Accurate disease prediction using clinical datasets is essential for improving early diagnosis and clinical decision-support systems; however, many existing deep learning approaches are disease-specific, computationally intensive, and difficult to generalize across heterogeneous biomedical datasets. This study addresses this challenge by proposing a unified and dataset-aware deep learning framework that enables accurate and interpretable disease prediction across diverse clinical datasets. The framework adopts a modular architecture that selects appropriate models based on dataset characteristics such as feature dimensionality, sample size, and class imbalance. It integrates multiple deep learning architectures, including MLP, one-dimensional CNN, FT-Transformer, autoencoder-based classifiers, and ensemble strategies. Robust preprocessing, fold-safe feature selection, and nested cross-validation are incorporated to ensure reliable performance evaluation. The framework is evaluated on three heterogeneous benchmark datasets: the UCI Heart Disease dataset (303 samples, 13 clinical features), the PIMA Indians Diabetes dataset (768 samples, 8 metabolic features), and the Parkinson's disease voice dataset (195 recordings, 22 acoustic features). Experimental results demonstrate competitive predictive performance relative to classical baselines across the diverse tasks. The FT-Transformer + autoencoder ensemble achieved an AUC of 0.8980 (±0.0483) for heart disease prediction, while the CNN + Autoencoder ensemble obtained an AUC of 0.8451 (±0.0270) for diabetes classification. For Parkinson's disease detection, the MLP achieved an AUC of 0.7538 with perfect specificity. Overall, all models achieved AUC values comparable to ML baselines. The study contributes a scalable and interpretable deep learning framework that improves reliability, generalization, and practical applicability for multi-disease prediction in real-world healthcare environments.

BACKGROUND Cutaneous allodynia is reported as a pain resulting from a stimulus that does not normally provoke or elicit pain and an altered quality of sensation characterized by extreme sensitivity to touch or mild temperature change. Glucagon-like peptide-1 receptor agonists (GLP-1RAs) are widely prescribed for management of diabetes mellitus and for obesity and weight management. Tirzepatide is a synthetic polypeptide and dual agonist for glucagon-like peptide-1 (GLP-1) and glucose-dependent insulinotropic polypeptide receptors. Clinical studies demonstrate that tirzepatide provides robust glycemic control and is highly effective for the treatment of obesity. Review of the tirzepatide U.S. prescribing information revealed hypersensitivity and injection site reactions as the only dermatological adverse events, with no reports of skin pain or allodynia. A recent pharmacovigilance analysis of the U.S. Food and Drug Administration Adverse Event Reporting System evaluating 6 GLP-1RAs, including tirzepatide, identified rare reports of allodynia; however, detailed case-level descriptions remain limited. CASE REPORT In this case series, we report moderate-to-severe allodynia in 2 patients with severe obesity treated with tirzepatide for weight management. Allodynia in these patients varied from static to dynamic in nature; the incidence of allodynia was temporally associated with dose escalation, related to higher doses, and resolved after termination of the drug, demonstrating strong association with tirzepatide treatment. CONCLUSIONS Although dermatologic adverse events, such as injection site reactions, hypersensitivity, and urticaria, have been reported, to the best of our knowledge, this is the first case series reporting allodynia specifically with tirzepatide therapy.

Hypertensive renal disease (HRD) is the second leading cause of end-stage renal disease (ESRD) following diabetes mellitus, with oxidative stress and inflammation serving as synergistic pathogenic drivers, the molecular mechanisms of which remain incompletely elucidated. Epigenetic regulation (especially histone methylation) is pivotal in chronic kidney disease (CKD). SMYD3, a histone methyltransferase, has been reported to modulate oxidative stress and inflammation via mediating H3K4me3 modification, while its specific role and regulatory mechanism in HRD remain largely unclear. This study explored SMYD3's function using angiotensin II (Ang II)-induced HRD models (28-day subcutaneous Ang II-infused mice in vivo; HK-2 cells in vitro). Mice were grouped into Control, HRD, HRD + MTA (an H3K4 methylation inhibitor), BCI-121 (a SMYD3 inhibitor), and HRD + BCI-121. In vivo, blood, urine, and kidney samples were analyzed via biochemical assays (creatinine, BUN, oxidative stress biomarkers) and histopathology (HE, PAS, Masson staining). In vitro, SMYD3 was inhibited by BCI-121 or siRNA, with Western blotting, co-IP, and ChIP detecting interactions among SMYD3, H3K4me3, TXNIP promoter, and JAK2/STAT3 pathway-related molecules. Ang II infusion aggravated renal dysfunction (elevated creatinine, BUN, urinary albumin), pathological damage, oxidative stress, inflammation, and cellular senescence, accompanied by increased SMYD3 and H3K4me3. Treatment with MTA/BCI-121 alleviated these changes, and SMYD3 knockdown/inhibition reversed Ang II-induced injuries in HK-2 cells. Mechanistically, SMYD3 was associated with enhanced TXNIP transcription via H3K4me3 methylation, activating NLRP3 inflammasome and oxidative stress pathways. SMYD3 was regulated by the JAK2/STAT3 pathway; STAT3 inhibitor S3I-201 reduced SMYD3 and H3K4me3, indicating that JAK2/STAT3 upregulates SMYD3 to exacerbate HRD. In conclusion, our findings demonstrate that SMYD3 acts as a key responsive mediator of Ang II-induced renal oxidative stress and inflammation, which is closely associated with the promotion of H3K4me3 enrichment at the TXNIP promoter. We also identify that the Ang II-activated JAK2/STAT3 axis may function as an upstream regulator of SMYD3 expression, thus providing novel insights and potential therapeutic targets for HRD.

Lung cancer remains the leading cause of cancer-related deaths worldwide. Advances in immunotherapy and targeted therapies have improved survival in stage IV non-small cell lung cancer (NSCLC), highlighting the importance of local control of the primary tumor. Hypofractionated radiotherapy (hypo-RT), delivering high biologically effective doses (BEDs) over shorter durations, shows potential in locally advanced NSCLC but is underexplored in metastatic settings.

This retrospective study evaluated 104 patients with stage IV NSCLC who received hypo-RT to the primary tumor at LMU Munich between December 2013 and June 2022. The primary endpoints were overall survival (OS) and progression-free survival (PFS), while the secondary endpoints included local failure-free survival, regional failure-free survival, distant failure-free survival, and the identification of prognostic factors.

Median OS and PFS were 15.4 months (95% confidence interval [CI]: 10.5-24.4) and 3.9 months (95% CI, 3.2-6.2), respectively. Significant predictors of OS included Eastern Cooperative Oncology Group performance status >1 (hazard ratio [HR]: 3.260, P = .0003), male sex (HR: 1.869, P = .038), metastases in >2 organ systems (HR: 2.014, P = .022), and BED < 58.5 Gy (HR: 2.117, P = .017). Predictors of PFS included smoking history <30 pack-years (HR: 1.912, P = .003) and BED < 58.5 Gy (HR: 1.816, P = .025).

Hypo-RT is effective and feasible in metastatic NSCLC, with higher BEDs improving survival. Findings highlight the importance of personalized treatment strategies integrating clinical, molecular, and therapeutic factors.

Timely palliative care can reduce the disease burden and improve quality of life in patients with cancer. Although several studies have developed assessment models for palliative care needs in cancer patients, the quality and clinical applicability of these models remain unclear.

To systematically review existing assessment models for palliative care needs in patients with cancer, with a focus on their characteristics, predictors, risk of bias, and applicability.

A systematic search was conducted in PubMed, Cochrane Library, Embase, Web of Science, CINAHL, Scopus, China National Knowledge Infrastructure (CNKI) through September 12, 2025. Data extraction and evaluation were rigorously performed by two researchers based on the Prediction Model Risk of Bias Assessment Tool (PROBAST).

A total of 5714 articles were identified, and eight studies were included, which covered 24 models for assessing palliative care needs. The sample size of the included studies ranged from 179 to 54,628, with areas under the curve ranging from 0.724 to 0.998. The models in all the included studies encompassed four categories of predictive factors: general demographic data, symptom/functional assessments, laboratory indicators, and treatment status. Five studies were rated as having a high risk of bias, primarily due to high risks associated with participants and conclusions, with generally low applicability.

Existing models demonstrate potential for identifying patients with cancer who have increased palliative care needs using routinely collected clinical data. Commonly included predictors were symptom burden, functional status, laboratory parameters, treatment-related factors, and demographic characteristics. However, the overall body of evidence is constrained by a substantial risk of bias, particularly arising from inappropriate data sources, limited sample sizes, suboptimal handling of continuous variables, insufficient reporting of missing data, and the lack of robust internal or external validation. In addition, many models adopted mortality-based surrogate outcomes rather than clinically meaningful indicators of palliative care needs. Therefore, the currently available models should be interpreted with caution, and further high-quality model development and external validation are required before they can support broader routine clinical implementation. Future research should prioritize clinically actionable outcomes and incorporate patient-, caregiver-, and family-level factors to improve the relevance of these models for referral decisions and care planning.

Meningiomas are the most common primary intracranial tumors, often treated surgically. However, complete resection is frequently limited by proximity to critical structures, necessitating adjuvant or definitive radiotherapy. Proton therapy offers dosimetric advantages over photon-based radiotherapy, particularly in sparing adjacent normal tissues. This study aims to systematically evaluate the effectiveness and safety of proton therapy for intracranial meningiomas across tumor grades and clinical scenarios. A systematic review and meta-analysis was conducted according to PRISMA 2020 guidelines using PubMed, EMBASE, Scopus, Web of Science, and Cochrane from inception to November 10, 2025. Studies were eligible if they reported clinical outcomes of proton therapy in ≥ 10 adult meningioma patients. Data extraction and risk-of-bias assessment were performed independently by two reviewers. Pooled complication rates and survival outcomes were calculated using random-effects models. Nineteen studies involving 1,431 patients were included. WHO Grade I tumors comprised 70.6% of cases; Grades II/III made up 25.2% and 4.2%, respectively. The most common proton dose regimens ranged from 13 to 70.2 Gy (RBE). The pooled complication rate was 16% (95% CI 5-27; p < 0.001; I² = 98.5%). Nine studies reported a statistically significant 5-year overall-survival proportion of 91% (95% CI 88-94; p < 0.001; I² = 49.3%). Radiologic local control averaged 71% (95% CI 50-86; I² = 88.2%). Proton therapy provides effective tumor control with acceptable toxicity, especially for low-grade or anatomically complex meningiomas. It is a valuable option for select patients, though further prospective studies are needed to optimize dosing and assess long-term outcomes.

No multicenter study has validated cochlear dose constraints for hearing preservation in pediatric brain tumor patients in a real-world setting. Although the PENTEC review proposed a 35 Gy mean cochlear dose threshold, supporting evidence from heterogeneous, multicenter pediatric cohorts remains scarce.

To evaluate dosimetric, therapeutic, and clinical risk factors for sensorineural hearing loss (SNHL) after pediatric cranial radiotherapy in a national multicenter cohort, with a specific focus on validating the clinical relevance of the 35 Gy mean cochlear dose threshold.

We retrospectively analyzed 88 children treated with cranial radiotherapy between 2014 and 2024 across four French pediatric radiotherapy centers participating in the national PediaRT registry. All patients had baseline and at least two post-radiotherapy audiograms graded according to the Chang Ototoxicity Scale, with SNHL defined as Chang grade ≥ 1a. Mean (Dmean) and minimum (Dmin) cochlear doses were extracted and analyzed using Kaplan-Meier estimates and Cox proportional hazards models.

Over a median follow-up of 37 months, 17 patients (19.3%) developed SNHL. In multivariate analysis, both a mean cochlear dose ≥ 35 Gy (HR = 3.90; 95% CI: 1.24-12.23; p = 0.020) and cisplatin exposure (HR = 3.85; 95% CI: 1.40-10.64; p = 0.009) were independently associated with SNHL. The 5-year cumulative incidence of SNHL was 43.4% for Dmean ≥ 35 Gy versus 13.1% for < 35 Gy (p = 0.0014), and 51.6% with cisplatin versus 12.4% without (p < 0.001). A significant monotonic relationship between cochlear dose and SNHL severity was observed (p = 0.0011).

This study provides the first national, multicenter validation of the 35 Gy mean cochlear dose threshold in a real-world pediatric population and identifies cisplatin exposure as a major independent risk factor. These findings support strict cochlear dose minimization during treatment planning and systematic long-term audiological surveillance in pediatric cancer survivorship care.

Kabuki syndrome is a rare disorder characterized by growth retardation, distinctive craniofacial features, intellectual disability, and congenital anomalies that can affect various systems. Disease-causing variants in KMT2D and KDM6A, two genes that regulate transcription via histone modifications, are responsible for most of the cases. We retrospectively reviewed the medical records and molecular studies of eight patients suspected of Kabuki syndrome in four centers in Turkey between 2016 and 2021. We detected eight patients with a definitive diagnosis of Kabuki syndrome. All patients had intellectual disability/developmental delay and some of the distinctive dysmorphic features. Most of the clinical presentations occurred at a frequency similar to the current literature. Seven patients had pathogenic variants in the KMT2D gene, while three variants were previously unreported. One was diagnosed with his characteristic clinical findings. A patient with a novel KMT2D variant had recurrent bone fractures. Kabuki syndrome affects various systems, necessitating a multidisciplinary approach to diagnosis and follow-up. Clinical findings, as well as molecular studies, are valuable in diagnosis. In this study, we clarify the signs and symptoms of the patients and contribute to the molecular spectrum of Kabuki syndrome.

This study´s aim was to analyse clinical outcomes of microvascular reconstructions performed for malignancy-caused midface defects of varying sizes at our department.

This retrospective study included 98 patients (38 women; 60 men; age range 16-94 years) who underwent microvascular reconstruction for a malignancy-related midface defect between April 2017 and August 2022.

Bony (fibula, scapula) and soft tissue (radial forearm, anterolateral thigh) flap techniques were used depending on defect size. Most defects requiring coverage were classified as class II (n = 70, 71.4%; 18 (25.7%) bony vs. 52 (74.3%) soft tissue) or class IV (n = 16, 16.3%; 4 (25.0%) bony vs. 12 (75.0%) soft tissue) according to Brown and Shaw. The proportion of other horizontal defect classes was significantly lower (n = 12, 12.2%; all soft tissue). Unilateral defects were present in 81 cases (82.7%; 15 (18.5%) bony vs. 66 (81.5%) soft tissue), while bilateral defects were observed in 17 cases (17.3%; 7 (42.2%) bony vs. 10 (57.8%) soft tissue). Postoperative complications were observed in 57 cases (58.2%), predominantly comprising minor issues such as bleeding or wound healing disorders. Major complications, including the need for transplant revisions, occurred in 9 cases (9.2%), with early transplant loss in three instances. No significant correlation was found between transplant loss and defect classification (p = 0.73) or transplant type (p = 0.07). However, bony flaps demonstrated a higher tendency towards flap loss compared to soft tissue flaps (18.2% vs 4.0%). Dental rehabilitation using dental implants was performed in only 18.2% of all patients who received bony reconstruction.

The complication rate associated with microvascular reconstructions was relatively high. Most complications were minor and could be managed therapeutically. Major complications, such as flap failure, remained relatively uncommon. Neither defect size nor transplant type used demonstrated a significant impact, though bony transplants showed a higher tendency towards flap loss. Despite the observed complication rate, free flaps continue to represent a reliable treatment option for patients with midfacial defects. In cases where bony reconstruction is not strictly necessary, dental rehabilitation using patient-specific preprosthetic implants may offer a viable alternative to bone grafts.

To evaluate the feasibility of thoracolaparoscopic resection for adenocarcinoma of the esophagogastric junction with left intrathoracic anastomosis performed without patient repositioning. A retrospective analysis was performed on the clinical data of patients who underwent thoracolaparoscopic resection for adenocarcinoma of the esophagogastric junction with left intrathoracic anastomosis without patient repositioning at the Cancer Hospital, Chinese Academy of Medical Sciences and Peking Union Medical College from August 2016 to January 2020. A total of 39 patients were enrolled (32 males and 7 females), with the age of (64.2±7.2) years. The patients' surgical status, postoperative pathological stage, postoperative complications, and follow-up information were analyzed. The survival curves were plotted using the Kaplan-Meier method. All 39 procedures were completed successfully with no conversion to thoracotomy or laparotomy. The operative time was 250.0 (223.5, 289.0) min, and intraoperative blood loss was (165±79) ml. Postoperative pathological stages were stage Ⅰ in 13, stage Ⅱ in 4, stage Ⅲ in 11, and stage Ⅳ in 11 patients. R0 resection was achieved in all patients. Postoperative complications included anastomotic leakage in 3 patients (7.7%, 3/39), pulmonary infection in 3 patients (7.7%, 3/39), arrhythmia in 3 patients (7.7%, 3/39), pleural effusion in 3 patients (7.7%, 3/39), incision infection in 1 patient (2.6%, 1/39), gastroesophageal reflux in 5 patients (12.8%, 5/39), and iatrogenic injury of the right mediastinal pleura in 3 patients (7.7%, 3/39). All complications were resolved after conservative treatment. There were no cases of chylothorax, reoperation, ICU admission, or perioperative death. The patients were followed up for a median of 55 (36, 60) months postoperatively. The 1-year and 3-year disease-free survival rates were 87.2% and 69.2%, respectively. The 1-year, 3-year, and 5-year overall survival rates were 92.3%, 74.4%, and 48.1%, respectively. Thoracolaparoscopic adenocarcinoma of the esophagogastric junction resection with left intrathoracic anastomosis performed without patient repositioning appears feasible and has clinical value for broader adoption.