Inter-Rater and Intra-Rater Agreement in Scoring Severity of Rodent Cardiomyopathy and Relation to Artificial Intelligence-Based Scoring.

We previously developed a computer-assisted image analysis algorithm to detect and quantify the microscopic features of rodent progressive cardiomyopathy (PCM) in rat heart histologic sections and validated the results with a panel of five veterinary toxicologic pathologists using a multinomial logistic model. In this study, we assessed both the inter-rater and intra-rater agreement of the pathologists and compared pathologists' ratings to the artificial intelligence (AI)-predicted scores. Pathologists and the AI algorithm were presented with 500 slides of rodent heart. They quantified the amount of cardiomyopathy in each slide. A total of 200 of these slides were novel to this study, whereas 100 slides were intentionally selected for repetition from the previous study. After a washout period of more than six months, the repeated slides were examined to assess intra-rater agreement among pathologists. We found the intra-rater agreement to be substantial, with weighted Cohen's kappa values ranging from k = 0.64 to 0.80. Intra-rater variability is not a concern for the deterministic AI. The inter-rater agreement across pathologists was moderate (Cohen's kappa k = 0.56). These results demonstrate the utility of AI algorithms as a tool for pathologists to increase sensitivity and specificity for the histopathologic assessment of the heart in toxicology studies.

The Role of IgM Antibodies in T Cell Lymphoma Protection in a Novel Model Resembling Anaplastic Large Cell Lymphoma.

MRL/lpr mice typically succumb to immune complex-mediated nephritis within the first year of life. However, MRL/lpr mice that only secrete IgM Abs because of activation-induced deaminase deficiency (AID-/-MRL/lpr mice) experienced a dramatic increase in survival. Further crossing of these mice to those incapable of making secretory IgM (µS mice) generated mice lacking any secreted Abs but with normal B cell receptors. Both strains revealed no kidney pathology, yet Ab-deficient mice still experienced high mortality. In this article, we report Ab-deficient MRL/lpr mice progressed to high-grade T cell lymphoma that can be reversed with injection of autoreactive IgM Abs or following adoptive transfer of IgM-secreting MRL/lpr B cells. Anti-nuclear Abs, particularly anti-dsDNA IgM Abs, exhibited tumor-killing activities against a murine T cell lymphoma cell line. Passive transfers of autoreactive IgM Abs into p53-deficient mice increased survival by delaying onset of T cell lymphoma. The lymphoma originated from a double-negative aberrant T cell population seen in MRL/lpr mice and most closely resembled human anaplastic large cell lymphoma. Combined, these results strongly implicate autoreactive IgM Abs in protection against T cell lymphoma.

Using Artificial Intelligence to Detect, Classify, and Objectively Score Severity of Rodent Cardiomyopathy.

Rodent progressive cardiomyopathy (PCM) encompasses a constellation of microscopic findings commonly seen as a spontaneous background change in rat and mouse hearts. Primary histologic features of PCM include varying degrees of cardiomyocyte degeneration/necrosis, mononuclear cell infiltration, and fibrosis. Mineralization can also occur. Cardiotoxicity may increase the incidence and severity of PCM, and toxicity-related morphologic changes can overlap with those of PCM. Consequently, sensitive and consistent detection and quantification of PCM features are needed to help differentiate spontaneous from test article-related findings. To address this, we developed a computer-assisted image analysis algorithm, facilitated by a fully convolutional network deep learning technique, to detect and quantify the microscopic features of PCM (degeneration/necrosis, fibrosis, mononuclear cell infiltration, mineralization) in rat heart histologic sections. The trained algorithm achieved high values for accuracy, intersection over union, and dice coefficient for each feature. Further, there was a strong positive correlation between the percentage area of the heart predicted to have PCM lesions by the algorithm and the median severity grade assigned by a panel of veterinary toxicologic pathologists following light microscopic evaluation. By providing objective and sensitive quantification of the microscopic features of PCM, deep learning algorithms could assist pathologists in discerning cardiotoxicity-associated changes.

Evaluation of Cystic Endometrial Hyperplasia and the Normal Estrous Cycle in Longitudinal Sections of Uterus from Female Harlan Sprague-Dawley Rats.

The National Toxicology Program (NTP) has recently introduced the practice of examining longitudinal histological sections of the female rodent uterus to improve the identification of non-neoplastic lesions, preneoplastic lesions, and uterine tumors. This practice has created a need for reference material that includes normal histology, spontaneous lesions, and inducible lesions in longitudinal as well as transverse sections of the body of the uterus, uterine horns, cervix and vagina. Using 3 archived NTP reproductive and developmental toxicity studies, the authors reviewed longitudinal and transverse sections of uteri from female Hsd:Sprague Dawley SD® (Hsd:SD) rats for cystic endometrial hyperplasia (CEH). The purposes of this review were to (1) evaluate if existing criteria for CEH in transverse uterine sections could be applied to longitudinal sections to develop diagnostic features of CEH in longitudinal uterine sections of rat uterus and (2) create an atlas of the normal estrous cycle phases in longitudinal sections of young and mature adult Hsd:SD rat uteri. The information provided in this original article should help facilitate the examination of longitudinal sections of the uterus in future commercial and governmental rodent studies.

Toward a molecular equivalent dose: use of the medaka model in comparative risk assessment.

Recent changes in the risk assessment landscape underscore the need to be able to compare the results of toxicity and dose-response testing between a growing list of animal models and, quite possibly, an array of in vitro screening assays. How do we compare test results for a given compound between vastly different species? For example, what dose level in the ambient water of a small fish model would be equivalent to 10 ppm of a given compound in the rat's drinking water? Where do we begin? To initially address these questions, and in order to compare dose-response tests in a standard rodent model with a fish model, we used the concept of molecular dose. Assays that quantify types of DNA damage that are directly relevant to carcinogenesis integrate the factors such as chemical exposure, uptake, distribution, metabolism, etc. that tend to vary so widely between different phyletic levels. We performed parallel exposures in F344 rats and Japanese medaka (Oryzias latipes) to the alkylating hepatocarcinogen, dimethylnitrosamine (DMN). In both models, we measured the DNA adducts 8-hydroxyguanine, N(7)-methylguanine and O(6)-methylguanine in the liver; mutation frequency using lambda cII transgenic medaka and lambda cII transgenic (Big Blue(R)) rats; and early morphological changes in the livers of both models using histopathology and immunohistochemistry. Pulse dose levels in fish were 0, 10, 25, 50, or 100 ppm DMN in the ambient water for 14 days. Since rats are reported to be especially sensitive to DMN, they received 0, 0.1, 1, 5, 10, or 25 ppm DMN in the drinking water for the same time period. While liver DNA adduct concentrations were similar in magnitude, mutant frequencies in the DMN-exposed medaka were up to 20 times higher than in the Big Blue rats. Future work with other compounds will generate a more complete picture of comparative dose response between different phyletic levels and will help guide risk assessors using "alternative" models.

Diet-induced obesity is associated with hyperleptinemia, hyperinsulinemia, hepatic steatosis, and glomerulopathy in C57Bl/6J mice.

Obesity and obesity-related illnesses are global epidemics impacting the health of adults and children. The purpose of the present work is to evaluate a genetically intact obese mouse model that more accurately reflects the impact of aging on diet-induced obesity and type 2 diabetes in humans. Male C57Bl/6J mice consumed either a control diet or one in which 60% kcal were due to lard beginning at 5-6 weeks of age. Body weight and fat measurements were obtained and necropsy performed at 15, 20, 30, and 40 weeks of age. Serum chemistry, histopathology, gene expression of the liver, and renal and hepatic function were also evaluated. In concert with significant increases in percent body fat and weight, mice fed the high-fat versus control diet had significantly increased levels of serum cholesterol. At ages 20 and 30 weeks, serum glucose was significantly higher in obese versus controls, while serum insulin levels were >/=4-fold higher in obese mice at ages 30 and 40 weeks. The effect of age exacerbated the effects of consuming a high-fat diet. In addition to being hyperinsulinemic and leptin resistant, older obese mice exhibited elevated hepatic PAI-1 and downregulation of GLUT4, G6PC, IGFBP-1, and leptin receptor mRNA in the liver, steatosis with subsequent inflammation, glomerular mesangial proliferation, elevated serum ALT, AST, and BUN, and increased numbers of pancreatic islets.