Datasets from 849 dogs ≤ 3 years of age were analysed retrospectively, considering breed, sex, anatomic sites, multiplicity, tumour diameter, grading, and immunohistochemistry by using IBM® SPSS Statistics (version 29.0.2.0). Danish–Swedish farm dogs (OR: 8.11, 95% CI 3.33–19.77, p < 0.001) and English Setters (OR: 6.68, 95% CI 3.20–13.97, p < 0.001) showed the highest odds ratios (ORs) compared to crossbreeds. German Shepherd Dogs showed lower ORs (OR: 0.05, 95% CI 0.007–0.364, p < 0.01). The most common sites affected in the young dogs were the trunk (39.2%), hind limbs (24.9%), head (10.1%), fore limbs (7.4%), and pinnae (5%). Patnaik grading revealed 330 grade I (35.7%), 580 grade II (62.8%), and 14 grade III MCTs (1.5%). Kiupel “low grade” was found in 903 MCTs (97%), and 21 MCTs were “high grade” (3%). The prevalence of grade I MCTs was higher in Boxers (48.0%, p < 0.05) and Pugs (67.9%, p < 0.01). Pinnal MCTs were mostly grade I (58.3%, p < 0.01). In 6% (n = 51), multilocalised MCTs were found, and Golden Retrievers (12.5%, p = 0.02) and Pugs (21.4%, p = 0.005) were more affected than crossbreeds. Out of 87 dogs, 20 showed a Ki-67 count above the cut-off as well as KIT staining pattern II. The findings warrant heightened vigilance in evaluating cutaneous lesions in young dogs and highlight similarities and differences compared to the general dog population. Analysis of survival data is required to evaluate how the findings translate into a clinical setting.

Canine mammary cancers (CMCs) are one of the most prevalent types of neoplasm in dogs, are frequently malignant, and display high tumour heterogeneity, making evaluating prognosis and predicting successful treatment outcomes difficult. In a previous pilot study, overexpression of the Wnt pathway-associated protein SFRP1 was found to correlate with negative metastasis status in CMCs at both mRNA and protein levels. To establish SFRP1 as a potential biomarker for CMC progression, additional verification of these results in an independent dataset is required, as well as an investigation as to whether SFRP1 expression in CMCs is associated with altered Wnt- or RANKL signalling pathways. In an independent verification cohort of 122 cases of archival CMC FFPE material, expression of SFRP1 was assessed by RT-qPCR and immunohistochemistry. The same tumours were further assessed for RANKL, phosphoROCK2, and NFkB-p65 protein expression. Our data verified that SFRP1 mRNA (p = 0.025) was negatively associated with metastasis status; however, differences in protein expression did not reach statistical significance (p = 0.139). Neither did SFRP1 significantly correlate with expression of any of the other proteins tested. Instead, a strong association was found for RANKL positivity with increased metastasis status (p < 0.001). Co-expression of SFRP1 significantly lowered the higher risk of metastatic spread when compared to RANKL(pos)/SFRP1(neg) CMCs (p = 0.033). Noticeably, all vascular-invasive cell clusters observed in tissue section vessels stained positive for RANKL. Our study identified RANKL expression as a strong marker for cancer progression with a strong link to vascular-invasive cells. However, SFRP1 expression may potentially suppress the pro-metastatic nature of RANKL(pos) CMCs.

Oncolytic coxsackievirus B3 (oCVB3) strain PD-H has shown potent oncolytic efficacy and a remarkable safety profile in the treatment of colorectal cancer in vivo after intratumoral (i.t.) injection. In this study, we investigated the safety and efficiency of PD-H following intravenous (i.v.) virus administration. When injected i.v. into Balb/C mice bearing subcutaneous Colon-26 tumors, PD-H led to slightly reduced tumor progression and a significant increase in animal survival, but it also caused multi-organ infection and tissue damage. To improve the safety profile of PD-H, we inserted microRNA target sites (miR-TS) of the heart-specific miR-1, pancreas-specific miR-375, liver-specific miR-122, and brain-specific miR-124 or the tumor-suppressor miR-145 into the genome of PD-H and generated the viruses PD-622TS and PD-145TS. Both viruses replicated similarly and induced cytotoxicity comparable to that of PD-H in the colorectal carcinoma cell lines Colon-26 and CT-26Luc. Their replication was inhibited in HEK293T cells transiently transfected with the cognate microRNAs. In vivo, i.v. administration of PD-145TS and PD-622TS to healthy Balb/C mouse resulted in significantly lower viral titers in the organs of mice and led to significantly less-intense pathological alterations compared to PD-H. PD-622TS injected i.v. into Balb/C mice with CT-26Luc-induced peritoneal carcinomatosis did not induce off-target alterations in normal organs, but it failed to induce a therapeutic effect. These data indicate that PD-H or microRNA-regulated PD derivatives exhibit only limited therapeutic efficacy following i.v. injection in colorectal tumor-bearing mice. However, the newly engineered microRNA-regulated PD-H variants demonstrate improved safety profiles.

Molecular genetic approaches have become indispensable in veterinary medicine for early tumour detection, confirming histomorphological diagnoses, and non-invasive diagnosis using urine samples. In canine urothelial carcinoma (UC), molecular testing includes BRAFV595E mutation analysis and the identification of specific copy number alterations (CNAs). CNAs affect multiple genes simultaneously, leading to their overexpression or underexpression. Recurrent gene gains on canine chromosomes (CFA)13 and CFA36 and losses on CFA19 are highly prevalent in canine UCs. This study evaluates CNA test performance and the influence of tumour purity on CNA test results using 76 histologically confirmed UC tissue samples. The tumour region of each histopathological section was measured, and its proportion relative to the total tissue area was calculated. In total, 58/76 cases (76.3%) were CNA-positive, defined by a ratio greater than 1.23 for both CFA13/19 and CFA36/19. A total of 14/18 negative cases had a ratio > 1.23 for CFA13/19 or CFA36/19. The sensitivity of this CNA assay depends on the applied thresholds, with a tumour-to-total tissue ratio of ≥20% yielding positive CNA results in 84% of cases, rising to 89% at ≥40%. These findings indicate that <20% tumour content and a moderate to high proportion of copy number-neutral cells may be associated with reduced CNA detectability.

Mechanical ventilation (MV), a common life-saving intervention in critical care medicine, can itself cause pulmonary damage. Intravascular hemolysis is common in sepsis and acute respiratory distress syndrome (ARDS). Although inflammation, infection, or atelectasis can enhance ventilator-induced lung injury (VILI), data are scarce on the interaction between hemolysis and VILI. Therefore, mice were ventilated with either lung-protective MV (LPV) or injurious MV (HPV), and selected mice received an intravascular infusion of hemolyzed murine red blood cells containing cell-free hemoglobin (CFH). Lung edema quantified by wet-to-dry weight ratio did not differ between mice receiving LPV or LPV + CFH but was significantly higher in mice receiving HPV + CFH compared with HPV alone. Pulmonary expression of proinflammatory cytokines such as interleukin-6 (IL-6) and tumor-necrosis factor alpha (TNF-α) did not differ between mice with LPV + CFH or LPV but increased significantly in mice receiving HPV. Norepinephrine (NE), used to simulate CFH-associated hypertension but not CFH itself, further increased IL-6 gene expression and concentration in bronchoalveolar lavage fluid in mice with HPV. However, mice with HPV + CFH showed significantly impaired lung mechanics compared with mice with HPV or HPV + NE. Analyzing 437 critically ill patients with severe ARDS and therapy with venovenous extracorporeal membrane oxygenation (ECMO) confirmed that increased plasma concentrations of CFH were associated with a reduced pulmonary compliance. The findings suggest that mice subjected to HPV + CFH show increased impairment of lung mechanics that is associated with lung edema but cannot be fully explained by the proinflammatory and proedematous effects of CFH-induced hypertension. Associated with reduced pulmonary compliance also in humans, increased CFH plasma concentrations might be a future therapeutical target.NEW & NOTEWORTHY Mechanical ventilation is frequently used in critically ill patients and can itself cause pulmonary damage. Furthermore, hemolysis is common in sepsis and ARDS. Mice receiving mechanical ventilation and an infusion of cell-free hemoglobin (CFH) demonstrated increased impairment of lung mechanics associated with lung edema not fully explainable by the proinflammatory and proedematous effects of CFH. Furthermore, increased plasma concentrations of CFH correlated with reduced pulmonary compliance in patients with severe ARDS and therapy with ECMO.

Automated mitosis detection is a well-established task in computational pathology. While previous benchmarks focused on scanner-induced domain shift, clinical "real-world" application requires models to be robust across the vast variance to be expected in the histological landscape. The MItosis DOmain Generalization (MIDOG) 2025 challenge was designed to evaluate algorithmic performance across unprecedented biological and contextual diversity. We curated a test dataset of 365 cases, encompassing 12 distinct human, canine and feline tumor types, digitized across multiple scanning platforms. Moving beyond hand-selected hotspots, the challenge required detection also in random tissue areas (representative of the whole slide detection situation) and challenging areas (areas rich in hard negatives). In the second track, we introduced the classification of atypical mitotic figures (AMFs). There were 18 teams submitting to the detection track, with F1 scores ranging up to 0.740. In the AMF detection track, we had 21 submissions with balanced accuracy values up to 0.908. Our analysis reveals that while most models perform reliably in traditional hotspots, significant performance degradation occurs in challenging ROIs, where false positive rates tripled. Furthermore, performance varied significantly across the 12 tumor types, highlighting "blind spots" in current state-of-the-art architectures when encountering rare or highly pleomorphic malignancies. Moreover, we evaluated the effectiveness of ensembling and found a mean increases of 1.5 and 1.3 percentage points in F1 score and balanced accuracy, respectively. In contrast, TTA showed no relevant improvement. MIDOG 2025 demonstrates that "in the wild" mitosis detection remains a significant hurdle. The transition from hotspot-only evaluation to a multi-contextual framework provides a more realistic proxy for clinical reliability.

Pathology foundation models (FMs) are models trained on vast amounts of typically unlabeled data and have been shown to yield regularized latent spaces that can be used effectively in downstream classification tasks. This is also true for the classification of mitotic figures vs. other cells. However, it is so far unclear if the latent space of current FMs provides features that are discriminant and spatially suitably resolved to also serve as a backbone for dense object detection paradigms. In this work, we investigate this question for common current pathology FMs (UNI, UNI2-h, Virchow, Virchow2, H-optimus-0, H-optimus-1) and compare their performance against a fully end-to-end trained baseline based on a ResNet50 architecture. We combine FM backbones with representatives of single stage, dual stage and self-attention-based detectors (RetinaNet, Faster R-CNN, Deformable DETR respectively) on the multi-domain MIDOG++ dataset, and on the TUPAC16 dataset as an out-of-domain case. We show that the H-optimus-0 and Virchow models yielded competitive performance, indicating that the latent spaces of current FMs, all trained on image-level self-supervision, are suitable for direct mitotic figure detection and may be slightly more robust on our out-of-domain test case. All code is made available publicly at https://github.com/DeepMicroscopy/FM4MFdet.

Automated measurements of anisokaryosis in canine cutaneous mast cell tumors (ccMCTs) have been shown to be predictive of survival, but questions remain regarding the intratumoral distribution of anisokaryosis. Whole-slide images of 96 ccMCTs were analyzed with a deep learning-based segmentation algorithm to quantify anisokaryosis using the standard deviation (SD) of the nuclear area. In 35/96 cases, >5% of the non-overlapping 256 × 256 µm2 regions were hotspots (SD ≥11.5 µm2). Regions selected by 7 pathologists within these 35 cases matched hotspots in 32% of the instances. Outcome analysis (tumor-related death) based on single tumor regions yielded an area under the curve (AUC) of 0.901 for pathologist-selected hotspots, falling between random region selection (AUC: 0.862) and 90th-percentile targeted selection (AUC: 0.956). Whole-slide analysis of the hotspot proportion predicted survival with an AUC of 0.956, with 20% of hotspots as a prognostically meaningful threshold. Whereas pathologists-selected tumor regions are prognostically meaningful for nuclear morphometry, whole-slide analysis may provide additional prognostic information.

Oncolytic viruses (OVs) represent an emerging class of cancer immunotherapeutics. Arming OVs with immunomodulatory transgenes is a promising strategy to reshape the tumor microenvironment (TME) and enhance therapeutic efficacy. We developed the oncolytic Coxsackievirus B3 (CVB3) strain PD-H, which has previously shown antitumor activity in colorectal and pancreatic cancer models. Here, we evaluated the capacity of PD-H to express transgenes and investigated whether its antitumor efficacy in colorectal cancer can be enhanced by insertion of the interleukin-2 (IL-2) mimetic Neoleukin-2/15 (Neo-2/15). Transgene insertion was best tolerated at the VP1-2A junction of the PD-H polyprotein, with viral replication and cytotoxicity inversely correlating with insert size. Genetic stability depended on the host cell line, transgene sequence and transgene length, with inserts up to 350 bp remaining stable for at least 10 viral passages. PD-H-derived Neo-2/15 was biologically active and induced proliferation of human CD4+ and CD8+ T cells in vitro. In vivo, intratumoral administration of PD-Neo-2/15 reduced the growth of subcutaneous Colon-26 tumors more effectively than PD-H and was associated with modulation of the TME, including an increased proportion of CD8+ T cells. Taken together, we demonstrate that arming PD-H with Neo-2/15 is a promising strategy to further enhance its anticancer efficacy.

Accurate detection and classification of mitotic figures in histopathology remains difficult due to their infrequent occurrence, morphological variability, and strong domain shifts across scanners, species, and tumor types. The Mitosis Domain Generalization (MIDOG) Challenge 2025 addresses these limitations through two complementary tasks: robust mitotic figure detection across different tumor regions (Track 1) and atypical vs. normal mitotic figure classification (Track 2). In this paper, we provide the official baseline methods for both tracks: a Fully Convolutional One-Stage (FCOS)-based detector for Track 1 and an EfficientNet-V2 classifier for Track 2. Both tasks use the multi-center, multi-species MIDOG++ training dataset and are evaluated on a diverse test set containing 12 tumor types, including domains absent from training, and three region-of-interest categories (hotspot, random, potentially challenging). The detection baseline achieved an overall F1 score of 0.6883 on the final test set, with strong performance in hotspot regions (F1 = 0.7472) but reduced robustness in random (F1 = 0.6500) and challenging (F1 = 0.4161) region of interests (RoIs). The classification baseline reached a balanced accuracy of 0.8274 and ROC AUC of 0.9067. These baselines offer a reproducible reference for the participants in the MIDOG 2025 challenge.

This retrospective study provides an extensive evaluation of feline squamous cell carcinoma (SCC), comparing oral and cutaneous SCC, as well as different oral/cutaneous sites, with respect to sex, age, breed, and coat length. It was based on 4300 SCCs submitted to LABOKLIN GmbH & Co. KG from 2017 to 2023. No sex predisposition was identified. Affected cats were predominantly older (median age 13 years). SCC risk increased with age, although cats of very young age were also affected. Breed predispositions were not found. However, compared to non-pedigree cats, Persians, Norwegian Forest Cats, British Shorthairs (BSH), Chartreux, and Siamese cats showed a reduced risk, as did longhaired cats. The predominant sites were the oral cavity (41.0%, 1762/4300) and skin (35.8%, 1540/4300). Maine Coons, BSHs, Persians, and Norwegian Forest Cats, as well as longhaired cats, developed oral SCC more frequently. Intraoral, gingival (36.2%, 637/1762) and lingual (19.0%, 334/1762) SCC predominated. Common cutaneous sites included the pinnae (35.0%, 539/1540), unspecified head/neck regions (8.8%, 135/1540), and the nose (8.7%, 134/1540). Maine Coons and BSHs showed less auricular SCC; Sphynx had more on the trunk. These findings emphasise SCC as a crucial differential diagnosis for oral and cutaneous lesions, even in young cats.

Whereas reproducibility of studies is a prerequisite for trustworthy deep learning (DL) in veterinary histopathology and microscopy, the actual degree of methodological transparency that exists in the literature remains uncertain. We performed a Preferred Reporting Items for Systematic Reviews and Meta-Analyses-guided systematic review to quantify the degree to which supervised DL and supervised machine learning studies report reproducibility-critical details. Using a veterinary-journal-restricted Boolean search executed in PubMed and Scopus, we screened 180 unique records and included 50 primary research articles for full-text analysis. Based on a recently published guideline for the development of DL models in veterinary pathology, we extracted information for each study across 5 dimensions: (1) study and task characterization, (2) data transparency, (3) experimental design and data-leakage control, (4) model and training details, and (5) performance evaluation and reporting. Among the included studies, private data sets predominated, with 90% of studies relying on private data. Sharing of code was uncommon (3%). Key training details such as augmentation and hyperparameters were often incompletely reported; augmentation was not reported in 56% of studies, and key hyperparameters were absent in 40% of studies. It was often not clear whether patient-level stratification (necessary to avoid data leakage) was performed. In summary, these results highlight major deficits in the reporting of details and experimental design necessary for reproducing DL results in veterinary histopathology. This review provides a practical baseline and reporting roadmap to support more transparent and reproducible research in veterinary computational pathology.

Zusammenfassung Ein 5 Jahre altes Zwergkaninchen (Oryctolagus cuniculus) mit einem Körpergewicht von 2,46 kg wurde aufgrund akuter Inappetenz, Apathie und Diarrhoe vorgestellt. Das Tier lebte mit einem Partnertier in einem Außengehege mit natürlichem Bodengrund und einem unterirdischen Tunnelsystem. Die klinische Allgemeinuntersuchung ergab eine hochgradige Apathie, Hypothermie (36,3°C) sowie eine verstärkte abdominale Atmung. Labordiagnostisch wurden eine nicht-regenerative Anämie sowie eine heterophile und monozytäre Leukozytose festgestellt. Röntgenologisch waren eine vergrößerte Leber sowie ein peritonealer Detailverlust erkennbar. Als Hauptbefunde konnten im abdominalen Ultraschall neben einer Vergrößerung von Leber und Milz mehrere hypoechogene Rundschatten in der Leber sowie deutlich vergrößerte mesenteriale Lymphknoten und freie Flüssigkeit nachgewiesen werden. Zudem waren mehrere Umfangsvermehrungen in den Metren und ein fokaler Niereninfarkt im Bereich des kraniodorsalen rechten Nierenkortex auffällig. In der Computertomografie wurden neben den bereits im Ultraschall festgestellten Veränderungen generalisierte Verdickungen der Dickdarmwände nachgewiesen. Aufgrund des schlechten Allgemeinbefindens und der Befunde wurde das Tier auf Wunsch des Besitzers euthanasiert und pathologisch untersucht. Hauptbefunde waren neben einer Lymphadenomegalie multifokaler Darmlymphknoten multiple nekrotisierende Entzündungen in Milz, Leber und Knochenmark. Aus der Milz konnte Yersinia pseudotuberculosis isoliert und die Diagnose Pseudotuberkulose gestellt werden. Diese potenzielle Zoonose wurde bisher nur in einem weiteren Fallbericht bei Heimtierkaninchen in der Fachliteratur dokumentiert und sollte differenzialdiagnostisch bei schwerer Allgemeinerkrankung von Kaninchen in Außenhaltung berücksichtigt werden.

Atypical mitosis marks a deviation in the cell division process that has been shown be an independent prognostic marker for tumor malignancy. However, atypical mitosis classification remains challenging due to low prevalence, at times subtle morphological differences from normal mitotic figures, low inter-rater agreement among pathologists, and class imbalance in datasets. Building on the Atypical Mitosis dataset for Breast Cancer (AMi-Br), this study presents a comprehensive benchmark comparing deep learning approaches for automated atypical mitotic figure (AMF) classification, including end-to-end fine-tuned deep learning models, foundation models with linear probing, and foundation models fine-tuned with low-rank adaptation (LoRA). For rigorous evaluation, we further introduce two new held-out AMF datasets - AtNorM-Br, a dataset of mitotic figures from the TCGA breast cancer cohort, and AtNorM-MD, a multi-domain dataset of mitotic figures from a subset of the MIDOG++ training set. We found average balanced accuracy values of up to 0.8135, 0.7788, and 0.7723 on the in-domain AMi-Br and the out-of-domain AtNorm-Br and AtNorM-MD datasets, respectively. Our work shows that atypical mitotic figure classification, while being a challenging problem, can be effectively addressed through the use of recent advances in transfer learning and model fine-tuning techniques. We make all code and data used in this paper available in this github repository: https://github.com/DeepMicroscopy/AMi-Br_Benchmark

Emerging and re-emerging viral infections demand the rapid development of broad-spectrum antivirals. Host-directed therapies targeting cellular metabolic pathways required for viral replication offer a promising strategy with a low risk of resistance. Dihydroorotate dehydrogenase (DHODH), the rate-limiting enzyme in de novo pyrimidine synthesis, is essential for the replication of many human pathogenic viruses. Several DHODH inhibitors, including vidofludimus calcium, are under clinical investigation for viral diseases, such as coronavirus disease 19 (COVID-19). We report medicinal chemistry optimization of vidofludimus, yielding more potent derivatives with improved DHODH inhibition across mammalian species and enhanced activity against severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2). Optimized compounds also showed broad-spectrum antiviral activity against enveloped and nonenveloped DNA and RNA viruses and a retrovirus, with single-digit nanomolar potency and no detectable cytotoxicity. Selectivity indices exceeded 50,000, highlighting the therapeutic potential of this target. These findings identify this compound class as a promising starting point for host-directed antivirals (HDAs), as well as for pandemic preparedness and applications in other disease areas.

The annotation of large scale histopathology image datasets remains a major bottleneck in developing robust deep learning models for clinically relevant tasks, such as mitotic figure classification. Folder-based annotation workflows are usually slow, fatiguing, and difficult to scale. To address these challenges, we introduce SWipeable ANnotations (SWAN), an open-source, MIT-licensed web application that enables intuitive image patch classification using a swiping gesture. SWAN supports both desktop and mobile platforms, offers real-time metadata capture, and allows flexible mapping of swipe gestures to class labels. In a pilot study with four pathologists annotating 600 mitotic figure image patches, we compared SWAN against a traditional folder-sorting workflow. SWAN enabled rapid annotations with pairwise percent agreement ranging from 86.52

Homozygous sauteur d'Alfort rabbits have a splicing-site mutation in the RAR-related orphan receptor B (RORB) gene that causes changes in their locomotion. These rabbits sometimes walk on their thoracic limbs, especially at higher speeds. They also experience additional abnormalities, including eye alterations such as blindness due to retinal dysplasia, cataract formation, lens luxation, coloboma, glaucoma and other complications. This report describes two 17-month-old intact female dwarf rabbits that have exhibited locomotory disturbances since their first few weeks of life and severe eye alterations since they were several months old. Both animals appeared smaller than their other siblings. The locomotory and eye alterations resemble those observed in sauteur d'Alfort rabbits. Genetic analysis showed that both rabbits were homozygous for the wild-type mutation in the RORB gene. This indicates that their behavioural and ocular alterations were attributable to a distinct genetic basis rather than the sauteur d'Alfort allele.

AIMS:The gastrointestinal (GI) tract is composed of distinct sub-regions, which exhibit segment-specific differences in microbial colonization and (patho)physiological characteristics. Gut microbes can be collectively considered as an active endocrine organ. Microbes produce metabolites, which can be taken up by the host and can actively communicate with the immune cells in the gut lamina propria with consequences for cardiovascular health. Variation in bacterial load and composition along the GI tract may influence the mucosal microenvironment and thus be reflected its interstitial fluid (IF). Characterization of the segment-specific microenvironment is challenging and largely unexplored because of lack of available tools. METHODS AND RESULTS:Here, we developed methods, namely tissue centrifugation and elution, to collect IF from the mucosa of different intestinal segments. These methods were first validated in rats and mice, and the tissue elution method was subsequently translated for use in humans. These new methods allowed us to quantify microbiota-derived metabolites, mucosa-derived cytokines, and proteins at their site-of-action. Quantification of short-chain fatty acids showed enrichment in the colonic IF. Metabolite and cytokine analyses revealed differential abundances within segments, often significantly increased compared to plasma, and proteomics revealed that proteins annotated to the extracellular phase were site-specifically identifiable in IF. Lipopolysaccharide injections in rats showed significantly higher ileal IL-1β levels in IF compared to the systemic circulation, suggesting the potential of local as well as systemic effect. CONCLUSION:Collection of IF from defined segments and the direct measurement of mediators at the site-of-action in rodents and humans bypasses the limitations of indirect analysis of faecal samples or serum, providing direct insight into this understudied compartment.

BACKGROUND:Liver tumours are common and deadly diseases in humans and are often associated with a poor prognosis. To better understand the mechanisms and potential therapies for hepatocellular carcinoma and cholangiocarcinoma, experimental animal models are used in preclinical research and spontaneously occurring tumours in companion animals can also be investigated as models. OBJECTIVE:We provide an overview of commonly used animal models in hepatic cancer research, including their applications, advantages, and disadvantages. Additionally, we present pathological findings of the most frequent spontaneous liver neoplasms in dogs and cats. MATERIALS AND METHODS:We performed a review of the scientific literature and of findings in animal models. Additionally, we performed an analysis of case statistics from a veterinary pathology submission laboratory. RESULTS:Mouse models are most frequently used in the study of hepatic tumours. These include xenografts, in which human tumour cells are injected, chemically and diet-induced models for studying environmental and metabolic factors, and genetically modified mice. In addition to mice, other species such as rats, rabbits, and woodchucks are also utilized. Woodchucks carry their own hepatitis virus, making them particularly suitable for investigating viral pathogenesis. Furthermore, domestic animals such as dogs and cats develop spontaneous liver neoplasms, which makes them valuable models. Dogs predominantly develop hepatocellular adenomas and carcinomas, whereas benign, cystic bile duct lesions are more common in cats. CONCLUSIONS:A wide range of animal models is available for various research questions. Choosing the appropriate model is crucial for the success of any study and depends on multiple factors.