AlzPED: Improving the predictive power and translational validity of preclinical testing of candidate therapeutics in Alzheimer’s disease animal models

Abstract

Background AlzPED is a publicly available database created by the National Institute on Aging to address key factors contributing to poor translation of preclinical efficacy from animal models to the clinic in Alzheimer’s disease (AD) therapy development. Specifically, AlzPED is missioned to identify critical experimental design elements and methodology missing from studies that make them susceptible to misinterpretation and reduce their reproducibility and translational value, as well as serve as a platform for reporting unpublished negative findings. AlzPED provides funding agencies with a tool for enforcement of the requirements for transparent reporting and rigorous study design. Through these capabilities, AlzPED is intended to guide the development and implementation of strategies and recommendations for standardized best practices for the rigorous preclinical testing of AD candidate therapeutics and mitigate the publication bias that favors the reporting of positive findings. Method Using key word‐driven literature searches published studies are acquired and curated by two experts for data on authors, funding source, AD animal models, therapeutic targets and therapeutic agents, study design, and outcome measures, prior to publication in the database. Rigor in study design is evaluated with an experimental design scorecard that summarizes the elements of experimental design reported in each study in the database. Result AlzPED currently houses curated summaries from 917 published studies. Summaries are searchable by author, funding source, animal model, therapeutic target and therapeutic agent. At present, the database contains data on 185 animal models, 175 therapeutic targets, 804 therapeutic agents and, more than 1500 AD‐related outcome measures. Evaluation of experimental design scorecards from each study demonstrates significant under‐reporting of critical elements of methodology even in high impact factor journals and highly cited published preclinical research. Conclusion Analysis of curated studies demonstrates serious deficiencies in reporting critical elements of methodology such as power/sample size calculation, blinding for treatment/outcomes, randomization, balancing for sex, animal genetic background and others, irrespective of journal impact factor and number of citations. These deficiencies diminish the scientific rigor, reproducibility and translational value of the preclinical studies. Thus, it is evident that a standardized set of best practices is required for successful translation of therapeutic efficacy in AD research.

Investigating the replicability of preclinical cancer biology

“…Research into replicability in other disciplines has also found that findings with stronger initial evidence (such as larger effect sizes and/or smaller p-values) is more likely to replicate ( Nosek et al, 2021 ; Open Science Collaboration, 2015 ), and it may be worth exploring if other findings from other disciplines – such as more surprising findings being less likely to replicate – can be generalized to cancer biology. There are also unique qualities of research in cancer biology that could be related to replicability, and a number of ongoing projects exploring replication in preclinical research ( Amaral et al, 2019 ; Drude et al, 2021 ) will add to the data presented here and increase our understanding of replication and translational success ( Chakroborty et al, 2020 ). To facilitate such investigation, we have made all of the data for our meta-analysis available for reanalysis.…”