Spindle cell/mesenchymal tumors of the kidney are rare. The diagnosis is supported mainly by the application of ancillary techniques such as immunohistochemistry (IH) and in-situ hybridization (FISH). An accurate diagnosis is essential because early management by complete resection and adjuvant chemotherapy improves the prognosis dramatically. Synovial sarcoma and primitive neuroectodermal tumor/Ewing sarcoma are infrequent malignancies which usually present in soft tissues but rarely in the kidney. The challenge for the pathologists is to histologically differentiate between different types of sarcomas such as PNET/Ewing's sarcoma, sarcomatous dedifferentiated renal cell carcinoma, metastasis, non-Hodgkin's lymphoma, nephroblastoma and angiomyolipoma.We report from our experience six exemplary rare cases that presented in the kidney as spindle/round cell tumors.We have arrived at the accurate diagnosis after performing a large panel of IH and FISH.In summary we advise an immunohistochemical panel for round/spindle cell tumors of the kidney and for unclear cases we advise to add (FISH) to get the correct diagnosis, as they are completely different regarding surgical approach and post-operative adjuvant therapy.
Different convolutional neural network (CNN) models have been tested for their application in histological image analyses. However, these models are prone to overfitting due to their large parameter capacity, requiring more data or valuable computational resources for model training. Given these limitations, we introduced a novel architecture (termed PlexusNet). We utilized 310 Hematoxylin and Eosin stained (H&E) annotated histological images of prostate cancer cases from TCGA-PRAD and Stanford University and 398 H&E whole slides images from the Camelyon 2016 challenge. PlexusNet-architecture -derived models were compared to models derived from several existing "state of the art" architectures. We measured discrimination accuracy, calibration, and clinical utility. An ablation study was conducted to study the effect of each component of PlexusNet on model performance. A well-fitted PlexusNet-based model delivered comparable classification performance (AUC: 0.963) in distinguishing prostate cancer from healthy tissues, although it was at least 23 times smaller, had a better model calibration and clinical utility than the comparison models. A separate smaller PlexusNet model accurately detected slides with breast cancer metastases (AUC: 0.978); it helped reduce the slide number to examine by 43.8% without consequences, although its parameter capacity was 200 times smaller than ResNet18. We found that the partitioning of the development set influences the model calibration for all models. However, with PlexusNet architecture, we could achieve comparable well-calibrated models trained on different partitions. In conclusion, PlexusNet represents a novel model architecture for histological image analysis that achieves classification performance comparable to other models while providing orders-of-magnitude parameter reduction.