Histiocytic sarcoma (HS) is an extremely rare but aggressive hematopoietic malignancy, and the prognosis has been reported to be rather unfavorable with a median overall survival of merely 6 months. We presented a 58-year-old female patient complaining of abdominal pain and fever, who was admitted to our institution in September 2021. Fluorine-18-fluorodeoxyglucose (FDG) positron emission tomography–computed tomography (PET/CT) scan showed enlargement of generalized multiple lymph nodes. Subsequently, laparoscopic retroperitoneal lesion biopsy and bone marrow aspiration were performed. The pathological findings indicated the diagnosis of HS concurrent with follicular lymphoma. The immunohistochemistry (IHC) staining of the tumor lesion revealed a high expression of CD38 and PD-L1 proteins. Furthermore, KRAS gene mutation was identified by means of next-generation sequencing. The patient exhibited poor treatment response to both first- and second-line cytotoxic chemotherapies. Therefore, she underwent six cycles of Daratumumab (anti-CD38 monoclonal antibody), Pazopanib (multi-target receptor tyrosine kinases inhibitor) combined with third-line chemotherapy, followed by involved-site radiotherapy and maintenance therapy with the PD-1 inhibitor Tislelizumab. Long-term partial remission was finally achieved after multi-modality treatment. Duration of remission and overall survival reached 22 and 32 months, respectively. Our case indicated that immuno-targeted treatment coupled with chemotherapy and radiotherapy might constitute a potential therapeutic option for HS.
Introduction: Programmed cell death ligand-1 (PD-L1) combined positive score (CPS) evaluation plays a pivotal role in predicting immunotherapy efficacy for gastric cancer. However, manual CPS assessment suffers from significant inter-observer variability among pathologists, leading to clinical inconsistencies. To address this limitation, we developed a deep learning-based artificial intelligence (AI) system that automates PD-L1 CPS quantification for patients with gastric cancer (GC) using whole slide images (WSIs).Methods: We developed a deep learning-based artificial intelligence (AI) system that automates PD-L1 CPS quantification for patients with gastric cancer (GC) using whole slide images (WSIs). Our pipeline firstly employs a dual-network architecture for tumor region detection: MobileNet for patch-level classification and U-Net for pixel-level segmentation. Followed by a YOLO-based cell detection model to compute PD-L1 expression on different cells for CPS calculation. A total of 308 GC WSIs were included, including 210 in the internal cohort and 98 in the external cohort. Within the internal cohort, 100 WSIs were utilized for the model development, while the remaining 110 WSIs served as an internal testing set for comparative analysis between AI-derived CPS values and pathologist-derived reference standards.Results: The AI-derived CPS demonstrated strong concordance with expert pathologists’ consensus in internal cohort (Cohen’s kappa = 0.782). Furthermore, the AI-based CPS prediction pipeline was evaluated for its performance in the external cohort, and showed robust performance (Cohen’s kappa = 0.737).Discussion: Our system provides a standardized decision-support tool for immunotherapy stratification in GC management, demonstrating potential to improve CPS assessment reproducibility.