Editorial: Verification of Animal Pain Models by Reverse Translation

Abstract

The validity of preclinical models in evaluating and developing potential analgesics has been an area of discussion for over a decade. The poor success of these models in predicting clinical efficacy (Woolf, 2010; Barrett, 2015; Burma et al., 2017) leads to the question as to whether the fault lies in the models not invoking the same pathology as the diseases being modeled, or if the failure is due to the differences in the physiology, anatomy, and pharmacology of the model species versus humans. An additional factor is that the design of proof-of-concept clinical trials of novel analgesic agents may have been sub-optimal (Hijma and Froeneveld, 2021). Ideally, the animal models being used for analgesic discovery would have the same pathophysiology as the human disease and would predict with 100 percent accuracy the efficacy of an agent in human trials. This ideal, however, is not realistic and compromises are necessary. However, repeated verification and optimization of the models can inform investigators of the limitations of the models, and by utilizing multiple models a realistic evaluation of the therapeutic potential of an agent can be realized. This research topic is devoted to outlining methods to increase confidence in the translatability of the data obtained in animal models of human pain conditions. An important aspect of this process is to use known information about human disease to improve the animal models, create new models, and to eliminate non-productive models. The papers in this research topic can be divided into three general categories of discussion (see Figure 1). The first category evaluates rodent models by reverse translation. Known therapies or known characteristics of the disease in humans are examined in established models to determine if there is a correlation between humans and rodents. The review by Fisher et al. evaluates neuropathic pain models and argues that matching quantifiable endpoints between the model and humans is critical to improving the translatability of the data obtained from the model. They further suggest, as do several of the authors in the topic, that pain suppressed behaviors may be superior to pain enhanced behaviors when determining therapeutic potential of an agent. The original research by Negus et al. addresses the affective/motivational aspect of pain when using a rodent model. In humans the distress induced by pain is often the most critically relevant feature of pain that determines the duration of disability or the quality of life for a patient. The authors present data indicating that rodents have a weak affective/motivational response to injury, suggesting that rodents may only model the sensory/discriminative aspect of pain; thus, making them less relevant as human models. This conclusion is countered by Cho et al. who suggest that a complete evaluation of the biopsychosocial model of pain in rodents is possible and can lead to a better understanding of pain and pain control in humans. They present methods of evaluating rodent behavior that provide insight into the psychological and social aspects of pain. The review by Pineda-Farias et al. evaluates models of cancer pain and provides hope that greater knowledge of how cancer and cancer therapy produce pain in humans will inform modeling in animals. The interesting review by Shen et al. Edited and reviewed by: Alastair George Stewart, The University of Melbourne, Australia