Renji Hospital

Imbalanced class distributions among different organs pose significant challenges in real-world semi-supervised multi-organ segmentation. Integrating anatomical priors offers a promising research direction to mitigate these imbalances. In this paper, we explore the capabilities of Multimodal Large Language Models (MLLM) to extract robust, generic textual anatomical insights serving as prior knowledge for segmentation model. Specifically, we employ GPT-4o to generate detailed textual descriptions of anatomical priors-including both inter-organ relative positional relationships and organ shape characteristics. These priors generated only once for the whole training and testing are then seamlessly integrated into the segmentation model as parameters within the segmentation head. Furthermore, we align the textual priors with visual features using contrastive learning. The inter-organ positional priors guide the model in localizing smaller organs relative to larger ones, while the organ shape priors help ensure that the learned morphological structures are more anatomically plausible. Extensive experiments demonstrate that our method significantly outperforms some state-of-the-art approaches. The source code is available at: https://github.com/Lunn88/TAK-Semi.

Malignant Triton Tumor (MTT) is a rare, highly aggressive peripheral nerve sheath tumor characterized by rhabdomyoblastic differentiation, seldom occurring primarily in the thoracic cavity. This report presents a rare case of MTT initially manifesting as chest pain, and reviews relevant literature to summarize its clinical features and therapeutic strategies. A 23-year-old woman experienced chest pain for ten days and her condition worsened within 3 days. The patient had a known history of neurofibromatosis type 1 (NF1) and exhibited classic café-au-lait spots on physical examination. Diagnosis was confirmed via chest CTA, PET-CT, bronchoscopy, ultrasound, and histopathology. After surgical resection, the tumor recurred rapidly, prompting multiple treatments including chemotherapy, targeted therapy, and combination immunotherapy. Pathology after right lower lobectomy and chest wall resection confirmed MTT (S-100+/Desmin+/Myogenin+). Recurrence in the mediastinum was detected two months postoperatively. Disease stabilization was achieved using cadonilimab, apatinib, and ifosfamide/etoposide. Notably, the first application of single-cell RNA sequencing (scRNA-seq) in MTT, combined with the CopyKAT algorithm, distinguished malignant from non-malignant cells and revealed a heterogeneous tumor microenvironment composed of distinct functional cell populations, highlighting the tumor's high degree of heterogeneity. This case underscores the importance of considering MTT in young NF1 patients presenting with intrathoracic masses. An individualized, multimodal treatment approach may extend survival. scRNA-seq provides valuable insights into the molecular landscape of MTT and hold promise for guiding precision therapy in the future.

Dysregulation of the CARD11-BCL10-MALT1 (CBM) complex is associated with a group of inborn errors of immunity termed “CBM-opathies,” which encompass a spectrum of clinical manifestations including combined immunodeficiency, autoimmune inflammation, atopic disorders, and lymphoproliferation. In this study, we identified novel compound heterozygous variants in the CBM complex in a patient with a family history of immune dysfunction. The patient inherited the variants CARD11 p.K215N and MALT1 p.K543R/p.M732T from asymptomatic carrier parents. Phenotypically, the patient exhibited a developmental arrest of B lymphocytes at the transitional/naïve B cell stage, accompanied by activation of virus-response pathways. Impaired development of T follicular helper cells was linked to defective germinal center formation and agammaglobulinemia. Furthermore, the patient showed expansion of T peripheral helper cells and a deficiency in regulatory T cells, both associated with autoimmunity and colitis. In vitro studies confirmed an imbalance in Tph/Tfh cell differentiation. Single-cell RNA sequencing further revealed a deficiency in B cell development and an enriched population of pro-inflammatory CD3dimCD4−CD8−CD247+ T cells, functionally enriched in the MAPK signaling pathway. Mechanistically, the MALT1 K543R and M732T variants attenuated MALT1’s enzymatic activity and compromised its protein stability, while the CARD11K215N variant disrupted CARD11-mediated promotion of BCL10 filament formation. We demonstrated that these three variants act synergistically to impair NF-κB activation. Specifically, CARD11K215N cooperates with the co-pathogenic MALT1M732T and the modifier variant MALT1K543R to destabilize the functional integrity of the CBM complex, thereby driving the patient’s phenotype. In summary, our study provides new insights into the pathogenesis of autoimmune inflammatory disorders within the spectrum of CBM-opathies and reveals a potential role for the CBM complex in regulating the balance between T peripheral helper and T follicular helper cells. Identified three novel variants in the CBM complex in a patient with a family history of immune dysfunction. CARD11K215N/MALT1K543R/M732T variants impaired B cell development and CD4+T cell differentiation, especially T peripheral helper and T follicular helper cells, by destabilizing the functional integrity of CBM complex.

Large-language models (LLMs) have demonstrated increasing potential in healthcare, including applications in clinical information retrieval, medical reasoning, and decision support. Autoimmune liver diseases (AILDs), including autoimmune hepatitis, primary biliary cholangitis, primary sclerosing cholangitis, and IgG4-related liver disease, are rare and heterogeneous conditions that often present with nonspecific features and require specialized expertise for accurate diagnosis and management. In resource-limited settings, limited access to hepatology subspecialists may contribute to delayed diagnoses and suboptimal care. Although reasoning-oriented LLMs are designed to support structured clinical inference, their performance on AILD-related tasks has not been systematically evaluated. To evaluate and compare the accuracy, safety, readability, comprehensiveness, and diagnostic performance of six large language models for addressing AILD-related clinical questions and real-world cases. We developed 26 clinically relevant questions spanning the key domains of AILDs, including pathogenesis, risk factors, clinical presentation, diagnosis, treatment, and prognosis. Six publicly available LLMs (o1-preview, Claude-3.5-Sonnet, GPT-4o, GPT-4o-mini, GPT-3.5-Turbo, and LLaMA-3.1-405B) generated responses that were independently evaluated by 11 board-certified hepatologists with more than 10years of specialty experience. Accuracy, safety, readability, comprehensiveness, and clinical helpfulness were assessed by using predefined rating scales. Diagnostic performance was further evaluated in three models (o1-preview, GPT-4o, and GPT-3.5-Turbo) using 21 confirmed real-world AILD cases, with accuracy determined by comparison with gold-standard clinical diagnoses. Among the six models, the o1-preview demonstrated the highest overall performance. It achieved the greatest proportion of accurate responses (78.3%), the highest mean safety score, and the most favorable readability profile, despite generating the longest responses. Comprehensiveness and helpfulness ratings were also highest for o1-preview. In diagnostic testing, the o1-preview achieved an overall accuracy of 81.0%, outperforming GPT-4o and GPT-3.5-Turbo, particularly in single-entity AILD diagnoses. All models demonstrated reduced diagnostic accuracy for autoimmune hepatitis-primary biliary cholangitis overlap syndrome. Reasoning-based LLMs, particularly o1-preview, demonstrate potential value in supporting clinical information delivery and diagnostic reasoning for AILDs, especially in resource-limited settings. However, LLM outputs should serve as adjunctive support rather than substitutes for specialist judgment. Further domain-specific refinement and multicenter validation are necessary to ensure safe and effective integration into hepatology and clinical pharmacy practices.

The authors declare no conflicts of interest.

Both the atherogenic index of plasma (AIP) and obesity have been implicated in cardiovascular disease (CVD) risk. This study aimed to investigate the associations of AIP and AIP-obesity indices (AIP-WC, AIP-BMI, AIP-WHtR, AIP-WWI, and AIP-BRI) with CVD and mortality. This study enrolled 21,944 individuals from the National Health and Nutrition Examination Survey (NHANES) fielded during the years 1999–2018. Cox and logistic regression modeling, along with restricted cubic splines (RCS), served to assess the associations of AIP and AIP-obesity indices with all-cause mortality, CVD mortality, and CVD prevalence. The incremental predictive value of AIP and AIP-obesity indices was further assessed. Throughout an average follow-up period of 116 months, 3,326 all-cause deaths were documented, among which 1,042 were CVD deaths. AIP, AIP-WC, AIP-BMI, AIP-WHtR, AIP-WWI, and AIP-BRI were significantly associated with all-cause mortality, CVD mortality, and CVD prevalence. In the fully adjusted model, relative to the lowest-tertile participants, the highest-tertile participants of these indices had an elevated risk of 17.7–20.3% for all-cause mortality, an elevated risk of 30.3–33.8% for CVD mortality, and a 103.7–122.2% increase in the odds of CVD. RCS analyses indicated non-linear associations of AIP, AIP-WC, AIP-BMI, AIP-WHtR, AIP-WWI, and AIP-BRI with all-cause mortality and CVD prevalence, whereas linear associations were observed with CVD mortality. Incremental predictive value analyses showed that compared with AIP alone, AIP-WC, AIP-BMI, AIP-WHtR, AIP-WWI, and AIP-BRI provided superior risk reclassification and discrimination across all outcomes. AIP and AIP-obesity indices were significantly associated with CVD and mortality risk. AIP combined with obesity-related indices, particularly BRI, provides additional value for the stratification of CVD and mortality risk.