Abstract
P01 | Chronic toxicity and carcinogenicity of AGIQ in Sprague Dawley rats by dietary exposure
Bob Maronpot 1 , Yuval Ramot 2 , Abraham Nyska 3 , Christopher Sproule 4 , Rebecca Moore 5 , Brad Bolon 6 , Shim-mo Hayashi 7
1 Maronpot Consulting, LLC, Raleigh, United States
2 BFaculty of Medicine, Hebrew University of Jerusalem, Jerusalem, Israel
3 Consultant, Tel Aviv, Israel
4 Integrated Laboratory Systems, LLC, Morrisville, United States
5 Integrated Laboratory Systems, Morrisville, United States
6 GEMpath, Inc., Longmont, United States
7 National Institute of Health, Kawasaki, Japan
alpha-Glycosyl isoquercitrin (AGIQ) is a flavonoid with antioxidative and tumor-suppressive capabilities that is marketed as a food additive in Japan. This study assessed the chronic toxicity and carcinogenicity of AGIQ in male and female SD rats following up to 5.0% dietary exposure.
Exposure for one (chronic toxicity study) or two (carcinogenicity studies) years.
No AGIQ-related toxicity was observed except for yellow discoloration of bone. In the carcinogenicity study, a statistically significant increase in malignant gliomas of the brain or spinal cord was observed in female rats exposed to 5.0% AGIQ compared to controls (0/50 and 5/49 [10.2%], respectively). Based on cytological features and immunohistochemical positivity for the microglial marker Iba1 (ionized calcium-binding adaptor molecule 1), but not for astrocytic (GFAP, glial fibrillary acidic protein) or oligodendrocyte (Olig2, oligodendrocyte transcription factor 2) markers, neoplasms were categorized as “microglial tumor, malignant”. Recent historical control data (HCD) showed a total glioma incidence in female SD rats of 4/550 (0.7%, range 0-3.3%).
A Scientific Advisory Panel concluded that the Iba1-positive glioma could not definitively be attributed to AGIQ exposure, and thus has limited implications in predicting human cancer risk. The biological significance of these apparently rat-specific malignant tumors remains questionable since they are so rare in humans that the current classification system for human neural tumors does not include a “microglial tumor” category.
HCD indicate that gliomas occur spontaneously in rats, thus supporting the interpretation that gliomas observed in the current study are unrelated to treatment.
P02 | Mineral Fiber-Reinforced Plate Implantation in a Sheep Model: Long Term Safety and Bio-Integration In-Vivo Study
Nicolette D. Jackson 1 , Ezequiel Palmanovich 2 , Abraham Nyska 3 , Meir Nyska 4
1 Inotiv, Fort Collins, United States
2 Meir Medical Center, Kfar Saba, Israel
3 Consultant, Tel Aviv, Israel
4 Meir Medical Center, Kfar Saba, Israel
Traditionally, orthopedic implants were made of metal to provide mechanical strength and durability, essential for stable bone fixation. However, complications including implant migration, stress concentrations and bone-tissue loss, resulting in the need for hardware removal, drove the use of non-permanent polymer-based implants, which carry other risks, of adverse inflammatory reactions to degradation by-products. In this study, we evaluated the biocompatibility and bio-integration profile of a new class of mineral-fiber reinforced plates in a sheep tibia model.
Plates, made of continuous reinforcing mineral fibers bound together by PLDLA (50%w/w), were implanted bilaterally over the medial surface of sheep tibiae. Left tibiae underwent periosteal elevation, and right tibiae had intact periosteum. MicroCT and histopathology were performed at 13,26,52,78,104,134-weeks(W) post-implantation. Overall cellular response, rate of bioabsorption (i.e., phagocytosis, M2-like macrophages/giant cell (MNGCs) infiltration), and mesenchymal ingrowth were graded according to ISO-10993-6(annex E).
Mesenchymal ingrowth into the device wall was similar for both groups, and the cellular response consisted of anti-inflammatory M2-like macrophages and MNGCs. Adverse inflammation was not observed. Phagocytic activity started at 13W, peaked at 52W-78W, and resolved by 104W. Fibers were still evident at 78W and fully remodeled by 104W. By 134W, implants were fully bio-integrated.
This study demonstrates the safe use of fiber-reinforced implants in a bone plate application, with complete bio-integration with surrounding tissue and no toxicological adverse events.
This new bone-fixation technology offers a solution to metal hardware complications and removals, while avoiding adverse inflammatory reactions reported for existing non-permanent implants.
P05 | Organ Weights and Spontaneous Microscopic Observations in the Testis from Sexually Mature Cynomolgus Macaques
Lars Mecklenburg, Annette Romeike
Labcorp, Muenster, Germany
When non-human primates represent the only relevant species for nonclinical safety evaluation of human medicinal products, potentially adverse effects on male fertility are typically assessed in repeat dose toxicity studies with sexually mature monkeys. As per regulatory guidance [ICH S6(R1)], default parameters for such an indirect assessment of fertility are reproductive organ weight and histopathology.
Data was collected from 69 vehicle-treated animals of 13 nonclinical safety studies that were conducted between 2016 and 2023. All animals were proven to be sexually mature by the presence of sperm in a semen sample. Their body weight was between 4.45 and 11.12 kg. 11 animals were of Mauritius origin, and 58 animals were of Asia mainland origin. Where appropriate, original diagnoses, that were reported in the study report, were translated into the current INHAND terminology.
Approx. 1/3 of sexually mature male Cynomolgus macaques show a microscopic observation in the testis. The most frequently found observation is Degeneration/atrophy, tubule which occurs at an incidence of 32% but in most cases is minimal or slight. Only 1 out of 69 animals (1.5%) showed a bilateral Degeneration/atrophy of marked severity. Testis weight in sexually mature Cynomolgus macaques shows high variability with a range of 26.7g to 88.6g (based on the 2.5 and 97.5 percentiles) and does not correlate well with body weight.
We did not observe any relevant difference between Asia origin and Mauritius origin animals (with the caveat that the number of Mauritius origin animals in our investigation was low).
P08 | Comparison of hepatocyte proliferation in different models of liver injury
Charlotte Lempp 1 , Anouk Oldenburger 1 , Steven Kerr 2 , Tanja Schönberger 1 , Fabian Heinemann 1 , Peter Groß 1
1 Boehringer Ingelheim Pharma GmbH & Co KG, Biberach/Riss, Germany
2 Boehringer Ingelheim Pharma GmbH & Co KG, Ridgefield, United States
Nonalcoholic fatty liver disease (NAFLD) is among the most common chronic liver diseases worldwide. Besides dietary models, chemical and surgical models mimic important disease hallmarks (i.e. steatosis, inflammation, fibrosis) for pharmaceutical research. Even though liver regeneration is known to occur in the cirrhotic stage and is involved in neoplastic transformation, regeneration enhancing liver function has not been systematically characterized in these models, especially in acute disease stages.
Three NAFLD mouse models were analyzed at different timepoints after injury (24-48h for APAP; 24- 96h for CCl4; 12-192h for hepatectomy). Using an AI system, immunohistochemical expression of proliferation markers Ki67, PCNA and BrdU was quantified specifically in hepatocytes and lesioned liver area was analyzed quantitatively based on Masson-trichrome stain.
In CCl4, the highest numbers of Ki67- and BrdU- positive hepatocytes occur at 72h post injury; a significant increase in Ki67- positive cells can be detected as early as 48h in CCl4 but not APAP. PCNA-labelling partially reflects these findings but shows overall higher positive detections, also in controls. In partial hepatectomy, proliferation starts to increase at 36h and peaks at 42h; hepatocellular degeneration and necrosis is most widespread at 72h.
Hepatocyte proliferation is most pronounced at 48-72h after CCl4 and at 36-72h after 2/3 hepatectomy. BrdU and Ki67 reveal similar specific labelling, whereas PCNA is less specific for proliferation. Widespread liver lesions in hepatectomy are described to negatively impact regeneration. These data help to select among the mechanistically different injury models, time points and methods to assess hepatocyte proliferation.
P09 | Adenosquamous carcinoma with sebaceous gland differentiation in the Rat Mammary Gland
Heike Marxfeld 2 , Rikke Kaae Kirk 1 , Mikala Skydsgaard 1
1 Novo Nordisk A/S, Maaloev, Denmark
2 BASF SE, Ludwigshafen, Germany
The morphology and hormone receptor pattern of a rare type of adenocarcinoma of the rat was investigated. The two female rats were control animals from a 52 week and a 104 week carcinogenicity study respectively.
Control animals from carcinogenicity study with Sprague-Dawley [Crl:CD(SD)] and HanBrl:WIST (SPF) rats were investigated to generate historical control data for these strains.
Tumor tissue of a female control rat from a 104 week carcinogenicity study which died spontaneously after 728 days on study was investigated. Macroscopically, a 15 mm diameter cyst was noted in the right inguinal side of the mammary gland. In another 52 weeks carcinogenicity study a female control rat was sacrificed on Day 372 (terminal kill) on study and a dark mass was observed in the left ventrocaudal side (19 mm in diameter). Hematoxylin-Eosin stained slides and serial sections stained immunohistochemically with antibodies against estrogen receptor, progesterone receptor and androgen receptor were evaluated.
A malignant mammary gland tumor with features of sebaceous gland differentiation is described with the expression pattern of nuclear hormone receptors in tumor cells.
A new diagnosis “Mammary Carcinoma, adenosquamous with sebaceous differentiation” in the rat is proposed to be included in the INHAND nomenclature.
P12 | Antisense oligonucleotide BMN 351 restores motor function in a hDMDdel52/mdx mouse model of exon 51 skip-amenable Duchenne muscular dystrophy
Todd OPPENEER 1 , Yulan QI 1 , Joshua HENSHAW 1 , Kevin LARIMORE 1 , Jukka PUOLIVALI 2 , Caitlyn CARTER 3 , Pierluigi FANT 4 , Laura WETZEL 1 , Monika SIGG 1 , Charles A. O'NEILL 1
1 BioMarin Pharmaceutical Inc., San Rafael, United States
2 Charles River Discovery Research Services, Kuopio, Finland
3 Charles River Laboratories, Mattawan, United States
4 Charles River Laboratories France Safety Assessment, St. Germain-Nuelles, France
Duchenne muscular dystrophy (DMD) is an X-linked recessive degenerative muscle disease caused by frame-shift mutations in the DMD gene encoding dystrophin. A novel antisense oligonucleotide (BMN 351) was designed to selectively induce skipping in a DMD exon neighboring the frame-shift mutation, allowing the synthesis of a shortened and partially functional dystrophin protein.
BMN 351 was given to hDMDdel52/mdx mice by intravenous injection once weekly for 25 weeks at the doses of 6 or 18 mg/kg. Motor function and clinical/anatomic pathology data were evaluated at weeks 29 and 37 (i.e., 4- or 12-weeks post-dosing, respectively). Vehicle-treated hDMDdel52/mdx mice and healthy C57BL/6J wild type (WT) mice were used for comparison.
Results were indicative of pharmacological efficacy of BMN 351 at each time point. Functionally, BMN 351 treatment at both doses improved overall gait scores, when compared with vehicle-treated hDMDdel52/mdx mice. Clinical pathology parameters such as AST, LDH and CK in BMN 351-treated mice at either dose were lower than in vehicle-treated hDMDdel52/mdx mice. Microscopic findings such as myofiber atrophy, necrosis, regeneration, mineralization, and inflammation showed a trend towards a dose-related lower incidence and severity in BMN 351-treated mice compared with vehicle-treated hDMD del52/mdx mice.
Vehicle-treated hDMDdel52/mdx mice had poor motor performance, high serum concentrations of AST, LDH, CK and microscopic muscle changes, consistent with the DMD phenotype. These changes were not observed in WT mice.
The functional and anatomic/clinical pathology improvement observed in hDMDdel52/mdx mice treated with BMN 351 are promising and support further development for clinical trials.
P14 | Recommendations for GLP-Conform Archiving of Whole Slide Images
Ute Bach 1 , Erio Barale-Thomas 2 , Brian Knight 3 , Sébastien Laurent 4 , Xavier Palazzi 5 , Gabriele Pohlmeyer-Esch 6
1 Bayer AG, Wuppertal, Germany
2 Janssen, Beerse, Belgium
3 Boehringer Ingelheim, Ridgefield, United States
4 Sanofi, Montpellier, France
5 Pfizer, Groton, United States
6 Boehringer Ingelheim, Biberach an der Riss, Germany
In recent years, working groups of toxicologic pathology societies (ESTP, STP), consortium-driven direct interactions with regulatory authorities and efforts made by individual institutions have helped to develop a general understanding of elements required for the GLP-conform use of digital pathology.
To date only a few institutions have claimed to use digital pathology under GLP, and only a minority of these have been inspected by national GLP authorities, generating little feedback to build upon.
An informal pre-competitive inter-laboratory exchange forum was set up in fall 2022 by interested toxicologic pathologists from the pharmaceutical and agrochemical industry and CROs, with the aim to share and convert the collective knowledge of theoretical requirements into aligned practical recommendations and solutions for the most critical aspects of digital pathology workstream GLP validation.
Archiving of whole slide images (WSI) must be considered in the context of raw data generation. If archival of WSI is required (primary evaluation, retrospective peer review), in addition to the archival of glass slides, the processes to retain original electronic data generated by the computerized system as well as procedures to assure data integrity need to be defined according to the respective OECD documents.
The group recommends using DICOM format for back compatibility reasons. Additionally, processes such as archival log, duration of archiving as well as recovery and disaster recovery need to be defined. Finally, the decision of using either cloud solution or physical address of the data center for archiving purposes needs to be based on a suitable risk assessment.
P15 | Recommendations for the Utility of Equivalency and Concordance Studies in Digital Pathology
Jonathan Carter 1 , Ute Bach 2 , Julie Boisclair 3 , Thomas Forest 4 , Peter Hall 5 , Charles Halsey 6 , Charlotte Lempp 7 , Pierre Maliver 8 , Patricia Neves 9 , Moritz Radbruch 2 , Frédéric Schorsch 9 , Charlotte Dalsgaard 10 , Xavier Palazzi 6 , Gabriele Pohlmeyer-Esch 7
1 Labcorp Early Development Laboratories Ltd., Harrogate, United Kingdom
2 Bayer AG, Wuppertal, Germany
3 Novartis, Basel, Switzerland
4 Merck & Co. Inc., Westpoint, United States
5 UCB Pharma, Slough, United Kingdom
6 Pfizer, Groton, United States
7 Boehringer Ingelheim, Biberach an der Riss, Germany
8 Roche Pharma Research and Early Development, Roche Innovation Center Basel, Basel, Switzerland
9 Bayer CropScience, Sophia-Antipolis, France
10 Novo Nordisk A/S, Måløv, Denmark
To date only a few institutions have claimed to use digital pathology under GLP, and only a minority of these have been inspected by national GLP authorities, generating little feedback to build upon.
An informal pre-competitive inter-laboratory exchange forum was set up in fall 2022 by interested toxicologic pathologists from the pharmaceutical and agrochemical industry and CROs, with the aim to share and convert the collective knowledge of theoretical requirements into aligned practical recommendations and solutions for the most critical aspects of digital pathology workstream GLP validation in common use cases.
With regard to the utilization of comparative examinations of glass slide versus whole slide image, it was this group’s opinion that a blinded concordance study (examination of glass slides and WSI, with an intervening wash-out period, and statistical analysis for concordance of the data generated from each media type) was primarily applicable as a proof of concept and not a prerequisite for GLP-compliant digital peer review or primary read. However, a need for published documentation of concordance studies from the toxicologic pathology industry to further support GLP digital primary read was identified.
The requirement for documented pathologist approval that WSI are fit for purpose (GLP digital peer review and primary read) was, therefore, considered to be met by an equivalency study (non-blinded comparison of glass slides and WSI to qualitatively confirm equivalence based on assessment of critical features and/or lesions).
N.A.
P16 | Recommendations for the Qualification of Instruments in Digital Pathology
Gundi Mueller 1 , Charles Halsey 2 , Moritz Radbruch 3 , Dirk Schaudien 4 , Xavier Palazzi 2 , Gabriele Pohlmeyer-Esch 5
1 Merck Healthcare KGaA, Darmstadt, Germany
2 Pfizer, Groton, United States
3 Bayer AG, Wuppertal, Germany
4 Fraunhofer Institute, Hanover, Germany
5 Boehringer Ingelheim, Biberach an der Riss, Germany
In recent years, working groups of toxicologic pathology societies (ESTP, STP), consortium-driven direct interactions with regulatory authorities and efforts made by individual institutions have helped to develop a general understanding of elements required for the GLP-conform use of digital pathology for peer reviews and primary reads. To date only a few institutions have claimed to use digital pathology under GLP, and only a minority of these have been inspected by national GLP authorities, generating little feedback to build upon.
An informal pre-competitive inter-laboratory exchange forum was set up in fall 2022 by interested toxicologic pathologists from the pharmaceutical and agrochemical industry and CROs, with the aim to share and convert the collective knowledge of theoretical requirements into aligned practical recommendations and solutions for the most critical aspects of digital pathology workstream GLP validation in common use cases.
The group recommends that instrument qualification of scanners and pathologist workstations be based on a set of predetermined specifications by both the vendor (installation and operational qualification) and end-user requirement specifications (user acceptance testing).
Spatial and color calibration of scanners (calibrating displayed lengths against known values) are discussed. The group felt that the monitor and the viewing environment are a point of concern. This can be addressed via a QC slide, “point of use QA” approaches, color calibration, and/or via equivalency studies. Options and open questions are discussed.
P17 | Pre-Requisites to the Conduct of a Non GLP Digital Peer Review on a GLP study
Julie Boisclair 1 , Jonathan Carter 2 , Peter Hall 3 , Matt Jacobsen 4 , Pierre Maliver 5 , Xavier Palazzi 6 , Gabriele Pohlmeyer-Esch 7
1 Novartis, Basel, Switzerland
2 Labcorp Early Development Laboratories Ltd., Harrogate, United Kingdom
3 UCB Pharma, Slough, United Kingdom
4 AstraZeneca, Cambridge, United Kingdom
5 Roche Pharma Research and Early Development, Roche Innovation Center Basel, Basel, Switzerland
6 Pfizer, Groton, United States
7 Boehringer Ingelheim, Biberach an der Riss, Germany
In recent years, working groups of toxicologic pathology societies, consortium-driven direct interactions with regulatory authorities, and efforts made by individual institutions have helped to develop a general understanding of the elements required for the GLP-conform use of digital pathology.
To date only a few institutions have claimed to use digital pathology under GLP, and only a minority of these have been inspected by national GLP authorities, generating little feedback to build upon.
An informal pre-competitive inter-laboratory exchange forum was set up in fall 2022 by interested toxicologic pathologists from the pharmaceutical and agrochemical industry and CROs, with the aim to share and convert the collective knowledge of theoretical requirements into aligned practical recommendations and solutions for the most critical aspects of digital pathology workstream GLP validation.
One scenario discussed was not claiming GLP compliance for digital peer review on a GLP study. An in-depth assessment of the existing published regulatory guidance and their requirements was performed. The analysis suggested that not claiming GLP compliance for a digital peer review should be acceptable as long as the overall integrity of the GLP study is maintained. It was considered this scenario represented a low regulatory risk.
The poster summarizes the conclusion of the regulatory review, possible digital workflows with their respective stakeholders, and lists the mandatory and optional, site- and study-specific requirements that need to be in place before, during and after the conduct of the digital peer review. Examples of wording for SOPs, study plans, and study reports are suggested.
P18 | MEN-like Syndrome: Does it occur in rats? - An Analysis of Data from the RITA Database
Rupert Kellner 2 , Maike Huisinga 1 , Susanne Rittinghausen 2 , Matthias Rinke 3 , Krystyna Siudak 3
1 BASF SE, Ludwigshafen, Germany
2 Fraunhofer ITEM, Hannover, Germany
3 Bayer AG, Wuppertal, Germany
Multiple endocrine neoplasia (MEN) is a condition which encompasses several distinct syndromes featuring at least two tumors of endocrine glands, mainly of neuroendocrine origin. While proliferations of neuroendocrine origin in humans are related to a genetic background (MEN Syndromes), the background in domestic and laboratory animals in most cases is unknown (MEN-like Syndromes). An analysis of RITA data was performed to investigate whether the condition of a MEN-like syndrome may occur also in Wistar- and Sprague Dawley (SD) rats.
Substrains of control Wistar and SD rats were analyzed for proliferative neuroendocrine lesions from adrenals, pancreas, thyroid, parathyroid and pituitary. The diagnoses followed the criteria as described in INHAND. Animals with single organ versus multiple organs affected were counted and compared to expected values. In case of significantly increased multiple affections the combinations of organs known from MEN Syndromes were evaluated in search of MEN-like Syndromes. Co-occurrence in two organs versus one organ was analyzed as compared to expected values.
In male Wistar rats, significant results were observed in pancreas and thyroid. In male SD rats, similar results were seen for the thyroid, adrenal, and pancreas. An increased co-occurrence of adrenal and thyroidal lesions was clearly evident for both male Wistar and SD control rats.
Even though no clear evidence for the existence of MEN1-like syndromes was given, the results from males of both breeds showed a MEN2-like pattern. This analysis indicates the value of detailed metadata of the RITA database.
P20 | Histology-guided engineering and characterization of complex in vitro models of blood-brain barrier for efficacy/toxicity in early drug development
Luisa Bell, Martina Pigoni, Roberto Villasenor, Christelle Zundel, Petra Stäuble, Nadine Stokar
Roche, Basel, Switzerland
Complex in vitro models (CIVM) of the blood-brain barrier (BBB) could facilitate effective testing of drug candidates targeting the brain early in the drug discovery process. Temporal and phenotypic histomorphological readouts of CIVMs offer an excellent opportunity to address discrete mechanistic questions at the cellular and/or tissue level and enable comparison to the standardized histomorphological readouts of in vivo studies.
Here, we present an array of tissue technologies (Haematoxylin Eosin staining, HE; immunohistochemistry, IHC; immunofluorescence, IF) that enabled characterization/validation and histology-guided engineering of CIVM of the BBB, that was previously developed for investigating receptor-mediated transcytosis (Simonneau et al., 2021).
Expression of relevant cellular BBB markers was observed both in human brain tissue samples and BBB organoids by IHC. Multiplex IF provided spatial information during the engineering steps for a correct assembly of the three cell types to a functional BBB. We could conclude that our BBB organoids did not recapitulate the morphology and cell composition of the human BBB, demanding an improved cell source for brain endothelial cells. To troubleshoot, different protocols of endothelial cell generation were developed with BBB transwells. Histologic evaluation of a HE staining was used to monitor the assembly of a monolayer of brain endothelial cells in the transwell.
Standard morphologic readouts provide a quick and simple overview during engineering of novel CIVM enabling their characterization towards their tissue of origin and input cell sources. In turn, pathologists can bring their expertise to improve the translatability of preclinical research and relate outcomes to human patients.
P21 | Pathology 2.0: a collaborative working group to increase awareness of cutting-edge technologies in experimental and toxicologic pathology
Josep Monné 2 , Dirk Schaudien 1
1 Fraunhofer ITEM, Hannover, Germany
2 Roche Innovation Center Basel, F. Hoffmann-La Roche Ltd, Basel, Switzerland
The ESTP Pathology 2.0 working group fosters a collaborative network of pathologists and non-pathologists from different organizations across Europe and US. The main purpose is to share knowledge, skills, and experience among the society members about relevant topics of interest for the pathology community.
Knowledge sharing comes in the form of webinars, publications, surveys, and presentations at international congresses. The partnership and coordinated collaboration between individuals from different organizations also creates a space that embrace innovation and supports the development and implementation of new technologies.
Pathology 2.0 is organized in 5 different subgroups which work synergistically in different goals and strategies. The subgroups are: 1) Molecular Pathology; 2) Multiplexing; 3) Spatial OMICS; 4) Mass Spectrometry Imaging; 5) Complex in vitro models & Pathology. Since first created, Pathology 2.0 has been continuously expanding and now counts more than 40 members. Most importantly, the working group is always open to new members and ideas to continue advancing knowledge, promoting collaboration and ultimately bringing a positive impact to the toxicological pathology field.
This poster gives an overview of the Pathology 2.0 activities and goals achieved during the last year.
P24 | Automated histomorphologic artificial intelligence (AI) readout for toxicity evaluation in blood-brain barrier organoids
Luisa Bell, Elena Kassianidou, Roberto Villasenor, Nadine Stokar
Roche, Basel, Switzerland
Human complex in vitro models (CIVM) address the need for more predictive toxicology models that might increase future clinical success in drug development. Digital tools enable high throughput and automation combined with evaluation of single cell morphology as it is the standard for histopathologic evaluation performed by pathologists today. As a proof-of-concept model, we used the blood-brain barrier (BBB) organoids, an excellent CIVM to screen for efficacy and toxicity in brain delivery of biologics.
We developed an artificial intelligence (AI) tool using Visiopharm software to automatically detect apoptotic/necrotic nuclei in BBB organoid 40x whole slide Haematoxylin and Eosin (HE) images. The ground truth included Caspase-3 immunohistochemistry (IHC) as well as evaluation of apoptotic/necrotic nuclei on HE slides (40 randomly chosen organoids) by three blinded ECVP board-certified veterinary anatomic pathologists. For validation, we treated BBB organoids with different concentrations of staurosporine to induce apoptosis and compared the algorithms’ performance towards a standard low magnification (5x) caspase 3/7 assay.
As a result, our AI algorithms detected the apoptosis as well as the 5x caspase assay, and additionally provided a percentage of apoptosis at the single cell level instead of an arbitrary fluorescence signal. It further added single cell and spatial resolution, digitalization, automation, high throughput and reproducible components.
Our proof-of-concept study showed that our label-free end-to-end workflow presents a novel tool to automatically screen for apoptosis/necrosis in BBB organoids at a single-cell level, and can be applied to future toxicity studies performed with BBB organoids.
P25 | Improved tolerability with the vascular access button (VAB) vs. vascular access harness (VAH) method in the rat
Pierluigi FANT, Joffrey DUCROQ, AMINA TIGHENIFI, Philippe VIGNAND, Veronica CROCCHIANTI
Charles River Laboratories France Safety Assessment, St. Germain-Nuelles, France
The vascular access harness (VAH) and vascular access button (VAB) are currently used in infusion studies in rats.
Thirty-six male Wistar rats were split into 4 groups and submitted to 13-week intermittent infusion of saline (0.9% NaCl) or mannitol (50 mg/mL) at 1 mL/kg/hour, 5 days a week. Each solution was administered either with VAH or VAB system. Clinical signs, body weight, food intake, infusion incidents, clinical pathology, gross and microscopic findings were evaluated.
There was no mortality. Independent of the formulation infused (saline or mannitol), catheter occlusion occurred in 11/18 VAH-equipped rats and 3/18 males VAB-equipped rats. Catheter disconnection occurred in 10/18 VAH-equipped rats. Skin ulceration due to the harness was noted in 3/18 VAH-equipped rats. At euthanasia, lower white blood cell counts were noted in VAB-equipped rats.
At necropsy, there were firm consistency of catheterized vena cava, enlargement of left iliac lymph nodes, cutaneous scab and spleen enlargement in VAH-equipped rats and not in VAB-equipped rats. Microscopically, findings were observed at the infusion site (thrombus and vascular/perivascular inflammation), iliac lymph nodes (increased cellularity of germinal centers and plasma cells), spleen (extramedullary hematopoiesis), kidneys (inflammation) and lungs (thrombus) at lower severity/incidence in VAB- than VAH-equipped rats. Bacterial colonies were observed at the infusion site in VAH-equipped rats only.
In-life and post-mortem findings were indicative of improved tolerability with the VAB perfusion system, which also had animal welfare advantages because it allowed animal group housing.
P28 | Real-time messaging of pathological diagnoses from toxicology laboratory information management system (LIMS) to cloud-based digital pathology web application
Pierre-Maxence Vaysse 1 , Patricia Neves 1 , Moritz Radbruch 2 , Hervé Lormeau 1 , Fabien Lopez 1 , Matthieu Bouillé 3 , Frédéric Schorsch 1
1 Bayer Crop Science, Sophia Antipolis, France
2 Bayer Pharmaceuticals, Wuppertal, France
3 Bayer Crop Science, Lyon, France
Exchanges between toxicologic pathologists have always been critical to establish diagnoses, especially to reinforce the interpretation during peer reviews. Access to Whole Slide Images (WSI) is now easier and can be done by different pathologists from different places at a chosen time.
Here we present an innovative feature which allows pathologists to have access simultaneously to the WSI and their pathological diagnoses updated in near real-time from the internal laboratory information management system (LIMS).
Toxicology studies were managed on Pristima (Xybion) LIMS. WSI were generated on an Aperio AT2 scanner (Leica Biosystems), filenames were automatically created from the glass slides’ barcodes to encoding relevant identifying metadata including study number, animal number and tissue abbreviations. WSI were uploaded to a Concentriq (Proscia), web application hosted on a Bayer-private Amazon web service (AWS) account.
First, automated population of the respective WSI metadata fields (study number, animal number, and tissue abbreviations) in Concentriq took less than a minute after WSI upload. Then, Pristima messaging module identified each WSI in Concentriq specific to this combination of study number/animal number/tissue. Finally, Pristima messaging module sent the correct pathological diagnoses to the dedicated metadata field next to each corresponding WSI in less than 2 minutes.
The developed interface Pristima-Concentriq offers an innovative way to exchange on interpreted WSI from a web application easily accessible to users, from anywhere around the world. This interface relies on a defined WSI naming convention and is expected to improve study case evaluation and image analysis research in the future.
P33 | Therapeutic Oligonucleotides: A Retrospective Evaluation of Non-Clinical Toxicology Data Obtained at Charles River Laboratories France
Pierluigi FANT 1 , Mylad DI CAMILLO 1 , Isabelle LECONTE 1 , Cécile SOBRY 2
1 Charles River Laboratories France Safety Assessment, St. Germain-Nuelles, France
2 Charles River Laboratories, Evreux, France
RNA-targeting oligonucleotides constitute a promising class of therapeutics for the potential treatment of otherwise undruggable diseases.
A retrospective review was carried out from approximately 90 studies performed at two Charles River Laboratories sites in France during the last 15 years using different species (rat, mouse and non-human primate). The database includes therapeutic areas, chemistry platform of the tested oligonucleotides, experimental study design (study duration and route of administration), and data collected from toxicokinetic, clinical and anatomic pathology evaluations, mainly with antisense oligonucleotides administered by the intravenous or subcutaneous route.
Unscheduled termination was reported in 14% of studies, generally related to kidney or liver findings (mostly in rodents) and systemic inflammation (mostly in NHPs).
Clinical pathology changes included liver findings (increased transaminases in 40% of the studies, mostly in NHP), renal findings (increased creatinine and urea in 30% of the studies, mostly in rats), and hematology findings (thrombocytopenia in 30% of the studies, mostly in rodents). Complement activation was noted in 80% of NHP studies.
Gross findings were observed at the injection site, kidney, liver, and spleen. Microscopically, basophilic granules were observed in the kidney, liver, lymph nodes and spleen, and were considered to reflect the accumulation of the oligonucleotides. Degenerative findings were observed in the liver (80% of NHP studies, 90% in mice, 80% in rats) and kidney (85% in NHP, 75% in mice, 90% in rats).
The results of this retrospective review confirm that there are both similarities and species differences in the toxicity profile in laboratory animal species.
P34 | Guide for Combining Primary Tumors for Statistical Analysis in Carcinogenicity Studies
Charlotte Keenan 1 , Alessandro Piaia 2 , Muthafar Al-Haddawi 3 , Jean-Guy Bienvenu 4 , Alys Bradley 5 , Paul Brown 6 , Hepei Chen 6 , Karyn Colman 7 , Michael Elwell 8 , Nicholas Gatto 9 , Dawn Goodman 10 , Binod Jacob 9 , Lynda Lanning 6 , LuAnn McKinney 6 , Erin Muhlbradt 11 , Rick Perry 12 , Karen Regan 13 , Benjamin Sefing 9 , Michael Thibodeau 14 , Erin Tibbs-Slone 15 , Jochen Woicke 16 , Craig Zwickl 17
1 C.M. Keenan ToxPath Consulting, Doylestown, United States
2 Novartis, Basel, Switzerland
3 Bristol-Myers Squibb, New Brunswick, United States
4 Charles River Laboratories, Senneville, Canada
5 Charles River Laboratories, Edinburgh, United Kingdom
6 US Food and Drug Administration, Silver Spring, United States
7 Novartis, Cambridge, United States
8 Independent Consultant, Apex, United States
9 Merck & Co, Inc, Kenilworth, United States
10 Independent Consultant, Providence, United States
11 Medical Science and Computing, a Guidehouse Company, Rockville, United States
12 Pfizer, Groton, United States
13 Regan Path/Tox Services, Kewadin, United States
14 Boehringer Ingelheim Pharmaceuticals, Inc, Ridgefield, United States
15 Charles River Laboratories, Mattawan, United States
16 Independent Consultant, Scottsdale, United States
17 Independent Consultant, Indianapolis, United States
This guide was created at the request of the U. S. Food and Drug Administration (FDA) by a Working Group of biopharmaceutical experts from the Society of Toxicologic Pathology (STP), FDA, and members of the Standard for Exchange of Nonclinical Data (SEND) initiative, with a primary goal of assisting pharmacology/toxicology reviewers and biostatisticians in statistical analysis of nonclinical tumor data. The guide will also be useful to study and peer review pathologists in interpreting tumor data. This guide provides a user friendly, higher-level hierarchy of tumor types or categories correlating the tumor names from the International Harmonization of Nomenclature and Diagnostic Criteria (INHAND) publications with those available in the NEOPLASM controlled terminology (CT) code list in SEND. The version of CT used in a study should be referenced in the nonclinical study data reviewer’s guide (SDRG) (section 3.1) of electronic submissions to the FDA. The tumor combination guide instructions and examples are in a tabular format to make informed decisions for combining tumor data for statistical analysis. The guide is targeted for publication on the Clinical Data Interchange Standards Consortium (CDISC) website by the end of 2023. The strategy for combining tumor types for statistical analysis is based on scientific criteria gleaned from the current scientific literature; as SEND and INHAND terminology and information evolve, this guide will be updated.
P35 | AI-based testicular staging in dogs from nonclinical toxicity studies
Leonie Herkommer 1 , Maximilian Gottschalk 2 , Miriam Hägele 2 , Katja Lingelbach 2 , Yumi Kangawa 1 , Stefanie Arms 1 , Cornelius Böhm 2 , Sharon Ruane 2 , Maximilian Alber 2 , Gabriele Pohlmeyer-Esch 1
1 Boehringer Ingelheim Pharma GmBH & Co.KG, Ingelheim am Rhein, Germany
2 Aignostics GmbH, Berlin, Germany
Spermatogenic staging of tubules via histologic examination is typically used to evaluate testicular toxicity in non-clinical studies. However, this process is time-consuming, non-exhaustive, and shows inter-observer variability between pathologists. Artificial intelligence (AI) offers a standardized, efficient method for performing this analysis at scale.
AI models for tubule detection and classification were developed using 65 whole slide images (WSIs) of only H&E-stained testes from sexually mature beagle dogs. To ensure robust model performance, control and treated dogs with diagnosed testicular findings were included. Two toxicologic pathologists provided labels for the model training and evaluation. 10 WSIs (7 control, 3 treated) were held out as an evaluation set. The classification model assigned each detected tubule to the partially pooled stages: stage 1/2, 3, 4, 5, 6/7, and 8, and additional categories: “artifact”, “tangential”, and “abnormal”. Additionally, the tubule detection model output was used to assess tubular dilatation.
The classification model performed on par with current research (0.87 F1 score, 0.87 precision, 0.88 recall). The model performed similarly across control and treated groups and generalized across varying staining intensities and tissue qualities. Application of the models showed significant differences between control and treated cases with increased tubular dilatation and a higher share of “abnormal” tubules.
This AI approach quantifies spermatogenic stages, artifacts, tangential and abnormal tubules in canine testes, and provides standardized assessment of useful metrics such as tubular diameter. This rapid, unbiased and scalable approach has the potential to significantly enhance testicular toxicity evaluation in non-clinical studies.
In the future the development of comparable open source models could significantly facilitate toxicological assessments.
P36 | Female Reproductive System Changes in Sprague-Dawley Rats Following Intraperitoneal Injection of Sesame Oil as Vehicle in Dual Route Studies
Anikethana Ramesh, Rajesh Karunanithi, Manohar Madhav Deshmukh, Kiran Hosaholalu Jayaram, Srinivasa Prahalada, Mohan Krishnappa, Kalaiselvan Ponnusamy
Syngene International Limited, Bangalore, India
Pre-clinical evaluation of certain medical devices requires administration of test item extract through intravenous (IV) and intraperitoneal (IP) routes. This study summarizes the microscopic changes noted in female reproductive system of Sprague-Dawley rats from 30-day dual route IV/IP studies in comparison with IV only studies.
Data was collected from 284 females aged 12–16 weeks across 11 studies for IV/IP route and 280 females aged 12–16 weeks across 10 studies for IV route. Vehicle for test item extract preparation for IV and IP studies were normal saline and sesame oil, respectively.
In dual route IV/IP studies, microscopic changes included decidual reaction (9.85%) in uterus, vaginal mucification (7.04%), mammary gland hyperplasia (12.67%), uterine endometrial folding (3.16%) and cervical mucification (2.11%).
In I/V only studies, microscopic changes included a single incidence of decidual reaction (0.35%) in uterus, vaginal mucification (2.5%), mammary gland hyperplasia (3.2%), uterine endometrial folding (0.7%) and cervical mucification (0.7%).
The increased incidences of uterine decidual reaction and related changes in IV/IP studies in vehicle control and test item groups suggest the role of sesame oil. As per literature, intrauterine instillation of sesame oil results in induction of deciduoma in rats. Sesame oil injected in the peritoneal cavity migrated to the uterine lumen through the ovarian bursa and fallopian tube. Repeated restraining of animals for test item extract administration may also lead to pseudopregnancy, which in turn may predispose to decidual reactions. Hence, the changes in female reproductive organs were primarily attributed to sesame oil in dual route studies.
P38 | CystAssist: Automated AI quantification of kidney features in preclinical murine model of ADPKD
Brandon Ginley 1 , Brendon Lutnick 1 , Tao Chen 2 , Mathieu Marella 3 , Wen Zhou 4 , Vaja Chipashvili 4 , Rajashree Rana 4 , Nikolay Bukanov 4 , Anna-Lena Frisk 5 , Albert Juan Ramon 1 , Stephen Yip 1 , Hoang Lu 1 , Kristopher Standish 1
1 Janssen R&D Data Sciences, Raritan, United States
2 Janssen Pathology, Shanghai, China
3 Janssen Pathobiology, La Jolla, United States
4 Janssen CVMRPH, Boston, United States
5 Janssen R&D Pathobiology, Beerse, Belgium
Autosomal dominant polycystic kidney disease (ADPKD) is one of the most common life-threatening genetic diseases, affecting 1 in 400-1000 people. We have developed an automated tool to segment and quantify kidney features in a murine model of ADPKD.
We used a juvenile cystic kidneys mouse model: a recessive mutation in the mouse which causes the development of polycystic kidney disease. To examine the relationship between disease progression by age and cystic index readout in mice, we collected kidneys at the age of 5-, 8-, 10-, 12-, 14-weeks. Kidneys were sectioned at 5 µm, stained with hematoxylin and eosin, and scanned with an Aperio whole slide scanner at 40x magnification. 58 sections and corresponding pathologist-based cyst annotations were used for training a neural network to detect cystic regions, papilla, and fat in each kidney; 32 sections were used for validation.
To validate the performance of this algorithm, kidney segmentations were reviewed and corrected manually by a pathologist. In this analysis, the CystAssist segmentation algorithm scored an average sensitivity of 0.92 and specificity 0.99.
We have developed a platform to precisely quantify kidney features. This tool saves the pathologist valuable work time, allowing them to concentrate on more difficult, scientifically challenging, and interesting tasks. The same platform can segment and phenotype glomeruli and tubules, and enumerate podocytes, forming a comprehensive platform for evaluating kidney tissue in health and disease. The segmentation tool described utilizes the open-source work: https://github.com/SarderLab/Histo-cloud.
P44 | Detection of mRNA in formalin-fixed paraffin-embedded cochlea from monkeys using in situ hybridization (ISH) techniques
Frédéric Gervais 1 , Alizée Hugedé 2 , Johanna Dauphin 2 , Sylvie Bonavaud 2 , Laurent Desire 3
1 Charles River Laboratories Evreux, Miserey, France
2 Charle River Laboratories, Evreux, France
3 Sensorion-Pharma, Montpellier, France
Toxicity and biodistribution of gene therapy products must be evaluated. While PCR techniques remain the gold standard to quantifying vector DNA / transgenic RNAs, ISH techniques have the advantage of providing a morphological context but are often considered as challenging to develop, poorly sensitive and not quantitative, especially in tissues difficult to process such as the cochlea.
Six non-human primates received a single intracochlear injection of an AAV vector engineered for the treatment of a specific condition leading to hearing loss and were terminated after a 3-month recovery period. The cochlea were sampled, fixed, decalcified, paraffin-embedded and processed to slides. Detection of endogenous and transgenic mRNAs were attempted by fluorescence using the RNAscope technology.
The mRNAs of interest were localized in the target cells as expected, providing valuable information. Several parameters had to be optimized, including the decalcification step, the ISH protocol, slide mounting and image acquisition, especially to adjust for tissue autofluorescence.
ISH assays can be performed on fresh frozen, fixed frozen and FFPE samples. Formalin and paraffin embedding, while providing excellent morphology, may cause up to 30% of RNA loss; which is detrimental to sensitivity. These drawbacks were addressed using the RNAscope technology; which is compatible with partially degraded RNAs and stands out for its high sensitivity.
RNAscope labeling allows localization, which is complementary to the in-solution genomic approaches, quantification if combined to image analysis, allows to visualize multiple transcripts simultaneously, and could be combined with IHC in order to visualized RNA and proteins in the same section.
P45 | Comparison of two methods (ISH versus IHC) to detect human cells in murine tissue sections
Frédéric Gervais 1 , Alizée Hugedé 2 , Catherine Rochereau 2 , Olivier Foulon 2
1 Charles River Laboratories Evreux, Miserey, France
2 Charle River Laboratories, Evreux, France
Before being marketed, safety and efficacy of any cell therapy must be assessed. Biodistribution and persistence of therapeutic cells can be investigated by sensitive molecular biology techniques but also by IHC or ISH assays that provide morphological context. The performance of such markers in FFPE tissues were compared in a toxicity/biodistribution study where human therapeutic cells were administered to NSG mice. More specifically, two markers were compared, including sensitivity, specificity, ease of use and cost.
32 mice were injected intravenously 0 or 5.106 human therapeutic cells developed to regulate the immune system. Mice were euthanized on Days 7 and 28 and several organs investigated for the presence of therapeutic cells using IHC or ISH markers (KU80 and ALU, respectively). All slides were evaluated histologically, digitized, and analyzed (Visiopharm software). The total number of positive cells was counted.
On Day 7, both ALU and KU80 were relevant in labelling the human cells in the lungs, whereas ALU was better on Day 28 to detect rare events. Image analysis confirmed that ALU hybridization was more sensitive and specific than KU80 immunolabeling.
In conclusion, KU80 and ALU detection were both helpful in assessing the limited persistence of human cells in mice over time. The choice of the marker depends on the number of cells, the quantification strategy, and the cost.
Supplemental Material
sj-zip-1-tpx-10.1177_01926233241236845 – Supplemental material for Poster Abstracts of the 20th Congress of the European Society of Toxicologic Pathology (Basel, Switzerland, 26th - 29th of September 2023, Emerging Therapeutic Modalities)
Supplemental material, sj-zip-1-tpx-10.1177_01926233241236845 for Poster Abstracts of the 20th Congress of the European Society of Toxicologic Pathology (Basel, Switzerland, 26th - 29th of September 2023, Emerging Therapeutic Modalities) by Marco Magnani in Toxicologic Pathology
Supplementary Material
Please find the following supplemental material available below.
For Open Access articles published under a Creative Commons License, all supplemental material carries the same license as the article it is associated with.
For non-Open Access articles published, all supplemental material carries a non-exclusive license, and permission requests for re-use of supplemental material or any part of supplemental material shall be sent directly to the copyright owner as specified in the copyright notice associated with the article.
