Humans have long sought to comprehend the etiology and mechanisms of disease, as well as the pathological alterations underlying clinical manifestations at the onset of illness. In ancient Chinese myths and legends, disease was commonly attributed to supernatural forces such as demons and malevolent spirits. With a better understanding of the natural world, these early superstitions gave way to explanations based on natural causation. Around the time that Hippocrates advanced the concept that natural factors were important in health and disease, Chinese medical scholars developed the theoretical framework of Yin-Yang and the Five Elements to explain the origins and progression of disease. Over time, theories of disease evolved, due to transformations in religion, philosophy, science, and technology. The study of disease transitioned from rudimentary and speculative accounts to systematic investigations encompassing organs, tissues, and cells. When missionaries introduced Western medicine to China, it was gradually gained acceptance among part of the medical community. In China today, both Western medicine and traditional Chinese medicine are actively supported and seen as complementary approaches to health care. The rich history of Chinese pathology is not well documented in Western medical literature. In this review, we explain the key theories, events, practitioners, and scholars in Chinese pathology, covering ancient China (before 221 bce) to the end of Imperial China (1912 ce).
Purpose Accurate prediction of lymph node metastasis (LNM) remains challenging in early gastric cancer (EGC), particularly when determining the need for additional surgery after endoscopic resection. We developed and evaluated a deep learning framework using multiple pathology foundation models for LNM prediction and explored histological features associated with model-predicted metastatic risk. Materials and Methods Foundation model ensembles were developed using whole-slide images from surgical EGC cases and evaluated in internal and external surgical cohorts. The model was further tested in an endoscopic submucosal dissection cohort in which all patients subsequently underwent additional gastrectomy with lymph node dissection. Performance was assessed by area under the curve (AUC) and compared with scratch-trained baseline models. At a fixed sensitivity of 1.00, the potential reduction in lymph node dissection among histologically node-negative patients was evaluated. Model-guided histological analysis was performed to interpret features associated with predicted metastatic risk. Results The final ensemble model achieved an AUC of 0.879 (95% CI, 0.799-0.944) in the internal and 0.801 (95% CI, 0.697-0.891) in the external cohort, representing improvements of 0.040 and 0.066, respectively, over the baseline model. In the endoscopic submucosal dissection cohort, the model achieved an AUC of 0.766 (95% CI, 0.635-0.882), representing an improvement of 0.189 over the baseline model. At a fixed sensitivity of 1.00, the model suggested avoiding lymph node dissection in approximately 5% of histologically node-negative patients. Model-guided histological analysis showed associations consistent with established pathological features, with exploratory observations including context-dependent associations between inflammatory infiltration and tumor differentiation, differences in predicted risk between well- and moderately differentiated tumors, and lower predicted risk in poorly cohesive carcinoma. Conclusions A multiple foundation model ensemble showed improved LNM risk prediction in EGC and provided interpretable histological insights supporting clinical understanding and risk stratification.
Purpose Consensus molecular classes of urothelial carcinoma (UC) include basal/squamous (Ba/Sq), luminal papillary (LumP), luminal nonspecified (LumNS), luminal unstable (LumU), stroma-rich, and neuroendocrine (NE)-like. We aimed to determine the consensus molecular classifications of micropapillary (MP), plasmacytoid (PC), and sarcomatoid (SM) subtypes of bladder UC and characterize their spatial transcriptomic profiles. Materials and Methods A tissue microarray of 17 FFPE cores (5 MP, 6 PC, 6 SM) was profiled for expression of 18,676 mRNAs using the Bruker GeoMx Digital Spatial Profiler. 38 areas of interest (AOIs) were selected based on H&E and fluorescence markers. Each AOI was assigned to a consensus molecular class using a published approach, and pathway analysis was performed using Reactome gene sets. Results Most MP and PC demonstrated luminal molecular classes, whereas SM was predominantly Ba/Sq. Adjacent mixed histological components within the same tumor frequently corresponded to distinct consensus molecular classes. Ba/Sq tumors upregulated invasion- and extracellular matrix (ECM) remodeling-associated genes, whereas luminal classes demonstrated distinct differentiation-associated signatures. MP demonstrated a predominantly luminal epithelial transcriptional program characterized by increased expression of luminal/urothelial differentiation markers (KRT20, UPK1A, PSCA, FOXA1), secretory-associated genes (LCN2, MUC1, MUC20, SPINK1), and relative ERBB2 enrichment, with suppression of basal/squamous-associated genes (KRT5, KRT6A, KRT14, TP63). In contrast, SM demonstrated a predominantly mesenchymal/stromal transcriptional program characterized by enrichment of mesenchymal/ECM-associated genes (FN1, VIM, COL6A1, TGFB1), and reduced expression of luminal/epithelial and adhesion-associated genes (TACSTD2, PSCA, NECTIN4, CDH1). PC appeared to demonstrate an intermediate phenotype between MP and SM, with partial preservation of epithelial differentiation and secretory programs, reduced adhesion programs, and moderate stromal-associated features. The therapeutic target NECTIN4 was relatively enriched in MP and PC compared with SM, whereas FGFR3 was relatively downregulated in MP and PC compared with CIUC. Pathway analysis demonstrated relative enrichment of IGF/IGFBP signaling across MP, PC, and SM, with MP enriched for protein secretory programs, PC for RTK/MAPK and WNT/β-catenin signaling, and SM for ECM remodeling. Conclusions Adjacent histological components frequently corresponded to different consensus molecular classes in UC. MP, PC, and SM demonstrated distinct transcriptomic profiles, with PC exhibiting an intermediate phenotype between MP and SM.
