Chao-Kuei Juan’s research while affiliated with Taichung Veterans General Hospital and other places

Background Psoriasis is associated with various cardiovascular diseases (CVDs). The aim of this study was to compare the risk of CVD in patients with psoriasis who were prescribed biologics or oral therapies, and to assess the association between different classes of biologics and CVD risk. Methods and Findings This retrospective cohort study utilized the TriNetX Global Collaborative Network (2014–2025). Patients with psoriasis newly prescribed biologics (BIO-cohort) and those newly initiating oral anti-psoriatic drugs without biologic exposure (Non-BIO-cohort) were enrolled. A propensity score-matched analysis was conducted, accounting for age, sex, race, comorbidities, body mass index, serum lipid profile, and inflammatory marker levels. Cardiovascular risk was compared between the BIO- and Non-BIO-cohorts using Cox regression to calculate hazard ratios (HRs) with 95% confidence intervals (CIs). After matching, each cohort comprised 12,732 patients, with approximately 50% being female, a mean age of 57 years, and 55% identifying as White. The 5-year cumulative incidence of any CVDs was significantly lower in the BIO-cohort (10.68%; 95% CI [10.03%, 11.36%]) than in the Non-BIO-cohort (16.17%; 95% CI: [15.34%, 17.05%]) ( p < 0.001). The BIO-cohort had attenuated risks of any CVDs (HR 0.621; 95% CI [0.571, 0.676]), cerebrovascular diseases (HR 0.616; 95% CI [0.519, 0.731]), arrhythmias (HR 0.632; 95% CI [0.565, 0.706]), inflammatory heart diseases (HR 0.566; 95% CI [0.360, 0.891]), ischemic heart diseases (HR 0.579; 95% CI [0.465, 0.721]), heart failure (HR 0.637; 95% CI [0.521, 0.780]), non-ischemic cardiomyopathy (HR 0.654; 95% CI [0.466, 0.918]), thrombotic disorders (HR 0.570; 95% CI [0.444, 0.733]), peripheral arterial occlusive diseases (HR 0.501; 95% CI [0.383, 0.656]), and major adverse cardiac events (HR 0.697; 95% CI [0.614, 0.792]). Receiving only anti-tumor necrosis factor (TNF)-α (HR 0.886; 95% CI [0.807, 0.973]), anti-interleukin (IL)-17 (HR 0.724; 95% CI [0.599, 0.875]), or anti-IL-23 (HR 0.739; 95% CI [0.598, 0.914]) was associated with reduced risks of any CVDs, whereas no significant association was observed for only anti-IL-12/23 (HR 0.915; 95% CI [0.742, 1.128]). This risk reduction remained consistent across various subgroups, including age (≤45 or >45 years), sex (male or female), regions of research data (the United States, Europe, Middle East and Africa, and Asia-Pacific), and comorbidities (psoriatic arthritis, hypertension, diabetes, hyperlipidemia, overweight or obesity). Eight sensitivity analyses, such as extending the washout period or tightening medication definitions, validated our findings. The main limitation of our study is the observational design, which can only establish associations, not causation. Conclusions Patients with psoriasis prescribed biologics exhibited a lower risk of CVDs versus those on oral therapy. Anti-TNF-α, anti-IL-17, and anti-IL-23 were associated with decreased cardiovascular hazards, while anti-IL-12/23 was not.

We aimed to develop an accurate and efficient skin cancer classification system using deep-learning technology with a relatively small dataset of clinical images. We proposed a novel skin cancer classification method, SkinFLNet, which utilizes model fusion and lifelong learning technologies. The SkinFLNet's deep convolutional neural networks were trained using a dataset of 1215 clinical images of skin tumors diagnosed at Taichung and Taipei Veterans General Hospital between 2015 and 2020. The dataset comprised five categories: benign nevus, seborrheic keratosis, basal cell carcinoma, squamous cell carcinoma, and malignant melanoma. The SkinFLNet's performance was evaluated using 463 clinical images between January and December 2021. SkinFLNet achieved an overall classification accuracy of 85%, precision of 85%, recall of 82%, F-score of 82%, sensitivity of 82%, and specificity of 93%, outperforming other deep convolutional neural network models. We also compared SkinFLNet's performance with that of three board-certified dermatologists, and the average overall performance of SkinFLNet was comparable to, or even better than, the dermatologists. Our study presents an efficient skin cancer classification system utilizing model fusion and lifelong learning technologies that can be trained on a relatively small dataset. This system can potentially improve skin cancer screening accuracy in clinical practice.

Background The risk of osteoporosis has been explored in atopic dermatitis (AD). The long-term risk of fractures in AD patients and the effects of various AD treatment on bone health remain to be elucidated. Objective To examine the long-term risk of fractures in AD patients. Methods This nationwide matched cohort study was conducted using the Taiwan's National Health Insurance Research Database, for the period 1997 to 2013. A total of 36,855 AD patients and 147,420 reference subjects without AD were identified. Demographic characteristics and comorbidities were compared, and cumulative incidence of fractures was examined. Adjusted hazard ratios for fracture risks of AD and various AD treatment were calculated using Cox proportional hazard model. Results: A total of 1,518 patients (4.12%) in the AD cohort and 5,579 patients (3.78%) in the reference cohort had fractures (p=.003). The mean ages were 22.6 years in both groups. The 16-year cumulative incidence of fractures in the AD cohort (8.043%) was significantly higher than that in the reference cohort (7.366%) (p=.002). Severe AD (adjusted hazard ratio [aHR] 1.31, 95% confidence interval [CI] 1.08-1.59) was independently associated with fractures. Other independent risk factors included exposure to topical (aHR 1.21, 95% CI 1.05-1.39) or systemic corticosteroids (≥10 mg/day, aHR 1.62, 95% CI 1.38-1.91). Use of disease modifying antirheumatic drugs (aHR 0.71, 95% CI 0.53-0.90) and phototherapy (aHR 0.73, 95% CI 0.56-0.95) were associated with a lower risk of fractures. The results were consistent across sensitivity analyses. Conclusion AD patients have a higher incidence of fractures. Severe AD is independently associated with fractures.

Objectives: The evolution of psoriasis (PsO) to psoriatic arthritis (PsA) has been proposed recently. There are three phases that occur in sequence prior to classifiable PsA: PsO patients, PsO patients with a positive imaging, and PsO patients with arthralgia not explained by other diagnosis. The purpose of this study was to compare the differences among preclinical phases using ultrasound and clinical assessment. Methods: Patients with psoriasis were recruited. Patients who had been previously diagnosed with psoriatic arthritis or who had used biologics were excluded. A 52-joint ultrasound (52j US) assessment and clinical assessments including the swollen joint count, tender joint count, erythrocyte sediment rate, C-reactive protein, dactylitis score, enthesitis score, psoriasis severity, and nail psoriasis severity, were performed. Results: A total of 188 eligible psoriasis patients were enrolled. Physical examination revealed 39 patients (20%) with at least one swollen joint. The 52j US assessment demonstrated 90 patients (47%) having at least one joint with grey-scale score 2-3. All patients were further stratified into PsO patients (n=58), PsO patients with a positive imaging, (n=59), PsO patients with arthralgia not explained by other diagnosis (n=27), and classifiable PsA (n=39). There were no differences in clinical characteristics other than tender joint count found among the three preclinical phases of PsA. Dactylitis score, swollen joint count and heatly assessment questionnaire score were significantly higher in classifiable PsA. Conclusions: Nearly half of the psoriasis patients without previously diagnosed psoriatic arthritis would be classified into the preclinical phases of psoriatic arthritis based on the 52j US and clinical assessments. Ultrasound assessment is helpful for identifying psoriasis patients who are in the preclinical phases of psoriatic arthritis, particularly for those without arthralgia.

Background: Up to 25% of patients with cutaneous lupus erythematosus (CLE) can develop systemic lupus erythematosus (SLE). However, the risk of autoimmune diseases other than SLE in CLE patients who have only skin manifestations (CLE-alone) has rarely been explored. Objective: To investigate the long-term risk and independent factors of non-SLE autoimmune diseases among CLE-alone patients. Method: A nationwide cohort study using the Taiwanese National Health Insurance Research Database 1997-2013. CLE patients and matched subjects were included. Cumulative incidences of autoimmune diseases after 1 year of CLE-alone diagnosis were compared. Cox proportional hazard model was also performed. Results: A total of 971 CLE-alone patients and 5,175 reference subjects were identified. The 10-year cumulative incidence of autoimmune diseases other than SLE was significantly elevated in the CLE-alone cohort (9.00%, 95% confidence interval [CI] 6.72-11.29) than in the reference cohort (4.20%, 95% CI 3.53-4.87%) (p < 0.001). CLE-alone was independently associated with non-SLE autoimmune diseases (adjusted hazard ratio 1.55, 95% CI 1.10-2.18). Among CLE-alone patients, females and those taking long-term systemic corticosteroids (a proxy for extensive disease) were associated with non-SLE autoimmune diseases after adjusting for the number of repeated autoimmune laboratory tests. Conclusion: CLE-alone is independently associated with future non-SLE autoimmune diseases.