biological pathways with clinical data. This addresses the need for precise, data-driven risk assessment without costly biomarkers.

Aims: The objective is to create ABioSPath to predict one-year IS risk in AF patients accurately. It aims to identify key molecular pathways and provide individualized risk profiles using routine clinical data.

Methods: We conducted a retrospective cohort study using electronic health records from 7,859 AF patients across 43 Hong Kong hospitals (January 2008–December 2009). ABioSPath employs a multilayer network of gene, protein, and chemical interactions, analyzed via graph convolutional networks, LSTM, and attention mechanisms. Patients with confirmed AF were included, excluding those with incomplete records.

Results: ABioSPath outperformed baseline models, achieving an AUROC of 0.7815 (95% CI: 0.7346–0.8283), positive predictive value of 0.430, negative predictive value of 0.870, sensitivity of 0.500, specificity of 0.885, average precision of 0.409, and Brier score of 0.195. Cohort analysis identified key proteins (CRP, REN, PTGS2) in prevalent pathways. Individual analysis highlighted PIK3/Akt and cytokine/chemokine signaling pathways and revealed IS risks linked to drugs like prochlorperazine maleate, demonstrating robust, interpretable predictions.

Conclusions: ABioSPath provides an effective, interpretable tool for IS risk prediction in AF patients, with potential for broader disease screening. Its use of routine data enhances accessibility, and pathway insights aid drug development. Future work should validate the model across diverse populations and diseases.

Keywords: Ischemic stroke, Atrial fibrillation, Deep learning

Introduction: Opioid deprescribing is recommended to reduce opioid-related harms; however, no Australian studies have assessed the clinical outcomes of opioid dose reductions.
Aims: This study aims to determine if opioid dose changes are associated with mental health-related or substance use-related emergency department (ED) presentations.
Method: This self-controlled case-crossover study used POpulation Level Analysis and Reporting (POLAR) primary care data linked with data from three metropolitan hospitals in Victoria, Australia. People who had an ED presentation between April 2018 and May 2022 and had received ≥four opioid prescriptions in the 12 months preceding their ED presentation were included. Adjusted odds ratios (aOR) for ED presentations were estimated using conditional logistic regression, comparing opioid dose change in the 30 days prior to ED presentation to that in five corresponding sets of control periods of equal length not immediately preceding an ED presentation.
Results: Of the 1,458 eligible patients, 75.9% experienced a >25% reduction in their prescribed opioid dose in the 30 days before ED presentation. Compared with receiving no opioid prescriptions in the 30 days prior, >25% reduction in prescribed opioid dose (aOR: 1.78; 95% CI: 1.44-2.21) or opioid discontinuation (aOR:2.04; 95% CI: 1.48-2.82) was linked to higher odds of ED presentation whilst 10-25% reduction (aOR:0.15; 95% CI: 0.10-0.23) and stable or increased dose (aOR:0.01; 95% CI: 0.008-0.022) was associated with lower odds of ED presentation.
Discussions and Conclusions: Larger opioid dose reduction or discontinuation is associated with increased risk of subsequent mental health-related and substance use-related ED presentations.
Keywords: opioid dose change, case-cross over, emergency department presentations

Introduction: Transvenous pacemakers (TV-VVI) are widely used but carry risks related to intravascular leads and subcutaneous pockets. Leadless pacemakers (LPM) were developed to minimize such complications, yet randomized comparisons with TV-VVI are lacking, and real-world evidence remains limited.
Aims: To compare 30-day readmission and procedural complication outcomes between leadless and single-chamber transvenous pacemakers in older adults using a nationally representative database.
Methods: We analyzed data from the National Readmissions Database (2016–2022) to identify patients aged ≥65 years who received either LPM or TV-VVI implants, as captured by ICD-10 procedural codes. High-dimensional 1:1 propensity score matching was applied to adjust for baseline confounders. Multivariable logistic regression was used to assess differences in readmission rates, procedural complications, and other outcomes.
Results: A total of 49,852 patients were included, with 44.8% receiving LPMs. After matching, LPM recipients had significantly lower odds of device-related complications (adjusted odds ratio [aOR]: 0.45; 95% CI: 0.30–0.65), revision or replacement procedures (aOR: 0.20; 95% CI: 0.11–0.36), implant-related complications (aOR: 0.58; 95% CI: 0.34–0.97), and device infection/inflammation (aOR: 0.48; 95% CI: 0.26–0.85). Conversely, LPMs were associated with higher odds of arteriovenous fistula (OR: 12.09; 95% CI: 2.01–232.52), vascular pseudoaneurysm (OR: 2.19; 95% CI: 1.08–4.58), and pericardial complications (OR: 1.74; 95% CI: 1.28–2.36). No significant differences were observed in 30-day readmission rates (15.5% vs. 15.9%; OR: 1.05; P = 0.079), in-hospital mortality (aOR: 0.94; P = 0.881), or prolonged hospitalization (aOR: 0.62; P = 0.107).
Conclusions: In this large national cohort of older adults, leadless pacemaker implantation was associated with fewer device-related and infectious complications compared to transvenous pacemakers. However, certain vascular and pericardial risks were elevated. These findings highlight the need for continued comparative research to guide optimal pacemaker selection.
Keywords: Leadless pacemaker; Readmission; Procedural complications

Introduction
According to previous evidence, many atherosclerotic cardiovascular disease (ASCVD) patients fail to achieve optimal lipid-lowering targets despite using oral hypolipidemic agents, and the addition of proprotein convertase subtilisin/kexin type 9 inhibitors (PCSK9i) is recommended. However, the real-world evidence remains unclear whether Asian population can achieve treatment goals within one year after receiving PCSK9i therapy.

Aim
This study was aimed to assess the lipid control rate among different ASCVD risk patients with PCSK9i uses.

Methods
We conducted a retrospective cohort study by using the largest multi-institutional electronic medical records in Taiwan. We included patients with hypercholesterolemia who newly received PCSK9i between 2017 and 2023. To outcome assessment, patients without baseline or follow-up LDL-C levels were further excluded. According to ESC guidelines, we defined patients into very high-risk, high-risk, moderate or lower-risk categories based on factors such as blood pressure, cholesterol levels, and the presence of conditions like ASCVD, diabetes mellitus, and chronic kidney disease. The above risk factors were defined by ICD-10-CM code, ICD-10-PCS and laboratory data.

Results
This study included a total of 215 people. Of whom, the average age was 57.0 years (SD: 13.2) and 68.8% were males. The average total cholesterol and LDL were 233 mg/dL, and 153 mg/dL, respectively. Among the patients receiving PCSK9i, 60.5% were categorized as very high risk, 17.7% as high risk, and 21.8% as moderate or low risk. After one year of follow-up, 37.7% of patients in the very high-risk group, 47.4% in the high-risk group, and 55.3% in the low-risk group successfully achieved their treatment goals.

Conclusion
These findings suggest that, in real-world practice, PCSK9 inhibitors used as adjunctive lipid-lowering therapy demonstrate limited effectiveness in patients at very high and high ASCVD risk. Enhanced lipid management strategies may be necessary for these populations.

Introduction: In January 2019, ticagrelor, a P2Y12 inhibitor, was added to the subsidy list for preventing thrombotic events in adult patients with acute coronary syndrome (ACS).

Aims: This study compared real-world outcomes between ticagrelor and clopidogrel and assessed the impact of the subsidy decision on the healthcare system.

Methods: This retrospective study utilised linked national health record databases, and included adult patients with a ACS-related hospital admission, who were initiated on ticagrelor or clopidogrel between January 2017 to December 2020 and on aspirin concurrently. Propensity score matching was employed to balance patient demographics, comorbidities, baseline laboratory values and concomitant medications. Poisson regression was used to estimate incidence rate ratio (IRR) with 95% confidence interval (CIs) for major adverse cardiovascular events (MACE), including acute myocardial infarction (AMI), stroke or transient ischemic attack (TIA), and all-cause death, and secondary outcomes.

Results: After propensity score matching, 3475 patients remained in each cohort. Ticagrelor was associated with a 14% reduction in the incidence of MACE compared to clopidogrel at 1-year follow up [IRR 0.86 (95% CI 0.74, 0.99)], primarily driven by a significant reduction in all-cause mortality among ticagrelor patients [IRR 0.59 (95% CI 0.45, 0.76)]. The incidence of AMI and stroke/TIA for ticagrelor were numerically lower but did not reach statistical significance [IRR 0.96 (95% CI 0.81, 1.14)]. For safety outcomes, no significant differences were observed in bleeding events [IRR 0.87 (95% CI 0.64, 1.17)] or dyspnea [IRR 0.95 (95% CI 0.61, 1.48)] between the two cohorts.

Conclusions: Ticagrelor was associated with a lower risk of MACE compared to clopidogrel, primarily due to a significant reduction in all-cause mortality. These findings align with the results of the pivotal trial and other real-world studies, suggesting ticagrelor as a more effective treatment option in this patient population.

Keywords: Ticagrelor, acute coronary syndrome, real-world outcomes

Introduction:
Real-world evidence (RWE) becomes increasingly important in regulatory and medical affairs in pharmaceutical industry, complementing randomized controlled trials (RCTs) by addressing gaps in long-term safety and effectiveness. This summary reviews how major regulatory agencies evaluate RWE during drug review, and clarifies the specific, practice-driven functions that RWE serves within Medical Affairs.
Aims:
1. Compare the roles of RWE as supportive or primary evidence across major regulatory frameworks.
2. Identify the functions of RWE in Medical Affairs and summarize methodological and procedural enablers for its adoption in routine practice.
Methods:
We conducted a structured review (2018–2025) of RWE guidelines and approval records from major regulatory agencies, including the U.S. FDA, EMA, NMPA, PMDA, and HSA. The analysis examined each agency’s use and approval trends of RWE, and primary applications in Medical Affairs. Quantitative data were sourced from published systematic reviews and agency reports.
Results:
Regulatory frameworks now accept RWE as both supportive and, in defined scenarios, primary evidence. Between 2019 and 2021, 116 FDA approvals cited RWE; 88 influenced benefit–risk decisions and five used RWE as the primary evidence. EMA reported RWE in 40% of 2018–2019 marketing applications, supporting safety and effectiveness. NMPA’s Boao Lecheng and Greater Bay Area pilots led to 13 RWE-supported approvals by 2023, formally recognizing RWE as primary, or extrapolated evidence. PMDA and HSA have also issued actionable RWE guidance. In Medical Affairs, RWE supports local evidence generation planning and gains insights from real-world practice. These impacts are enabled by strong organizational planning and effective collaboration across stakeholders.
Conclusions:
RWE is now globally recognized as supportive and primary evidence in drug approval, with adoption accelerating in Asia into routine decision-making. This shift doesn't only impact regulatory pathways but also reshapes the way Medical Affairs teams drive scientific communication with stakeholders.
Keywords: RWE, drug approval, medical affairs

Introduction: Warm Autoimmune Hemolytic Anemia(wAIHA) is a rare disease categorized into primary(pwAIHA) and secondary wAIHA(swAIHA). Due to its rarity, treatment often relies on physician experience, resulting in significant uncertainty surrounding its characteristics and management in China.

Aim: This study aims to describe the clinical features of adult wAIHA patients.

Method:This retrospective, observational cohort study utilized electronic health records from three hospitals in Shanghai. Adult patients hospitalized with wAIHA between January 2014 and March 2024 were included. Patients exhibiting evidence of acquired or hereditary causes of hemolytic anemia were excluded. Follow-up phone calls by physicians were conducted until death or November 15, 2024, whichever occurred first.

Results:A total of 202 wAIHA patients were identified, with 51.5% classified as pwAIHA. The median age at diagnosis was 59.0 years for pwAIHA and 56.0 years for swAIHA, with 58.7% and 59.2% female, respectively. Notably, 63.5% of pwAIHA and 100.0% of swAIHA patients had comorbidities, primarily hypertension. Severe or very severe anemia(Hb ≤ 60 g/L) was present in 28.8%of pwAIHA and 50.0% of swAIHA patients. Most presented elevated reticulocytes (≥4%)(85.3% of pwAIHA and 85.1% of swAIHA), total bilirubin(≥17.1 µmol/L) (91.2% and 78.3%), and lactate dehydrogenase(>250 U/L) (84.5% and 84.1%), indicating increased red blood cell destruction. In swAIHA underlying disease, autoimmune diseases accounted for 43.9%, hematological diseases for 26.5%, infections for 13.3%, and drug-induced cases for 8.2%. Among pwAIHA patients, 94.2% received corticosteroids as first-line treatment, with 18.3% combining immunoglobulins. The overall response rate was 94.6%, but half of the patients experienced relapse after a median of 130 days(IQR 47.5–480.8). During median follow-up of 31.6 months, complications occurred in 61.5% of patients, with infections and steroid-induced diabetes being the most common treatment-related complications.

Conclusion:This study enhances the understanding of wAIHA's clinical characteristics, emphasizing the need for improved medical interventions and support for affected populations in China.

Introduction
Japan faces a serious issue of remarkably long psychiatric hospital stays, with an average hospitalization period of 299 days in 2022, a duration notably longer than other countries. In response to this situation, the Mental Health and Welfare Act and related legislation were revised in 2022. These revisions were conducted to shorten hospital stays in psychiatric beds.

Aims
This study evaluated the effectiveness of the 2022 revisions to the Mental Health and Welfare Act and related legislation.

Method
A nationwide serial cross-sectional study was conducted using publicly available data from the Japanese Mental Health and Welfare Survey (630 Survey).The analysis was conducted based on patient counts annually for 4 years (June 2021 to June 2024). Long-term hospitalization was defined as continuous hospitalization for one year or more in this study, and the reduction rate of long-term hospitalized psychiatric patients was set as the outcome. The mean percentage (95% confidence interval) of reduction rate for long-term hospitalized psychiatric patients by prefecture (n=47) were calculated.
Paired t-tests were performed to confirm the difference of long-term hospitalized psychiatric patients between 2021-2024. P-value < 0.05 was considered statistically significant.

Result
From 2021 to 2024, there has been a decreasing trend, with the number of long-term hospitalized psychiatric patients across Japan decreasing by 8.97%(164,196 people in 2021 and 149,462 people in 2024) over the four years.
Additionally, means (95% CI) of the number of long-term hospitalized psychiatric patients in 47 prefectures are 3,421 people（2,700 people-4,141 people）in 2021 and 3,180 people (2,526 people-3,834 people) in 2024, a statistically significant decrease were observed.

Conclusion
A substantial decrease in the long-term psychiatric hospitalization rate was observed in Japan after the revision of the Mental Health and Welfare Act and related legislation in 2022.

Keywords
long-term psychiatric hospitalization, policy assessment, serial cross-sectional study

Introduction:
Studies have shown that younger children in a classroom are more likely to be diagnosed with and treated for ADHD. However, the magnitude and implications vary across countries, reflecting differences in educational systems, diagnostic criteria, and prescribing practices. In a Canadian context, previous research on relative age effect has focused exclusively on elementary-school aged students and predated DSM-5.

Aims:
This study investigates whether relative age continues to influence ADHD diagnosis and treatment in the DSM-5 era and examines whether this effect emerges during preschool-age and persists into high-school age among boys and girls.

Methods:
We included British Columbian children aged 3-17 at any point between Jan 1, 2014, to Dec 31, 2023. We estimated period prevalence of ADHD diagnosis and treatment by birth month and calculated the absolute and relative risk of diagnosis and treatment for children born in December vs. January, stratified by age group: preschool (3-5), elementary school (6-12), and high school (13-17).

Results:
2,112,662 children were included in the cohort, and 51.4% were boys. Elementary and high school boys born in December were more likely to receive ADHD diagnosis (relative risk [RR] 1.27, 95% confidence interval [CI] 1.21, 1.32, RR 1.28, 95% CI 1.21, 1.27, respectively). Meanwhile, when comparing girls born in December vs. January, the relative-age effect was stronger in elementary school (RR 1.36 95% CI 1.27, 1.47), but became weaker in high school (RR 1.18, 95% CI 1.11, 1.26). This effect was not present among preschool-aged children. Similar relative-age effects were observed for the use of ADHD pharmacological treatment.

Conclusions:
Relative-age effect in ADHD diagnosis and treatment persisted into high school among boys while declining among girls. Future ADHD assessment should carefully avoid overdiagnosis due to misinterpretation of immaturity related to younger relative age in the classroom.

Background:
Immediate hypersensitivity reactions (IHRs) to systemic cancer therapies are influenced by both patient-specific and treatment-related factors. Identifying these risks is vital for safer chemotherapy delivery.
Objective:
To evaluate demographic, clinical, and regimen-related risk factors associated with the occurrence and severity of IHRs in hospitalized chemotherapy patients.
Methods:
A retrospective review was conducted on 71 cancer patients (43 females, 28 males) who developed IHRs during systemic therapy under clinical pharmacist supervision. Variables analyzed included age, sex, BMI, comorbidities (hypertension, diabetes, renal impairment), and chemotherapy regimens. Reactions were graded using CTCAE v5.0 criteria.
Results:
The median age of IHR cases ranged from 49–73 years, with older adults (60–74 years) showing a higher incidence of Grade II/III reactions—particularly among males receiving ECF or Rituximab-based regimens. While IHRs were more frequent in females, males experienced a higher proportion of severe (Grade III) reactions, especially when comorbidities were present.
BMI distribution showed 60% of patients were of normal weight and 40% overweight/obese, but no direct correlation with reaction severity was found. However, severe reactions occurred across BMI categories in patients with additional risk factors. Hypertension (20% females, 34% males) and type 2 diabetes (10% males) were linked to higher reaction severity. Renal dysfunction was observed in a few male patients with moderate reactions.
High-risk regimens included Paclitaxel ± Carboplatin (females), ECF, and Rituximab (males), while FOLFOX and Oxaliplatin regimens caused mostly mild reactions. Most IHRs occurred within 30 minutes of infusion initiation.
Conclusion:
Advanced age, comorbidities, and certain multi-agent regimens increase IHR risk. BMI alone was not predictive, but overlapping risks worsened outcomes. Clinical pharmacist oversight enhanced early detection. A pre-infusion checklist assessing age, comorbidities, and regimen type is recommended for safer administration.
Keywords: Hypersensitivity, chemotherapy, risk factors, age, comorbidities, paclitaxel, rituximab, hypertension, pharmacist intervention.

Introduction: Diabetic retinopathy (DR), a primary microvascular complication of diabetes mellitus, is leading a cause of new-onset blindness among adults in developed countries, imposing significant economic and societal burdens. The risks of DR associated with glucagon-like peptide-1 receptor agonists (GLP-1RA), sodium-glucose cotransporter 2 inhibitors (SGLT2i), and dipeptidyl peptidase-4 inhibitors (DPP-4i) remain unclear based on findings from large-scale randomized controlled trials.
Aims: This study aimed to compare the effects of GLP-1RA, SGLT2i, and DPP-4i on the risk of DR in patients with type 2 diabetes mellitus (T2DM).
Methods: We conducted a retrospective cohort study using data from the Ningbo Regional Health Information Platform (NRHIP), which includes approximately 2.1 million diabetic patients in Ningbo, China. An active comparator, new-user cohort design was employed to perform pairwise comparisons among new users of GLP-1RA, SGLT2i, and DPP-4i. Propensity score matching (1:1) was used to balance baseline characteristics, including demographics, comorbidities, medication use, and laboratory results. Cox regression models were used to calculate hazard ratios (HRs) and 95% confidence intervals (CIs) for DR.
Results: Three cohorts were established: SGLT2i (n=37,070), GLP-1RA (n=9,596), and DPP-4i (n=56,711). After propensity score matching and multivariate Cox regression analyses with DPP-4i as the reference, GLP-1RA significantly reduced the risk of DR (HR=0.86, 95% CI: 0.75–0.98, p=0.024). Similarly, SGLT2i was associated with a reduced DR risk (HR=0.90, 95% CI: 0.85–0.97, p=0.007). No statistically significant difference in DR risk was observed between GLP-1RA and SGLT2i (HR=0.98, 95% CI: 0.85–1.13, p=0.766).
Conclusions: Our retrospective cohort study of adults with T2DM suggest that the use of GLP-1RA and SGLT2i was associated with a lower risk of developing DR compared to DPP-4i. Clinically, GLP-1RA and SGLT2i can be recommended to reduce DR risk, although regular monitoring is necessary to identify potential vision-threatening complications.
Keywords: Diabetic retinopathy; GLP-1RA; SGLT2i

Introduction: Pediatric patients face elevated ADR rates due to developmental pharmacokinetics, limited clinical data, and prevalent off-label drug use. Hospitalized children show 16.4% cumulative ADR incidence, with anti-infective-related ADRs at 38% (vs. 14.9% in adults), causing prolonged hospitalization and increased costs. Prediction models are critical to identify high-risk children and prevent harm, especially given 20% of pediatric ADRs are preventable. Despite existing systematic reviews assessing ADR prediction models for adult inpatients and older patients, pediatric-specific ADR prediction models lack systematic evaluation.
Aims: To identify published ADR prediction models for pediatric patients and secondarily evaluate their methodological quality.
Methods: Embase, PubMed, CNKI, Wanfang, VIP, and SinoMed were systematically searched from inception to January 9, 2025. Two independent reviewers conducted screening, full-text review, and data extraction, with disagreements resolved by consensus or third-author arbitration. Methodological quality was assessed using the Checklist for Critical Appraisal and Data Extraction for Systematic Reviews of Prediction Modelling Studies (CHARMS), the Prediction Model Risk of Bias Assessment Tool (PROBAST) for bias risk, and the Transparent Reporting of a Multivariable Prediction Model for Individual Prognosis or Diagnosis (TRIPOD) adherence. Predictive ability (AUROC, sensitivity, specificity) were reported.
Results: From 12,667 screened studies, 7 articles (describing 10 models) met the inclusion criteria. Logistic regression was the primary modeling method, with one study using machine learning. Common methodological limitations included unreported handling of missing data and univariable predictor screening. Models predicted heterogeneous ADRs across diverse settings, discrimination (AUROC) ranged from 0.63 to 0.97, with sensitivity and specificity between 52.0%-98.5% and 33.3%-98.8%, respectively. TRIPOD adherence varied from 62.16% to 86.49%. Critically, no models underwent external validation.
Conclusion: This review reveals an urgent need for validated pediatric ADR prediction models. Pervasive methodological limitations and absent external validation preclude clinical use. Prioritizing development and rigorous validation against TRIPOD/PROBAST standards is essential.

The presence of chronic kidney disease (CKD) among patients with heart failure with reduced ejection fraction (HFrEF) often leads to conservative uptitration or underutilization of sacubitril/valsartan and sodium–glucose cotransporter-2 inhibitors (SGLT2i). This hesitancy stems from their pivotal trials that largely excluded patients with severe CKD, with those having an estimated glomerular filtration rate (eGFR) of <30 mL/min/1.73 m2 being particularly underrepresented, thus limiting their real-world applicability and clinical impact.

To address these gaps, this study evaluated the risk-benefit tradeoffs of sacubitril/valsartan and SGLT2i coadministration in patients with HFrEF across the spectrum of time-varying eGFR.

Time-dependent Cox’s models were developed using a counting process framework to accommodate time-varying covariates and interaction terms. Data were retrospectively extracted from electronic health records at National Taiwan University Hospital, comprising 501 patients initiated on sacubitril/valsartan between March 2017 and January 2020, with follow-up until September 2022. Stepwise variable selection was applied, and penalized smoothing splines were used to model nonlinear trends and optimize dose-response thresholds.

A total of 59,814 time-dependent observations were captured, including medication use, echocardiographic parameters, and laboratory findings. The results indicated that patients with HFrEF and an eGFR of <30 mL/min/1.73 m2 derived significant benefits from sacubitril/valsartan (with a cutoff dose of >194 mg daily) and SGLT2i coadministration, regardless of whether renal dysfunction was preexisting or developed during treatment. Model robustness was confirmed by a generalized R2 of 0.45 and an exceptionally high concordance index (0.9340), indicating strong predictive accuracy.

This study challenges outdated treatment hesitancy and provides compelling real-world evidence advocating aggressive yet tailored S/V and SGLT2i coadministration in this high-risk population; these findings call for a shift in clinical practice, ensuring that renal dysfunction is no longer a contraindication but rather an indication for optimized therapy to improve survival outcomes.

heart failure with reduced ejection fraction, sacubitril/valsartan, sodium–glucose cotransporter-2 inhibitors

Introduction: Chiglitazar is a PPAR agonist used for treating type 2 diabetes (T2D). It was developed in China and has been marketed in China since 2021. Systematic review evidence on chiglitazar is lacking.
Aims: To evaluate study quality and describe the safety and efficacy of chiglitazar.
Methods: A systematic review was conducted in PubMed and China National Knowledge Infrastructure (CNKI) (through April 2025). Randomized controlled trials (RCTs) and observational studies in English and Chinese were identified by searching “chiglitazar” (or its Chinese name) and selecting studies evaluating safety, efficacy or effectiveness of chiglitazar. Both the initial search and study selection were performed by 2 independent researchers. Study quality was assessed by Jadad scale (for RCTs) and Newcastle-Ottawa scale (for observational studies). Random-effect models were used for meta-analysis.
Results: Thirteen studies were identified (all in China; 7 in English and 6 in Chinese; 1 observational and 12 RCTs [including 9 post-approval RCTs]). Treatment duration ranged from 4-24 weeks. Five studies were low quality, primarily due to blinding and withdrawal issues.
Eight studies (1402 patients) reported safety (all RCTs). The pooled RRs for AEs and SAEs were 1.00 (95% CI: 0.95, 1.06) and 0.98 (95% CI: 0.91, 1.06), respectively, for 7 placebo-controlled studies; one study showed similar AE/SAE frequencies between chiglitazar and sitagliptin. Among 5 post-approval studies, the pooled RR for AEs was 0.97 (95% CI: 0.87, 1.07) versus placebo. Six studies reported hypoglycemia and 5 studies reported edema (all <5%).
Chiglitazar demonstrated better efficacy (fasting blood glucose and/or HbA1c) versus placebo (7 studies; including 2 pre-approval studies) and was non-inferior or superior to active comparators (sitagliptin, pioglitazone, semaglutide) (6 studies; including 1 pre-approval study).
Conclusions: Chiglitazar is safe and effective for T2D treatment, though study quality varied. Future studies should employ rigorous methodology to strengthen these findings.
Keywords: Chiglitazar, safety, efficacy

Introduction:
With the expanding use of immune checkpoint inhibitors (ICIs) and the increasing reliance on corticosteroids to manage immune-related adverse events, safety data on the concomitant use of corticosteroids with ICIs remain insufficient.
Aims:
To identify safety signals of ICIs when used concomitantly with corticosteroids in VigiBase.
Methods:
The Individual Case Safety Reports (ICSRs) for anti-PD-1, anti-PD-L1, and anti-CTLA-4 inhibitors were extracted from VigiBase through December 2024. We performed a preferred term level disproportionality analysis, comparing reports with ICIs co-administration with corticosteroid against ICI alone as the main analysis. Concomitant use was defined by suspected, interacting, or concomitant events in the report. Signals were defined when having at least three reports and any of the following criteria were met: proportional reporting ratio or reporting odds ratio ≥ 2 with chi-square ≥ 4, or information component 95% confidence intervals ≥ 0. Detected signals were grouped by the system organ class (SOC). The subgroup analyses were performed by ICI monotherapy versus ICI combination therapy.
Results:
Among the 445,721 ICSRs for ICI, 55,050 (12.3%) were reported concomitantly with corticosteroids. Concomitant users were more likely to be male (56.9% vs 53.8%), aged 45–64 years (31.1% vs 25.9%), and had serious events (86.2% vs 77.5%). Disproportionality analysis identified 300 signals for concomitant use, and the most frequently reported SOC was “infections and infestations”, followed by “gastrointestinal disorders and Respiratory”. In the subgroup analysis, the leading SOC remained consistent with the main analysis, but the next most common SOC differed with “Respiratory, thoracic and mediastinal disorders” for monotherapy and “Nervous system disorders” for ICI combination therapy.
Conclusion:
Concomitant use of ICIs and corticosteroids may modify the overall safety profile. These findings demonstrate the need for additional research to confirm the safety of concomitant ICI and corticosteroids.

Key words: Immune checkpoint inhibitors, VigiBase, Corticosteroids

Introduction:
Polypharmacy is a global health issue among older adults, and Japan is not an exception. In March 2021, operational instructions to introduce polypharmacy management systems in hospitals were released. Additionally, revisions to the medical reimbursement system, including provisions related to polypharmacy, were enacted by the Ministry of Health, Labour and Welfare (MHLW) in April 2022.

Aims:
This study aimed to analyse the nationwide polypharmacy reduction in Japan after newly implemented national policies during the COVID-19 pandemic.

Methods:
This study employed a serial cross-sectional design using annual prescription-based reimbursement claim data from a national database covering fiscal years (FY) 2020 (April 2020–March 2021), 2021 (April 2021–March 2022), and 2022 (April 2022–March 2023). Older adults were defined as 75 years and older, according to the academic definition in Japan. The primary outcome was the polypharmacy proportion (PP), calculated as the percentage of annual prescriptions containing seven or more medications out of the total number of annual prescriptions. Data from 335 secondary medical areas (SMAs), officially designated by the MHLW as regional healthcare units, were used as samples.

Results:
The mean (95% CI) PP across all SMAs (n=335) was 4.16% (4.04–4.27) in FY2020 and 3.86% (3.76–3.97) in FY2022. A statistically significant decrease was observed (p < 0.001, paired t-test), and the reduction ratio of PP is 7.21% from FY2020 to FY2022.
Stratification analysis by sex and age group showed that the PP decreased from FY2020 to FY2022 among males (4.14% to 3.90%), females (3.31% to 2.96%), and those aged 75 years and older (6.41% to 6.04%), respectively.

Conclusions:
A significant reduction in nationwide polypharmacy was observed after the implementation of new policies during the COVID-19 pandemic in Japan.

Keywords: polypharmacy, policy, COVID, older adults

Introduction: While sex differences in suicidal behavior and drug response are known, sex-specific patterns in suicide-related adverse drug events (ADEs) among children and adolescents remain insufficiently characterized.
Aims: To investigate sex-specific differences in the reporting patterns of suicide-related ADEs in children and adolescents.
Methods: We compared female and male reports of suicide-related ADEs among individuals aged 2–17 years in VigiBase from 2000 to 2024. vigiPoint is a data-driven pharmacovigilance method that highlights distinguishing features of a data subset using shrinkage-adjusted log-odds ratios (logORs). Assessed features included age group, region, report type, reporter, subtypes of suicide-related ADE and drug. Features with 99% two-sided Bayesian credibility intervals of shrinkage-adjusted logORs exceeding ±0.5 were identified as key. Drug categories with sex-specific reporting differences identified by vigiPoint were further assessed through disproportionality analysis for potential safety signals. Signals were defined when reporting odds ratio or proportional reporting ratio ≥ 2, chi-square ≥ 4, and information component 95% confidence intervals ≥ 0.
Results: Among 31,881 suicide-related ADE reports, 64.8% were reported for females and 32.3% for males. Suicidal attempt (73.4% vs 50.6%) was more frequent in female reports, whereas suicide ideation (14.6% vs 33.8%) was less frequent. Of the drugs identified with sex-specific differences in suicide-related ADEs, paracetamol (9.1% vs 2.9%) and ibuprofen (4.5% vs 2.2%) were more reported in females, while atomoxetine (1.2% vs 7.8%), isotretinoin (2.5% vs 8.1%), methylphenidate (1.2% vs 4.7%), montelukast (2.3% vs 6.4%), and risperidone (1.4% vs 2.7%) in males. Disproportionality analysis detected signals for atomoxetine, isotretinoin, montelukast, and paracetamol in both sexes, and methylphenidate in males only.
Conclusions: A signal for methylphenidate in relation to suicide-related ADEs was detected only in males. Further study is needed to evaluate the potential association between methylphenidate and suicide in male children and adolescents.

Keywords: Pharmacovigilance, Suicide-related adverse drug events, Sex-specific differences

Introduction: Adolescent suicide is the leading cause of death in South Korea, with habitual drug use identified as a significant risk factor for suicide-related behaviors (SRBs). Understanding sex-specific differences in the association is essential for developing targeted prevention strategies.
Aims: To investigate sex differences in the association between habitual drug use and SRBs among Korean adolescents.
Methods: This cross-sectional study utilized data from the 2024 Korea Youth Risk Behavior Web-based Survey of adolescents aged 14 to 19. Habitual drug use was defined as self-reported habitual or intentional use of drugs (e.g., tranquilizers, stimulants, opioid analgesics) or inhalants (e.g., glue, cannabis, cocaine). SRBs, including ideation, planning, and attempts, were assessed based on experiences within the past 12 months. Descriptive statistics for adolescent characteristics were presented as unweighted frequencies and weighted percentages. Adjusted odds ratios (aORs) and 95% confidence intervals (CIs) were calculated using sex-stratified multiple logistic regression with adjustment for covariates.
Results: Among 54,653 adolescents, 1.2% of males and 1.6% of females reported habitual drug use. The prevalence of SRBs was higher among females than males, with rates of suicidal ideation (16.2% vs. 9.4%), planning (6.0% vs. 3.7%), and attempts (3.3% vs. 2.2%). Habitual drug use was significantly associated with an increased likelihood of SRBs in both sexes. Specifically, males with habitual drug use exhibited higher aORs for suicidal ideation (2.04 [1.49–2.79]) and suicidal attempts (3.38 [2.29–5.01]) than females (1.80 [1.42–2.28] for ideation; 2.35 [1.77–3.11] for attempts). Conversely, females had a slightly higher aOR for suicidal planning related to habitual drug use (2.66 [2.02–3.50]) than males (2.37 [1.72–3.26]).
Conclusions: Both males and females who used drugs habitually showed a significant increase in SRBs, although the strength of these associations varied by sex. These findings underscore the need for sex-specific strategies for suicide prevention.
Keywords: Habitual Drug Use, Suicide-Related Behaviors, Adolescents

Introduction:
Amyotrophic lateral sclerosis (ALS) is a progressive and fatal neurodegenerative disorder. Elevated circulating low-density lipoprotein (LDL) cholesterol is associated with an increased risk of ALS onset. Previous studies have explored the relationship between statin use and ALS onset. However, findings have been conflicting, possibly due to methodological limitations, such as confounding by indication and failure to account for baseline differences in LDL cholesterol levels. Robust comparative studies are warranted to clarify the role of statins in ALS onset.

Aims:
To evaluate the effect of statin use on ALS onset in patients with hypercholesterolemia.

Methods:
We conducted an active-comparator, new-user cohort study using two Japanese administrative claims databases from April 2012 to February 2024. Patients with hypercholesterolemia who initiated statins or ezetimibe after at least 365 days of baseline observation and without prior ALS diagnosis were included. The outcome was incident ALS defined as a first definitive diagnosis of ALS. A Cox proportional hazards model was used to estimate the hazard ratio of statins versus ezetimibe. Inverse probability of treatment weighting using propensity scores was applied to adjust for confounding. A fibrate user cohort was analyzed for benchmarking.

Results:
The study included 607,292 statin users, 26,963 ezetimibe users and 114,871 fibrate users. The incidence rate per 100,000 person-years of ALS was 6.8, 15.9, and 4.3, respectively. The adjusted hazard ratio for statins compared with ezetimibe was 0.42 (95% CI, 0.19–0.92). Prior to treatment initiation, mean LDL cholesterol levels exceeded 160 mg/dL equally in both the statin and ezetimibe users.

Conclusions:
This study suggests that statins lower the risk of ALS onset compared with ezetimibe in patients with hypercholesterolemia. This potential protective effect of statins may not be solely attributable to their lowering effect on circulating LDL cholesterol.

Keywords:
Amyotrophic lateral sclerosis; statins; administrative claims database

Introduction: Chronic kidney disease (CKD) increases the risk of cardiovascular disease (CVD), particularly in older adults. However, there is controversy in statin use in older adults with CKD due to their underrepresentation in clinical trials. The effectiveness and safety of statins for primary prevention in elderly CKD patients require further investigation.

Aims: This study aimed to evaluate the effectiveness of statin therapy in reducing CVD and all-cause mortality among older CKD patients. It also assessed the safety of statins by examining the risk of major adverse events.

Methods: We analyzed electronic health records from Hong Kong, identifying 22,565 CKD patients aged 75–84 years and 8,811 aged ≥85 years with elevated LDL-C (≥2.6 mmol/L) between 2008 and 2015. Target trial emulation was used to estimate associations between statin use and incidence of CVD, all-cause mortality and major adverse events. Pooled logistic regression models were used to estimate intention-to-treat (ITT) and per-protocol (PP) effects.

Results: Statin therapy significantly reduced the risk of overall CVDs among patients aged 75–84 years (ITT: HR 0.94 [95% CI: 0.89–0.99]; PP: 0.86 [0.80–0.92]). A substantial risk reduction in all-cause mortality was also observed in this age group (ITT: 0.87 [0.82–0.91]; PP: 0.78 [0.72–0.84]). Among patients aged ≥85 years, statin use was similarly associated with lower CVD risk (ITT: 0.88 [0.79–0.99]; PP: 0.81 [0.71–0.92]). All-cause mortality was likewise reduced in this older cohort (ITT: 0.89 [0.81–0.98]; PP: 0.80 [0.71–0.91]). No significant increase in the incidence of myopathies or liver dysfunction was observed in either age group.

Conclusion: Statins are effective and safe for primary prevention of CVD and all-cause mortality in CKD patients aged ≥75 years, including those ≥85 years. Further research could be conducted to investigate the effects of statins on kidney function in this population.

Keywords: chronic kidney disease, elderly, statins

Introduction: Emerging evidence suggests potential hepatic benefits from both renin-angiotensin-aldosterone system inhibitors (RAASi) and sodium-glucose cotransporter 2 inhibitors (SGLT-2i), but their combined effect is unclear.
Aims: To evaluate the synergistic association of concomitant RAASi and SGLT-2i therapy with the risk of major hepatic events.
Methods: We utilized nationwide claims data (2014-2023) of Korea to establish two active-comparator cohorts of patients with type 2 diabetes and hypertension. In Cohort 1, RAASi users initiated either SGLT-2i or dipeptidyl-peptidase 4 inhibitor (DPP4i). In Cohort 2, SGLT-2i users initiated either RAASi or calcium-channel blocker (CCB). After inverse probability of treatment weighting, we used weighted Cox models to estimate hazard ratios (HRs) and 95% confidence intervals (CI) for a composite of major hepatic events (ascites, variceal bleeding, hepatic failure, or liver transplant) under an as-treated approach.
Results: In Cohort 1, among 1,533,726 patients receiving RAASi, 244,702 initiators of SGLT-2i and 1,289,024 initiators of DPP4i were identified. Adding SGLT-2i to existing RAASi therapy was associated with a lower risk of major hepatic events compared to adding DPP4i (HR 0.66, 95% CI 0.63–0.69). In Cohort 2, among 49,239 patients receiving SGLT-2i, 39,979 initiators of RAASi and 9,260 initiators of CCB were identified. Adding RAASi to SGLT-2i therapy showed a non-significant risk reduction versus adding CCB (HR 0.81, 95% CI 0.64–1.04). However, a subgroup analysis of Cohort 2 revealed that the addition of angiotensin receptor blocker (ARB) was associated with a reduced risk (HR 0.78, 95% CI 0.61–1.00; p=0.047), whereas angiotensin-converting enzyme inhibitor (ACEi) was not (HR 1.77, 95% CI 1.08–2.89).
Discussion: The combination of SGLT-2i and RAASi therapy was associated with a reduced risk of major hepatic events. This benefit was most pronounced when adding SGLT-2i to ongoing RAASi therapy and when ARB, specifically, was added to ongoing SGLT-2i therapy.

Introduction:
Parkinson’s disease (PD) is a progressive neurodegenerative disorder with rising prevalence in type 2 diabetes mellitus (T2DM). Previous studies suggest that certain antidiabetic drug classes, such as GLP-1 receptor agonists, SGLT2 inhibitors, and DPP-4 inhibitors, may have neuroprotective effects beyond glycemic control, potentially lowering PD risk.
Aims:
The relationship between dipeptidyl peptidase-4 (DPP-4) inhibitors and Parkinson’s disease (PD) risk remains unclear. Therefore, we aimed to examine the association between DPP-4 inhibitor use and PD incidence in South Korean adults with type 2 diabetes mellitus (T2DM).
Methods:
Using the Korean National Health Insurance Service-National Sample Cohort (2002–2019), we conducted a retrospective cohort study. Patients with T2DM initiating DPP-4 inhibitors were propensity score-matched 1:1 with controls by age, sex, Charlson Comorbidity Index, and index year. Each cohort included 3,961 patients. Statistical analyses included t-tests, chi-square tests, and Cox proportional-hazard models to estimate hazard ratios (HRs) with 95% confidence intervals (CIs).
Results:
During follow-up, Parkinson’s disease occurred in 17 patients (0.43%) in the DPP-4 inhibitor cohort and 69 patients (1.74%) in the matched control cohort. DPP-4 inhibitor users had a significantly lower risk of PD than matched controls (HR = 0.573, p<0.05). Cardiovascular disease (CVD) was associated with increased PD risk (HR=2.785, 95% CI=1.783–4.350), as was peptic ulcer disease (PUD; HR=2.272, 95% CI=1.473–3.505). Meglitinide use was also linked to higher PD risk (HR=10.517, 95% CI=1.410–79.946). Other antidiabetic classes, such as SGLT2 inhibitors and GLP-1 receptor agonists, showed no significant associations.
Conclusions:
DPP-4 inhibitor use was associated with lower PD incidence in patients with T2DM., suggesting potential neuroprotection. Comorbid CVD, PUD, and meglitinide use were associated with increased PD risk. Prospective studies are needed to confirm these findings and clarify underlying mechanisms.
Keywords: DPP-4 inhibitors, Parkinson’s disease, type 2 diabetes mellitus

Introduction: Biosimilars are cost-saving alternatives to original biologics, but their uptake remains limited in Japan. Understanding prescription and switching trends is key to promoting biosimilar use.
Aims: To describe the temporal trends in the prescription of biologics in Japan, with additional analysis focusing on switching from original biologics to biosimilars.
Methods: Using the JMDC claims database (Jan 2005‐May 2024), we identified patients who received at least one prescription for 17 biologics (original biologics or biosimilars). We examined monthly trends in the proportions of original biologics vs biosimilars. Furthermore, we assessed the proportion of patients by usage pattern: original-only, biosimilar-only, and switchers. Biosimilar adoption by medical institution type was also estimated.
Results: Temporal trends in the proportions of original biologics and biosimilars varied widely. In May 2024, the proportion of biosimilar prescriptions was 13.6% for somatropin and 92.5% for filgrastim. At the individual level, the proportion of patients switching from original biologics to biosimilars was low (1.2–14.0%), indicating that switches do not often occur within the same patient, while more recent new users of biologics start biosimilars. At the institutional level, university-related hospitals were more likely and clinics were less likely to introduce biosimilars than public and other types of hospitals.
Conclusion: Temporal trends and switching patterns of biosimilars vary by the type of biologics. The type of medical institution influences biosimilar use and should be considered in assessments. Further research and strategies are needed to promote biosimilar use in clinics.
Keywords: biological products, biosimilar pharmaceuticals, pharmacoepidemiology

Introduction: Overall survival (OS) is often considered gold standard outcome in clinical trials for oncology treatments in patients with locally advanced (LA) head and neck squamous cell carcinoma (HNSCC). However, the need for long-term follow-up significantly increases the duration of the trials and thereby delaying patient access to new treatments. Using event-free survival (EFS) as an early clinical endpoint would potentially be more efficient than OS. Therefore, there is a need to evaluate the relationship between EFS and OS.
Aims: To describe and compare the EFS and OS patterns of LA HNSCC by different stages at diagnosis and by resectable/unresectable diseases, and to evaluate the correlations between EFS and OS using patient-level real-world clinical data.
Methods: This is a retrospective cohort study using electronic health records in Hong Kong. Patients with LA HNSCC and started primary treatment between 2006 and 2017 were included. The patients were followed for 5 years. Events for OS included all-cause deaths, while events for EFS included locoregional progression, recurrent, distal metastasis, and all-cause deaths. Patterns for EFS and OS were evaluated using Kaplan–Meier curves, and correlations between EFS and OS (in days) were measured by correlation coefficient, R.
Results: A total of 1,811 eligible patients was included. EFS and OS showed similar survival patterns: better survival was observed in patients with lower stages at diagnosis, or with resectable tumours (log-rank test p<0.05). EFS showed good correlation with OS (R=0.86). Subgroup analysis considering only either locoregional progression, recurrent, or distal metastasis events showed good correlation between distal metastasis-free survival and OS, but moderate correlations for the other two.
Conclusions: EFS showed good correlation with OS, suggesting that it could potentially be used as an early endpoint in trials for LA HNSCC.
Keywords: locally advanced head and neck squamous cell carcinoma, event-free survival, overall survival

Introduction: Patients with haematological malignancies (HM) are at high risk for drug–drug interactions (DDIs) due to complex treatment regimens and frequent antibiotic use. Not all drug interactions shown by databases results in clinically significant outcomes. Differentiating actionable interactions from theoretical alerts is essential for optimizing therapeutic decisions through antimicrobial stewardship.

Aim: To evaluate the frequency, severity and clinical significance of potential antibiotic-drug interactions and to implement targeted interventions in HM patients.

Methodology: A prospective observational study was carried out in oncology wards for the duration of two years (January 2023-December 2024). DDIs were identified using Micromedex database. Data was analyzed using descriptive statistics and Man-Whitney U test.

Results: Among the 393 HM patients. A total of 1,712 DDIs were identified, of which 604 (35.2%) involved at least one antibiotic. Of these: 76.9% (n=465) were classified as major, 18.3% (n=111) as moderate, 4.4% (n=27) as minor, and 0.1% (n=1) as contraindicated. The most frequently observed high risk combinations included: Azithromycin–Ondansetron (n=48) associated with QT prolongation, Trimethoprim-sulfamethoxazole–Fluconazole (n=17) associated with cardiotoxicity risk. A total of 12.5% (n=76) antibiotic drug interactions were diagnostically validated and clinically managed. The median length of stay was significantly higher among patients with diagnostically validated drug interactions compared to those with possible interactions (Mann–Whitney U = 3.5, Z = -6.88, p < 0.001).

Discussion: Diagnostic validated antibiotic-drug interactions had a notably longer hospital stay. These findings indicates that although interaction databases identify many potential DDIs only a portion of this results in actual clinical consequences. The prolonged length of stay may reflect the impact of adverse effects, diagnostic work-up, therapeutic changes, or complications arising from interaction-related toxicity. This study highlights the importance of distinguishing clinically actionable interactions and reinforce the role of prospective validation and antimicrobial stewardship in optimizing patient outcomes.

Keywords:
Antibiotic interactions, drug-drug interactions, antimicrobial stewardship

Introduction: Heterogeneity in baseline conditions, clinical characteristics, and intraoperative management may predispose patients to differential risks for adverse postoperative outcomes. Understanding this phenotypic heterogeneity is crucial for trajectory pattern analysis and for tailoring individualized perioperative care strategies.
Aims: This study aimed to identify distinct cardiac surgery patient phenotypes from multimodal, high-dimensional data to enhance risk stratification.
Methods: We conducted a multicenter, retrospective study included 10,847 adult patients undergoing cardiac surgery with cardiopulmonary bypass at three tertiary hospitals in eastern China (2013-2024) and an external cohort from a publicly available research dataset (2011-2020). A high-dimensional dataset was constructed by integrating demographic and surgical information, clinical features, and high-resolution intraoperative vital signs, resulting in a comprehensive feature set composed of 1,006 key parameters. Phenotypes were derived via unsupervised agglomerative hierarchical clustering, validated across three external datasets.
Results: Five distinct phenotypes were identified from the derivation cohort and clinically confirmed by expertise, with median ages ranging from 52 to 67 years and female representation varying from 24.8% to 44.0%. Phenotypes A (stable perfusion) and B (mild stress response) represented healthier subsets with stable intraoperative profiles, and phenotype A had the lowest risk of acute organ injury. Phenotype C (cardiovascular compromise), despite comprising older patients with extensive coronary artery disease, exhibited milder hemodynamic stress and lower risk of organ injury. Phenotypes D (coagulopathy and inflammation) and E (severe hemodynamic instability) were associated with significantly higher incidences of 7-day acute kidney injury (46.0% and 66.9%) and 48-hour acute liver failure (37.9% and 38.8%), greater mortality risk and prolonged ICU/hospital stays.
Conclusions: Data-driven phenotyping revealed distinct subgroups within heterogeneous surgery patients, each exhibiting unique characteristics linked to adverse outcomes. The integration of dynamic intraoperative vital signs with perioperative data enhances risk stratification and supports individualized perioperative management strategies.
Keywords: clinical phenotyping, unsupervised learning, multimodal data

Background
There are conflicting findings regarding the risk of acute pancreatitis (AP) associated with use of dipeptidyl peptidase-4 (DPP-4) inhibitors among prior studies, primarily due to influence of confounding factors.
Objectives
To evaluate the risk of AP associated with use of DPP-4 inhibitors using trend-in-trend (TiT) design.
Methods
We conducted the study using data derived from the National Health Insurance Database. We applied TiT design to evaluate the risk of AP associated with the use of DPP-4 inhibitors. Patients with any diagnosis of type 2 diabetes mellitus (T2DM) between 2013 to 2021 were enrolled. Index date was defined as either the date of the first DPP-4 inhibitors prescription or a randomly assigned date for patients without exposure to DPP-4 inhibitors. TiT is a hybrid epidemiologic‐ecologic research design that examining trends in the frequency of the outcome of interest as a function of trends in exposure. TiT method has been shown to estimate the causal effect without being influenced by unmeasured confounders. Propensity score (PS) overlap weighting, adjusted for measured confounders, was employed for comparison with TiT analysis. Sensitivity analyses included suicide attempts and acute kidney injury as negative control outcomes.
Results
We identified 2,089,710 T2DM patients with 536,284 patients exposed to DPP-4 inhibitors during the study period. Under TiT analysis, use of DPP-4 inhibitors was not associated with AP (odds ratio: 0.57; 95% CI: 0.54–2.02). Results from the PS approach indicated a higher risk of AP (1.36; 95% CI 1.30–1.42). Additionally, under the same approach, DPP-4 inhibitors were associated with increased risk of suicide attempts (1.33; 95% CI 1.24–1.42) and acute kidney injury (1.76; 95% CI 1.71–1.82).
Conclusions
Our findings did not suggest an increased risk of AP following the use of DPP-4 inhibitors. Potential unmeasured confounders should be considered and controlled using appropriate methods.

Keyword: Trend-in-Trend, Confounder, DPP-4 inhibitors

Introduction:
Sodium-glucose co-transporter 2 (SGLT2) inhibitors have proven benefits in the management of Type 2 Diabetes Mellitus (T2DM) but concerns regarding an increased risk of urinary tract infections (UTIs) persist.
Aims:
To assess and compare the incidence and clinical profile of UTIs among T2DM patients receiving SGLT2 inhibitors versus those on other anti-diabetic therapies.
Methods:
A retrospective observational study was conducted based on medical records of 1,129 T2DM patients treated between January 2020 and December 2023. Patients were categorized into the SGLT2 inhibitor group (n = 541) and non-SGLT2 group (n = 588). Data on UTI incidence, patient demographics, comorbidities, microbial patterns, and clinical outcomes were analyzed.
Results:
UTI incidence was significantly higher in the SGLT2 inhibitor group (45.84%) compared to the non-SGLT2 group (27.21%) (p < 0.0001). HbA1c levels were significantly lower in the SGLT2 group (mean 6.90±1.27) compared to the non-SGLT2 group (7.20±1.56; p = 0.001). Male patients in the SGLT2 group had a higher rate of UTI (54.43%) than those in the non-SGLT2 group (41.25%; p = 0.0093), while females in the non-SGLT2 group showed a higher UTI incidence. Common comorbidities included hypertension (74.66%), chronic kidney disease (70.15%), and ischemic heart disease (48.36%). Escherichia coli was the most frequently isolated pathogen. Most UTI cases were uncomplicated and responded well to standard antimicrobial therapy. No significant differences were observed between groups regarding UTI severity, recurrence, hospitalization rates, or renal complications.
Conclusions:
The study found a significantly higher incidence of UTI among T2DM patients using SGLT2 inhibitors compared to those on other anti-diabetic therapies, although most infections were uncomplicated and manageable. SGLT2 inhibitors remain a valuable therapeutic option, but careful monitoring for UTI symptoms is recommended, particularly in high-risk patients.
Keywords:
SGLT2 inhibitors; Type 2 Diabetes Mellitus; Urinary Tract Infections

Introduction: With the growing availability of biologics and small molecule therapies, treatment options for inflammatory bowel disease (IBD) have expanded considerably. Understanding real-world utilization patterns of advanced therapies is essential to inform clinical decision-making.
Aims: To describe patient characteristics and treatment patterns of advanced therapies in IBD.
Methods: We conducted a descriptive cohort study using South Korea’s national health insurance claims database. Adults aged ≥19 years newly diagnosed with IBD between January 2014 and December 2022 were identified. Patients were stratified by IBD subtype—ulcerative colitis (UC) or Crohn’s disease (CD). We assessed baseline characteristics at initiation of first advanced therapy, treatment persistence, and switching patterns.
Results: Among 53,552 patients (UC: 40,288; CD: 13,264), 8,847 (UC: 4,044; CD: 4,803) initiated advanced therapies. The median age at initiation was 43 years (IQR 30–56) for UC and 29 years (IQR 23–38) for CD. Median time-to-initiation was shorter for CD (9.6 months; IQR 3.0–27.3) compared to UC (16.9 months; IQR 6.0–36.5). In both subtypes, infliximab was the most frequently used first-line agent (UC: 42.3%, CD: 55.9%), followed by adalimumab (UC: 19.8%, CD: 24.6%). Vedolizumab (21.0%) and ustekinumab (14.6%) were also commonly used in UC and CD, respectively. Persistence with initial therapy was longer in CD (33.2 months; IQR 18.0–58.1) than in UC (23.4 months; IQR 11.2–43.6). Upon switching to second-line therapy, most patients transitioned to a different class (UC: n=256, 67.2%; CD: n=178, 71.5%). The most frequent second-line agents were tofacitinib in UC (24.9%) and ustekinumab in CD (40.2%).
Conclusions: TNF-α inhibitors remain the most widely used advanced therapies, largely due to their early introduction to the market. However, recent approvals of additional biologics and small molecule agents have expanded options. As the therapeutic landscape evolves, further studies are warranted to guide optimal treatment selection for initiating patients.
Keywords: Drug utilization; IBD; advanced treatment

INTRODUCTION: Dapagliflozin received funding in September 2022 for treating heart failure (HF) with reduced ejection fraction and chronic kidney disease. However, following a revised pricing proposal, empagliflozin was funded in November 2023 while dapagliflozin was delisted in August 2024 due to unfavourable cost-effectiveness compared with empagliflozin.

AIMS: To evaluate the impact of drug listing changes on SGLT2i utilisation and prescribing among HF patients.

METHODS: An interrupted time series analysis of public healthcare institutions’ drug dispensing data was used to evaluate changes in SGLT2i utilisation before and after listing changes. Segmented regression models assessed level changes (LC) and trend changes (TC) in defined daily doses (DDDs), accounting for autocorrelation and excluding two “phase-in” months for anticipatory policy responses. Prescribing patterns in adult HF patients were compared using linked national health records with prevalent and incident SGLT2i use in 2021 and 2023. Prescribing rates between these years were compared, with rate ratios (RR) calculated from Poisson regression models.

RESULTS: After dapagliflozin listing in September 2022, its use increased [LC +73.5K DDDs/month (95% CI 14.6K, 132.4K)] while empagliflozin use declined [LC -116.0K DDDs/month (95% CI -185.8K, -46.3K)]. The utilisation shifted significantly towards empagliflozin after its listing in November 2023 [LC +106.0K DDDs/month (95% CI 35.1K, 176.8K); TC +105.4K DDDs/month (95% CI 95.0K, 115.7K)], with corresponding decreases in dapagliflozin [LC -164.2K DDDs/month (95% CI -225.6K, -102.8K); TC -100.1K DDDs/month (95% CI -108.9K, -91.3K)], partly due to early notice of dapagliflozin's delisting as well. SGLT2i prescribing rates in HF patients increased by 109% [RR 2.09 (95% CI 2.04-2.15)] for prevalent users and 59% [RR 1.59 (95% CI 1.52-1.65)] for initiators.

CONCLUSIONS: Drug listing changes significantly impacted SGLT2i utilisation and prescribing, underscoring the role of drug pricing policies in shaping funding decisions and medication use in the public healthcare sector.

Keywords: SGLT2i, drug utilisation, prescribing patterns

Introduction
Cardiovascular disease (CVD) is a major cause of mortality and comorbidity in Hong Kong. To alleviate the healthcare burden and improve CVD prevention, we developed P-CARDIAC, a population-based risk model incorporating over 120 risk factors. Validation of P-CARDIAC in local clinical settings is essential for its feasibility.
Aims
The objective of this study is to evaluate the accuracy of the P-CARDIAC. We anticipate that the findings will provide valuable insights for effective service mapping.
Methods
We conducted a prospective population-based cohort study in a cardiac specialist outpatient clinic under Hospital Authority. Eligible participants, aged 18 to 80 years with prior CVD events, were recruited. Their electronic health records were extracted to observe recurrent CVD outcomes and calculate P-CARDIAC scores. P-CARDIAC scores at CV event occurrence, with and without CV-related treatment, were compared. A one-way ANOVA was performed to compare the scores.
Results
The preliminary results of this study were updated till mid-April, 2025, and included data with a mean age of 65.8 years among 859 participants. There were 33 participants experienced recurrent CVD events among 503 with prior CVD event records. All participants were undergoing at least three polypharmacy treatments during both the period before and after their recurrent CVD date. Among the participants with recurrent CVD events, 93.9% were male, with coronary heart disease being the most common type of recurrent event (72.7%). P-CARDIAC scores in both previous and recurrent CVD dates were statistically significant higher than those scores at times without a CV event. Participants receiving treatment showed lower estimated risk.
Conclusions
P-CARDIAC scores were significantly higher for participants with CVD events, demonstrating its sensitivity to event occurrences. We expect P-CARDIAC to be a frontline tool to stratified different risk patients and enable cost-effective allocation of healthcare resources.
Keywords: Cardiovascular disease, Risk prediction tool, Recurrent CVD event

Introduction: Evidence indicates that vitamin D is essential for enhancing immunity and supporting wound healing by influencing the vitamin D receptor (VDR) and cathelicidin (LL-37). On the other hand, a deficiency in vitamin D may result in reduced production of VDR and LL-37, which could worsen the severity of diabetic foot infections (DFI) and increase the chances of requiring surgical interventions.

Aim: To determine the association between vitamin D status, VDR protein, and LL-37 levels with various treatment strategies in DFI patients.

Methods: A hospital-based cross-sectional study involving 169 DFI patients was conducted in a tertiary care facility in South India. Patient demographics and details of the treatment (surgical or only medical) administered was collected from the hospital's medical records. Serum vitamin D levels were measured using an electrochemiluminescence immunoassay. VDR and LL-37 were estimated in serum using ELISA kits. Data were analysed using Jamovi software version 2.4.8.0.

Results: Among the 169 patients with DFI, 139 (82.2%) received surgical interventions, while 30 (17.8%) were managed conservatively with medications. Patients who underwent surgical management showed lower median levels of vitamin D, VDR, and LL-37 in comparison to those who received drug therapy (13.2 ng/mL, 0.704 ng/mL, and 1.29 ng/mL vs. 25.1 ng/mL, 1.34 ng/mL, and 2.46 ng/mL, respectively). Additionally, in former patients a significant increase in levels of HbA1c, C-reactive protein, neutrophil-lymphocyte ratio, and systemic immune-inflammatory index, along with an extended duration of hospital stay (p < 0.001). Furthermore, levels of vitamin D, VDR, and LL-37 were lower in patients who underwent amputation (p < 0.001).

Conclusion: Our findings indicate that patients with DFI who received surgical management exhibited a significant decrease in vitamin D, VDR protein, and LL-37 levels compared to patients who received only drug therapy.

Introduction:
Glucagon-like peptide-1 receptor agonists, including semaglutide, are widely used to manage type 2 diabetes and obesity. Although their effects on individuals with obesity are well documented, evidence regarding their impact in individuals with normal weight is limited. Some randomized trials have included participants with body mass index (BMI) below 25 kg/m², but outcomes in this group remain under-investigated. To address this gap, we evaluated semaglutide’s weight loss effect in normal-weight adults.
Aims:
This study aimed to assess the weight loss effect of semaglutide in normal-weight adults through a meta-analysis of randomized controlled trials (RCTs).
Methods:
We conducted a systematic review on April 24, 2024, using PubMed, Embase, and the Cochrane Library, yielding 1,157 records. After screening and eligibility assessment, four RCTs comprising ten treatment arms were included. Trials were eligible if they reported weight outcomes for participants with BMI < 25 kg/m² and included placebo control arms. Separate meta-analyses were performed by dosage (0.5 mg and 1.0 mg) using a random-effects model.
Results:
In trials using pure placebo comparators (SUSTAIN-1 and Ikushima et al.), the 0.5 mg and 1.0 mg groups showed pooled mean differences in weight of −2.87 kg (95% CI: −4.53, −1.20) and −4.23 kg (95% CI: −7.67, −0.79), respectively. When all eligible placebo-controlled trials were assessed, some of which allowed the use of background antidiabetic medications, the 0.5 mg group showed −3.14 kg (95% CI: −4.41, −1.87) and the 1.0 mg group −3.37 kg (95% CI: −4.87, −1.86). Between-study heterogeneity was minimal (I² = 0–25.8%).
Conclusions:
Semaglutide significantly reduced weight in adults with BMI < 25 kg/m². These findings provide clinical insight into semaglutide’s effect across BMI ranges and support the need for further research on its use in non-obese populations.
Keywords:
Semaglutide, weight loss, meta-analysis