1 December 2017 Meeting Minutes
A teleconference/webex was held on 1 December 2017. The meeting reviewed and discussed current activities in which histopathology data from 2 different compounds (both mTORC1/C2 inhibitors) from the eTOX database were obtained and entered into the Histographic visual display tool. Both compounds were from AstraZeneca and provided to eTOX (non-confidential), with pharmacology and Phase 1 cancer clinical trial data published. A Brown gave an overview of the eTOX data sets and published references (which will be shared with working group). K Snyder gave an introduction to the software and provided demonstration of the visual displays produced. A unique figure was created for each compound (AZD2014 and AZD8055). Two rat and 3 dog studies are present for compound AZD8055, and 11 rat and 2 dog studies are present for AZD2014. The following hierarchy was generated, from inner ring to outer rings: 1) tissue, 2) histo finding, 3) species, 4) dose, 5) sex, 6) period. Data from eTOX spreadsheet was copied and pasted into software in categories 1, 2, 3, etc, based on hierarchy from left to right. The last column represents the outer edge. The figures/display needs to be viewed using Google Chrome or Firefox web browser. Errors exist in eTOX database regarding number of animals. Kevin working on calculating average incidence for findings. Incidence data currently displayed based on ratios of animals with a tissue finding to total number of animals per group. Control groups were included if histopath findings present for a tissue. The eTOX data does not clearly differentiate between post-treatment and reversal phase cohorts. Question arose as to whether Histographic tool can utilize SEND data. Kevin evaluating this question. Question arose as to whether tissues can be de-selected from the display (such as those typically affected by stress). Can in-life observations or clinical signs data be displayed using Histographic ? Depends upon time-course element, and whether aggregate incidence data can be obtained, ie, number of animals per treatment group with a particular observation. The Histographic tool provides a significant advantage for displaying pathology findings from the eTOX database as currently there is no visual display tool available for eTOX. In addition, proof of concept for aggregating and displaying pathology data from multiple toxicology studies was achieved with these case examples. A suggestion was made that scientific questions should be identified, followed by generation of a visual display to help address those questions. A list of questions was generated. Questions Toxicologists Ask of Histopathology Datasets • What are the target organs? o Can we rank order them? • What findings are present in these organs? o Was there a dose-relationship? o Was there a difference between treated and controls? o Was the finding reversible? o Was the finding present in both males and females? o Which species did these findings occur in? o Is there a relationship with study duration? • Are there dose-related findings of great importance at relatively low incidence? • How do the findings compare across compounds? • Can we differentiate stress-related findings (potentially remove/isolate these findings from analysis)? o Typical Stress Organs: Thymus, lymphoid tissues, adrenal glands (chronic stress), stomach ulceration/erosion, bone marrow o Can we select tissues to display ? Next steps for the project team are to evaluate the figures for each compound to solicit further input, and to compare data with published Phase 1 cancer trial adverse events. The team's goal is to prepare an abstract for a poster presentation at the 2018 FDA PhUSE CSS in Silver Spring, MD. A proposed focus of the poster will be use of Histographic for displaying pathology data from the eTOX database, and proof of concept for aggregating and displaying data from multiple studies in a singe tool/figure.