Comparative data on sequential transarterial chemoembolization (TACE) after stereotactic body radiation therapy (SBRT) in recurrent hepatocellular carcinoma (HCC) remain limited. This study aimed to evaluate the efficacy of this combination.

We retrospectively reviewed 152 patients with recurrent HCC who met predefined eligibility criteria; 109 received SBRT alone and 43 received SBRT plus TACE. To minimize selection bias, a 2:1 propensity score matching was performed, resulting in 68 patients in the SBRT-alone group and 36 in the SBRT plus TACE group for the final comparative analysis. Overall survival, progression-free survival, and local control were assessed using the Kaplan-Meier method.

The SBRT plus TACE group was associated with numerically higher survival rates, although this difference did not reach statistical significance. The cumulative one-, three-, and five-year overall survival rates were 91.2%, 76.3%, and 61.8% for SBRT alone, compared to 100.0%, 86.1%, and 77.5% for the combination therapy ( p = 0.069). The corresponding progression-free survival rates were 73.1%, 51.1%, and 32.3% versus 88.9%, 58.1%, and 52.3% ( p = 0.091). No acute grade ≥3 toxicities were observed in either group.

In this exploratory analysis of recurrent HCC, the combination of SBRT and TACE demonstrated a favorable trend toward improved survival compared with SBRT alone, without an increase in severe toxicity. While these findings did not reach statistical significance, they establish the safety profile of the combined approach and provide preliminary evidence supporting its potential therapeutic role. This hypothesis-generating study justifies and informs the design of larger, prospective trials to definitively evaluate the efficacy of this regimen.

Early and accurate prognostic assessment is crucial to avoid serious disease progression in patients with liver failure. Thyroid hormone is an important metabolic regulator involved in hepatic function. This review examines in detail the pathophysiological regulation of the hypothalamic-pituitary-thyroid axis in patients with liver failure and emphasizes the importance of thyroid profiling (thyroid-stimulating hormone, T3, and T4) in prognostic assessment and risk stratification. T3 can enhance liver regeneration. The clinical application of thyroid hormone replacement therapy in patients with acute-on-chronic liver failure complicated by non-thyroidal illness syndrome is controversial. This review aims to inform clinical practice regarding the relevance of TH level assessment in liver failure and to provide novel insights into the prognostic evaluation and comprehensive care of liver failure complicated by thyroid dysfunction.

Diabetic cardiomyopathy (DCM), a diabetes-specific cardiovascular complication, is pathologically characterized by cardiomyocyte apoptosis, oxidative stress, inflammatory responses, and myocardial fibrosis, distinguishing it from other cardiac disorders, such as hypertension and coronary artery disease. Challenges in early diagnosis, coupled with the limited efficacy and adverse effects of current treatments, have made DCM a significant contributor to heart failure and mortality in patients with diabetes. Natural products, recognized for their diverse sources, structural variety, and multitarget therapeutic potential, have shown promise in preventing and treating DCM. Drawing on advances over the past five years, this review systematically summarizes the pharmacological effects and molecular mechanisms of natural products (e.g., flavonoids, terpenoids, phenylpropanoids, alkaloids, and polysaccharides) in the treatment of DCM, with the aim of providing a theoretical foundation for further research and drug development.

The history of screening for cervical cancer is rich with implementing cutting-edge ideas and technologies. From the very first “Pap smear” to the semi-automated and computerized systems of today, the way we screen for cervical cancer has changed dramatically in the past 75 years. With the advent of new techniques and more advanced machine learning algorithms, we sought to understand the current and future applications of artificial intelligence in clinical pathology around cervical cancer screening, diagnosis, and treatment.

A structured narrative review was conducted to examine the historical evolution and contemporary advances in cervical cancer screening, diagnosis, excisional management, and artificial intelligence applications. Peer-reviewed articles, consensus guidelines, and global policy documents published between January 1990 and March 2025 were identified through targeted searches of PubMed and review of reference lists from relevant publications. Search terms included combinations of “cervical cancer screening,” “Papanicolaou test,” “liquid-based cytology,” “HPV testing,” “colposcopy,” “loop electrosurgical excision procedure,” “digital pathology,” “deep learning,” and “artificial intelligence.” Emphasis was placed on multi-center validation studies, systematic reviews, regulatory and implementation guidance, and global health frameworks. Publications lacking methodological transparency or direct relevance to clinical or translational practice were excluded.

Through a review of the literature, we describe how innovations in conventional and liquid-based cytology, human papillomavirus testing, and organized screening programs established the current prevention framework. Building on this foundation, recent studies demonstrate promising performance of deep learning algorithms applied to conventionally prepared cervical cytology slides, with systems capable of binary normal versus abnormal classification as well as more granular diagnostic categorization. Artificial intelligence-assisted colposcopy and computer-vision approaches have also shown improved diagnostic concordance, workflow efficiency, and potential to expand screening capacity in resource-limited environments.

There has been much work done in the past several years surrounding the implementation of deep learning algorithms in regard to cervical cancer screening. The work in this field shows promise in enhancing diagnostic accuracy, streamlining diagnostic workflow, and decreasing turnaround times from specimen collection to rendering a diagnosis. However, there are still many technical, legal, and ethical questions that must be answered prior to widespread adoption of these algorithms for patient care.

Colorectal cancer histopathological grading relies on the accurate segmentation of glandular structures. Current deep learning–based methods depend heavily on large-scale pixel-level annotations that are labor-intensive and not amenable to clinical practice. Weakly supervised semantic segmentation offers a promising alternative; yet, existing class activation map–based weakly supervised semantic segmentation approaches often produce incomplete, low-quality pseudo-masks that overemphasize discriminative regions and fail to provide reliable supervision for unannotated glandular structures, limiting their suitability for dense histopathology segmentation under sparse supervision. We propose a novel weakly supervised teacher–student framework that leverages sparse pathologists’ annotations and an Exponential Moving Average–stabilized teacher network to generate refined pseudo-masks.

Our framework integrates confidence-based filtering, adaptive fusion of teacher predictions with limited ground truth, and curriculum-guided refinement, enabling the student network to progressively delineate and accurately segment unannotated glandular regions. We validated our framework on an institutional colorectal cancer cohort from The Ohio State University Wexner Medical Center, consisting of 60 hematoxylin and eosin-stained whole-slide images from independent patients with varying degrees of gland differentiation, as well as on public benchmarks including the Gland Segmentation dataset (derived from stage T3–T4 colorectal adenocarcinomas), TCGA-COAD, TCGA-READ, and SPIDER.

The proposed framework achieved strong performance on the institutional dataset despite limited annotations. On the Gland Segmentation dataset, it demonstrated competitive performance compared to both weakly and fully supervised approaches, achieving a mean Intersection over Union of 80.10% ± 1.52 and a mean Dice coefficient of 89.10% ± 2.10. Moreover, cross-cohort evaluations showed robust generalization on TCGA-COAD and TCGA-READ without requiring additional annotations, while reduced performance on SPIDER reflected pronounced domain shift.

Our framework provides an annotation-efficient and generalizable paradigm for accurate gland segmentation in colorectal histopathology, offering a practical pathway toward significantly reducing annotation burdens while preserving high segmentation fidelity.

Normalization can standardize and improve machine learning (ML) performance on omics data. However, it is unclear whether normalization is associated with overfitting (i.e., worse cross-dataset performance than intra-dataset performance). Therefore, we aimed to examine associations of normalization and regularization with overfitting of ML on omics data.

Using three paired transcriptomic and clinical datasets (lung adenocarcinoma: the Cancer Genome Atlas (TCGA)/Oncology Singapore; melanoma: TCGA/Dana-Farber Cancer Institute; glioblastoma: TCGA/Clinical Proteomic Tumor Analysis Consortium), we applied ANOVA-based gene selection methods, six normalization methods, and six ML models to classify cancer patients’ deaths. Balanced accuracy (BA) and area under the curve (AUC) in intra- and cross-dataset settings were compared using inferential analyses.

Normalization consistently improved intra-dataset performance (median BA/AUC changes: 0.035–0.214/0.115–0.279) on all data, particularly with Z_Raw, but decreased or slightly increased cross-dataset performance (median BA/AUC changes: −0.029–0.079/0.029–0.064). Least Absolute Shrinkage and Selection Operator (LASSO) model without normalization consistently outperformed most of the ML models in cross-dataset testing across cancer types. ML models on all and molecular-alone data showed similar best performances.

Normalization increases ML’s intra-dataset performance and overfitting in three paired cancer transcriptomic and clinical datasets. Regularized models such as LASSO appear to mitigate overfitting and achieve robust cross-dataset performance. Therefore, cross-dataset evaluation and regularized models are recommended to assess and reduce overfitting, while normalization should be used cautiously. Adding clinical data seems to have little impact on ML models’ performance. However, future work on other diseases and datasets is warranted.

Chronic stress-induced hypercortisolism causes diminished ovarian reserve (DOR), contributing to infertility and miscarriage. Androgen supplementation is an emerging therapeutic approach for DOR. The traditional Chinese herbal decoction modified Gengnianchun formula (MGNC) has shown clinical efficacy in treating DOR. This study aimed to compare the effectiveness of MGNC with that of androgens in a stress-induced DOR mouse model.

Sexually mature female C57 mice aged six weeks were randomly assigned to six groups (n = 10 per group, with 3 independent replicates per group), including the control, model, low-dose testosterone (LT), medium-dose testosterone (MT), high-dose testosterone (HT), and MGNC groups. This sample size and study design were determined based on preliminary experimental data. Chronic stress was induced in mice, except for the control group, by daily glucocorticoid injection, and the mice in the LT, MT, HT, and MGNC groups were treated at the same time with testosterone (low, medium, or high dose) or MGNC for six weeks. Body weight, estrous cycles, ovarian follicle counts, hormone profiles, including follicle-stimulating hormone (FSH), luteinizing hormone (LH), estradiol, and testosterone, and in vitro preantral follicle growth rates (via MGNC-enriched or androgen-treated serum) were assessed.

All groups presented stable body weights. MGNC ameliorated estrous cycle irregularities caused by stress, while testosterone exacerbated the abnormality. Moreover, MGNC outperformed LT in improving primordial/primary/antral follicle counts and corpus luteum formation, while MT and HT did not improve ovarian follicle reserve. LT was associated with the highest serum estradiol level, but none of the testosterone doses reduced FSH levels or the FSH/LH ratio, whereas MGNC lowered FSH and the FSH/LH ratio. Additionally, MGNC-enriched serum significantly enhanced the in vitro follicular growth rate in corticosterone-supplemented culture medium, and this effect was superior to that observed with testosterone-pretreated serum.

MGNC demonstrates superior efficacy over androgen therapy in treating chronic stress-induced DOR in mice, supporting further investigations into its clinical potential and mechanisms.

Mediastinal germ cell tumors (GCTs) are rare malignant neoplasms that occasionally develop somatic-type malignancies (SMs), such as sarcomas, carcinomas, and hematologic malignancies.

We report a unique case of a 16-year-old male patient with a mediastinal GCT that simultaneously developed two different SMs: well-differentiated angiosarcoma and acute megakaryoblastic leukemia (AML). The patient initially presented with left shoulder pain and intermittent shortness of breath. The imaging study demonstrated a 12.5 × 9.0 × 8.5 cm heterogeneous mass in the left anterior mediastinum. The mediastinal mass was resected and showed a cystic mature teratoma with somatic transformation into well-differentiated angiosarcoma and AML. A subsequent bone marrow biopsy confirmed the diagnosis of AML, and next-generation sequencing demonstrated the presence of PTEN and TP53 gene mutations in the AML. Despite aggressive chemotherapy and allogeneic stem cell transplantation, the patient died 10 months after diagnosis.

Our report demonstrates the unique capability of mediastinal GCTs to simultaneously develop two different SMs. The presence of two different SMs in mediastinal GCTs is associated with extremely aggressive behavior and a poor prognosis.