Automated diagnosis of 7 canine skin tumors using machine learning on H&E-stained whole slide images.

Microscopic evaluation of hematoxylin and eosin-stained slides is still the diagnostic gold standard for a variety of diseases, including neoplasms. Nevertheless, intra- and interrater variability are well documented among pathologists. So far, computer assistance via automated image analysis has shown potential to support pathologists in improving accuracy and reproducibility of quantitative tasks. In this proof of principle study, we describe a machine-learning-based algorithm for the automated diagnosis of 7 of the most common canine skin tumors: trichoblastoma, squamous cell carcinoma, peripheral nerve sheath tumor, melanoma, histiocytoma, mast cell tumor, and plasmacytoma. We selected, digitized, and annotated 350 hematoxylin and eosin-stained slides (50 per tumor type) to create a database divided into training, n = 245 whole-slide images (WSIs), validation (n = 35 WSIs), and test sets (n = 70 WSIs). Full annotations included the 7 tumor classes and 6 normal skin structures. The data set was used to train a convolutional neural network (CNN) for the automatic segmentation of tumor and nontumor classes. Subsequently, the detected tumor regions were classified patch-wise into 1 of the 7 tumor classes. A majority of patches-approach led to a tumor classification accuracy of the network on the slide-level of 95% (133/140 WSIs), with a patch-level precision of 85%. The same 140 WSIs were provided to 6 experienced pathologists for diagnosis, who achieved a similar slide-level accuracy of 98% (137/140 correct majority votes). Our results highlight the feasibility of artificial intelligence-based methods as a support tool in diagnostic oncologic pathology with future applications in other species and tumor types.

Pan-tumor CAnine cuTaneous Cancer Histology (CATCH) dataset.

Due to morphological similarities, the differentiation of histologic sections of cutaneous tumors into individual subtypes can be challenging. Recently, deep learning-based approaches have proven their potential for supporting pathologists in this regard. However, many of these supervised algorithms require a large amount of annotated data for robust development. We present a publicly available dataset of 350 whole slide images of seven different canine cutaneous tumors complemented by 12,424 polygon annotations for 13 histologic classes, including seven cutaneous tumor subtypes. In inter-rater experiments, we show a high consistency of the provided labels, especially for tumor annotations. We further validate the dataset by training a deep neural network for the task of tissue segmentation and tumor subtype classification. We achieve a class-averaged Jaccard coefficient of 0.7047, and 0.9044 for tumor in particular. For classification, we achieve a slide-level accuracy of 0.9857. Since canine cutaneous tumors possess various histologic homologies to human tumors the added value of this dataset is not limited to veterinary pathology but extends to more general fields of application.

Co-culture of primary human T cells with leukemia cells to measure regulatory T cell expansion.

The expansion of regulatory T cells (Tregs) is known to be mediated by cytokines including IL-10 and TGFß but has additionally been shown to depend on the interaction of the immune receptors ICOSLG and ICOS. Here, we describe a co-culture system which enables quantification of the ability of leukemia cells to induce Treg expansion through secreted cytokines and direct receptor interactions. The protocol is applicable for MHC-matched and -unmatched experiments and allows assessment of Treg expansion without using a mouse model. For complete details on the use and execution of this protocol, please refer to Külp et al. (2022).

The immune checkpoint ICOSLG is a relapse-predicting biomarker and therapeutic target in infant t(4;11) acute lymphoblastic leukemia.

The most frequent genetic aberration leading to infant ALL (iALL) is the chromosomal translocation t(4;11), generating the fusion oncogenes KMT2A:AFF1 and AFF1:KMT2A, respectively. KMT2A-r iALL displays a dismal prognosis through high relapse rates and relapse-associated mortality. Relapse occurs frequently despite ongoing chemotherapy and without the accumulation of secondary mutations. A rational explanation for the observed chemo-resistance and satisfactory treatment options remain to be elucidated. We found that elevated ICOSLG expression level at diagnosis was associated with inferior event free survival (EFS) in a cohort of 43 patients with t(4;-11) iALL and that a cohort of 18 patients with iALL at relapse displayed strongly increased ICOSLG expression. Furthermore, co-culturing t(4;11) ALL cells (ICOSLGhi) with primary T-cells resulted in the development of Tregs. This was impaired through treatment with a neutralizing ICOSLG antibody. These findings imply ICOSLG (1) as a relapse-predicting biomarker, and (2) as a therapeutic target involved in a potential immune evasion relapse-mechanism of infant t(4;11) ALL.

Inhibition of bone morphogenetic protein signaling reduces viability, growth and migratory potential of non-small cell lung carcinoma cells.

PURPOSE: BMP signaling has an oncogenic and tumor-suppressing activity in lung cancer that makes the prospective therapeutic utility of BMP signaling in lung cancer treatment complex. A more in-depth analysis of lung cancer subtypes is needed to identify BMP-related therapeutic targets. We sought to examine the influence of BMP signaling on the viability, growth and migration properties of the cell line LCLC-103H, which originates from a large cell lung carcinoma with giant cells and an extended aneuploidy. METHODS: We used BMP-4 and LDN-214117 as agonist/antagonist system for the BMP receptor type I signaling. Using flow cytometry, wound healing assay, trans-well assay and spheroid culture, we examined the influence of BMP signaling on cell viability, growth and migration. Molecular mechanisms underlying observed changes in cell migration were investigated via gene expression analysis of epithelial-mesenchymal transition (EMT) markers. RESULTS: BMP signaling inhibition resulted in LCLC-103H cell apoptosis and necrosis 72 h after LDN-214117 treatment. Cell growth and proliferation are markedly affected by BMP signaling inhibition. Chemotactic motility and migratory ability of LCLC-103H cells were clearly hampered by LDN-214117 treatment. Cell migration changes after BMP signaling inhibition were shown to be coupled with considerable down-regulation of transcription factors involved in EMT, especially Snail. CONCLUSIONS: BMP signaling inhibition in LCLC-103H cells leads to reduced growth and proliferation, hindered migration and accelerated cell death. The findings contribute to the pool of evidence on BMP signaling in lung cancer with a possibility of introducing BMP signaling inhibition as a novel therapeutic approach for the disease.